diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,140033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 25.6,
+  "eval_steps": 500,
+  "global_step": 20000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00128,
+      "grad_norm": 0.6006259322166443,
+      "learning_rate": 5e-06,
+      "loss": 1.1793,
+      "step": 1
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 0.6071895956993103,
+      "learning_rate": 1e-05,
+      "loss": 1.2464,
+      "step": 2
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 0.584945797920227,
+      "learning_rate": 1.5e-05,
+      "loss": 1.2501,
+      "step": 3
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 0.6185263395309448,
+      "learning_rate": 2e-05,
+      "loss": 1.285,
+      "step": 4
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 0.6382588744163513,
+      "learning_rate": 2.5e-05,
+      "loss": 1.2345,
+      "step": 5
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 0.6335554122924805,
+      "learning_rate": 3e-05,
+      "loss": 1.2112,
+      "step": 6
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 0.619415283203125,
+      "learning_rate": 3.5e-05,
+      "loss": 1.2691,
+      "step": 7
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 0.6417765617370605,
+      "learning_rate": 4e-05,
+      "loss": 1.2458,
+      "step": 8
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 0.5834643244743347,
+      "learning_rate": 4.5e-05,
+      "loss": 1.119,
+      "step": 9
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.5374929308891296,
+      "learning_rate": 5e-05,
+      "loss": 1.0867,
+      "step": 10
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 0.5630415678024292,
+      "learning_rate": 4.9997999199679874e-05,
+      "loss": 1.1491,
+      "step": 11
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 0.5184459090232849,
+      "learning_rate": 4.9995998399359746e-05,
+      "loss": 1.1131,
+      "step": 12
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 0.5558237433433533,
+      "learning_rate": 4.999399759903962e-05,
+      "loss": 1.1601,
+      "step": 13
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 0.5303581953048706,
+      "learning_rate": 4.999199679871949e-05,
+      "loss": 1.0869,
+      "step": 14
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.570815920829773,
+      "learning_rate": 4.998999599839936e-05,
+      "loss": 1.074,
+      "step": 15
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 0.5760554075241089,
+      "learning_rate": 4.9987995198079234e-05,
+      "loss": 1.1381,
+      "step": 16
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 0.565275251865387,
+      "learning_rate": 4.9985994397759105e-05,
+      "loss": 1.1634,
+      "step": 17
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 0.5491355061531067,
+      "learning_rate": 4.998399359743898e-05,
+      "loss": 1.0564,
+      "step": 18
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 0.5034635663032532,
+      "learning_rate": 4.998199279711885e-05,
+      "loss": 0.9781,
+      "step": 19
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.5160910487174988,
+      "learning_rate": 4.997999199679872e-05,
+      "loss": 1.0313,
+      "step": 20
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 0.5402336716651917,
+      "learning_rate": 4.99779911964786e-05,
+      "loss": 1.1054,
+      "step": 21
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.4777592420578003,
+      "learning_rate": 4.9975990396158465e-05,
+      "loss": 0.9927,
+      "step": 22
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.4950423240661621,
+      "learning_rate": 4.9973989595838337e-05,
+      "loss": 1.0234,
+      "step": 23
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 0.4756661057472229,
+      "learning_rate": 4.997198879551821e-05,
+      "loss": 0.9862,
+      "step": 24
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.47625282406806946,
+      "learning_rate": 4.996998799519808e-05,
+      "loss": 0.9818,
+      "step": 25
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 0.4371047914028168,
+      "learning_rate": 4.996798719487795e-05,
+      "loss": 0.9961,
+      "step": 26
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 0.4368005692958832,
+      "learning_rate": 4.9965986394557824e-05,
+      "loss": 0.9859,
+      "step": 27
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 0.4649926424026489,
+      "learning_rate": 4.99639855942377e-05,
+      "loss": 1.0016,
+      "step": 28
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 0.45246022939682007,
+      "learning_rate": 4.9961984793917574e-05,
+      "loss": 1.016,
+      "step": 29
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.41020557284355164,
+      "learning_rate": 4.995998399359744e-05,
+      "loss": 0.9839,
+      "step": 30
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.39636656641960144,
+      "learning_rate": 4.995798319327731e-05,
+      "loss": 0.9041,
+      "step": 31
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 0.4260682165622711,
+      "learning_rate": 4.995598239295718e-05,
+      "loss": 1.0016,
+      "step": 32
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 0.4026011526584625,
+      "learning_rate": 4.9953981592637055e-05,
+      "loss": 0.8832,
+      "step": 33
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 0.3997635841369629,
+      "learning_rate": 4.995198079231693e-05,
+      "loss": 0.968,
+      "step": 34
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.37634411454200745,
+      "learning_rate": 4.9949979991996806e-05,
+      "loss": 0.9207,
+      "step": 35
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 0.3642372786998749,
+      "learning_rate": 4.994797919167668e-05,
+      "loss": 0.8979,
+      "step": 36
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 0.3647741675376892,
+      "learning_rate": 4.994597839135655e-05,
+      "loss": 0.8897,
+      "step": 37
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 0.3974230885505676,
+      "learning_rate": 4.9943977591036414e-05,
+      "loss": 0.9336,
+      "step": 38
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 0.3915899693965912,
+      "learning_rate": 4.9941976790716286e-05,
+      "loss": 0.9408,
+      "step": 39
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.40685927867889404,
+      "learning_rate": 4.993997599039616e-05,
+      "loss": 0.9166,
+      "step": 40
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 0.35324081778526306,
+      "learning_rate": 4.993797519007603e-05,
+      "loss": 0.8683,
+      "step": 41
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 0.36212658882141113,
+      "learning_rate": 4.993597438975591e-05,
+      "loss": 0.8828,
+      "step": 42
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 0.38610896468162537,
+      "learning_rate": 4.993397358943578e-05,
+      "loss": 0.9092,
+      "step": 43
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 0.36929619312286377,
+      "learning_rate": 4.993197278911565e-05,
+      "loss": 0.9132,
+      "step": 44
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.3837580680847168,
+      "learning_rate": 4.9929971988795524e-05,
+      "loss": 0.9606,
+      "step": 45
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 0.36300909519195557,
+      "learning_rate": 4.992797118847539e-05,
+      "loss": 0.8827,
+      "step": 46
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 0.38039928674697876,
+      "learning_rate": 4.992597038815526e-05,
+      "loss": 0.8996,
+      "step": 47
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.3811199367046356,
+      "learning_rate": 4.992396958783513e-05,
+      "loss": 0.9114,
+      "step": 48
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 0.36827829480171204,
+      "learning_rate": 4.992196878751501e-05,
+      "loss": 0.8199,
+      "step": 49
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.35475796461105347,
+      "learning_rate": 4.9919967987194883e-05,
+      "loss": 0.8741,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 0.3906758725643158,
+      "learning_rate": 4.9917967186874755e-05,
+      "loss": 0.9037,
+      "step": 51
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 0.3880329430103302,
+      "learning_rate": 4.991596638655463e-05,
+      "loss": 0.9474,
+      "step": 52
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 0.38061219453811646,
+      "learning_rate": 4.99139655862345e-05,
+      "loss": 0.894,
+      "step": 53
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.39929622411727905,
+      "learning_rate": 4.9911964785914364e-05,
+      "loss": 0.9322,
+      "step": 54
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.38368162512779236,
+      "learning_rate": 4.9909963985594236e-05,
+      "loss": 0.9286,
+      "step": 55
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 0.3783194124698639,
+      "learning_rate": 4.9907963185274115e-05,
+      "loss": 0.9337,
+      "step": 56
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.3646315932273865,
+      "learning_rate": 4.9905962384953986e-05,
+      "loss": 0.8836,
+      "step": 57
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 0.3632262349128723,
+      "learning_rate": 4.990396158463386e-05,
+      "loss": 0.8382,
+      "step": 58
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 0.41220006346702576,
+      "learning_rate": 4.990196078431373e-05,
+      "loss": 0.9255,
+      "step": 59
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.39080944657325745,
+      "learning_rate": 4.98999599839936e-05,
+      "loss": 0.8664,
+      "step": 60
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 0.37954285740852356,
+      "learning_rate": 4.9897959183673474e-05,
+      "loss": 0.8716,
+      "step": 61
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 0.40623852610588074,
+      "learning_rate": 4.989595838335334e-05,
+      "loss": 0.9211,
+      "step": 62
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.3674965500831604,
+      "learning_rate": 4.989395758303322e-05,
+      "loss": 0.8427,
+      "step": 63
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 0.3909313380718231,
+      "learning_rate": 4.989195678271309e-05,
+      "loss": 0.8804,
+      "step": 64
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.397886723279953,
+      "learning_rate": 4.988995598239296e-05,
+      "loss": 0.8772,
+      "step": 65
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.38570407032966614,
+      "learning_rate": 4.988795518207283e-05,
+      "loss": 0.8593,
+      "step": 66
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 0.4147164225578308,
+      "learning_rate": 4.9885954381752705e-05,
+      "loss": 0.9182,
+      "step": 67
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 0.3853755593299866,
+      "learning_rate": 4.988395358143258e-05,
+      "loss": 0.8583,
+      "step": 68
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.3911557197570801,
+      "learning_rate": 4.988195278111245e-05,
+      "loss": 0.8066,
+      "step": 69
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.3869986832141876,
+      "learning_rate": 4.9879951980792314e-05,
+      "loss": 0.8863,
+      "step": 70
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 0.425576388835907,
+      "learning_rate": 4.987795118047219e-05,
+      "loss": 0.9285,
+      "step": 71
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.4300587773323059,
+      "learning_rate": 4.9875950380152064e-05,
+      "loss": 0.8604,
+      "step": 72
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 0.4201204478740692,
+      "learning_rate": 4.9873949579831936e-05,
+      "loss": 0.8818,
+      "step": 73
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 0.41632699966430664,
+      "learning_rate": 4.987194877951181e-05,
+      "loss": 0.8951,
+      "step": 74
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.398713082075119,
+      "learning_rate": 4.986994797919168e-05,
+      "loss": 0.8571,
+      "step": 75
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 0.40266332030296326,
+      "learning_rate": 4.986794717887155e-05,
+      "loss": 0.8546,
+      "step": 76
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 0.41864725947380066,
+      "learning_rate": 4.9865946378551424e-05,
+      "loss": 0.8228,
+      "step": 77
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.3853597044944763,
+      "learning_rate": 4.9863945578231295e-05,
+      "loss": 0.8282,
+      "step": 78
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 0.44597557187080383,
+      "learning_rate": 4.986194477791117e-05,
+      "loss": 0.9163,
+      "step": 79
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.4048195481300354,
+      "learning_rate": 4.985994397759104e-05,
+      "loss": 0.8456,
+      "step": 80
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.4152766466140747,
+      "learning_rate": 4.985794317727091e-05,
+      "loss": 0.8562,
+      "step": 81
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 0.4202408492565155,
+      "learning_rate": 4.985594237695078e-05,
+      "loss": 0.8878,
+      "step": 82
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 0.4306046962738037,
+      "learning_rate": 4.9853941576630655e-05,
+      "loss": 0.8816,
+      "step": 83
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.4268186688423157,
+      "learning_rate": 4.9851940776310527e-05,
+      "loss": 0.7838,
+      "step": 84
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.4262542128562927,
+      "learning_rate": 4.98499399759904e-05,
+      "loss": 0.812,
+      "step": 85
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 0.44477730989456177,
+      "learning_rate": 4.984793917567027e-05,
+      "loss": 0.8684,
+      "step": 86
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.4398692548274994,
+      "learning_rate": 4.984593837535014e-05,
+      "loss": 0.8871,
+      "step": 87
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.4244236350059509,
+      "learning_rate": 4.9843937575030014e-05,
+      "loss": 0.8801,
+      "step": 88
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 0.47705894708633423,
+      "learning_rate": 4.9841936774709886e-05,
+      "loss": 0.8609,
+      "step": 89
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.44810983538627625,
+      "learning_rate": 4.983993597438976e-05,
+      "loss": 0.8376,
+      "step": 90
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 0.41785943508148193,
+      "learning_rate": 4.983793517406963e-05,
+      "loss": 0.8113,
+      "step": 91
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 0.44845548272132874,
+      "learning_rate": 4.98359343737495e-05,
+      "loss": 0.8438,
+      "step": 92
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.4400695562362671,
+      "learning_rate": 4.983393357342937e-05,
+      "loss": 0.8242,
+      "step": 93
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.4483565092086792,
+      "learning_rate": 4.9831932773109245e-05,
+      "loss": 0.8877,
+      "step": 94
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.46313560009002686,
+      "learning_rate": 4.982993197278912e-05,
+      "loss": 0.8162,
+      "step": 95
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.4211414158344269,
+      "learning_rate": 4.982793117246899e-05,
+      "loss": 0.807,
+      "step": 96
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.44126561284065247,
+      "learning_rate": 4.982593037214886e-05,
+      "loss": 0.8403,
+      "step": 97
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.4854923486709595,
+      "learning_rate": 4.982392957182873e-05,
+      "loss": 0.8869,
+      "step": 98
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.5096859335899353,
+      "learning_rate": 4.982192877150861e-05,
+      "loss": 0.9123,
+      "step": 99
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.4502103328704834,
+      "learning_rate": 4.9819927971188476e-05,
+      "loss": 0.834,
+      "step": 100
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 0.4698023796081543,
+      "learning_rate": 4.981792717086835e-05,
+      "loss": 0.8685,
+      "step": 101
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.464958131313324,
+      "learning_rate": 4.981592637054822e-05,
+      "loss": 0.8724,
+      "step": 102
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 0.44018036127090454,
+      "learning_rate": 4.981392557022809e-05,
+      "loss": 0.8166,
+      "step": 103
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 0.4710625112056732,
+      "learning_rate": 4.9811924769907964e-05,
+      "loss": 0.9041,
+      "step": 104
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.45761746168136597,
+      "learning_rate": 4.9809923969587836e-05,
+      "loss": 0.8414,
+      "step": 105
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.43286818265914917,
+      "learning_rate": 4.9807923169267714e-05,
+      "loss": 0.827,
+      "step": 106
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 0.4683651030063629,
+      "learning_rate": 4.9805922368947586e-05,
+      "loss": 0.8831,
+      "step": 107
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.4688320457935333,
+      "learning_rate": 4.980392156862745e-05,
+      "loss": 0.8693,
+      "step": 108
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.4783925712108612,
+      "learning_rate": 4.980192076830732e-05,
+      "loss": 0.8086,
+      "step": 109
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.463559091091156,
+      "learning_rate": 4.9799919967987195e-05,
+      "loss": 0.8274,
+      "step": 110
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.5109454989433289,
+      "learning_rate": 4.979791916766707e-05,
+      "loss": 0.9021,
+      "step": 111
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 0.44130197167396545,
+      "learning_rate": 4.979591836734694e-05,
+      "loss": 0.8358,
+      "step": 112
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.45340853929519653,
+      "learning_rate": 4.979391756702682e-05,
+      "loss": 0.8054,
+      "step": 113
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.496158629655838,
+      "learning_rate": 4.979191676670669e-05,
+      "loss": 0.8812,
+      "step": 114
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.4708779454231262,
+      "learning_rate": 4.978991596638656e-05,
+      "loss": 0.7753,
+      "step": 115
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.5166846513748169,
+      "learning_rate": 4.9787915166066426e-05,
+      "loss": 0.9508,
+      "step": 116
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.4957675635814667,
+      "learning_rate": 4.97859143657463e-05,
+      "loss": 0.8593,
+      "step": 117
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 0.4897170066833496,
+      "learning_rate": 4.978391356542617e-05,
+      "loss": 0.8419,
+      "step": 118
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.4425666630268097,
+      "learning_rate": 4.978191276510604e-05,
+      "loss": 0.8184,
+      "step": 119
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.4843791723251343,
+      "learning_rate": 4.977991196478592e-05,
+      "loss": 0.8335,
+      "step": 120
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.4781690239906311,
+      "learning_rate": 4.977791116446579e-05,
+      "loss": 0.8835,
+      "step": 121
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.5279680490493774,
+      "learning_rate": 4.9775910364145664e-05,
+      "loss": 0.8627,
+      "step": 122
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.525348961353302,
+      "learning_rate": 4.9773909563825536e-05,
+      "loss": 0.9173,
+      "step": 123
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.4671289324760437,
+      "learning_rate": 4.97719087635054e-05,
+      "loss": 0.8189,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48094815015792847,
+      "learning_rate": 4.976990796318527e-05,
+      "loss": 0.8514,
+      "step": 125
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.49200230836868286,
+      "learning_rate": 4.9767907162865145e-05,
+      "loss": 0.8669,
+      "step": 126
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 0.45223578810691833,
+      "learning_rate": 4.976590636254502e-05,
+      "loss": 0.811,
+      "step": 127
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.48548707365989685,
+      "learning_rate": 4.9763905562224895e-05,
+      "loss": 0.8733,
+      "step": 128
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.4744742214679718,
+      "learning_rate": 4.976190476190477e-05,
+      "loss": 0.82,
+      "step": 129
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.47322678565979004,
+      "learning_rate": 4.975990396158464e-05,
+      "loss": 0.8076,
+      "step": 130
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.5006943345069885,
+      "learning_rate": 4.975790316126451e-05,
+      "loss": 0.9173,
+      "step": 131
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.5181525945663452,
+      "learning_rate": 4.9755902360944376e-05,
+      "loss": 0.8946,
+      "step": 132
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.4706552028656006,
+      "learning_rate": 4.975390156062425e-05,
+      "loss": 0.8297,
+      "step": 133
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.5205573439598083,
+      "learning_rate": 4.9751900760304126e-05,
+      "loss": 0.9313,
+      "step": 134
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.5010756850242615,
+      "learning_rate": 4.9749899959984e-05,
+      "loss": 0.7662,
+      "step": 135
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.5203558206558228,
+      "learning_rate": 4.974789915966387e-05,
+      "loss": 0.8708,
+      "step": 136
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 0.5031847357749939,
+      "learning_rate": 4.974589835934374e-05,
+      "loss": 0.8258,
+      "step": 137
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.5000547170639038,
+      "learning_rate": 4.9743897559023614e-05,
+      "loss": 0.804,
+      "step": 138
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.5364178419113159,
+      "learning_rate": 4.9741896758703485e-05,
+      "loss": 0.8398,
+      "step": 139
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.4716164171695709,
+      "learning_rate": 4.973989595838335e-05,
+      "loss": 0.8043,
+      "step": 140
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.48213207721710205,
+      "learning_rate": 4.973789515806323e-05,
+      "loss": 0.803,
+      "step": 141
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.49524998664855957,
+      "learning_rate": 4.97358943577431e-05,
+      "loss": 0.8401,
+      "step": 142
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.5219356417655945,
+      "learning_rate": 4.973389355742297e-05,
+      "loss": 0.8474,
+      "step": 143
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.5039879679679871,
+      "learning_rate": 4.9731892757102845e-05,
+      "loss": 0.8219,
+      "step": 144
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.48044779896736145,
+      "learning_rate": 4.9729891956782717e-05,
+      "loss": 0.8212,
+      "step": 145
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 0.5111274123191833,
+      "learning_rate": 4.972789115646259e-05,
+      "loss": 0.8266,
+      "step": 146
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.5483076572418213,
+      "learning_rate": 4.972589035614246e-05,
+      "loss": 0.8127,
+      "step": 147
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.4989505708217621,
+      "learning_rate": 4.972388955582233e-05,
+      "loss": 0.8285,
+      "step": 148
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.5818595886230469,
+      "learning_rate": 4.9721888755502204e-05,
+      "loss": 0.8504,
+      "step": 149
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.530348539352417,
+      "learning_rate": 4.9719887955182076e-05,
+      "loss": 0.8316,
+      "step": 150
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.50677090883255,
+      "learning_rate": 4.971788715486195e-05,
+      "loss": 0.8048,
+      "step": 151
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.5426594614982605,
+      "learning_rate": 4.971588635454182e-05,
+      "loss": 0.8929,
+      "step": 152
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.5290346145629883,
+      "learning_rate": 4.971388555422169e-05,
+      "loss": 0.8454,
+      "step": 153
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.49939247965812683,
+      "learning_rate": 4.971188475390156e-05,
+      "loss": 0.8309,
+      "step": 154
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.527169406414032,
+      "learning_rate": 4.9709883953581435e-05,
+      "loss": 0.8509,
+      "step": 155
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.525046169757843,
+      "learning_rate": 4.970788315326131e-05,
+      "loss": 0.8518,
+      "step": 156
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 0.5176852345466614,
+      "learning_rate": 4.970588235294118e-05,
+      "loss": 0.8062,
+      "step": 157
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.5134345889091492,
+      "learning_rate": 4.970388155262105e-05,
+      "loss": 0.8723,
+      "step": 158
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.5024152994155884,
+      "learning_rate": 4.970188075230092e-05,
+      "loss": 0.8053,
+      "step": 159
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.5135634541511536,
+      "learning_rate": 4.9699879951980794e-05,
+      "loss": 0.7935,
+      "step": 160
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.5161725878715515,
+      "learning_rate": 4.9697879151660666e-05,
+      "loss": 0.808,
+      "step": 161
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.5259711146354675,
+      "learning_rate": 4.969587835134054e-05,
+      "loss": 0.8209,
+      "step": 162
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.4903233051300049,
+      "learning_rate": 4.969387755102041e-05,
+      "loss": 0.793,
+      "step": 163
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 0.5051981806755066,
+      "learning_rate": 4.969187675070028e-05,
+      "loss": 0.7971,
+      "step": 164
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.49914005398750305,
+      "learning_rate": 4.9689875950380154e-05,
+      "loss": 0.8292,
+      "step": 165
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.5912098288536072,
+      "learning_rate": 4.9687875150060026e-05,
+      "loss": 0.8573,
+      "step": 166
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.5313730239868164,
+      "learning_rate": 4.96858743497399e-05,
+      "loss": 0.8106,
+      "step": 167
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.5088040232658386,
+      "learning_rate": 4.968387354941977e-05,
+      "loss": 0.8159,
+      "step": 168
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.5306310057640076,
+      "learning_rate": 4.968187274909965e-05,
+      "loss": 0.7539,
+      "step": 169
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.4969288110733032,
+      "learning_rate": 4.967987194877951e-05,
+      "loss": 0.7364,
+      "step": 170
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.5466715097427368,
+      "learning_rate": 4.9677871148459385e-05,
+      "loss": 0.7734,
+      "step": 171
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 0.5333448052406311,
+      "learning_rate": 4.967587034813926e-05,
+      "loss": 0.7622,
+      "step": 172
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.49693411588668823,
+      "learning_rate": 4.967386954781913e-05,
+      "loss": 0.7967,
+      "step": 173
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.5356098413467407,
+      "learning_rate": 4.9671868747499e-05,
+      "loss": 0.8457,
+      "step": 174
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.5146982073783875,
+      "learning_rate": 4.966986794717887e-05,
+      "loss": 0.8258,
+      "step": 175
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.5248112082481384,
+      "learning_rate": 4.966786714685875e-05,
+      "loss": 0.7929,
+      "step": 176
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.5330566763877869,
+      "learning_rate": 4.966586634653862e-05,
+      "loss": 0.7916,
+      "step": 177
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.5528807044029236,
+      "learning_rate": 4.966386554621849e-05,
+      "loss": 0.8199,
+      "step": 178
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.5558779835700989,
+      "learning_rate": 4.966186474589836e-05,
+      "loss": 0.7826,
+      "step": 179
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.5845450758934021,
+      "learning_rate": 4.965986394557823e-05,
+      "loss": 0.8208,
+      "step": 180
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.565697431564331,
+      "learning_rate": 4.9657863145258103e-05,
+      "loss": 0.8423,
+      "step": 181
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.49734389781951904,
+      "learning_rate": 4.9655862344937975e-05,
+      "loss": 0.7908,
+      "step": 182
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.5148658156394958,
+      "learning_rate": 4.965386154461785e-05,
+      "loss": 0.7928,
+      "step": 183
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 0.5373660922050476,
+      "learning_rate": 4.9651860744297726e-05,
+      "loss": 0.794,
+      "step": 184
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.5560426712036133,
+      "learning_rate": 4.96498599439776e-05,
+      "loss": 0.8215,
+      "step": 185
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.5648142695426941,
+      "learning_rate": 4.964785914365746e-05,
+      "loss": 0.8003,
+      "step": 186
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.6051068902015686,
+      "learning_rate": 4.9645858343337335e-05,
+      "loss": 0.8386,
+      "step": 187
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.5387214422225952,
+      "learning_rate": 4.9643857543017206e-05,
+      "loss": 0.8217,
+      "step": 188
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.5685967803001404,
+      "learning_rate": 4.964185674269708e-05,
+      "loss": 0.8346,
+      "step": 189
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.5596044659614563,
+      "learning_rate": 4.963985594237695e-05,
+      "loss": 0.8023,
+      "step": 190
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.556212842464447,
+      "learning_rate": 4.963785514205683e-05,
+      "loss": 0.8171,
+      "step": 191
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.5799182653427124,
+      "learning_rate": 4.96358543417367e-05,
+      "loss": 0.7663,
+      "step": 192
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 0.5294367074966431,
+      "learning_rate": 4.963385354141657e-05,
+      "loss": 0.7845,
+      "step": 193
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.5038023591041565,
+      "learning_rate": 4.963185274109644e-05,
+      "loss": 0.7471,
+      "step": 194
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.5687416195869446,
+      "learning_rate": 4.962985194077631e-05,
+      "loss": 0.8603,
+      "step": 195
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.5455114245414734,
+      "learning_rate": 4.962785114045618e-05,
+      "loss": 0.8024,
+      "step": 196
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.5598348379135132,
+      "learning_rate": 4.962585034013605e-05,
+      "loss": 0.7709,
+      "step": 197
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.5542442798614502,
+      "learning_rate": 4.962384953981593e-05,
+      "loss": 0.7776,
+      "step": 198
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 0.5675541162490845,
+      "learning_rate": 4.9621848739495804e-05,
+      "loss": 0.8026,
+      "step": 199
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.5367075204849243,
+      "learning_rate": 4.9619847939175676e-05,
+      "loss": 0.8048,
+      "step": 200
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.5945519804954529,
+      "learning_rate": 4.961784713885555e-05,
+      "loss": 0.8014,
+      "step": 201
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.5422332882881165,
+      "learning_rate": 4.961584633853541e-05,
+      "loss": 0.8761,
+      "step": 202
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.5434099435806274,
+      "learning_rate": 4.9613845538215284e-05,
+      "loss": 0.793,
+      "step": 203
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.49900439381599426,
+      "learning_rate": 4.9611844737895156e-05,
+      "loss": 0.7663,
+      "step": 204
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.5360503792762756,
+      "learning_rate": 4.9609843937575035e-05,
+      "loss": 0.8549,
+      "step": 205
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.5519281625747681,
+      "learning_rate": 4.960784313725491e-05,
+      "loss": 0.8166,
+      "step": 206
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.5933863520622253,
+      "learning_rate": 4.960584233693478e-05,
+      "loss": 0.7875,
+      "step": 207
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 0.5276301503181458,
+      "learning_rate": 4.960384153661465e-05,
+      "loss": 0.798,
+      "step": 208
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.5891780853271484,
+      "learning_rate": 4.960184073629452e-05,
+      "loss": 0.7636,
+      "step": 209
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.6080418825149536,
+      "learning_rate": 4.959983993597439e-05,
+      "loss": 0.8056,
+      "step": 210
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 0.5800920724868774,
+      "learning_rate": 4.959783913565426e-05,
+      "loss": 0.7786,
+      "step": 211
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.544090747833252,
+      "learning_rate": 4.959583833533414e-05,
+      "loss": 0.807,
+      "step": 212
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.5422959923744202,
+      "learning_rate": 4.959383753501401e-05,
+      "loss": 0.8109,
+      "step": 213
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.5759320855140686,
+      "learning_rate": 4.959183673469388e-05,
+      "loss": 0.8202,
+      "step": 214
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.6158757209777832,
+      "learning_rate": 4.958983593437375e-05,
+      "loss": 0.8622,
+      "step": 215
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.5751588344573975,
+      "learning_rate": 4.9587835134053625e-05,
+      "loss": 0.8078,
+      "step": 216
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.5633991956710815,
+      "learning_rate": 4.95858343337335e-05,
+      "loss": 0.8264,
+      "step": 217
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.5534703135490417,
+      "learning_rate": 4.958383353341336e-05,
+      "loss": 0.8046,
+      "step": 218
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.5665048956871033,
+      "learning_rate": 4.958183273309324e-05,
+      "loss": 0.8078,
+      "step": 219
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.5370583534240723,
+      "learning_rate": 4.957983193277311e-05,
+      "loss": 0.7598,
+      "step": 220
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.5629434585571289,
+      "learning_rate": 4.9577831132452985e-05,
+      "loss": 0.8066,
+      "step": 221
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.5847271680831909,
+      "learning_rate": 4.9575830332132856e-05,
+      "loss": 0.8446,
+      "step": 222
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.5916556119918823,
+      "learning_rate": 4.957382953181273e-05,
+      "loss": 0.802,
+      "step": 223
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.6037752628326416,
+      "learning_rate": 4.95718287314926e-05,
+      "loss": 0.8271,
+      "step": 224
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.5692646503448486,
+      "learning_rate": 4.956982793117247e-05,
+      "loss": 0.8059,
+      "step": 225
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.516044020652771,
+      "learning_rate": 4.9567827130852344e-05,
+      "loss": 0.7265,
+      "step": 226
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.5782244801521301,
+      "learning_rate": 4.9565826330532216e-05,
+      "loss": 0.7515,
+      "step": 227
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.5291789174079895,
+      "learning_rate": 4.956382553021209e-05,
+      "loss": 0.7896,
+      "step": 228
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 0.5594845414161682,
+      "learning_rate": 4.956182472989196e-05,
+      "loss": 0.8484,
+      "step": 229
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.5874782800674438,
+      "learning_rate": 4.955982392957183e-05,
+      "loss": 0.8865,
+      "step": 230
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.5848100781440735,
+      "learning_rate": 4.95578231292517e-05,
+      "loss": 0.8274,
+      "step": 231
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 0.5242554545402527,
+      "learning_rate": 4.9555822328931575e-05,
+      "loss": 0.7192,
+      "step": 232
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.5743463039398193,
+      "learning_rate": 4.955382152861145e-05,
+      "loss": 0.82,
+      "step": 233
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.5418503880500793,
+      "learning_rate": 4.955182072829132e-05,
+      "loss": 0.7837,
+      "step": 234
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.5967192053794861,
+      "learning_rate": 4.954981992797119e-05,
+      "loss": 0.7851,
+      "step": 235
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.6640393137931824,
+      "learning_rate": 4.954781912765106e-05,
+      "loss": 0.868,
+      "step": 236
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.5320351123809814,
+      "learning_rate": 4.9545818327330934e-05,
+      "loss": 0.7881,
+      "step": 237
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.5583452582359314,
+      "learning_rate": 4.9543817527010806e-05,
+      "loss": 0.7926,
+      "step": 238
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.5930384993553162,
+      "learning_rate": 4.954181672669068e-05,
+      "loss": 0.8088,
+      "step": 239
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.5465940833091736,
+      "learning_rate": 4.953981592637055e-05,
+      "loss": 0.8136,
+      "step": 240
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.5826738476753235,
+      "learning_rate": 4.953781512605042e-05,
+      "loss": 0.8333,
+      "step": 241
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.5693926811218262,
+      "learning_rate": 4.9535814325730293e-05,
+      "loss": 0.8225,
+      "step": 242
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.5517392158508301,
+      "learning_rate": 4.9533813525410165e-05,
+      "loss": 0.797,
+      "step": 243
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.5277729034423828,
+      "learning_rate": 4.953181272509004e-05,
+      "loss": 0.74,
+      "step": 244
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.5529365539550781,
+      "learning_rate": 4.952981192476991e-05,
+      "loss": 0.785,
+      "step": 245
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.5595861673355103,
+      "learning_rate": 4.952781112444978e-05,
+      "loss": 0.8319,
+      "step": 246
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.5934503674507141,
+      "learning_rate": 4.952581032412966e-05,
+      "loss": 0.7452,
+      "step": 247
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.5756224989891052,
+      "learning_rate": 4.9523809523809525e-05,
+      "loss": 0.7893,
+      "step": 248
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.6283906698226929,
+      "learning_rate": 4.9521808723489396e-05,
+      "loss": 0.9012,
+      "step": 249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5654857158660889,
+      "learning_rate": 4.951980792316927e-05,
+      "loss": 0.7181,
+      "step": 250
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 0.5287903547286987,
+      "learning_rate": 4.951780712284914e-05,
+      "loss": 0.7838,
+      "step": 251
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.5523903369903564,
+      "learning_rate": 4.951580632252901e-05,
+      "loss": 0.8112,
+      "step": 252
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 0.5311181545257568,
+      "learning_rate": 4.9513805522208884e-05,
+      "loss": 0.7635,
+      "step": 253
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 0.5373832583427429,
+      "learning_rate": 4.951180472188876e-05,
+      "loss": 0.8084,
+      "step": 254
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.506324052810669,
+      "learning_rate": 4.9509803921568634e-05,
+      "loss": 0.7529,
+      "step": 255
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 0.5614925622940063,
+      "learning_rate": 4.95078031212485e-05,
+      "loss": 0.7682,
+      "step": 256
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 0.5953945517539978,
+      "learning_rate": 4.950580232092837e-05,
+      "loss": 0.7948,
+      "step": 257
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.5029836297035217,
+      "learning_rate": 4.950380152060824e-05,
+      "loss": 0.7703,
+      "step": 258
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 0.591628909111023,
+      "learning_rate": 4.9501800720288115e-05,
+      "loss": 0.8775,
+      "step": 259
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.5612605214118958,
+      "learning_rate": 4.949979991996799e-05,
+      "loss": 0.8241,
+      "step": 260
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.5402380228042603,
+      "learning_rate": 4.9497799119647866e-05,
+      "loss": 0.7018,
+      "step": 261
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 0.5295984148979187,
+      "learning_rate": 4.949579831932774e-05,
+      "loss": 0.8048,
+      "step": 262
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 0.5453588962554932,
+      "learning_rate": 4.949379751900761e-05,
+      "loss": 0.8198,
+      "step": 263
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.5718904137611389,
+      "learning_rate": 4.9491796718687474e-05,
+      "loss": 0.744,
+      "step": 264
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.5623178482055664,
+      "learning_rate": 4.9489795918367346e-05,
+      "loss": 0.7693,
+      "step": 265
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 0.5621166229248047,
+      "learning_rate": 4.948779511804722e-05,
+      "loss": 0.7858,
+      "step": 266
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.5423117876052856,
+      "learning_rate": 4.948579431772709e-05,
+      "loss": 0.7858,
+      "step": 267
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 0.5965234637260437,
+      "learning_rate": 4.948379351740697e-05,
+      "loss": 0.7983,
+      "step": 268
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 0.5860886573791504,
+      "learning_rate": 4.948179271708684e-05,
+      "loss": 0.7928,
+      "step": 269
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.5337466597557068,
+      "learning_rate": 4.947979191676671e-05,
+      "loss": 0.792,
+      "step": 270
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 0.6088325381278992,
+      "learning_rate": 4.9477791116446584e-05,
+      "loss": 0.8432,
+      "step": 271
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 0.5539262890815735,
+      "learning_rate": 4.947579031612645e-05,
+      "loss": 0.8019,
+      "step": 272
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.5926856398582458,
+      "learning_rate": 4.947378951580632e-05,
+      "loss": 0.7869,
+      "step": 273
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 0.5703017711639404,
+      "learning_rate": 4.947178871548619e-05,
+      "loss": 0.8028,
+      "step": 274
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.5751794576644897,
+      "learning_rate": 4.946978791516607e-05,
+      "loss": 0.8149,
+      "step": 275
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.5824299454689026,
+      "learning_rate": 4.9467787114845943e-05,
+      "loss": 0.7349,
+      "step": 276
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 0.5836261510848999,
+      "learning_rate": 4.9465786314525815e-05,
+      "loss": 0.8029,
+      "step": 277
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 0.6032565236091614,
+      "learning_rate": 4.946378551420569e-05,
+      "loss": 0.806,
+      "step": 278
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.5672261118888855,
+      "learning_rate": 4.946178471388556e-05,
+      "loss": 0.8044,
+      "step": 279
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.5680912733078003,
+      "learning_rate": 4.9459783913565424e-05,
+      "loss": 0.7874,
+      "step": 280
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 0.5770596861839294,
+      "learning_rate": 4.9457783113245296e-05,
+      "loss": 0.7395,
+      "step": 281
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.5634730458259583,
+      "learning_rate": 4.9455782312925175e-05,
+      "loss": 0.7117,
+      "step": 282
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 0.6248441934585571,
+      "learning_rate": 4.9453781512605046e-05,
+      "loss": 0.7941,
+      "step": 283
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 0.5985219478607178,
+      "learning_rate": 4.945178071228492e-05,
+      "loss": 0.7826,
+      "step": 284
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.5826301574707031,
+      "learning_rate": 4.944977991196479e-05,
+      "loss": 0.7999,
+      "step": 285
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 0.5907866954803467,
+      "learning_rate": 4.944777911164466e-05,
+      "loss": 0.7725,
+      "step": 286
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 0.6153370141983032,
+      "learning_rate": 4.9445778311324534e-05,
+      "loss": 0.8651,
+      "step": 287
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.5640325546264648,
+      "learning_rate": 4.94437775110044e-05,
+      "loss": 0.7517,
+      "step": 288
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 0.5785256028175354,
+      "learning_rate": 4.944177671068427e-05,
+      "loss": 0.8121,
+      "step": 289
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.5664944052696228,
+      "learning_rate": 4.943977591036415e-05,
+      "loss": 0.7345,
+      "step": 290
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.5458950996398926,
+      "learning_rate": 4.943777511004402e-05,
+      "loss": 0.73,
+      "step": 291
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 0.5893869996070862,
+      "learning_rate": 4.943577430972389e-05,
+      "loss": 0.7927,
+      "step": 292
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 0.6073876619338989,
+      "learning_rate": 4.9433773509403765e-05,
+      "loss": 0.7971,
+      "step": 293
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.5632811784744263,
+      "learning_rate": 4.943177270908364e-05,
+      "loss": 0.7442,
+      "step": 294
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.5832781791687012,
+      "learning_rate": 4.942977190876351e-05,
+      "loss": 0.8098,
+      "step": 295
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 0.5741657018661499,
+      "learning_rate": 4.9427771108443374e-05,
+      "loss": 0.7586,
+      "step": 296
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.6392624974250793,
+      "learning_rate": 4.942577030812325e-05,
+      "loss": 0.7998,
+      "step": 297
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 0.5990936160087585,
+      "learning_rate": 4.9423769507803124e-05,
+      "loss": 0.8197,
+      "step": 298
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 0.5655567049980164,
+      "learning_rate": 4.9421768707482996e-05,
+      "loss": 0.7936,
+      "step": 299
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.5669339895248413,
+      "learning_rate": 4.941976790716287e-05,
+      "loss": 0.7923,
+      "step": 300
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 0.5685401558876038,
+      "learning_rate": 4.941776710684274e-05,
+      "loss": 0.8257,
+      "step": 301
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 0.581586480140686,
+      "learning_rate": 4.941576630652261e-05,
+      "loss": 0.7951,
+      "step": 302
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.5897729992866516,
+      "learning_rate": 4.9413765506202484e-05,
+      "loss": 0.7391,
+      "step": 303
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 0.5444718599319458,
+      "learning_rate": 4.9411764705882355e-05,
+      "loss": 0.7317,
+      "step": 304
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.5707020163536072,
+      "learning_rate": 4.940976390556223e-05,
+      "loss": 0.7453,
+      "step": 305
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.5731601119041443,
+      "learning_rate": 4.94077631052421e-05,
+      "loss": 0.7614,
+      "step": 306
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 0.5668581128120422,
+      "learning_rate": 4.940576230492197e-05,
+      "loss": 0.7696,
+      "step": 307
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 0.6144934892654419,
+      "learning_rate": 4.940376150460184e-05,
+      "loss": 0.7765,
+      "step": 308
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.6028556823730469,
+      "learning_rate": 4.9401760704281715e-05,
+      "loss": 0.8292,
+      "step": 309
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.5614394545555115,
+      "learning_rate": 4.9399759903961587e-05,
+      "loss": 0.7867,
+      "step": 310
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 0.6021682620048523,
+      "learning_rate": 4.939775910364146e-05,
+      "loss": 0.7805,
+      "step": 311
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.6178930401802063,
+      "learning_rate": 4.939575830332133e-05,
+      "loss": 0.846,
+      "step": 312
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 0.6026805639266968,
+      "learning_rate": 4.93937575030012e-05,
+      "loss": 0.7695,
+      "step": 313
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 0.6167937517166138,
+      "learning_rate": 4.9391756702681074e-05,
+      "loss": 0.7939,
+      "step": 314
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.5787027478218079,
+      "learning_rate": 4.9389755902360946e-05,
+      "loss": 0.7763,
+      "step": 315
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 0.5984156131744385,
+      "learning_rate": 4.938775510204082e-05,
+      "loss": 0.8166,
+      "step": 316
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 0.6511149406433105,
+      "learning_rate": 4.938575430172069e-05,
+      "loss": 0.8453,
+      "step": 317
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.6076487898826599,
+      "learning_rate": 4.938375350140057e-05,
+      "loss": 0.7335,
+      "step": 318
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 0.5762828588485718,
+      "learning_rate": 4.938175270108043e-05,
+      "loss": 0.7503,
+      "step": 319
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.6130267381668091,
+      "learning_rate": 4.9379751900760305e-05,
+      "loss": 0.7707,
+      "step": 320
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.6319014430046082,
+      "learning_rate": 4.937775110044018e-05,
+      "loss": 0.7817,
+      "step": 321
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 0.5863409638404846,
+      "learning_rate": 4.937575030012005e-05,
+      "loss": 0.8031,
+      "step": 322
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 0.5963563323020935,
+      "learning_rate": 4.937374949979992e-05,
+      "loss": 0.8168,
+      "step": 323
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.5877537131309509,
+      "learning_rate": 4.937174869947979e-05,
+      "loss": 0.8302,
+      "step": 324
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.5984660387039185,
+      "learning_rate": 4.936974789915967e-05,
+      "loss": 0.8747,
+      "step": 325
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 0.5737277269363403,
+      "learning_rate": 4.936774709883954e-05,
+      "loss": 0.8451,
+      "step": 326
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.5724412798881531,
+      "learning_rate": 4.936574629851941e-05,
+      "loss": 0.8066,
+      "step": 327
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 0.5816782116889954,
+      "learning_rate": 4.936374549819928e-05,
+      "loss": 0.756,
+      "step": 328
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 0.592029333114624,
+      "learning_rate": 4.936174469787915e-05,
+      "loss": 0.7409,
+      "step": 329
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.5857312083244324,
+      "learning_rate": 4.9359743897559024e-05,
+      "loss": 0.7277,
+      "step": 330
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 0.6745790839195251,
+      "learning_rate": 4.9357743097238896e-05,
+      "loss": 0.8739,
+      "step": 331
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 0.5804886221885681,
+      "learning_rate": 4.9355742296918774e-05,
+      "loss": 0.7704,
+      "step": 332
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.6009431481361389,
+      "learning_rate": 4.9353741496598646e-05,
+      "loss": 0.7817,
+      "step": 333
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 0.5622819066047668,
+      "learning_rate": 4.935174069627852e-05,
+      "loss": 0.7781,
+      "step": 334
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.6032870411872864,
+      "learning_rate": 4.934973989595838e-05,
+      "loss": 0.8291,
+      "step": 335
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.6172161102294922,
+      "learning_rate": 4.9347739095638255e-05,
+      "loss": 0.8107,
+      "step": 336
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 0.5511362552642822,
+      "learning_rate": 4.934573829531813e-05,
+      "loss": 0.7283,
+      "step": 337
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 0.5732323527336121,
+      "learning_rate": 4.9343737494998e-05,
+      "loss": 0.694,
+      "step": 338
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.5522581934928894,
+      "learning_rate": 4.934173669467788e-05,
+      "loss": 0.8047,
+      "step": 339
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.5704146027565002,
+      "learning_rate": 4.933973589435775e-05,
+      "loss": 0.789,
+      "step": 340
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 0.564039409160614,
+      "learning_rate": 4.933773509403762e-05,
+      "loss": 0.764,
+      "step": 341
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.610349714756012,
+      "learning_rate": 4.933573429371749e-05,
+      "loss": 0.7716,
+      "step": 342
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 0.5944207310676575,
+      "learning_rate": 4.933373349339736e-05,
+      "loss": 0.7811,
+      "step": 343
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 0.5716763734817505,
+      "learning_rate": 4.933173269307723e-05,
+      "loss": 0.7904,
+      "step": 344
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.6018866896629333,
+      "learning_rate": 4.93297318927571e-05,
+      "loss": 0.783,
+      "step": 345
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 0.5952413082122803,
+      "learning_rate": 4.932773109243698e-05,
+      "loss": 0.8399,
+      "step": 346
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 0.5757237672805786,
+      "learning_rate": 4.932573029211685e-05,
+      "loss": 0.829,
+      "step": 347
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.6060802340507507,
+      "learning_rate": 4.9323729491796724e-05,
+      "loss": 0.8134,
+      "step": 348
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 0.6192966103553772,
+      "learning_rate": 4.9321728691476596e-05,
+      "loss": 0.8245,
+      "step": 349
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.5844107270240784,
+      "learning_rate": 4.931972789115647e-05,
+      "loss": 0.7463,
+      "step": 350
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.6216725707054138,
+      "learning_rate": 4.931772709083633e-05,
+      "loss": 0.7797,
+      "step": 351
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 0.6098612546920776,
+      "learning_rate": 4.9315726290516205e-05,
+      "loss": 0.7743,
+      "step": 352
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 0.5867882966995239,
+      "learning_rate": 4.931372549019608e-05,
+      "loss": 0.7186,
+      "step": 353
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.5984062552452087,
+      "learning_rate": 4.9311724689875955e-05,
+      "loss": 0.7579,
+      "step": 354
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.6255276203155518,
+      "learning_rate": 4.930972388955583e-05,
+      "loss": 0.8468,
+      "step": 355
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 0.5381520986557007,
+      "learning_rate": 4.93077230892357e-05,
+      "loss": 0.7261,
+      "step": 356
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.5565890073776245,
+      "learning_rate": 4.930572228891557e-05,
+      "loss": 0.7827,
+      "step": 357
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 0.6231935620307922,
+      "learning_rate": 4.930372148859544e-05,
+      "loss": 0.8469,
+      "step": 358
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 0.6231086850166321,
+      "learning_rate": 4.930172068827531e-05,
+      "loss": 0.8246,
+      "step": 359
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.5462002158164978,
+      "learning_rate": 4.9299719887955186e-05,
+      "loss": 0.7182,
+      "step": 360
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 0.5747554302215576,
+      "learning_rate": 4.929771908763506e-05,
+      "loss": 0.7594,
+      "step": 361
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 0.5723777413368225,
+      "learning_rate": 4.929571828731493e-05,
+      "loss": 0.7787,
+      "step": 362
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.6416032910346985,
+      "learning_rate": 4.92937174869948e-05,
+      "loss": 0.8315,
+      "step": 363
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 0.603192925453186,
+      "learning_rate": 4.9291716686674674e-05,
+      "loss": 0.7719,
+      "step": 364
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.6231751441955566,
+      "learning_rate": 4.9289715886354545e-05,
+      "loss": 0.8031,
+      "step": 365
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.6328988671302795,
+      "learning_rate": 4.928771508603442e-05,
+      "loss": 0.8228,
+      "step": 366
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 0.5893206000328064,
+      "learning_rate": 4.928571428571429e-05,
+      "loss": 0.7112,
+      "step": 367
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 0.5830476880073547,
+      "learning_rate": 4.928371348539416e-05,
+      "loss": 0.7569,
+      "step": 368
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.5955057740211487,
+      "learning_rate": 4.928171268507403e-05,
+      "loss": 0.7747,
+      "step": 369
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.5977997183799744,
+      "learning_rate": 4.9279711884753905e-05,
+      "loss": 0.7904,
+      "step": 370
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 0.5674768090248108,
+      "learning_rate": 4.9277711084433777e-05,
+      "loss": 0.752,
+      "step": 371
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.6169261932373047,
+      "learning_rate": 4.927571028411365e-05,
+      "loss": 0.7574,
+      "step": 372
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 0.618199348449707,
+      "learning_rate": 4.927370948379352e-05,
+      "loss": 0.7908,
+      "step": 373
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 0.5316541194915771,
+      "learning_rate": 4.927170868347339e-05,
+      "loss": 0.8064,
+      "step": 374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5974490642547607,
+      "learning_rate": 4.9269707883153264e-05,
+      "loss": 0.7936,
+      "step": 375
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 0.6208757758140564,
+      "learning_rate": 4.9267707082833136e-05,
+      "loss": 0.7862,
+      "step": 376
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 0.6077625155448914,
+      "learning_rate": 4.926570628251301e-05,
+      "loss": 0.7473,
+      "step": 377
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.591814398765564,
+      "learning_rate": 4.926370548219288e-05,
+      "loss": 0.7455,
+      "step": 378
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 0.6128320097923279,
+      "learning_rate": 4.926170468187275e-05,
+      "loss": 0.8214,
+      "step": 379
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.5914111733436584,
+      "learning_rate": 4.925970388155262e-05,
+      "loss": 0.8374,
+      "step": 380
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.5888047218322754,
+      "learning_rate": 4.9257703081232495e-05,
+      "loss": 0.7639,
+      "step": 381
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 0.5357518196105957,
+      "learning_rate": 4.925570228091237e-05,
+      "loss": 0.7629,
+      "step": 382
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 0.5791406631469727,
+      "learning_rate": 4.925370148059224e-05,
+      "loss": 0.77,
+      "step": 383
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.5658066272735596,
+      "learning_rate": 4.925170068027211e-05,
+      "loss": 0.8229,
+      "step": 384
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.5416905283927917,
+      "learning_rate": 4.924969987995198e-05,
+      "loss": 0.7255,
+      "step": 385
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 0.5073053240776062,
+      "learning_rate": 4.9247699079631854e-05,
+      "loss": 0.7436,
+      "step": 386
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.5704997777938843,
+      "learning_rate": 4.9245698279311726e-05,
+      "loss": 0.8392,
+      "step": 387
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 0.6032180786132812,
+      "learning_rate": 4.9243697478991605e-05,
+      "loss": 0.7793,
+      "step": 388
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 0.6236636638641357,
+      "learning_rate": 4.924169667867147e-05,
+      "loss": 0.8586,
+      "step": 389
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.5844170451164246,
+      "learning_rate": 4.923969587835134e-05,
+      "loss": 0.7596,
+      "step": 390
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 0.5753904581069946,
+      "learning_rate": 4.9237695078031214e-05,
+      "loss": 0.7724,
+      "step": 391
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 0.5851483941078186,
+      "learning_rate": 4.9235694277711086e-05,
+      "loss": 0.809,
+      "step": 392
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.6062473058700562,
+      "learning_rate": 4.923369347739096e-05,
+      "loss": 0.7379,
+      "step": 393
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 0.5940774083137512,
+      "learning_rate": 4.923169267707083e-05,
+      "loss": 0.7636,
+      "step": 394
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.6108981370925903,
+      "learning_rate": 4.922969187675071e-05,
+      "loss": 0.7407,
+      "step": 395
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.5814557671546936,
+      "learning_rate": 4.922769107643058e-05,
+      "loss": 0.7893,
+      "step": 396
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 0.5524982810020447,
+      "learning_rate": 4.9225690276110445e-05,
+      "loss": 0.7972,
+      "step": 397
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 0.6131865382194519,
+      "learning_rate": 4.922368947579032e-05,
+      "loss": 0.8582,
+      "step": 398
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.5640206336975098,
+      "learning_rate": 4.922168867547019e-05,
+      "loss": 0.7221,
+      "step": 399
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.5649920701980591,
+      "learning_rate": 4.921968787515006e-05,
+      "loss": 0.7759,
+      "step": 400
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 0.6427083611488342,
+      "learning_rate": 4.921768707482993e-05,
+      "loss": 0.8388,
+      "step": 401
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.6196634769439697,
+      "learning_rate": 4.9215686274509804e-05,
+      "loss": 0.8383,
+      "step": 402
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 0.5725811123847961,
+      "learning_rate": 4.921368547418968e-05,
+      "loss": 0.7602,
+      "step": 403
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 0.6151264309883118,
+      "learning_rate": 4.9211684673869555e-05,
+      "loss": 0.7759,
+      "step": 404
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.6025890707969666,
+      "learning_rate": 4.920968387354942e-05,
+      "loss": 0.7741,
+      "step": 405
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 0.5948590636253357,
+      "learning_rate": 4.920768307322929e-05,
+      "loss": 0.7644,
+      "step": 406
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 0.5557575821876526,
+      "learning_rate": 4.9205682272909163e-05,
+      "loss": 0.7398,
+      "step": 407
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.6094503402709961,
+      "learning_rate": 4.9203681472589035e-05,
+      "loss": 0.7671,
+      "step": 408
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 0.6672384738922119,
+      "learning_rate": 4.920168067226891e-05,
+      "loss": 0.7938,
+      "step": 409
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.6114287376403809,
+      "learning_rate": 4.9199679871948786e-05,
+      "loss": 0.8331,
+      "step": 410
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.5579826831817627,
+      "learning_rate": 4.919767907162866e-05,
+      "loss": 0.7256,
+      "step": 411
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 0.5519280433654785,
+      "learning_rate": 4.919567827130853e-05,
+      "loss": 0.755,
+      "step": 412
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 0.5924770832061768,
+      "learning_rate": 4.9193677470988395e-05,
+      "loss": 0.7783,
+      "step": 413
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.5979952812194824,
+      "learning_rate": 4.9191676670668266e-05,
+      "loss": 0.7305,
+      "step": 414
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.6058392524719238,
+      "learning_rate": 4.918967587034814e-05,
+      "loss": 0.7754,
+      "step": 415
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 0.5810478925704956,
+      "learning_rate": 4.918767507002801e-05,
+      "loss": 0.7785,
+      "step": 416
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.5745630860328674,
+      "learning_rate": 4.918567426970789e-05,
+      "loss": 0.7649,
+      "step": 417
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 0.6739091277122498,
+      "learning_rate": 4.918367346938776e-05,
+      "loss": 0.8086,
+      "step": 418
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 0.6409255266189575,
+      "learning_rate": 4.918167266906763e-05,
+      "loss": 0.7439,
+      "step": 419
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.6018539667129517,
+      "learning_rate": 4.9179671868747504e-05,
+      "loss": 0.8367,
+      "step": 420
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 0.6245967745780945,
+      "learning_rate": 4.917767106842737e-05,
+      "loss": 0.8294,
+      "step": 421
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 0.5872459411621094,
+      "learning_rate": 4.917567026810724e-05,
+      "loss": 0.7508,
+      "step": 422
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.6180544495582581,
+      "learning_rate": 4.917366946778711e-05,
+      "loss": 0.8241,
+      "step": 423
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 0.56467604637146,
+      "learning_rate": 4.917166866746699e-05,
+      "loss": 0.7355,
+      "step": 424
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.5801830887794495,
+      "learning_rate": 4.9169667867146864e-05,
+      "loss": 0.7786,
+      "step": 425
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.5825265049934387,
+      "learning_rate": 4.9167667066826735e-05,
+      "loss": 0.7876,
+      "step": 426
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 0.5895298719406128,
+      "learning_rate": 4.916566626650661e-05,
+      "loss": 0.8246,
+      "step": 427
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 0.5841552019119263,
+      "learning_rate": 4.916366546618648e-05,
+      "loss": 0.7721,
+      "step": 428
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.6116175651550293,
+      "learning_rate": 4.9161664665866344e-05,
+      "loss": 0.7855,
+      "step": 429
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.6221590042114258,
+      "learning_rate": 4.9159663865546216e-05,
+      "loss": 0.7947,
+      "step": 430
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 0.5872395634651184,
+      "learning_rate": 4.9157663065226095e-05,
+      "loss": 0.7433,
+      "step": 431
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.5529791712760925,
+      "learning_rate": 4.9155662264905967e-05,
+      "loss": 0.7531,
+      "step": 432
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 0.5675722360610962,
+      "learning_rate": 4.915366146458584e-05,
+      "loss": 0.7426,
+      "step": 433
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 0.5776454210281372,
+      "learning_rate": 4.915166066426571e-05,
+      "loss": 0.8128,
+      "step": 434
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.5539517402648926,
+      "learning_rate": 4.914965986394558e-05,
+      "loss": 0.806,
+      "step": 435
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 0.612989068031311,
+      "learning_rate": 4.9147659063625454e-05,
+      "loss": 0.7703,
+      "step": 436
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 0.6443326473236084,
+      "learning_rate": 4.914565826330532e-05,
+      "loss": 0.7351,
+      "step": 437
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.5801988244056702,
+      "learning_rate": 4.91436574629852e-05,
+      "loss": 0.792,
+      "step": 438
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 0.5797520875930786,
+      "learning_rate": 4.914165666266507e-05,
+      "loss": 0.7547,
+      "step": 439
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.6442052721977234,
+      "learning_rate": 4.913965586234494e-05,
+      "loss": 0.7916,
+      "step": 440
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.5808482766151428,
+      "learning_rate": 4.913765506202481e-05,
+      "loss": 0.7912,
+      "step": 441
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 0.6839514970779419,
+      "learning_rate": 4.9135654261704685e-05,
+      "loss": 0.8279,
+      "step": 442
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 0.6102442145347595,
+      "learning_rate": 4.913365346138456e-05,
+      "loss": 0.8261,
+      "step": 443
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.5921707153320312,
+      "learning_rate": 4.913165266106443e-05,
+      "loss": 0.7572,
+      "step": 444
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.6068132519721985,
+      "learning_rate": 4.91296518607443e-05,
+      "loss": 0.6949,
+      "step": 445
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 0.5876120924949646,
+      "learning_rate": 4.912765106042417e-05,
+      "loss": 0.7906,
+      "step": 446
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.587684154510498,
+      "learning_rate": 4.9125650260104044e-05,
+      "loss": 0.7745,
+      "step": 447
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 0.6146616339683533,
+      "learning_rate": 4.9123649459783916e-05,
+      "loss": 0.7909,
+      "step": 448
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 0.5915645360946655,
+      "learning_rate": 4.912164865946379e-05,
+      "loss": 0.7361,
+      "step": 449
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.5974079966545105,
+      "learning_rate": 4.911964785914366e-05,
+      "loss": 0.7355,
+      "step": 450
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 0.6230568885803223,
+      "learning_rate": 4.911764705882353e-05,
+      "loss": 0.7886,
+      "step": 451
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 0.604843020439148,
+      "learning_rate": 4.9115646258503404e-05,
+      "loss": 0.7817,
+      "step": 452
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.6246700882911682,
+      "learning_rate": 4.9113645458183276e-05,
+      "loss": 0.7952,
+      "step": 453
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 0.5888630747795105,
+      "learning_rate": 4.911164465786315e-05,
+      "loss": 0.7482,
+      "step": 454
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.5762295722961426,
+      "learning_rate": 4.910964385754302e-05,
+      "loss": 0.7301,
+      "step": 455
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.6081997156143188,
+      "learning_rate": 4.910764305722289e-05,
+      "loss": 0.7295,
+      "step": 456
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 0.6235604286193848,
+      "learning_rate": 4.910564225690276e-05,
+      "loss": 0.8245,
+      "step": 457
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 0.6132957339286804,
+      "learning_rate": 4.9103641456582635e-05,
+      "loss": 0.7289,
+      "step": 458
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.6033244132995605,
+      "learning_rate": 4.910164065626251e-05,
+      "loss": 0.8088,
+      "step": 459
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.6521779894828796,
+      "learning_rate": 4.909963985594238e-05,
+      "loss": 0.8683,
+      "step": 460
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 0.6089750528335571,
+      "learning_rate": 4.909763905562225e-05,
+      "loss": 0.7418,
+      "step": 461
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.6081743836402893,
+      "learning_rate": 4.909563825530212e-05,
+      "loss": 0.7847,
+      "step": 462
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 0.5884706377983093,
+      "learning_rate": 4.9093637454981994e-05,
+      "loss": 0.7429,
+      "step": 463
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 0.5694007873535156,
+      "learning_rate": 4.9091636654661866e-05,
+      "loss": 0.7418,
+      "step": 464
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.5716164112091064,
+      "learning_rate": 4.908963585434174e-05,
+      "loss": 0.7683,
+      "step": 465
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 0.5955536365509033,
+      "learning_rate": 4.9087635054021617e-05,
+      "loss": 0.7523,
+      "step": 466
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 0.5757789611816406,
+      "learning_rate": 4.908563425370148e-05,
+      "loss": 0.7261,
+      "step": 467
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.5787334442138672,
+      "learning_rate": 4.9083633453381353e-05,
+      "loss": 0.7195,
+      "step": 468
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 0.568333625793457,
+      "learning_rate": 4.9081632653061225e-05,
+      "loss": 0.7462,
+      "step": 469
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.5744010806083679,
+      "learning_rate": 4.90796318527411e-05,
+      "loss": 0.8024,
+      "step": 470
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.6531276702880859,
+      "learning_rate": 4.907763105242097e-05,
+      "loss": 0.8175,
+      "step": 471
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 0.6218701004981995,
+      "learning_rate": 4.907563025210084e-05,
+      "loss": 0.7496,
+      "step": 472
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 0.6153817772865295,
+      "learning_rate": 4.907362945178072e-05,
+      "loss": 0.8135,
+      "step": 473
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.6718766689300537,
+      "learning_rate": 4.907162865146059e-05,
+      "loss": 0.8049,
+      "step": 474
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.6055019497871399,
+      "learning_rate": 4.9069627851140456e-05,
+      "loss": 0.7722,
+      "step": 475
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 0.5902823805809021,
+      "learning_rate": 4.906762705082033e-05,
+      "loss": 0.7665,
+      "step": 476
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.5945876836776733,
+      "learning_rate": 4.90656262505002e-05,
+      "loss": 0.7589,
+      "step": 477
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 0.6314772963523865,
+      "learning_rate": 4.906362545018007e-05,
+      "loss": 0.8462,
+      "step": 478
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 0.6117784380912781,
+      "learning_rate": 4.9061624649859944e-05,
+      "loss": 0.7623,
+      "step": 479
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.5828944444656372,
+      "learning_rate": 4.905962384953982e-05,
+      "loss": 0.752,
+      "step": 480
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 0.6389551162719727,
+      "learning_rate": 4.9057623049219694e-05,
+      "loss": 0.7396,
+      "step": 481
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 0.5840926170349121,
+      "learning_rate": 4.9055622248899566e-05,
+      "loss": 0.7754,
+      "step": 482
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.6001207232475281,
+      "learning_rate": 4.905362144857943e-05,
+      "loss": 0.7678,
+      "step": 483
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 0.5531839728355408,
+      "learning_rate": 4.90516206482593e-05,
+      "loss": 0.7174,
+      "step": 484
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.6437347531318665,
+      "learning_rate": 4.9049619847939175e-05,
+      "loss": 0.7857,
+      "step": 485
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.5961335301399231,
+      "learning_rate": 4.904761904761905e-05,
+      "loss": 0.754,
+      "step": 486
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 0.5557673573493958,
+      "learning_rate": 4.9045618247298926e-05,
+      "loss": 0.7418,
+      "step": 487
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 0.6520072221755981,
+      "learning_rate": 4.90436174469788e-05,
+      "loss": 0.8426,
+      "step": 488
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.5974238514900208,
+      "learning_rate": 4.904161664665867e-05,
+      "loss": 0.7502,
+      "step": 489
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.6083403825759888,
+      "learning_rate": 4.903961584633854e-05,
+      "loss": 0.7499,
+      "step": 490
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 0.5932098031044006,
+      "learning_rate": 4.9037615046018406e-05,
+      "loss": 0.7344,
+      "step": 491
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.6413688659667969,
+      "learning_rate": 4.903561424569828e-05,
+      "loss": 0.7552,
+      "step": 492
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 0.6206446290016174,
+      "learning_rate": 4.903361344537815e-05,
+      "loss": 0.7687,
+      "step": 493
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 0.5776639580726624,
+      "learning_rate": 4.903161264505803e-05,
+      "loss": 0.8371,
+      "step": 494
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.6182857155799866,
+      "learning_rate": 4.90296118447379e-05,
+      "loss": 0.7361,
+      "step": 495
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 0.6016165018081665,
+      "learning_rate": 4.902761104441777e-05,
+      "loss": 0.8338,
+      "step": 496
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 0.6279016733169556,
+      "learning_rate": 4.9025610244097644e-05,
+      "loss": 0.756,
+      "step": 497
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.6500204205513,
+      "learning_rate": 4.9023609443777516e-05,
+      "loss": 0.8069,
+      "step": 498
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 0.6177827715873718,
+      "learning_rate": 4.902160864345738e-05,
+      "loss": 0.7365,
+      "step": 499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5693172216415405,
+      "learning_rate": 4.901960784313725e-05,
+      "loss": 0.6754,
+      "step": 500
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.5654707551002502,
+      "learning_rate": 4.901760704281713e-05,
+      "loss": 0.7525,
+      "step": 501
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 0.648030161857605,
+      "learning_rate": 4.9015606242497e-05,
+      "loss": 0.764,
+      "step": 502
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 0.6066122651100159,
+      "learning_rate": 4.9013605442176875e-05,
+      "loss": 0.7487,
+      "step": 503
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.6264938712120056,
+      "learning_rate": 4.901160464185675e-05,
+      "loss": 0.8013,
+      "step": 504
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.60635906457901,
+      "learning_rate": 4.900960384153662e-05,
+      "loss": 0.7421,
+      "step": 505
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 0.5976872444152832,
+      "learning_rate": 4.900760304121649e-05,
+      "loss": 0.733,
+      "step": 506
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.5609980225563049,
+      "learning_rate": 4.9005602240896356e-05,
+      "loss": 0.7141,
+      "step": 507
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 0.6109974980354309,
+      "learning_rate": 4.9003601440576234e-05,
+      "loss": 0.7301,
+      "step": 508
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 0.6050900220870972,
+      "learning_rate": 4.9001600640256106e-05,
+      "loss": 0.7644,
+      "step": 509
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.571101188659668,
+      "learning_rate": 4.899959983993598e-05,
+      "loss": 0.7736,
+      "step": 510
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 0.5456352829933167,
+      "learning_rate": 4.899759903961585e-05,
+      "loss": 0.7738,
+      "step": 511
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 0.5890512466430664,
+      "learning_rate": 4.899559823929572e-05,
+      "loss": 0.6723,
+      "step": 512
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.5800885558128357,
+      "learning_rate": 4.8993597438975594e-05,
+      "loss": 0.7153,
+      "step": 513
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 0.6010210514068604,
+      "learning_rate": 4.8991596638655466e-05,
+      "loss": 0.7784,
+      "step": 514
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.6430292129516602,
+      "learning_rate": 4.898959583833533e-05,
+      "loss": 0.7992,
+      "step": 515
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.5662366151809692,
+      "learning_rate": 4.898759503801521e-05,
+      "loss": 0.7449,
+      "step": 516
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 0.6085848212242126,
+      "learning_rate": 4.898559423769508e-05,
+      "loss": 0.7972,
+      "step": 517
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 0.5684623122215271,
+      "learning_rate": 4.898359343737495e-05,
+      "loss": 0.7667,
+      "step": 518
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.5847243666648865,
+      "learning_rate": 4.8981592637054825e-05,
+      "loss": 0.7233,
+      "step": 519
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.6326200366020203,
+      "learning_rate": 4.89795918367347e-05,
+      "loss": 0.7656,
+      "step": 520
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 0.6132087707519531,
+      "learning_rate": 4.897759103641457e-05,
+      "loss": 0.7842,
+      "step": 521
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.5856083631515503,
+      "learning_rate": 4.897559023609444e-05,
+      "loss": 0.7745,
+      "step": 522
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 0.6147796511650085,
+      "learning_rate": 4.897358943577431e-05,
+      "loss": 0.7964,
+      "step": 523
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 0.6051590442657471,
+      "learning_rate": 4.8971588635454184e-05,
+      "loss": 0.7573,
+      "step": 524
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.6189625263214111,
+      "learning_rate": 4.8969587835134056e-05,
+      "loss": 0.8481,
+      "step": 525
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 0.5812535881996155,
+      "learning_rate": 4.896758703481393e-05,
+      "loss": 0.8213,
+      "step": 526
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 0.5988802909851074,
+      "learning_rate": 4.89655862344938e-05,
+      "loss": 0.6807,
+      "step": 527
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.5860095620155334,
+      "learning_rate": 4.896358543417367e-05,
+      "loss": 0.7816,
+      "step": 528
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 0.5878039598464966,
+      "learning_rate": 4.8961584633853543e-05,
+      "loss": 0.7771,
+      "step": 529
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.6504335403442383,
+      "learning_rate": 4.8959583833533415e-05,
+      "loss": 0.82,
+      "step": 530
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.5590863227844238,
+      "learning_rate": 4.895758303321329e-05,
+      "loss": 0.7686,
+      "step": 531
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 0.6558895707130432,
+      "learning_rate": 4.895558223289316e-05,
+      "loss": 0.8116,
+      "step": 532
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 0.6584180593490601,
+      "learning_rate": 4.895358143257303e-05,
+      "loss": 0.7509,
+      "step": 533
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.5821325778961182,
+      "learning_rate": 4.89515806322529e-05,
+      "loss": 0.7289,
+      "step": 534
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.6342353820800781,
+      "learning_rate": 4.8949579831932775e-05,
+      "loss": 0.7905,
+      "step": 535
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 0.595000684261322,
+      "learning_rate": 4.8947579031612646e-05,
+      "loss": 0.7344,
+      "step": 536
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.6255506277084351,
+      "learning_rate": 4.894557823129252e-05,
+      "loss": 0.7994,
+      "step": 537
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 0.6021307110786438,
+      "learning_rate": 4.894357743097239e-05,
+      "loss": 0.8138,
+      "step": 538
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 0.5913227796554565,
+      "learning_rate": 4.894157663065226e-05,
+      "loss": 0.7822,
+      "step": 539
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.6197808980941772,
+      "learning_rate": 4.8939575830332134e-05,
+      "loss": 0.7853,
+      "step": 540
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 0.5861679911613464,
+      "learning_rate": 4.8937575030012006e-05,
+      "loss": 0.7293,
+      "step": 541
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 0.5833053588867188,
+      "learning_rate": 4.893557422969188e-05,
+      "loss": 0.7416,
+      "step": 542
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.5544353723526001,
+      "learning_rate": 4.893357342937175e-05,
+      "loss": 0.7278,
+      "step": 543
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 0.5617831945419312,
+      "learning_rate": 4.893157262905163e-05,
+      "loss": 0.7219,
+      "step": 544
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.6455240845680237,
+      "learning_rate": 4.892957182873149e-05,
+      "loss": 0.7656,
+      "step": 545
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.63936448097229,
+      "learning_rate": 4.8927571028411365e-05,
+      "loss": 0.8029,
+      "step": 546
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 0.5780683755874634,
+      "learning_rate": 4.892557022809124e-05,
+      "loss": 0.7684,
+      "step": 547
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 0.571263313293457,
+      "learning_rate": 4.892356942777111e-05,
+      "loss": 0.6714,
+      "step": 548
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.6045280694961548,
+      "learning_rate": 4.892156862745098e-05,
+      "loss": 0.7374,
+      "step": 549
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.5887953639030457,
+      "learning_rate": 4.891956782713085e-05,
+      "loss": 0.7341,
+      "step": 550
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 0.6200657486915588,
+      "learning_rate": 4.891756702681073e-05,
+      "loss": 0.7835,
+      "step": 551
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.6312881112098694,
+      "learning_rate": 4.89155662264906e-05,
+      "loss": 0.8279,
+      "step": 552
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 0.6151871681213379,
+      "learning_rate": 4.891356542617047e-05,
+      "loss": 0.7467,
+      "step": 553
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 0.5729169249534607,
+      "learning_rate": 4.891156462585034e-05,
+      "loss": 0.7405,
+      "step": 554
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.5994395017623901,
+      "learning_rate": 4.890956382553021e-05,
+      "loss": 0.8057,
+      "step": 555
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 0.564448893070221,
+      "learning_rate": 4.8907563025210084e-05,
+      "loss": 0.7428,
+      "step": 556
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 0.6355032920837402,
+      "learning_rate": 4.8905562224889955e-05,
+      "loss": 0.79,
+      "step": 557
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.5675954222679138,
+      "learning_rate": 4.8903561424569834e-05,
+      "loss": 0.7338,
+      "step": 558
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 0.5565802454948425,
+      "learning_rate": 4.8901560624249706e-05,
+      "loss": 0.7997,
+      "step": 559
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.5979486107826233,
+      "learning_rate": 4.889955982392958e-05,
+      "loss": 0.7952,
+      "step": 560
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.5932238101959229,
+      "learning_rate": 4.889755902360944e-05,
+      "loss": 0.7106,
+      "step": 561
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 0.659325361251831,
+      "learning_rate": 4.8895558223289315e-05,
+      "loss": 0.703,
+      "step": 562
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 0.6149618029594421,
+      "learning_rate": 4.889355742296919e-05,
+      "loss": 0.762,
+      "step": 563
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.5758309364318848,
+      "learning_rate": 4.889155662264906e-05,
+      "loss": 0.7698,
+      "step": 564
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.5630546808242798,
+      "learning_rate": 4.888955582232894e-05,
+      "loss": 0.7165,
+      "step": 565
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 0.6149067282676697,
+      "learning_rate": 4.888755502200881e-05,
+      "loss": 0.8136,
+      "step": 566
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.6086531281471252,
+      "learning_rate": 4.888555422168868e-05,
+      "loss": 0.7992,
+      "step": 567
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 0.6159600615501404,
+      "learning_rate": 4.888355342136855e-05,
+      "loss": 0.8229,
+      "step": 568
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 0.5952989459037781,
+      "learning_rate": 4.888155262104842e-05,
+      "loss": 0.7035,
+      "step": 569
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.6345956921577454,
+      "learning_rate": 4.887955182072829e-05,
+      "loss": 0.8317,
+      "step": 570
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 0.6076996326446533,
+      "learning_rate": 4.887755102040816e-05,
+      "loss": 0.7937,
+      "step": 571
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 0.5981006622314453,
+      "learning_rate": 4.887555022008804e-05,
+      "loss": 0.6843,
+      "step": 572
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.5810612440109253,
+      "learning_rate": 4.887354941976791e-05,
+      "loss": 0.6974,
+      "step": 573
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 0.5900784730911255,
+      "learning_rate": 4.8871548619447784e-05,
+      "loss": 0.7718,
+      "step": 574
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.5939428210258484,
+      "learning_rate": 4.8869547819127656e-05,
+      "loss": 0.7176,
+      "step": 575
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.6226340532302856,
+      "learning_rate": 4.886754701880753e-05,
+      "loss": 0.728,
+      "step": 576
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 0.6123561859130859,
+      "learning_rate": 4.886554621848739e-05,
+      "loss": 0.7569,
+      "step": 577
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 0.6113354563713074,
+      "learning_rate": 4.8863545418167264e-05,
+      "loss": 0.7683,
+      "step": 578
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.6293046474456787,
+      "learning_rate": 4.886154461784714e-05,
+      "loss": 0.8001,
+      "step": 579
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.5882707238197327,
+      "learning_rate": 4.8859543817527015e-05,
+      "loss": 0.6901,
+      "step": 580
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 0.5842380523681641,
+      "learning_rate": 4.885754301720689e-05,
+      "loss": 0.7069,
+      "step": 581
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.6349176168441772,
+      "learning_rate": 4.885554221688676e-05,
+      "loss": 0.7323,
+      "step": 582
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 0.5746724605560303,
+      "learning_rate": 4.885354141656663e-05,
+      "loss": 0.7153,
+      "step": 583
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 0.5570270419120789,
+      "learning_rate": 4.88515406162465e-05,
+      "loss": 0.7684,
+      "step": 584
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.6300098896026611,
+      "learning_rate": 4.884953981592637e-05,
+      "loss": 0.7983,
+      "step": 585
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 0.6173220872879028,
+      "learning_rate": 4.8847539015606246e-05,
+      "loss": 0.8115,
+      "step": 586
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 0.6043545007705688,
+      "learning_rate": 4.884553821528612e-05,
+      "loss": 0.7274,
+      "step": 587
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.6310338377952576,
+      "learning_rate": 4.884353741496599e-05,
+      "loss": 0.7706,
+      "step": 588
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 0.6150200366973877,
+      "learning_rate": 4.884153661464586e-05,
+      "loss": 0.7907,
+      "step": 589
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.5880206823348999,
+      "learning_rate": 4.8839535814325734e-05,
+      "loss": 0.7382,
+      "step": 590
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.6823338866233826,
+      "learning_rate": 4.8837535014005605e-05,
+      "loss": 0.7575,
+      "step": 591
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 0.6266196370124817,
+      "learning_rate": 4.883553421368548e-05,
+      "loss": 0.7685,
+      "step": 592
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 0.637712836265564,
+      "learning_rate": 4.883353341336535e-05,
+      "loss": 0.7326,
+      "step": 593
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.5898723006248474,
+      "learning_rate": 4.883153261304522e-05,
+      "loss": 0.749,
+      "step": 594
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.6104643940925598,
+      "learning_rate": 4.882953181272509e-05,
+      "loss": 0.7539,
+      "step": 595
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 0.615857720375061,
+      "learning_rate": 4.8827531012404965e-05,
+      "loss": 0.7938,
+      "step": 596
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.633080780506134,
+      "learning_rate": 4.8825530212084837e-05,
+      "loss": 0.8089,
+      "step": 597
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 0.615404486656189,
+      "learning_rate": 4.882352941176471e-05,
+      "loss": 0.8339,
+      "step": 598
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 0.5635229349136353,
+      "learning_rate": 4.882152861144458e-05,
+      "loss": 0.6853,
+      "step": 599
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.579932689666748,
+      "learning_rate": 4.881952781112445e-05,
+      "loss": 0.7536,
+      "step": 600
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 0.5975157618522644,
+      "learning_rate": 4.8817527010804324e-05,
+      "loss": 0.7634,
+      "step": 601
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 0.5809814929962158,
+      "learning_rate": 4.8815526210484196e-05,
+      "loss": 0.7197,
+      "step": 602
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.5941184163093567,
+      "learning_rate": 4.881352541016407e-05,
+      "loss": 0.7532,
+      "step": 603
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 0.629640519618988,
+      "learning_rate": 4.881152460984394e-05,
+      "loss": 0.7685,
+      "step": 604
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.6509769558906555,
+      "learning_rate": 4.880952380952381e-05,
+      "loss": 0.738,
+      "step": 605
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.5938847661018372,
+      "learning_rate": 4.880752300920368e-05,
+      "loss": 0.7364,
+      "step": 606
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 0.5702788829803467,
+      "learning_rate": 4.8805522208883555e-05,
+      "loss": 0.7057,
+      "step": 607
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 0.5880799889564514,
+      "learning_rate": 4.880352140856343e-05,
+      "loss": 0.6999,
+      "step": 608
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.6025612950325012,
+      "learning_rate": 4.88015206082433e-05,
+      "loss": 0.7189,
+      "step": 609
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.6370987296104431,
+      "learning_rate": 4.879951980792317e-05,
+      "loss": 0.7947,
+      "step": 610
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 0.5612797737121582,
+      "learning_rate": 4.879751900760304e-05,
+      "loss": 0.6949,
+      "step": 611
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.6249443888664246,
+      "learning_rate": 4.8795518207282914e-05,
+      "loss": 0.7593,
+      "step": 612
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 0.6149105429649353,
+      "learning_rate": 4.8793517406962786e-05,
+      "loss": 0.8293,
+      "step": 613
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 0.6441102623939514,
+      "learning_rate": 4.8791516606642665e-05,
+      "loss": 0.7343,
+      "step": 614
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.5925320386886597,
+      "learning_rate": 4.878951580632253e-05,
+      "loss": 0.7466,
+      "step": 615
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 0.6006816029548645,
+      "learning_rate": 4.87875150060024e-05,
+      "loss": 0.8136,
+      "step": 616
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 0.6164388060569763,
+      "learning_rate": 4.8785514205682274e-05,
+      "loss": 0.7459,
+      "step": 617
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.6219509840011597,
+      "learning_rate": 4.8783513405362146e-05,
+      "loss": 0.7392,
+      "step": 618
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 0.6259922981262207,
+      "learning_rate": 4.878151260504202e-05,
+      "loss": 0.7768,
+      "step": 619
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.5972528457641602,
+      "learning_rate": 4.877951180472189e-05,
+      "loss": 0.7749,
+      "step": 620
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.5769963264465332,
+      "learning_rate": 4.877751100440177e-05,
+      "loss": 0.7324,
+      "step": 621
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 0.5725414752960205,
+      "learning_rate": 4.877551020408164e-05,
+      "loss": 0.7089,
+      "step": 622
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 0.6348921656608582,
+      "learning_rate": 4.8773509403761505e-05,
+      "loss": 0.8087,
+      "step": 623
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.6241890788078308,
+      "learning_rate": 4.877150860344138e-05,
+      "loss": 0.7478,
+      "step": 624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6045325398445129,
+      "learning_rate": 4.876950780312125e-05,
+      "loss": 0.7607,
+      "step": 625
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 0.6093403100967407,
+      "learning_rate": 4.876750700280112e-05,
+      "loss": 0.7503,
+      "step": 626
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.6271349787712097,
+      "learning_rate": 4.876550620248099e-05,
+      "loss": 0.7631,
+      "step": 627
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 0.5978842377662659,
+      "learning_rate": 4.8763505402160864e-05,
+      "loss": 0.75,
+      "step": 628
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 0.5923817157745361,
+      "learning_rate": 4.876150460184074e-05,
+      "loss": 0.6844,
+      "step": 629
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.604054868221283,
+      "learning_rate": 4.8759503801520615e-05,
+      "loss": 0.7224,
+      "step": 630
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 0.645494282245636,
+      "learning_rate": 4.875750300120048e-05,
+      "loss": 0.7912,
+      "step": 631
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 0.6240355968475342,
+      "learning_rate": 4.875550220088035e-05,
+      "loss": 0.7429,
+      "step": 632
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.6154593229293823,
+      "learning_rate": 4.875350140056022e-05,
+      "loss": 0.7693,
+      "step": 633
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 0.6306690573692322,
+      "learning_rate": 4.8751500600240095e-05,
+      "loss": 0.733,
+      "step": 634
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.5954400897026062,
+      "learning_rate": 4.874949979991997e-05,
+      "loss": 0.8013,
+      "step": 635
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.6127607226371765,
+      "learning_rate": 4.8747498999599846e-05,
+      "loss": 0.7676,
+      "step": 636
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 0.6720303893089294,
+      "learning_rate": 4.874549819927972e-05,
+      "loss": 0.7987,
+      "step": 637
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 0.5854519605636597,
+      "learning_rate": 4.874349739895959e-05,
+      "loss": 0.769,
+      "step": 638
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.6683208346366882,
+      "learning_rate": 4.8741496598639455e-05,
+      "loss": 0.7729,
+      "step": 639
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.5927050709724426,
+      "learning_rate": 4.8739495798319326e-05,
+      "loss": 0.7671,
+      "step": 640
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 0.5779833793640137,
+      "learning_rate": 4.87374949979992e-05,
+      "loss": 0.6955,
+      "step": 641
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.6800764799118042,
+      "learning_rate": 4.873549419767907e-05,
+      "loss": 0.7995,
+      "step": 642
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 0.5779314041137695,
+      "learning_rate": 4.873349339735895e-05,
+      "loss": 0.7296,
+      "step": 643
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 0.5840075016021729,
+      "learning_rate": 4.873149259703882e-05,
+      "loss": 0.7412,
+      "step": 644
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.5941234230995178,
+      "learning_rate": 4.872949179671869e-05,
+      "loss": 0.8138,
+      "step": 645
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 0.5719572305679321,
+      "learning_rate": 4.8727490996398564e-05,
+      "loss": 0.7713,
+      "step": 646
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 0.6207135915756226,
+      "learning_rate": 4.872549019607843e-05,
+      "loss": 0.7417,
+      "step": 647
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.6460603475570679,
+      "learning_rate": 4.87234893957583e-05,
+      "loss": 0.8659,
+      "step": 648
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 0.5850974321365356,
+      "learning_rate": 4.872148859543817e-05,
+      "loss": 0.7083,
+      "step": 649
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.5269494652748108,
+      "learning_rate": 4.871948779511805e-05,
+      "loss": 0.6849,
+      "step": 650
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.6789783239364624,
+      "learning_rate": 4.8717486994797924e-05,
+      "loss": 0.8186,
+      "step": 651
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 0.5852586030960083,
+      "learning_rate": 4.8715486194477795e-05,
+      "loss": 0.759,
+      "step": 652
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 0.661586582660675,
+      "learning_rate": 4.871348539415767e-05,
+      "loss": 0.8003,
+      "step": 653
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.6112899780273438,
+      "learning_rate": 4.871148459383754e-05,
+      "loss": 0.7195,
+      "step": 654
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.6239529252052307,
+      "learning_rate": 4.8709483793517404e-05,
+      "loss": 0.6815,
+      "step": 655
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 0.6071779131889343,
+      "learning_rate": 4.8707482993197276e-05,
+      "loss": 0.6849,
+      "step": 656
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.6142399311065674,
+      "learning_rate": 4.8705482192877155e-05,
+      "loss": 0.7898,
+      "step": 657
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 0.5916489958763123,
+      "learning_rate": 4.8703481392557027e-05,
+      "loss": 0.7373,
+      "step": 658
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 0.6097525954246521,
+      "learning_rate": 4.87014805922369e-05,
+      "loss": 0.751,
+      "step": 659
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.5901023149490356,
+      "learning_rate": 4.869947979191677e-05,
+      "loss": 0.6901,
+      "step": 660
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 0.5702456831932068,
+      "learning_rate": 4.869747899159664e-05,
+      "loss": 0.7518,
+      "step": 661
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 0.606547474861145,
+      "learning_rate": 4.8695478191276514e-05,
+      "loss": 0.7508,
+      "step": 662
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.6051300764083862,
+      "learning_rate": 4.869347739095638e-05,
+      "loss": 0.7659,
+      "step": 663
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 0.6050798296928406,
+      "learning_rate": 4.869147659063626e-05,
+      "loss": 0.7429,
+      "step": 664
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.6163650751113892,
+      "learning_rate": 4.868947579031613e-05,
+      "loss": 0.7387,
+      "step": 665
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.6229680180549622,
+      "learning_rate": 4.8687474989996e-05,
+      "loss": 0.7673,
+      "step": 666
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 0.6289007663726807,
+      "learning_rate": 4.868547418967587e-05,
+      "loss": 0.738,
+      "step": 667
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 0.609407901763916,
+      "learning_rate": 4.8683473389355745e-05,
+      "loss": 0.7457,
+      "step": 668
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.6026242971420288,
+      "learning_rate": 4.868147258903562e-05,
+      "loss": 0.7617,
+      "step": 669
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.6227514147758484,
+      "learning_rate": 4.867947178871549e-05,
+      "loss": 0.7938,
+      "step": 670
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 0.6264092922210693,
+      "learning_rate": 4.867747098839536e-05,
+      "loss": 0.7826,
+      "step": 671
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.5935805439949036,
+      "learning_rate": 4.867547018807523e-05,
+      "loss": 0.7143,
+      "step": 672
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 0.6176136136054993,
+      "learning_rate": 4.8673469387755104e-05,
+      "loss": 0.7985,
+      "step": 673
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 0.6620816588401794,
+      "learning_rate": 4.8671468587434976e-05,
+      "loss": 0.7727,
+      "step": 674
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.6603094935417175,
+      "learning_rate": 4.866946778711485e-05,
+      "loss": 0.7125,
+      "step": 675
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 0.6638785600662231,
+      "learning_rate": 4.866746698679472e-05,
+      "loss": 0.7305,
+      "step": 676
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 0.6319957375526428,
+      "learning_rate": 4.866546618647459e-05,
+      "loss": 0.7595,
+      "step": 677
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.5931079387664795,
+      "learning_rate": 4.8663465386154464e-05,
+      "loss": 0.7415,
+      "step": 678
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 0.6137109398841858,
+      "learning_rate": 4.8661464585834336e-05,
+      "loss": 0.7438,
+      "step": 679
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.5985450744628906,
+      "learning_rate": 4.865946378551421e-05,
+      "loss": 0.7409,
+      "step": 680
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.62948077917099,
+      "learning_rate": 4.865746298519408e-05,
+      "loss": 0.8124,
+      "step": 681
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 0.5815073251724243,
+      "learning_rate": 4.865546218487395e-05,
+      "loss": 0.73,
+      "step": 682
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 0.5903521180152893,
+      "learning_rate": 4.865346138455382e-05,
+      "loss": 0.7071,
+      "step": 683
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.6045056581497192,
+      "learning_rate": 4.8651460584233695e-05,
+      "loss": 0.7571,
+      "step": 684
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.6277945041656494,
+      "learning_rate": 4.8649459783913573e-05,
+      "loss": 0.7475,
+      "step": 685
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 0.6049608588218689,
+      "learning_rate": 4.864745898359344e-05,
+      "loss": 0.7857,
+      "step": 686
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.6526109576225281,
+      "learning_rate": 4.864545818327331e-05,
+      "loss": 0.7536,
+      "step": 687
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 0.6064791679382324,
+      "learning_rate": 4.864345738295318e-05,
+      "loss": 0.7731,
+      "step": 688
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 0.5711452960968018,
+      "learning_rate": 4.8641456582633054e-05,
+      "loss": 0.7581,
+      "step": 689
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.603302538394928,
+      "learning_rate": 4.8639455782312926e-05,
+      "loss": 0.7323,
+      "step": 690
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 0.6006559729576111,
+      "learning_rate": 4.86374549819928e-05,
+      "loss": 0.7074,
+      "step": 691
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 0.5934480428695679,
+      "learning_rate": 4.8635454181672676e-05,
+      "loss": 0.73,
+      "step": 692
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.5922892093658447,
+      "learning_rate": 4.863345338135255e-05,
+      "loss": 0.7075,
+      "step": 693
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 0.5788107514381409,
+      "learning_rate": 4.8631452581032413e-05,
+      "loss": 0.7307,
+      "step": 694
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.5999823808670044,
+      "learning_rate": 4.8629451780712285e-05,
+      "loss": 0.811,
+      "step": 695
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.6252598762512207,
+      "learning_rate": 4.862745098039216e-05,
+      "loss": 0.7221,
+      "step": 696
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 0.6279902458190918,
+      "learning_rate": 4.862545018007203e-05,
+      "loss": 0.7428,
+      "step": 697
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 0.6252185702323914,
+      "learning_rate": 4.86234493797519e-05,
+      "loss": 0.7786,
+      "step": 698
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.6247672438621521,
+      "learning_rate": 4.862144857943178e-05,
+      "loss": 0.7813,
+      "step": 699
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.6344950199127197,
+      "learning_rate": 4.861944777911165e-05,
+      "loss": 0.7709,
+      "step": 700
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 0.6928293704986572,
+      "learning_rate": 4.861744697879152e-05,
+      "loss": 0.7902,
+      "step": 701
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.5995805859565735,
+      "learning_rate": 4.861544617847139e-05,
+      "loss": 0.7567,
+      "step": 702
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 0.6318543553352356,
+      "learning_rate": 4.861344537815126e-05,
+      "loss": 0.7079,
+      "step": 703
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 0.5997210144996643,
+      "learning_rate": 4.861144457783113e-05,
+      "loss": 0.7424,
+      "step": 704
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.6286918520927429,
+      "learning_rate": 4.8609443777511004e-05,
+      "loss": 0.7609,
+      "step": 705
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 0.5855825543403625,
+      "learning_rate": 4.860744297719088e-05,
+      "loss": 0.7097,
+      "step": 706
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 0.6897373795509338,
+      "learning_rate": 4.8605442176870754e-05,
+      "loss": 0.8089,
+      "step": 707
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.6260521411895752,
+      "learning_rate": 4.8603441376550626e-05,
+      "loss": 0.7375,
+      "step": 708
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 0.5700168013572693,
+      "learning_rate": 4.86014405762305e-05,
+      "loss": 0.7927,
+      "step": 709
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.6189404726028442,
+      "learning_rate": 4.859943977591036e-05,
+      "loss": 0.7137,
+      "step": 710
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.6204996109008789,
+      "learning_rate": 4.8597438975590235e-05,
+      "loss": 0.7494,
+      "step": 711
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 0.5846236944198608,
+      "learning_rate": 4.859543817527011e-05,
+      "loss": 0.7146,
+      "step": 712
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 0.6079381704330444,
+      "learning_rate": 4.8593437374949985e-05,
+      "loss": 0.7166,
+      "step": 713
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.6260196566581726,
+      "learning_rate": 4.859143657462986e-05,
+      "loss": 0.7639,
+      "step": 714
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.6194947361946106,
+      "learning_rate": 4.858943577430973e-05,
+      "loss": 0.7466,
+      "step": 715
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 0.5631256699562073,
+      "learning_rate": 4.85874349739896e-05,
+      "loss": 0.7732,
+      "step": 716
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.6009548306465149,
+      "learning_rate": 4.858543417366947e-05,
+      "loss": 0.7793,
+      "step": 717
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 0.6175373792648315,
+      "learning_rate": 4.858343337334934e-05,
+      "loss": 0.7672,
+      "step": 718
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 0.6090966463088989,
+      "learning_rate": 4.858143257302921e-05,
+      "loss": 0.7359,
+      "step": 719
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.5752703547477722,
+      "learning_rate": 4.857943177270909e-05,
+      "loss": 0.7208,
+      "step": 720
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 0.5956680774688721,
+      "learning_rate": 4.857743097238896e-05,
+      "loss": 0.7028,
+      "step": 721
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 0.6232988238334656,
+      "learning_rate": 4.857543017206883e-05,
+      "loss": 0.759,
+      "step": 722
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.6512808799743652,
+      "learning_rate": 4.8573429371748704e-05,
+      "loss": 0.7903,
+      "step": 723
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 0.6665694117546082,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.7636,
+      "step": 724
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.5766971111297607,
+      "learning_rate": 4.856942777110845e-05,
+      "loss": 0.7045,
+      "step": 725
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.610283613204956,
+      "learning_rate": 4.856742697078831e-05,
+      "loss": 0.7486,
+      "step": 726
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 0.6264128088951111,
+      "learning_rate": 4.856542617046819e-05,
+      "loss": 0.7867,
+      "step": 727
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 0.6030863523483276,
+      "learning_rate": 4.856342537014806e-05,
+      "loss": 0.75,
+      "step": 728
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.6077870726585388,
+      "learning_rate": 4.8561424569827935e-05,
+      "loss": 0.795,
+      "step": 729
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.554818332195282,
+      "learning_rate": 4.855942376950781e-05,
+      "loss": 0.7304,
+      "step": 730
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 0.5891968607902527,
+      "learning_rate": 4.855742296918768e-05,
+      "loss": 0.7437,
+      "step": 731
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.5956717133522034,
+      "learning_rate": 4.855542216886755e-05,
+      "loss": 0.7851,
+      "step": 732
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 0.6364977359771729,
+      "learning_rate": 4.855342136854742e-05,
+      "loss": 0.7926,
+      "step": 733
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 0.6908712387084961,
+      "learning_rate": 4.8551420568227294e-05,
+      "loss": 0.7565,
+      "step": 734
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.5801886320114136,
+      "learning_rate": 4.8549419767907166e-05,
+      "loss": 0.7251,
+      "step": 735
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 0.6222661137580872,
+      "learning_rate": 4.854741896758704e-05,
+      "loss": 0.6952,
+      "step": 736
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 0.6288057565689087,
+      "learning_rate": 4.854541816726691e-05,
+      "loss": 0.7691,
+      "step": 737
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.6167927384376526,
+      "learning_rate": 4.854341736694678e-05,
+      "loss": 0.7479,
+      "step": 738
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 0.5813360214233398,
+      "learning_rate": 4.8541416566626654e-05,
+      "loss": 0.7309,
+      "step": 739
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.6290002465248108,
+      "learning_rate": 4.8539415766306526e-05,
+      "loss": 0.7725,
+      "step": 740
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.6278480291366577,
+      "learning_rate": 4.85374149659864e-05,
+      "loss": 0.762,
+      "step": 741
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 0.5977178812026978,
+      "learning_rate": 4.853541416566627e-05,
+      "loss": 0.7463,
+      "step": 742
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 0.6514163613319397,
+      "learning_rate": 4.853341336534614e-05,
+      "loss": 0.7873,
+      "step": 743
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.5944337844848633,
+      "learning_rate": 4.853141256502601e-05,
+      "loss": 0.7534,
+      "step": 744
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.605015218257904,
+      "learning_rate": 4.8529411764705885e-05,
+      "loss": 0.7595,
+      "step": 745
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 0.6433084607124329,
+      "learning_rate": 4.852741096438576e-05,
+      "loss": 0.8321,
+      "step": 746
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.5718684792518616,
+      "learning_rate": 4.852541016406563e-05,
+      "loss": 0.7344,
+      "step": 747
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 0.6268803477287292,
+      "learning_rate": 4.85234093637455e-05,
+      "loss": 0.772,
+      "step": 748
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 0.6136980056762695,
+      "learning_rate": 4.852140856342537e-05,
+      "loss": 0.7513,
+      "step": 749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6517449617385864,
+      "learning_rate": 4.8519407763105244e-05,
+      "loss": 0.7786,
+      "step": 750
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 0.5884830355644226,
+      "learning_rate": 4.8517406962785116e-05,
+      "loss": 0.7912,
+      "step": 751
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 0.6458854675292969,
+      "learning_rate": 4.851540616246499e-05,
+      "loss": 0.7881,
+      "step": 752
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.6329476237297058,
+      "learning_rate": 4.851340536214486e-05,
+      "loss": 0.7819,
+      "step": 753
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 0.5884618163108826,
+      "learning_rate": 4.851140456182473e-05,
+      "loss": 0.7786,
+      "step": 754
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.5475010871887207,
+      "learning_rate": 4.8509403761504603e-05,
+      "loss": 0.7172,
+      "step": 755
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.6388839483261108,
+      "learning_rate": 4.8507402961184475e-05,
+      "loss": 0.7659,
+      "step": 756
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 0.6201974153518677,
+      "learning_rate": 4.850540216086435e-05,
+      "loss": 0.7723,
+      "step": 757
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 0.6046715974807739,
+      "learning_rate": 4.850340136054422e-05,
+      "loss": 0.7486,
+      "step": 758
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.6070242524147034,
+      "learning_rate": 4.850140056022409e-05,
+      "loss": 0.763,
+      "step": 759
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.6031877994537354,
+      "learning_rate": 4.849939975990396e-05,
+      "loss": 0.7415,
+      "step": 760
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 0.5899202823638916,
+      "learning_rate": 4.8497398959583835e-05,
+      "loss": 0.7867,
+      "step": 761
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.6172879338264465,
+      "learning_rate": 4.8495398159263706e-05,
+      "loss": 0.6983,
+      "step": 762
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 0.639527440071106,
+      "learning_rate": 4.8493397358943585e-05,
+      "loss": 0.7358,
+      "step": 763
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 0.6101019382476807,
+      "learning_rate": 4.849139655862345e-05,
+      "loss": 0.7267,
+      "step": 764
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.6153104305267334,
+      "learning_rate": 4.848939575830332e-05,
+      "loss": 0.7648,
+      "step": 765
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 0.5717357993125916,
+      "learning_rate": 4.8487394957983194e-05,
+      "loss": 0.7197,
+      "step": 766
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 0.6005909442901611,
+      "learning_rate": 4.8485394157663066e-05,
+      "loss": 0.6786,
+      "step": 767
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.6198951005935669,
+      "learning_rate": 4.848339335734294e-05,
+      "loss": 0.759,
+      "step": 768
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 0.6164072155952454,
+      "learning_rate": 4.848139255702281e-05,
+      "loss": 0.757,
+      "step": 769
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.5673977136611938,
+      "learning_rate": 4.847939175670269e-05,
+      "loss": 0.7746,
+      "step": 770
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.6161131262779236,
+      "learning_rate": 4.847739095638256e-05,
+      "loss": 0.77,
+      "step": 771
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 0.623842716217041,
+      "learning_rate": 4.8475390156062425e-05,
+      "loss": 0.7528,
+      "step": 772
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 0.6155294179916382,
+      "learning_rate": 4.84733893557423e-05,
+      "loss": 0.7517,
+      "step": 773
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.6041366457939148,
+      "learning_rate": 4.847138855542217e-05,
+      "loss": 0.7704,
+      "step": 774
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.6505810618400574,
+      "learning_rate": 4.846938775510204e-05,
+      "loss": 0.8253,
+      "step": 775
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 0.6835885643959045,
+      "learning_rate": 4.846738695478191e-05,
+      "loss": 0.8475,
+      "step": 776
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.6628032922744751,
+      "learning_rate": 4.846538615446179e-05,
+      "loss": 0.7722,
+      "step": 777
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 0.5734715461730957,
+      "learning_rate": 4.846338535414166e-05,
+      "loss": 0.7254,
+      "step": 778
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 0.6373343467712402,
+      "learning_rate": 4.8461384553821535e-05,
+      "loss": 0.7545,
+      "step": 779
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.6454253196716309,
+      "learning_rate": 4.84593837535014e-05,
+      "loss": 0.7757,
+      "step": 780
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 0.6249713897705078,
+      "learning_rate": 4.845738295318127e-05,
+      "loss": 0.6912,
+      "step": 781
+    },
+    {
+      "epoch": 1.00096,
+      "grad_norm": 1.1785544157028198,
+      "learning_rate": 4.8455382152861144e-05,
+      "loss": 1.291,
+      "step": 782
+    },
+    {
+      "epoch": 1.00224,
+      "grad_norm": 0.6030226349830627,
+      "learning_rate": 4.8453381352541015e-05,
+      "loss": 0.794,
+      "step": 783
+    },
+    {
+      "epoch": 1.00352,
+      "grad_norm": 0.5952867865562439,
+      "learning_rate": 4.8451380552220894e-05,
+      "loss": 0.7109,
+      "step": 784
+    },
+    {
+      "epoch": 1.0048,
+      "grad_norm": 0.595658004283905,
+      "learning_rate": 4.8449379751900766e-05,
+      "loss": 0.7261,
+      "step": 785
+    },
+    {
+      "epoch": 1.00608,
+      "grad_norm": 0.6058535575866699,
+      "learning_rate": 4.844737895158064e-05,
+      "loss": 0.7573,
+      "step": 786
+    },
+    {
+      "epoch": 1.00736,
+      "grad_norm": 0.5743350982666016,
+      "learning_rate": 4.844537815126051e-05,
+      "loss": 0.6788,
+      "step": 787
+    },
+    {
+      "epoch": 1.00864,
+      "grad_norm": 0.6272056102752686,
+      "learning_rate": 4.8443377350940375e-05,
+      "loss": 0.7366,
+      "step": 788
+    },
+    {
+      "epoch": 1.00992,
+      "grad_norm": 0.6139172315597534,
+      "learning_rate": 4.8441376550620247e-05,
+      "loss": 0.7665,
+      "step": 789
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 0.6504907011985779,
+      "learning_rate": 4.843937575030012e-05,
+      "loss": 0.7647,
+      "step": 790
+    },
+    {
+      "epoch": 1.01248,
+      "grad_norm": 0.6235556602478027,
+      "learning_rate": 4.843737494998e-05,
+      "loss": 0.7775,
+      "step": 791
+    },
+    {
+      "epoch": 1.01376,
+      "grad_norm": 0.5893948078155518,
+      "learning_rate": 4.843537414965987e-05,
+      "loss": 0.7583,
+      "step": 792
+    },
+    {
+      "epoch": 1.01504,
+      "grad_norm": 0.6206098198890686,
+      "learning_rate": 4.843337334933974e-05,
+      "loss": 0.7281,
+      "step": 793
+    },
+    {
+      "epoch": 1.01632,
+      "grad_norm": 0.639717161655426,
+      "learning_rate": 4.843137254901961e-05,
+      "loss": 0.7478,
+      "step": 794
+    },
+    {
+      "epoch": 1.0176,
+      "grad_norm": 0.6055462956428528,
+      "learning_rate": 4.8429371748699484e-05,
+      "loss": 0.698,
+      "step": 795
+    },
+    {
+      "epoch": 1.01888,
+      "grad_norm": 0.6277878880500793,
+      "learning_rate": 4.842737094837935e-05,
+      "loss": 0.7602,
+      "step": 796
+    },
+    {
+      "epoch": 1.02016,
+      "grad_norm": 0.6001823544502258,
+      "learning_rate": 4.842537014805922e-05,
+      "loss": 0.6931,
+      "step": 797
+    },
+    {
+      "epoch": 1.02144,
+      "grad_norm": 0.6369767785072327,
+      "learning_rate": 4.84233693477391e-05,
+      "loss": 0.7378,
+      "step": 798
+    },
+    {
+      "epoch": 1.02272,
+      "grad_norm": 0.6533408761024475,
+      "learning_rate": 4.842136854741897e-05,
+      "loss": 0.7738,
+      "step": 799
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.6154987812042236,
+      "learning_rate": 4.8419367747098844e-05,
+      "loss": 0.7081,
+      "step": 800
+    },
+    {
+      "epoch": 1.02528,
+      "grad_norm": 0.5977925062179565,
+      "learning_rate": 4.8417366946778716e-05,
+      "loss": 0.6907,
+      "step": 801
+    },
+    {
+      "epoch": 1.02656,
+      "grad_norm": 0.5921528935432434,
+      "learning_rate": 4.841536614645859e-05,
+      "loss": 0.7432,
+      "step": 802
+    },
+    {
+      "epoch": 1.02784,
+      "grad_norm": 0.5911890268325806,
+      "learning_rate": 4.841336534613846e-05,
+      "loss": 0.6892,
+      "step": 803
+    },
+    {
+      "epoch": 1.02912,
+      "grad_norm": 0.6107017993927002,
+      "learning_rate": 4.8411364545818324e-05,
+      "loss": 0.7641,
+      "step": 804
+    },
+    {
+      "epoch": 1.0304,
+      "grad_norm": 0.6192318797111511,
+      "learning_rate": 4.84093637454982e-05,
+      "loss": 0.7411,
+      "step": 805
+    },
+    {
+      "epoch": 1.03168,
+      "grad_norm": 0.5901237726211548,
+      "learning_rate": 4.8407362945178075e-05,
+      "loss": 0.7253,
+      "step": 806
+    },
+    {
+      "epoch": 1.03296,
+      "grad_norm": 0.6684927344322205,
+      "learning_rate": 4.840536214485795e-05,
+      "loss": 0.7597,
+      "step": 807
+    },
+    {
+      "epoch": 1.03424,
+      "grad_norm": 0.6069619059562683,
+      "learning_rate": 4.840336134453782e-05,
+      "loss": 0.709,
+      "step": 808
+    },
+    {
+      "epoch": 1.03552,
+      "grad_norm": 0.6064357757568359,
+      "learning_rate": 4.840136054421769e-05,
+      "loss": 0.7321,
+      "step": 809
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.5926334261894226,
+      "learning_rate": 4.839935974389756e-05,
+      "loss": 0.7346,
+      "step": 810
+    },
+    {
+      "epoch": 1.03808,
+      "grad_norm": 0.6023288369178772,
+      "learning_rate": 4.8397358943577434e-05,
+      "loss": 0.7672,
+      "step": 811
+    },
+    {
+      "epoch": 1.03936,
+      "grad_norm": 0.6245213747024536,
+      "learning_rate": 4.8395358143257306e-05,
+      "loss": 0.7458,
+      "step": 812
+    },
+    {
+      "epoch": 1.04064,
+      "grad_norm": 0.601436197757721,
+      "learning_rate": 4.839335734293718e-05,
+      "loss": 0.7848,
+      "step": 813
+    },
+    {
+      "epoch": 1.04192,
+      "grad_norm": 0.6355320811271667,
+      "learning_rate": 4.839135654261705e-05,
+      "loss": 0.8105,
+      "step": 814
+    },
+    {
+      "epoch": 1.0432,
+      "grad_norm": 0.5975676774978638,
+      "learning_rate": 4.838935574229692e-05,
+      "loss": 0.692,
+      "step": 815
+    },
+    {
+      "epoch": 1.04448,
+      "grad_norm": 0.6143369674682617,
+      "learning_rate": 4.8387354941976793e-05,
+      "loss": 0.7551,
+      "step": 816
+    },
+    {
+      "epoch": 1.04576,
+      "grad_norm": 0.6289949417114258,
+      "learning_rate": 4.8385354141656665e-05,
+      "loss": 0.6912,
+      "step": 817
+    },
+    {
+      "epoch": 1.04704,
+      "grad_norm": 0.6380932331085205,
+      "learning_rate": 4.838335334133654e-05,
+      "loss": 0.6864,
+      "step": 818
+    },
+    {
+      "epoch": 1.04832,
+      "grad_norm": 0.6500244736671448,
+      "learning_rate": 4.838135254101641e-05,
+      "loss": 0.7205,
+      "step": 819
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.6350643038749695,
+      "learning_rate": 4.837935174069628e-05,
+      "loss": 0.7197,
+      "step": 820
+    },
+    {
+      "epoch": 1.05088,
+      "grad_norm": 0.6777990460395813,
+      "learning_rate": 4.837735094037615e-05,
+      "loss": 0.8366,
+      "step": 821
+    },
+    {
+      "epoch": 1.05216,
+      "grad_norm": 0.6251175999641418,
+      "learning_rate": 4.8375350140056025e-05,
+      "loss": 0.7125,
+      "step": 822
+    },
+    {
+      "epoch": 1.05344,
+      "grad_norm": 0.6193746328353882,
+      "learning_rate": 4.8373349339735896e-05,
+      "loss": 0.798,
+      "step": 823
+    },
+    {
+      "epoch": 1.05472,
+      "grad_norm": 0.6032739877700806,
+      "learning_rate": 4.837134853941577e-05,
+      "loss": 0.7334,
+      "step": 824
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 0.6197022199630737,
+      "learning_rate": 4.836934773909564e-05,
+      "loss": 0.737,
+      "step": 825
+    },
+    {
+      "epoch": 1.05728,
+      "grad_norm": 0.646630585193634,
+      "learning_rate": 4.836734693877551e-05,
+      "loss": 0.7796,
+      "step": 826
+    },
+    {
+      "epoch": 1.05856,
+      "grad_norm": 0.5807000398635864,
+      "learning_rate": 4.8365346138455384e-05,
+      "loss": 0.7328,
+      "step": 827
+    },
+    {
+      "epoch": 1.05984,
+      "grad_norm": 0.6248646974563599,
+      "learning_rate": 4.8363345338135256e-05,
+      "loss": 0.7364,
+      "step": 828
+    },
+    {
+      "epoch": 1.06112,
+      "grad_norm": 0.6000564098358154,
+      "learning_rate": 4.836134453781513e-05,
+      "loss": 0.6898,
+      "step": 829
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.603438138961792,
+      "learning_rate": 4.8359343737495e-05,
+      "loss": 0.7358,
+      "step": 830
+    },
+    {
+      "epoch": 1.06368,
+      "grad_norm": 0.605436384677887,
+      "learning_rate": 4.835734293717487e-05,
+      "loss": 0.7493,
+      "step": 831
+    },
+    {
+      "epoch": 1.06496,
+      "grad_norm": 0.6158525347709656,
+      "learning_rate": 4.835534213685474e-05,
+      "loss": 0.743,
+      "step": 832
+    },
+    {
+      "epoch": 1.06624,
+      "grad_norm": 0.6049724817276001,
+      "learning_rate": 4.835334133653462e-05,
+      "loss": 0.7252,
+      "step": 833
+    },
+    {
+      "epoch": 1.06752,
+      "grad_norm": 0.5752277970314026,
+      "learning_rate": 4.835134053621449e-05,
+      "loss": 0.6848,
+      "step": 834
+    },
+    {
+      "epoch": 1.0688,
+      "grad_norm": 0.6556239128112793,
+      "learning_rate": 4.834933973589436e-05,
+      "loss": 0.7814,
+      "step": 835
+    },
+    {
+      "epoch": 1.07008,
+      "grad_norm": 0.6204028129577637,
+      "learning_rate": 4.834733893557423e-05,
+      "loss": 0.7009,
+      "step": 836
+    },
+    {
+      "epoch": 1.07136,
+      "grad_norm": 0.5991467833518982,
+      "learning_rate": 4.83453381352541e-05,
+      "loss": 0.7251,
+      "step": 837
+    },
+    {
+      "epoch": 1.07264,
+      "grad_norm": 0.5856754779815674,
+      "learning_rate": 4.8343337334933974e-05,
+      "loss": 0.7455,
+      "step": 838
+    },
+    {
+      "epoch": 1.07392,
+      "grad_norm": 0.6454278826713562,
+      "learning_rate": 4.8341336534613846e-05,
+      "loss": 0.7808,
+      "step": 839
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.6240683197975159,
+      "learning_rate": 4.8339335734293725e-05,
+      "loss": 0.7066,
+      "step": 840
+    },
+    {
+      "epoch": 1.07648,
+      "grad_norm": 0.6424701809883118,
+      "learning_rate": 4.83373349339736e-05,
+      "loss": 0.7447,
+      "step": 841
+    },
+    {
+      "epoch": 1.07776,
+      "grad_norm": 0.5974248051643372,
+      "learning_rate": 4.833533413365346e-05,
+      "loss": 0.735,
+      "step": 842
+    },
+    {
+      "epoch": 1.07904,
+      "grad_norm": 0.6363813281059265,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.7302,
+      "step": 843
+    },
+    {
+      "epoch": 1.08032,
+      "grad_norm": 0.6119391918182373,
+      "learning_rate": 4.8331332533013205e-05,
+      "loss": 0.7222,
+      "step": 844
+    },
+    {
+      "epoch": 1.0816,
+      "grad_norm": 0.5887010097503662,
+      "learning_rate": 4.832933173269308e-05,
+      "loss": 0.7207,
+      "step": 845
+    },
+    {
+      "epoch": 1.08288,
+      "grad_norm": 0.6205199360847473,
+      "learning_rate": 4.832733093237295e-05,
+      "loss": 0.72,
+      "step": 846
+    },
+    {
+      "epoch": 1.08416,
+      "grad_norm": 0.570545494556427,
+      "learning_rate": 4.832533013205282e-05,
+      "loss": 0.7201,
+      "step": 847
+    },
+    {
+      "epoch": 1.08544,
+      "grad_norm": 0.6642616987228394,
+      "learning_rate": 4.83233293317327e-05,
+      "loss": 0.8064,
+      "step": 848
+    },
+    {
+      "epoch": 1.08672,
+      "grad_norm": 0.650346577167511,
+      "learning_rate": 4.832132853141257e-05,
+      "loss": 0.7717,
+      "step": 849
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.5787684917449951,
+      "learning_rate": 4.831932773109244e-05,
+      "loss": 0.6377,
+      "step": 850
+    },
+    {
+      "epoch": 1.08928,
+      "grad_norm": 0.6291863322257996,
+      "learning_rate": 4.831732693077231e-05,
+      "loss": 0.7332,
+      "step": 851
+    },
+    {
+      "epoch": 1.09056,
+      "grad_norm": 0.6485406160354614,
+      "learning_rate": 4.831532613045218e-05,
+      "loss": 0.7445,
+      "step": 852
+    },
+    {
+      "epoch": 1.09184,
+      "grad_norm": 0.6354153156280518,
+      "learning_rate": 4.831332533013205e-05,
+      "loss": 0.7272,
+      "step": 853
+    },
+    {
+      "epoch": 1.09312,
+      "grad_norm": 0.5846593379974365,
+      "learning_rate": 4.8311324529811924e-05,
+      "loss": 0.7165,
+      "step": 854
+    },
+    {
+      "epoch": 1.0944,
+      "grad_norm": 0.6126893758773804,
+      "learning_rate": 4.83093237294918e-05,
+      "loss": 0.6913,
+      "step": 855
+    },
+    {
+      "epoch": 1.09568,
+      "grad_norm": 0.631846010684967,
+      "learning_rate": 4.8307322929171675e-05,
+      "loss": 0.7566,
+      "step": 856
+    },
+    {
+      "epoch": 1.09696,
+      "grad_norm": 0.6731998920440674,
+      "learning_rate": 4.8305322128851546e-05,
+      "loss": 0.7488,
+      "step": 857
+    },
+    {
+      "epoch": 1.09824,
+      "grad_norm": 0.6166826486587524,
+      "learning_rate": 4.830332132853141e-05,
+      "loss": 0.7271,
+      "step": 858
+    },
+    {
+      "epoch": 1.09952,
+      "grad_norm": 0.5922271013259888,
+      "learning_rate": 4.830132052821128e-05,
+      "loss": 0.6596,
+      "step": 859
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.6259810328483582,
+      "learning_rate": 4.8299319727891155e-05,
+      "loss": 0.7917,
+      "step": 860
+    },
+    {
+      "epoch": 1.10208,
+      "grad_norm": 0.6506845355033875,
+      "learning_rate": 4.829731892757103e-05,
+      "loss": 0.7029,
+      "step": 861
+    },
+    {
+      "epoch": 1.10336,
+      "grad_norm": 0.6502883434295654,
+      "learning_rate": 4.8295318127250906e-05,
+      "loss": 0.6832,
+      "step": 862
+    },
+    {
+      "epoch": 1.10464,
+      "grad_norm": 0.6381411552429199,
+      "learning_rate": 4.829331732693078e-05,
+      "loss": 0.7275,
+      "step": 863
+    },
+    {
+      "epoch": 1.10592,
+      "grad_norm": 0.629329264163971,
+      "learning_rate": 4.829131652661065e-05,
+      "loss": 0.7542,
+      "step": 864
+    },
+    {
+      "epoch": 1.1072,
+      "grad_norm": 0.6186816692352295,
+      "learning_rate": 4.828931572629052e-05,
+      "loss": 0.7195,
+      "step": 865
+    },
+    {
+      "epoch": 1.10848,
+      "grad_norm": 0.5985262393951416,
+      "learning_rate": 4.8287314925970386e-05,
+      "loss": 0.7651,
+      "step": 866
+    },
+    {
+      "epoch": 1.10976,
+      "grad_norm": 0.6066650152206421,
+      "learning_rate": 4.828531412565026e-05,
+      "loss": 0.7572,
+      "step": 867
+    },
+    {
+      "epoch": 1.11104,
+      "grad_norm": 0.6421188712120056,
+      "learning_rate": 4.828331332533013e-05,
+      "loss": 0.7796,
+      "step": 868
+    },
+    {
+      "epoch": 1.11232,
+      "grad_norm": 0.6231709122657776,
+      "learning_rate": 4.828131252501001e-05,
+      "loss": 0.7564,
+      "step": 869
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.6402696967124939,
+      "learning_rate": 4.827931172468988e-05,
+      "loss": 0.7095,
+      "step": 870
+    },
+    {
+      "epoch": 1.11488,
+      "grad_norm": 0.5962916612625122,
+      "learning_rate": 4.827731092436975e-05,
+      "loss": 0.7085,
+      "step": 871
+    },
+    {
+      "epoch": 1.11616,
+      "grad_norm": 0.6202763319015503,
+      "learning_rate": 4.8275310124049624e-05,
+      "loss": 0.7063,
+      "step": 872
+    },
+    {
+      "epoch": 1.11744,
+      "grad_norm": 0.626986026763916,
+      "learning_rate": 4.8273309323729496e-05,
+      "loss": 0.7705,
+      "step": 873
+    },
+    {
+      "epoch": 1.11872,
+      "grad_norm": 0.6388139724731445,
+      "learning_rate": 4.827130852340936e-05,
+      "loss": 0.754,
+      "step": 874
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6084983348846436,
+      "learning_rate": 4.826930772308923e-05,
+      "loss": 0.7012,
+      "step": 875
+    },
+    {
+      "epoch": 1.12128,
+      "grad_norm": 0.6018005609512329,
+      "learning_rate": 4.826730692276911e-05,
+      "loss": 0.7186,
+      "step": 876
+    },
+    {
+      "epoch": 1.12256,
+      "grad_norm": 0.6124555468559265,
+      "learning_rate": 4.8265306122448984e-05,
+      "loss": 0.7023,
+      "step": 877
+    },
+    {
+      "epoch": 1.12384,
+      "grad_norm": 0.610120415687561,
+      "learning_rate": 4.8263305322128855e-05,
+      "loss": 0.6897,
+      "step": 878
+    },
+    {
+      "epoch": 1.12512,
+      "grad_norm": 0.6392123103141785,
+      "learning_rate": 4.826130452180873e-05,
+      "loss": 0.7632,
+      "step": 879
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.6437667608261108,
+      "learning_rate": 4.82593037214886e-05,
+      "loss": 0.8141,
+      "step": 880
+    },
+    {
+      "epoch": 1.12768,
+      "grad_norm": 0.6260906457901001,
+      "learning_rate": 4.825730292116847e-05,
+      "loss": 0.7017,
+      "step": 881
+    },
+    {
+      "epoch": 1.12896,
+      "grad_norm": 0.6186864972114563,
+      "learning_rate": 4.8255302120848336e-05,
+      "loss": 0.6796,
+      "step": 882
+    },
+    {
+      "epoch": 1.13024,
+      "grad_norm": 0.6303755044937134,
+      "learning_rate": 4.8253301320528215e-05,
+      "loss": 0.7379,
+      "step": 883
+    },
+    {
+      "epoch": 1.13152,
+      "grad_norm": 0.5899752378463745,
+      "learning_rate": 4.8251300520208087e-05,
+      "loss": 0.7175,
+      "step": 884
+    },
+    {
+      "epoch": 1.1328,
+      "grad_norm": 0.5789473652839661,
+      "learning_rate": 4.824929971988796e-05,
+      "loss": 0.6513,
+      "step": 885
+    },
+    {
+      "epoch": 1.13408,
+      "grad_norm": 0.6217113733291626,
+      "learning_rate": 4.824729891956783e-05,
+      "loss": 0.7529,
+      "step": 886
+    },
+    {
+      "epoch": 1.13536,
+      "grad_norm": 0.6412451267242432,
+      "learning_rate": 4.82452981192477e-05,
+      "loss": 0.7977,
+      "step": 887
+    },
+    {
+      "epoch": 1.13664,
+      "grad_norm": 0.6278442144393921,
+      "learning_rate": 4.8243297318927574e-05,
+      "loss": 0.7078,
+      "step": 888
+    },
+    {
+      "epoch": 1.13792,
+      "grad_norm": 0.6618484854698181,
+      "learning_rate": 4.8241296518607446e-05,
+      "loss": 0.7483,
+      "step": 889
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.603378176689148,
+      "learning_rate": 4.823929571828732e-05,
+      "loss": 0.7063,
+      "step": 890
+    },
+    {
+      "epoch": 1.14048,
+      "grad_norm": 0.6106945276260376,
+      "learning_rate": 4.823729491796719e-05,
+      "loss": 0.7299,
+      "step": 891
+    },
+    {
+      "epoch": 1.14176,
+      "grad_norm": 0.6208406090736389,
+      "learning_rate": 4.823529411764706e-05,
+      "loss": 0.7362,
+      "step": 892
+    },
+    {
+      "epoch": 1.14304,
+      "grad_norm": 0.63667231798172,
+      "learning_rate": 4.823329331732693e-05,
+      "loss": 0.744,
+      "step": 893
+    },
+    {
+      "epoch": 1.14432,
+      "grad_norm": 0.5953905582427979,
+      "learning_rate": 4.8231292517006805e-05,
+      "loss": 0.7047,
+      "step": 894
+    },
+    {
+      "epoch": 1.1456,
+      "grad_norm": 0.6081019043922424,
+      "learning_rate": 4.822929171668668e-05,
+      "loss": 0.7546,
+      "step": 895
+    },
+    {
+      "epoch": 1.14688,
+      "grad_norm": 0.6807315349578857,
+      "learning_rate": 4.822729091636655e-05,
+      "loss": 0.7761,
+      "step": 896
+    },
+    {
+      "epoch": 1.14816,
+      "grad_norm": 0.6396183967590332,
+      "learning_rate": 4.822529011604642e-05,
+      "loss": 0.7651,
+      "step": 897
+    },
+    {
+      "epoch": 1.14944,
+      "grad_norm": 0.6671329736709595,
+      "learning_rate": 4.822328931572629e-05,
+      "loss": 0.8052,
+      "step": 898
+    },
+    {
+      "epoch": 1.15072,
+      "grad_norm": 0.6230823397636414,
+      "learning_rate": 4.8221288515406164e-05,
+      "loss": 0.7748,
+      "step": 899
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.6346234083175659,
+      "learning_rate": 4.8219287715086036e-05,
+      "loss": 0.7335,
+      "step": 900
+    },
+    {
+      "epoch": 1.15328,
+      "grad_norm": 0.6477737426757812,
+      "learning_rate": 4.821728691476591e-05,
+      "loss": 0.7561,
+      "step": 901
+    },
+    {
+      "epoch": 1.15456,
+      "grad_norm": 0.6157343983650208,
+      "learning_rate": 4.821528611444578e-05,
+      "loss": 0.7422,
+      "step": 902
+    },
+    {
+      "epoch": 1.15584,
+      "grad_norm": 0.6227743625640869,
+      "learning_rate": 4.821328531412565e-05,
+      "loss": 0.7081,
+      "step": 903
+    },
+    {
+      "epoch": 1.15712,
+      "grad_norm": 0.6442050337791443,
+      "learning_rate": 4.8211284513805524e-05,
+      "loss": 0.7545,
+      "step": 904
+    },
+    {
+      "epoch": 1.1584,
+      "grad_norm": 0.5827674865722656,
+      "learning_rate": 4.8209283713485396e-05,
+      "loss": 0.6298,
+      "step": 905
+    },
+    {
+      "epoch": 1.15968,
+      "grad_norm": 0.6147791147232056,
+      "learning_rate": 4.820728291316527e-05,
+      "loss": 0.7058,
+      "step": 906
+    },
+    {
+      "epoch": 1.16096,
+      "grad_norm": 0.6040810942649841,
+      "learning_rate": 4.820528211284514e-05,
+      "loss": 0.7474,
+      "step": 907
+    },
+    {
+      "epoch": 1.16224,
+      "grad_norm": 0.6344786286354065,
+      "learning_rate": 4.820328131252501e-05,
+      "loss": 0.732,
+      "step": 908
+    },
+    {
+      "epoch": 1.16352,
+      "grad_norm": 0.6694657802581787,
+      "learning_rate": 4.820128051220488e-05,
+      "loss": 0.7453,
+      "step": 909
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.6136607527732849,
+      "learning_rate": 4.8199279711884755e-05,
+      "loss": 0.7458,
+      "step": 910
+    },
+    {
+      "epoch": 1.16608,
+      "grad_norm": 0.6418918371200562,
+      "learning_rate": 4.8197278911564633e-05,
+      "loss": 0.723,
+      "step": 911
+    },
+    {
+      "epoch": 1.16736,
+      "grad_norm": 0.6443724036216736,
+      "learning_rate": 4.81952781112445e-05,
+      "loss": 0.6964,
+      "step": 912
+    },
+    {
+      "epoch": 1.16864,
+      "grad_norm": 0.6705388426780701,
+      "learning_rate": 4.819327731092437e-05,
+      "loss": 0.7893,
+      "step": 913
+    },
+    {
+      "epoch": 1.16992,
+      "grad_norm": 0.6297101974487305,
+      "learning_rate": 4.819127651060424e-05,
+      "loss": 0.7399,
+      "step": 914
+    },
+    {
+      "epoch": 1.1712,
+      "grad_norm": 0.6698152422904968,
+      "learning_rate": 4.8189275710284114e-05,
+      "loss": 0.7513,
+      "step": 915
+    },
+    {
+      "epoch": 1.17248,
+      "grad_norm": 0.6113006472587585,
+      "learning_rate": 4.8187274909963986e-05,
+      "loss": 0.7495,
+      "step": 916
+    },
+    {
+      "epoch": 1.17376,
+      "grad_norm": 0.6909430027008057,
+      "learning_rate": 4.818527410964386e-05,
+      "loss": 0.795,
+      "step": 917
+    },
+    {
+      "epoch": 1.17504,
+      "grad_norm": 0.6630053520202637,
+      "learning_rate": 4.8183273309323736e-05,
+      "loss": 0.7305,
+      "step": 918
+    },
+    {
+      "epoch": 1.17632,
+      "grad_norm": 0.6622920036315918,
+      "learning_rate": 4.818127250900361e-05,
+      "loss": 0.7672,
+      "step": 919
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.6172346472740173,
+      "learning_rate": 4.817927170868347e-05,
+      "loss": 0.6956,
+      "step": 920
+    },
+    {
+      "epoch": 1.17888,
+      "grad_norm": 0.6722678542137146,
+      "learning_rate": 4.8177270908363345e-05,
+      "loss": 0.7633,
+      "step": 921
+    },
+    {
+      "epoch": 1.1801599999999999,
+      "grad_norm": 0.6573194861412048,
+      "learning_rate": 4.817527010804322e-05,
+      "loss": 0.7313,
+      "step": 922
+    },
+    {
+      "epoch": 1.18144,
+      "grad_norm": 0.598237931728363,
+      "learning_rate": 4.817326930772309e-05,
+      "loss": 0.6963,
+      "step": 923
+    },
+    {
+      "epoch": 1.18272,
+      "grad_norm": 0.6254197359085083,
+      "learning_rate": 4.817126850740296e-05,
+      "loss": 0.727,
+      "step": 924
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 0.6584347486495972,
+      "learning_rate": 4.816926770708284e-05,
+      "loss": 0.7281,
+      "step": 925
+    },
+    {
+      "epoch": 1.1852800000000001,
+      "grad_norm": 0.6371662020683289,
+      "learning_rate": 4.816726690676271e-05,
+      "loss": 0.7286,
+      "step": 926
+    },
+    {
+      "epoch": 1.18656,
+      "grad_norm": 0.6230917572975159,
+      "learning_rate": 4.816526610644258e-05,
+      "loss": 0.725,
+      "step": 927
+    },
+    {
+      "epoch": 1.18784,
+      "grad_norm": 0.613785445690155,
+      "learning_rate": 4.816326530612245e-05,
+      "loss": 0.7596,
+      "step": 928
+    },
+    {
+      "epoch": 1.18912,
+      "grad_norm": 0.6123350262641907,
+      "learning_rate": 4.816126450580232e-05,
+      "loss": 0.7271,
+      "step": 929
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.604227602481842,
+      "learning_rate": 4.815926370548219e-05,
+      "loss": 0.72,
+      "step": 930
+    },
+    {
+      "epoch": 1.19168,
+      "grad_norm": 0.6275907754898071,
+      "learning_rate": 4.8157262905162064e-05,
+      "loss": 0.6766,
+      "step": 931
+    },
+    {
+      "epoch": 1.19296,
+      "grad_norm": 0.6855752468109131,
+      "learning_rate": 4.815526210484194e-05,
+      "loss": 0.7628,
+      "step": 932
+    },
+    {
+      "epoch": 1.19424,
+      "grad_norm": 0.6172601580619812,
+      "learning_rate": 4.8153261304521814e-05,
+      "loss": 0.7254,
+      "step": 933
+    },
+    {
+      "epoch": 1.19552,
+      "grad_norm": 0.6443458795547485,
+      "learning_rate": 4.8151260504201686e-05,
+      "loss": 0.7489,
+      "step": 934
+    },
+    {
+      "epoch": 1.1968,
+      "grad_norm": 0.6204928755760193,
+      "learning_rate": 4.814925970388156e-05,
+      "loss": 0.7606,
+      "step": 935
+    },
+    {
+      "epoch": 1.19808,
+      "grad_norm": 0.6195954084396362,
+      "learning_rate": 4.814725890356142e-05,
+      "loss": 0.7121,
+      "step": 936
+    },
+    {
+      "epoch": 1.19936,
+      "grad_norm": 0.6397810578346252,
+      "learning_rate": 4.8145258103241295e-05,
+      "loss": 0.7392,
+      "step": 937
+    },
+    {
+      "epoch": 1.20064,
+      "grad_norm": 0.6325953602790833,
+      "learning_rate": 4.814325730292117e-05,
+      "loss": 0.7099,
+      "step": 938
+    },
+    {
+      "epoch": 1.2019199999999999,
+      "grad_norm": 0.5596916079521179,
+      "learning_rate": 4.8141256502601045e-05,
+      "loss": 0.6416,
+      "step": 939
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.6255161762237549,
+      "learning_rate": 4.813925570228092e-05,
+      "loss": 0.6897,
+      "step": 940
+    },
+    {
+      "epoch": 1.20448,
+      "grad_norm": 0.6093683242797852,
+      "learning_rate": 4.813725490196079e-05,
+      "loss": 0.7421,
+      "step": 941
+    },
+    {
+      "epoch": 1.20576,
+      "grad_norm": 0.6216311454772949,
+      "learning_rate": 4.813525410164066e-05,
+      "loss": 0.6835,
+      "step": 942
+    },
+    {
+      "epoch": 1.2070400000000001,
+      "grad_norm": 0.671823263168335,
+      "learning_rate": 4.813325330132053e-05,
+      "loss": 0.7734,
+      "step": 943
+    },
+    {
+      "epoch": 1.20832,
+      "grad_norm": 0.7053853869438171,
+      "learning_rate": 4.81312525010004e-05,
+      "loss": 0.7447,
+      "step": 944
+    },
+    {
+      "epoch": 1.2096,
+      "grad_norm": 0.5679812431335449,
+      "learning_rate": 4.812925170068027e-05,
+      "loss": 0.6654,
+      "step": 945
+    },
+    {
+      "epoch": 1.21088,
+      "grad_norm": 0.5954858660697937,
+      "learning_rate": 4.812725090036015e-05,
+      "loss": 0.6828,
+      "step": 946
+    },
+    {
+      "epoch": 1.21216,
+      "grad_norm": 0.6449232697486877,
+      "learning_rate": 4.812525010004002e-05,
+      "loss": 0.7228,
+      "step": 947
+    },
+    {
+      "epoch": 1.21344,
+      "grad_norm": 0.6288172006607056,
+      "learning_rate": 4.812324929971989e-05,
+      "loss": 0.7043,
+      "step": 948
+    },
+    {
+      "epoch": 1.21472,
+      "grad_norm": 0.6665524244308472,
+      "learning_rate": 4.8121248499399764e-05,
+      "loss": 0.7963,
+      "step": 949
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.6399210095405579,
+      "learning_rate": 4.8119247699079636e-05,
+      "loss": 0.7559,
+      "step": 950
+    },
+    {
+      "epoch": 1.21728,
+      "grad_norm": 0.611251711845398,
+      "learning_rate": 4.811724689875951e-05,
+      "loss": 0.7279,
+      "step": 951
+    },
+    {
+      "epoch": 1.21856,
+      "grad_norm": 0.6060947775840759,
+      "learning_rate": 4.811524609843937e-05,
+      "loss": 0.7066,
+      "step": 952
+    },
+    {
+      "epoch": 1.21984,
+      "grad_norm": 0.660533607006073,
+      "learning_rate": 4.811324529811925e-05,
+      "loss": 0.7361,
+      "step": 953
+    },
+    {
+      "epoch": 1.22112,
+      "grad_norm": 0.6510452032089233,
+      "learning_rate": 4.811124449779912e-05,
+      "loss": 0.7185,
+      "step": 954
+    },
+    {
+      "epoch": 1.2224,
+      "grad_norm": 0.649373471736908,
+      "learning_rate": 4.8109243697478995e-05,
+      "loss": 0.7748,
+      "step": 955
+    },
+    {
+      "epoch": 1.2236799999999999,
+      "grad_norm": 0.6497557759284973,
+      "learning_rate": 4.810724289715887e-05,
+      "loss": 0.7367,
+      "step": 956
+    },
+    {
+      "epoch": 1.22496,
+      "grad_norm": 0.6782010197639465,
+      "learning_rate": 4.810524209683874e-05,
+      "loss": 0.6939,
+      "step": 957
+    },
+    {
+      "epoch": 1.22624,
+      "grad_norm": 0.647323727607727,
+      "learning_rate": 4.810324129651861e-05,
+      "loss": 0.6888,
+      "step": 958
+    },
+    {
+      "epoch": 1.22752,
+      "grad_norm": 0.6662117838859558,
+      "learning_rate": 4.810124049619848e-05,
+      "loss": 0.7344,
+      "step": 959
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.6221624612808228,
+      "learning_rate": 4.809923969587835e-05,
+      "loss": 0.7277,
+      "step": 960
+    },
+    {
+      "epoch": 1.23008,
+      "grad_norm": 0.6320923566818237,
+      "learning_rate": 4.8097238895558226e-05,
+      "loss": 0.7305,
+      "step": 961
+    },
+    {
+      "epoch": 1.23136,
+      "grad_norm": 0.6177527904510498,
+      "learning_rate": 4.80952380952381e-05,
+      "loss": 0.7247,
+      "step": 962
+    },
+    {
+      "epoch": 1.23264,
+      "grad_norm": 0.6447619199752808,
+      "learning_rate": 4.809323729491797e-05,
+      "loss": 0.739,
+      "step": 963
+    },
+    {
+      "epoch": 1.23392,
+      "grad_norm": 0.6220279335975647,
+      "learning_rate": 4.809123649459784e-05,
+      "loss": 0.7493,
+      "step": 964
+    },
+    {
+      "epoch": 1.2352,
+      "grad_norm": 0.6294159889221191,
+      "learning_rate": 4.8089235694277714e-05,
+      "loss": 0.7178,
+      "step": 965
+    },
+    {
+      "epoch": 1.23648,
+      "grad_norm": 0.6304751634597778,
+      "learning_rate": 4.8087234893957586e-05,
+      "loss": 0.7529,
+      "step": 966
+    },
+    {
+      "epoch": 1.23776,
+      "grad_norm": 0.654147744178772,
+      "learning_rate": 4.808523409363746e-05,
+      "loss": 0.7369,
+      "step": 967
+    },
+    {
+      "epoch": 1.23904,
+      "grad_norm": 0.6700395345687866,
+      "learning_rate": 4.808323329331733e-05,
+      "loss": 0.7395,
+      "step": 968
+    },
+    {
+      "epoch": 1.24032,
+      "grad_norm": 0.6803591251373291,
+      "learning_rate": 4.80812324929972e-05,
+      "loss": 0.815,
+      "step": 969
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.6240243911743164,
+      "learning_rate": 4.807923169267707e-05,
+      "loss": 0.7645,
+      "step": 970
+    },
+    {
+      "epoch": 1.24288,
+      "grad_norm": 0.6563379764556885,
+      "learning_rate": 4.8077230892356945e-05,
+      "loss": 0.7685,
+      "step": 971
+    },
+    {
+      "epoch": 1.24416,
+      "grad_norm": 0.6459086537361145,
+      "learning_rate": 4.807523009203682e-05,
+      "loss": 0.7109,
+      "step": 972
+    },
+    {
+      "epoch": 1.2454399999999999,
+      "grad_norm": 0.6518861651420593,
+      "learning_rate": 4.807322929171669e-05,
+      "loss": 0.7225,
+      "step": 973
+    },
+    {
+      "epoch": 1.24672,
+      "grad_norm": 0.6392287015914917,
+      "learning_rate": 4.807122849139656e-05,
+      "loss": 0.7454,
+      "step": 974
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 0.639618456363678,
+      "learning_rate": 4.806922769107643e-05,
+      "loss": 0.7345,
+      "step": 975
+    },
+    {
+      "epoch": 1.24928,
+      "grad_norm": 0.6422220468521118,
+      "learning_rate": 4.8067226890756304e-05,
+      "loss": 0.7571,
+      "step": 976
+    },
+    {
+      "epoch": 1.2505600000000001,
+      "grad_norm": 0.6119744777679443,
+      "learning_rate": 4.8065226090436176e-05,
+      "loss": 0.7367,
+      "step": 977
+    },
+    {
+      "epoch": 1.25184,
+      "grad_norm": 0.6147500872612,
+      "learning_rate": 4.806322529011605e-05,
+      "loss": 0.6999,
+      "step": 978
+    },
+    {
+      "epoch": 1.25312,
+      "grad_norm": 0.6690681576728821,
+      "learning_rate": 4.806122448979592e-05,
+      "loss": 0.7657,
+      "step": 979
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.6589951515197754,
+      "learning_rate": 4.805922368947579e-05,
+      "loss": 0.6975,
+      "step": 980
+    },
+    {
+      "epoch": 1.25568,
+      "grad_norm": 0.6869469285011292,
+      "learning_rate": 4.8057222889155663e-05,
+      "loss": 0.7403,
+      "step": 981
+    },
+    {
+      "epoch": 1.25696,
+      "grad_norm": 0.6274691224098206,
+      "learning_rate": 4.805522208883554e-05,
+      "loss": 0.7046,
+      "step": 982
+    },
+    {
+      "epoch": 1.25824,
+      "grad_norm": 0.6332339644432068,
+      "learning_rate": 4.805322128851541e-05,
+      "loss": 0.7729,
+      "step": 983
+    },
+    {
+      "epoch": 1.25952,
+      "grad_norm": 0.6267856955528259,
+      "learning_rate": 4.805122048819528e-05,
+      "loss": 0.75,
+      "step": 984
+    },
+    {
+      "epoch": 1.2608,
+      "grad_norm": 0.6330248117446899,
+      "learning_rate": 4.804921968787515e-05,
+      "loss": 0.7085,
+      "step": 985
+    },
+    {
+      "epoch": 1.26208,
+      "grad_norm": 0.6453738212585449,
+      "learning_rate": 4.804721888755502e-05,
+      "loss": 0.7433,
+      "step": 986
+    },
+    {
+      "epoch": 1.26336,
+      "grad_norm": 0.6724556684494019,
+      "learning_rate": 4.8045218087234895e-05,
+      "loss": 0.7734,
+      "step": 987
+    },
+    {
+      "epoch": 1.26464,
+      "grad_norm": 0.6251585483551025,
+      "learning_rate": 4.8043217286914766e-05,
+      "loss": 0.6533,
+      "step": 988
+    },
+    {
+      "epoch": 1.26592,
+      "grad_norm": 0.6305882930755615,
+      "learning_rate": 4.8041216486594645e-05,
+      "loss": 0.7489,
+      "step": 989
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.6422156095504761,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.7583,
+      "step": 990
+    },
+    {
+      "epoch": 1.26848,
+      "grad_norm": 0.6040436029434204,
+      "learning_rate": 4.803721488595438e-05,
+      "loss": 0.6738,
+      "step": 991
+    },
+    {
+      "epoch": 1.26976,
+      "grad_norm": 0.6677682399749756,
+      "learning_rate": 4.8035214085634254e-05,
+      "loss": 0.7181,
+      "step": 992
+    },
+    {
+      "epoch": 1.27104,
+      "grad_norm": 0.6229025721549988,
+      "learning_rate": 4.8033213285314126e-05,
+      "loss": 0.6651,
+      "step": 993
+    },
+    {
+      "epoch": 1.2723200000000001,
+      "grad_norm": 0.6611243486404419,
+      "learning_rate": 4.8031212484994e-05,
+      "loss": 0.787,
+      "step": 994
+    },
+    {
+      "epoch": 1.2736,
+      "grad_norm": 0.6305986642837524,
+      "learning_rate": 4.802921168467387e-05,
+      "loss": 0.763,
+      "step": 995
+    },
+    {
+      "epoch": 1.27488,
+      "grad_norm": 0.6216512322425842,
+      "learning_rate": 4.802721088435375e-05,
+      "loss": 0.7323,
+      "step": 996
+    },
+    {
+      "epoch": 1.27616,
+      "grad_norm": 0.6225135922431946,
+      "learning_rate": 4.802521008403362e-05,
+      "loss": 0.7429,
+      "step": 997
+    },
+    {
+      "epoch": 1.27744,
+      "grad_norm": 0.6232655048370361,
+      "learning_rate": 4.802320928371349e-05,
+      "loss": 0.7163,
+      "step": 998
+    },
+    {
+      "epoch": 1.27872,
+      "grad_norm": 0.585407555103302,
+      "learning_rate": 4.802120848339336e-05,
+      "loss": 0.6991,
+      "step": 999
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6161494255065918,
+      "learning_rate": 4.801920768307323e-05,
+      "loss": 0.7389,
+      "step": 1000
+    },
+    {
+      "epoch": 1.28128,
+      "grad_norm": 0.5975562930107117,
+      "learning_rate": 4.80172068827531e-05,
+      "loss": 0.7076,
+      "step": 1001
+    },
+    {
+      "epoch": 1.28256,
+      "grad_norm": 0.6090825200080872,
+      "learning_rate": 4.801520608243297e-05,
+      "loss": 0.724,
+      "step": 1002
+    },
+    {
+      "epoch": 1.28384,
+      "grad_norm": 0.6444687843322754,
+      "learning_rate": 4.801320528211285e-05,
+      "loss": 0.7562,
+      "step": 1003
+    },
+    {
+      "epoch": 1.28512,
+      "grad_norm": 0.6110180616378784,
+      "learning_rate": 4.801120448179272e-05,
+      "loss": 0.6253,
+      "step": 1004
+    },
+    {
+      "epoch": 1.2864,
+      "grad_norm": 0.683491051197052,
+      "learning_rate": 4.8009203681472595e-05,
+      "loss": 0.794,
+      "step": 1005
+    },
+    {
+      "epoch": 1.28768,
+      "grad_norm": 0.6164990663528442,
+      "learning_rate": 4.8007202881152467e-05,
+      "loss": 0.6802,
+      "step": 1006
+    },
+    {
+      "epoch": 1.2889599999999999,
+      "grad_norm": 0.6573185920715332,
+      "learning_rate": 4.800520208083233e-05,
+      "loss": 0.7123,
+      "step": 1007
+    },
+    {
+      "epoch": 1.29024,
+      "grad_norm": 0.6175433993339539,
+      "learning_rate": 4.8003201280512204e-05,
+      "loss": 0.6906,
+      "step": 1008
+    },
+    {
+      "epoch": 1.29152,
+      "grad_norm": 0.6465235948562622,
+      "learning_rate": 4.8001200480192075e-05,
+      "loss": 0.6843,
+      "step": 1009
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.617728590965271,
+      "learning_rate": 4.7999199679871954e-05,
+      "loss": 0.7187,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2940800000000001,
+      "grad_norm": 0.660306990146637,
+      "learning_rate": 4.7997198879551826e-05,
+      "loss": 0.7547,
+      "step": 1011
+    },
+    {
+      "epoch": 1.29536,
+      "grad_norm": 0.6021302938461304,
+      "learning_rate": 4.79951980792317e-05,
+      "loss": 0.6895,
+      "step": 1012
+    },
+    {
+      "epoch": 1.29664,
+      "grad_norm": 0.6052932143211365,
+      "learning_rate": 4.799319727891157e-05,
+      "loss": 0.7183,
+      "step": 1013
+    },
+    {
+      "epoch": 1.29792,
+      "grad_norm": 0.6130071878433228,
+      "learning_rate": 4.799119647859144e-05,
+      "loss": 0.7347,
+      "step": 1014
+    },
+    {
+      "epoch": 1.2992,
+      "grad_norm": 0.659451961517334,
+      "learning_rate": 4.7989195678271307e-05,
+      "loss": 0.7356,
+      "step": 1015
+    },
+    {
+      "epoch": 1.30048,
+      "grad_norm": 0.6528733372688293,
+      "learning_rate": 4.798719487795118e-05,
+      "loss": 0.6957,
+      "step": 1016
+    },
+    {
+      "epoch": 1.30176,
+      "grad_norm": 0.6614980697631836,
+      "learning_rate": 4.798519407763106e-05,
+      "loss": 0.7466,
+      "step": 1017
+    },
+    {
+      "epoch": 1.30304,
+      "grad_norm": 0.6211322546005249,
+      "learning_rate": 4.798319327731093e-05,
+      "loss": 0.7367,
+      "step": 1018
+    },
+    {
+      "epoch": 1.30432,
+      "grad_norm": 0.6198946833610535,
+      "learning_rate": 4.79811924769908e-05,
+      "loss": 0.7108,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.6257739067077637,
+      "learning_rate": 4.797919167667067e-05,
+      "loss": 0.7444,
+      "step": 1020
+    },
+    {
+      "epoch": 1.30688,
+      "grad_norm": 0.6333909630775452,
+      "learning_rate": 4.7977190876350544e-05,
+      "loss": 0.6794,
+      "step": 1021
+    },
+    {
+      "epoch": 1.30816,
+      "grad_norm": 0.6291680335998535,
+      "learning_rate": 4.7975190076030416e-05,
+      "loss": 0.7711,
+      "step": 1022
+    },
+    {
+      "epoch": 1.30944,
+      "grad_norm": 0.6223838925361633,
+      "learning_rate": 4.797318927571028e-05,
+      "loss": 0.7198,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3107199999999999,
+      "grad_norm": 0.6127943992614746,
+      "learning_rate": 4.797118847539016e-05,
+      "loss": 0.6988,
+      "step": 1024
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 0.6955495476722717,
+      "learning_rate": 4.796918767507003e-05,
+      "loss": 0.7987,
+      "step": 1025
+    },
+    {
+      "epoch": 1.31328,
+      "grad_norm": 0.6549299359321594,
+      "learning_rate": 4.7967186874749904e-05,
+      "loss": 0.7838,
+      "step": 1026
+    },
+    {
+      "epoch": 1.31456,
+      "grad_norm": 0.5967473387718201,
+      "learning_rate": 4.7965186074429776e-05,
+      "loss": 0.7516,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3158400000000001,
+      "grad_norm": 0.6326315999031067,
+      "learning_rate": 4.796318527410965e-05,
+      "loss": 0.7809,
+      "step": 1028
+    },
+    {
+      "epoch": 1.31712,
+      "grad_norm": 0.6114999651908875,
+      "learning_rate": 4.796118447378952e-05,
+      "loss": 0.7084,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.6324224472045898,
+      "learning_rate": 4.795918367346939e-05,
+      "loss": 0.7194,
+      "step": 1030
+    },
+    {
+      "epoch": 1.31968,
+      "grad_norm": 0.6108016967773438,
+      "learning_rate": 4.795718287314926e-05,
+      "loss": 0.7008,
+      "step": 1031
+    },
+    {
+      "epoch": 1.32096,
+      "grad_norm": 0.6326055526733398,
+      "learning_rate": 4.7955182072829135e-05,
+      "loss": 0.747,
+      "step": 1032
+    },
+    {
+      "epoch": 1.32224,
+      "grad_norm": 0.5859755277633667,
+      "learning_rate": 4.795318127250901e-05,
+      "loss": 0.7171,
+      "step": 1033
+    },
+    {
+      "epoch": 1.32352,
+      "grad_norm": 0.6281324028968811,
+      "learning_rate": 4.795118047218888e-05,
+      "loss": 0.7921,
+      "step": 1034
+    },
+    {
+      "epoch": 1.3248,
+      "grad_norm": 0.5920976996421814,
+      "learning_rate": 4.794917967186875e-05,
+      "loss": 0.6885,
+      "step": 1035
+    },
+    {
+      "epoch": 1.32608,
+      "grad_norm": 0.6835001111030579,
+      "learning_rate": 4.794717887154862e-05,
+      "loss": 0.7674,
+      "step": 1036
+    },
+    {
+      "epoch": 1.32736,
+      "grad_norm": 0.6190808415412903,
+      "learning_rate": 4.7945178071228494e-05,
+      "loss": 0.7408,
+      "step": 1037
+    },
+    {
+      "epoch": 1.32864,
+      "grad_norm": 0.6121822595596313,
+      "learning_rate": 4.7943177270908366e-05,
+      "loss": 0.6875,
+      "step": 1038
+    },
+    {
+      "epoch": 1.32992,
+      "grad_norm": 0.6977471709251404,
+      "learning_rate": 4.794117647058824e-05,
+      "loss": 0.8305,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.6396989822387695,
+      "learning_rate": 4.793917567026811e-05,
+      "loss": 0.705,
+      "step": 1040
+    },
+    {
+      "epoch": 1.3324799999999999,
+      "grad_norm": 0.6352539658546448,
+      "learning_rate": 4.793717486994798e-05,
+      "loss": 0.8305,
+      "step": 1041
+    },
+    {
+      "epoch": 1.33376,
+      "grad_norm": 0.6385099291801453,
+      "learning_rate": 4.7935174069627853e-05,
+      "loss": 0.723,
+      "step": 1042
+    },
+    {
+      "epoch": 1.33504,
+      "grad_norm": 0.6024138927459717,
+      "learning_rate": 4.7933173269307725e-05,
+      "loss": 0.684,
+      "step": 1043
+    },
+    {
+      "epoch": 1.33632,
+      "grad_norm": 0.6106967329978943,
+      "learning_rate": 4.79311724689876e-05,
+      "loss": 0.762,
+      "step": 1044
+    },
+    {
+      "epoch": 1.3376000000000001,
+      "grad_norm": 0.6504346132278442,
+      "learning_rate": 4.792917166866747e-05,
+      "loss": 0.6937,
+      "step": 1045
+    },
+    {
+      "epoch": 1.33888,
+      "grad_norm": 0.6468571424484253,
+      "learning_rate": 4.792717086834734e-05,
+      "loss": 0.7919,
+      "step": 1046
+    },
+    {
+      "epoch": 1.34016,
+      "grad_norm": 0.6253741979598999,
+      "learning_rate": 4.792517006802721e-05,
+      "loss": 0.7041,
+      "step": 1047
+    },
+    {
+      "epoch": 1.34144,
+      "grad_norm": 0.6577421426773071,
+      "learning_rate": 4.7923169267707085e-05,
+      "loss": 0.8185,
+      "step": 1048
+    },
+    {
+      "epoch": 1.34272,
+      "grad_norm": 0.6654506325721741,
+      "learning_rate": 4.7921168467386956e-05,
+      "loss": 0.7253,
+      "step": 1049
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.6202994585037231,
+      "learning_rate": 4.791916766706683e-05,
+      "loss": 0.6825,
+      "step": 1050
+    },
+    {
+      "epoch": 1.34528,
+      "grad_norm": 0.6256506443023682,
+      "learning_rate": 4.79171668667467e-05,
+      "loss": 0.6949,
+      "step": 1051
+    },
+    {
+      "epoch": 1.34656,
+      "grad_norm": 0.6409286856651306,
+      "learning_rate": 4.791516606642658e-05,
+      "loss": 0.72,
+      "step": 1052
+    },
+    {
+      "epoch": 1.34784,
+      "grad_norm": 0.6390777230262756,
+      "learning_rate": 4.7913165266106444e-05,
+      "loss": 0.7029,
+      "step": 1053
+    },
+    {
+      "epoch": 1.34912,
+      "grad_norm": 0.620672345161438,
+      "learning_rate": 4.7911164465786316e-05,
+      "loss": 0.7373,
+      "step": 1054
+    },
+    {
+      "epoch": 1.3504,
+      "grad_norm": 0.6143883466720581,
+      "learning_rate": 4.790916366546619e-05,
+      "loss": 0.7371,
+      "step": 1055
+    },
+    {
+      "epoch": 1.35168,
+      "grad_norm": 0.6494544744491577,
+      "learning_rate": 4.790716286514606e-05,
+      "loss": 0.7677,
+      "step": 1056
+    },
+    {
+      "epoch": 1.35296,
+      "grad_norm": 0.6207596659660339,
+      "learning_rate": 4.790516206482593e-05,
+      "loss": 0.7106,
+      "step": 1057
+    },
+    {
+      "epoch": 1.3542399999999999,
+      "grad_norm": 0.6054299473762512,
+      "learning_rate": 4.79031612645058e-05,
+      "loss": 0.7055,
+      "step": 1058
+    },
+    {
+      "epoch": 1.35552,
+      "grad_norm": 0.6713016629219055,
+      "learning_rate": 4.790116046418568e-05,
+      "loss": 0.7229,
+      "step": 1059
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.6039015054702759,
+      "learning_rate": 4.7899159663865554e-05,
+      "loss": 0.6853,
+      "step": 1060
+    },
+    {
+      "epoch": 1.35808,
+      "grad_norm": 0.6022716164588928,
+      "learning_rate": 4.789715886354542e-05,
+      "loss": 0.7085,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3593600000000001,
+      "grad_norm": 0.6193332076072693,
+      "learning_rate": 4.789515806322529e-05,
+      "loss": 0.7405,
+      "step": 1062
+    },
+    {
+      "epoch": 1.36064,
+      "grad_norm": 0.6295928359031677,
+      "learning_rate": 4.789315726290516e-05,
+      "loss": 0.7142,
+      "step": 1063
+    },
+    {
+      "epoch": 1.36192,
+      "grad_norm": 0.6390275359153748,
+      "learning_rate": 4.7891156462585034e-05,
+      "loss": 0.7402,
+      "step": 1064
+    },
+    {
+      "epoch": 1.3632,
+      "grad_norm": 0.6579723954200745,
+      "learning_rate": 4.7889155662264906e-05,
+      "loss": 0.7671,
+      "step": 1065
+    },
+    {
+      "epoch": 1.36448,
+      "grad_norm": 0.6290743350982666,
+      "learning_rate": 4.7887154861944785e-05,
+      "loss": 0.7414,
+      "step": 1066
+    },
+    {
+      "epoch": 1.3657599999999999,
+      "grad_norm": 0.6132893562316895,
+      "learning_rate": 4.788515406162466e-05,
+      "loss": 0.6352,
+      "step": 1067
+    },
+    {
+      "epoch": 1.36704,
+      "grad_norm": 0.6277240514755249,
+      "learning_rate": 4.788315326130453e-05,
+      "loss": 0.6889,
+      "step": 1068
+    },
+    {
+      "epoch": 1.36832,
+      "grad_norm": 0.6457761526107788,
+      "learning_rate": 4.7881152460984394e-05,
+      "loss": 0.7332,
+      "step": 1069
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.6207125186920166,
+      "learning_rate": 4.7879151660664265e-05,
+      "loss": 0.711,
+      "step": 1070
+    },
+    {
+      "epoch": 1.37088,
+      "grad_norm": 0.662109375,
+      "learning_rate": 4.787715086034414e-05,
+      "loss": 0.8474,
+      "step": 1071
+    },
+    {
+      "epoch": 1.37216,
+      "grad_norm": 0.6627180576324463,
+      "learning_rate": 4.787515006002401e-05,
+      "loss": 0.7317,
+      "step": 1072
+    },
+    {
+      "epoch": 1.37344,
+      "grad_norm": 0.6555449962615967,
+      "learning_rate": 4.787314925970388e-05,
+      "loss": 0.767,
+      "step": 1073
+    },
+    {
+      "epoch": 1.37472,
+      "grad_norm": 0.619805634021759,
+      "learning_rate": 4.787114845938376e-05,
+      "loss": 0.6963,
+      "step": 1074
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 0.6373822093009949,
+      "learning_rate": 4.786914765906363e-05,
+      "loss": 0.7016,
+      "step": 1075
+    },
+    {
+      "epoch": 1.37728,
+      "grad_norm": 0.6532891392707825,
+      "learning_rate": 4.78671468587435e-05,
+      "loss": 0.694,
+      "step": 1076
+    },
+    {
+      "epoch": 1.37856,
+      "grad_norm": 0.6646358370780945,
+      "learning_rate": 4.786514605842337e-05,
+      "loss": 0.7097,
+      "step": 1077
+    },
+    {
+      "epoch": 1.37984,
+      "grad_norm": 0.6382097601890564,
+      "learning_rate": 4.786314525810324e-05,
+      "loss": 0.7613,
+      "step": 1078
+    },
+    {
+      "epoch": 1.3811200000000001,
+      "grad_norm": 0.5848649144172668,
+      "learning_rate": 4.786114445778311e-05,
+      "loss": 0.7163,
+      "step": 1079
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.6216439008712769,
+      "learning_rate": 4.7859143657462984e-05,
+      "loss": 0.708,
+      "step": 1080
+    },
+    {
+      "epoch": 1.38368,
+      "grad_norm": 0.6398138999938965,
+      "learning_rate": 4.785714285714286e-05,
+      "loss": 0.7093,
+      "step": 1081
+    },
+    {
+      "epoch": 1.38496,
+      "grad_norm": 0.601285994052887,
+      "learning_rate": 4.7855142056822734e-05,
+      "loss": 0.7506,
+      "step": 1082
+    },
+    {
+      "epoch": 1.38624,
+      "grad_norm": 0.5848278999328613,
+      "learning_rate": 4.7853141256502606e-05,
+      "loss": 0.7303,
+      "step": 1083
+    },
+    {
+      "epoch": 1.3875199999999999,
+      "grad_norm": 0.6471388339996338,
+      "learning_rate": 4.785114045618248e-05,
+      "loss": 0.7204,
+      "step": 1084
+    },
+    {
+      "epoch": 1.3888,
+      "grad_norm": 0.6106283068656921,
+      "learning_rate": 4.784913965586234e-05,
+      "loss": 0.6882,
+      "step": 1085
+    },
+    {
+      "epoch": 1.39008,
+      "grad_norm": 0.6192972660064697,
+      "learning_rate": 4.7847138855542215e-05,
+      "loss": 0.7204,
+      "step": 1086
+    },
+    {
+      "epoch": 1.39136,
+      "grad_norm": 0.6262179613113403,
+      "learning_rate": 4.784513805522209e-05,
+      "loss": 0.7779,
+      "step": 1087
+    },
+    {
+      "epoch": 1.39264,
+      "grad_norm": 0.5966930985450745,
+      "learning_rate": 4.7843137254901966e-05,
+      "loss": 0.7018,
+      "step": 1088
+    },
+    {
+      "epoch": 1.39392,
+      "grad_norm": 0.5986944437026978,
+      "learning_rate": 4.784113645458184e-05,
+      "loss": 0.7918,
+      "step": 1089
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.649640679359436,
+      "learning_rate": 4.783913565426171e-05,
+      "loss": 0.8004,
+      "step": 1090
+    },
+    {
+      "epoch": 1.39648,
+      "grad_norm": 0.6140245199203491,
+      "learning_rate": 4.783713485394158e-05,
+      "loss": 0.7116,
+      "step": 1091
+    },
+    {
+      "epoch": 1.39776,
+      "grad_norm": 0.6077390909194946,
+      "learning_rate": 4.783513405362145e-05,
+      "loss": 0.7018,
+      "step": 1092
+    },
+    {
+      "epoch": 1.39904,
+      "grad_norm": 0.6407623887062073,
+      "learning_rate": 4.783313325330132e-05,
+      "loss": 0.7147,
+      "step": 1093
+    },
+    {
+      "epoch": 1.40032,
+      "grad_norm": 0.6380225419998169,
+      "learning_rate": 4.783113245298119e-05,
+      "loss": 0.7089,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4016,
+      "grad_norm": 0.5923783183097839,
+      "learning_rate": 4.782913165266107e-05,
+      "loss": 0.6903,
+      "step": 1095
+    },
+    {
+      "epoch": 1.4028800000000001,
+      "grad_norm": 0.6157001852989197,
+      "learning_rate": 4.782713085234094e-05,
+      "loss": 0.7125,
+      "step": 1096
+    },
+    {
+      "epoch": 1.40416,
+      "grad_norm": 0.6462724208831787,
+      "learning_rate": 4.782513005202081e-05,
+      "loss": 0.7091,
+      "step": 1097
+    },
+    {
+      "epoch": 1.40544,
+      "grad_norm": 0.6279814839363098,
+      "learning_rate": 4.7823129251700684e-05,
+      "loss": 0.7286,
+      "step": 1098
+    },
+    {
+      "epoch": 1.40672,
+      "grad_norm": 0.6624937057495117,
+      "learning_rate": 4.7821128451380556e-05,
+      "loss": 0.7458,
+      "step": 1099
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.5872619152069092,
+      "learning_rate": 4.781912765106043e-05,
+      "loss": 0.6632,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4092799999999999,
+      "grad_norm": 0.6288279294967651,
+      "learning_rate": 4.781712685074029e-05,
+      "loss": 0.7713,
+      "step": 1101
+    },
+    {
+      "epoch": 1.41056,
+      "grad_norm": 0.6384450793266296,
+      "learning_rate": 4.781512605042017e-05,
+      "loss": 0.7663,
+      "step": 1102
+    },
+    {
+      "epoch": 1.41184,
+      "grad_norm": 0.6555289626121521,
+      "learning_rate": 4.7813125250100043e-05,
+      "loss": 0.7339,
+      "step": 1103
+    },
+    {
+      "epoch": 1.41312,
+      "grad_norm": 0.6459196209907532,
+      "learning_rate": 4.7811124449779915e-05,
+      "loss": 0.7491,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4144,
+      "grad_norm": 0.6260380148887634,
+      "learning_rate": 4.780912364945979e-05,
+      "loss": 0.7219,
+      "step": 1105
+    },
+    {
+      "epoch": 1.41568,
+      "grad_norm": 0.602838933467865,
+      "learning_rate": 4.780712284913966e-05,
+      "loss": 0.6817,
+      "step": 1106
+    },
+    {
+      "epoch": 1.41696,
+      "grad_norm": 0.6588259935379028,
+      "learning_rate": 4.780512204881953e-05,
+      "loss": 0.7484,
+      "step": 1107
+    },
+    {
+      "epoch": 1.41824,
+      "grad_norm": 0.6435507535934448,
+      "learning_rate": 4.78031212484994e-05,
+      "loss": 0.7629,
+      "step": 1108
+    },
+    {
+      "epoch": 1.41952,
+      "grad_norm": 0.6327363848686218,
+      "learning_rate": 4.7801120448179275e-05,
+      "loss": 0.7366,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.7007765769958496,
+      "learning_rate": 4.7799119647859146e-05,
+      "loss": 0.7356,
+      "step": 1110
+    },
+    {
+      "epoch": 1.42208,
+      "grad_norm": 0.6083685159683228,
+      "learning_rate": 4.779711884753902e-05,
+      "loss": 0.7023,
+      "step": 1111
+    },
+    {
+      "epoch": 1.42336,
+      "grad_norm": 0.6256844997406006,
+      "learning_rate": 4.779511804721889e-05,
+      "loss": 0.709,
+      "step": 1112
+    },
+    {
+      "epoch": 1.4246400000000001,
+      "grad_norm": 0.6442143321037292,
+      "learning_rate": 4.779311724689876e-05,
+      "loss": 0.6688,
+      "step": 1113
+    },
+    {
+      "epoch": 1.42592,
+      "grad_norm": 0.6138997673988342,
+      "learning_rate": 4.7791116446578634e-05,
+      "loss": 0.7211,
+      "step": 1114
+    },
+    {
+      "epoch": 1.4272,
+      "grad_norm": 0.6425867676734924,
+      "learning_rate": 4.7789115646258506e-05,
+      "loss": 0.7207,
+      "step": 1115
+    },
+    {
+      "epoch": 1.42848,
+      "grad_norm": 0.6066508889198303,
+      "learning_rate": 4.778711484593838e-05,
+      "loss": 0.7242,
+      "step": 1116
+    },
+    {
+      "epoch": 1.42976,
+      "grad_norm": 0.5861117243766785,
+      "learning_rate": 4.778511404561825e-05,
+      "loss": 0.6949,
+      "step": 1117
+    },
+    {
+      "epoch": 1.4310399999999999,
+      "grad_norm": 0.6210717558860779,
+      "learning_rate": 4.778311324529812e-05,
+      "loss": 0.7018,
+      "step": 1118
+    },
+    {
+      "epoch": 1.43232,
+      "grad_norm": 0.6068648099899292,
+      "learning_rate": 4.778111244497799e-05,
+      "loss": 0.7291,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.5990647673606873,
+      "learning_rate": 4.7779111644657865e-05,
+      "loss": 0.7016,
+      "step": 1120
+    },
+    {
+      "epoch": 1.43488,
+      "grad_norm": 0.5916838049888611,
+      "learning_rate": 4.777711084433774e-05,
+      "loss": 0.6895,
+      "step": 1121
+    },
+    {
+      "epoch": 1.43616,
+      "grad_norm": 0.6179905533790588,
+      "learning_rate": 4.777511004401761e-05,
+      "loss": 0.6555,
+      "step": 1122
+    },
+    {
+      "epoch": 1.43744,
+      "grad_norm": 0.626921534538269,
+      "learning_rate": 4.777310924369748e-05,
+      "loss": 0.7184,
+      "step": 1123
+    },
+    {
+      "epoch": 1.43872,
+      "grad_norm": 0.6084997057914734,
+      "learning_rate": 4.777110844337735e-05,
+      "loss": 0.6929,
+      "step": 1124
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6844547390937805,
+      "learning_rate": 4.7769107643057224e-05,
+      "loss": 0.7536,
+      "step": 1125
+    },
+    {
+      "epoch": 1.44128,
+      "grad_norm": 0.6585316061973572,
+      "learning_rate": 4.7767106842737096e-05,
+      "loss": 0.7163,
+      "step": 1126
+    },
+    {
+      "epoch": 1.44256,
+      "grad_norm": 0.6486889123916626,
+      "learning_rate": 4.776510604241697e-05,
+      "loss": 0.7305,
+      "step": 1127
+    },
+    {
+      "epoch": 1.44384,
+      "grad_norm": 0.5947285890579224,
+      "learning_rate": 4.776310524209684e-05,
+      "loss": 0.7082,
+      "step": 1128
+    },
+    {
+      "epoch": 1.44512,
+      "grad_norm": 0.6200329065322876,
+      "learning_rate": 4.776110444177671e-05,
+      "loss": 0.7104,
+      "step": 1129
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.6373615860939026,
+      "learning_rate": 4.775910364145659e-05,
+      "loss": 0.7293,
+      "step": 1130
+    },
+    {
+      "epoch": 1.44768,
+      "grad_norm": 0.6362440586090088,
+      "learning_rate": 4.7757102841136455e-05,
+      "loss": 0.7632,
+      "step": 1131
+    },
+    {
+      "epoch": 1.44896,
+      "grad_norm": 0.6665335893630981,
+      "learning_rate": 4.775510204081633e-05,
+      "loss": 0.7378,
+      "step": 1132
+    },
+    {
+      "epoch": 1.45024,
+      "grad_norm": 0.6526025533676147,
+      "learning_rate": 4.77531012404962e-05,
+      "loss": 0.7462,
+      "step": 1133
+    },
+    {
+      "epoch": 1.45152,
+      "grad_norm": 0.5860944986343384,
+      "learning_rate": 4.775110044017607e-05,
+      "loss": 0.6819,
+      "step": 1134
+    },
+    {
+      "epoch": 1.4527999999999999,
+      "grad_norm": 0.6410444378852844,
+      "learning_rate": 4.774909963985594e-05,
+      "loss": 0.7381,
+      "step": 1135
+    },
+    {
+      "epoch": 1.45408,
+      "grad_norm": 0.6227912306785583,
+      "learning_rate": 4.7747098839535815e-05,
+      "loss": 0.6726,
+      "step": 1136
+    },
+    {
+      "epoch": 1.45536,
+      "grad_norm": 0.633658230304718,
+      "learning_rate": 4.774509803921569e-05,
+      "loss": 0.705,
+      "step": 1137
+    },
+    {
+      "epoch": 1.45664,
+      "grad_norm": 0.6156677007675171,
+      "learning_rate": 4.7743097238895565e-05,
+      "loss": 0.6913,
+      "step": 1138
+    },
+    {
+      "epoch": 1.45792,
+      "grad_norm": 0.6303706169128418,
+      "learning_rate": 4.774109643857543e-05,
+      "loss": 0.721,
+      "step": 1139
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.5811339020729065,
+      "learning_rate": 4.77390956382553e-05,
+      "loss": 0.727,
+      "step": 1140
+    },
+    {
+      "epoch": 1.46048,
+      "grad_norm": 0.6627085208892822,
+      "learning_rate": 4.7737094837935174e-05,
+      "loss": 0.7594,
+      "step": 1141
+    },
+    {
+      "epoch": 1.46176,
+      "grad_norm": 0.640595555305481,
+      "learning_rate": 4.7735094037615046e-05,
+      "loss": 0.7016,
+      "step": 1142
+    },
+    {
+      "epoch": 1.46304,
+      "grad_norm": 0.623562216758728,
+      "learning_rate": 4.773309323729492e-05,
+      "loss": 0.7622,
+      "step": 1143
+    },
+    {
+      "epoch": 1.46432,
+      "grad_norm": 0.6135162711143494,
+      "learning_rate": 4.7731092436974796e-05,
+      "loss": 0.6906,
+      "step": 1144
+    },
+    {
+      "epoch": 1.4656,
+      "grad_norm": 0.6718834042549133,
+      "learning_rate": 4.772909163665467e-05,
+      "loss": 0.7215,
+      "step": 1145
+    },
+    {
+      "epoch": 1.46688,
+      "grad_norm": 0.6534852981567383,
+      "learning_rate": 4.772709083633454e-05,
+      "loss": 0.7794,
+      "step": 1146
+    },
+    {
+      "epoch": 1.4681600000000001,
+      "grad_norm": 0.6073404550552368,
+      "learning_rate": 4.7725090036014405e-05,
+      "loss": 0.6651,
+      "step": 1147
+    },
+    {
+      "epoch": 1.46944,
+      "grad_norm": 0.6143197417259216,
+      "learning_rate": 4.772308923569428e-05,
+      "loss": 0.7012,
+      "step": 1148
+    },
+    {
+      "epoch": 1.47072,
+      "grad_norm": 0.6453569531440735,
+      "learning_rate": 4.772108843537415e-05,
+      "loss": 0.7459,
+      "step": 1149
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.6633815169334412,
+      "learning_rate": 4.771908763505402e-05,
+      "loss": 0.7477,
+      "step": 1150
+    },
+    {
+      "epoch": 1.47328,
+      "grad_norm": 0.6284077167510986,
+      "learning_rate": 4.77170868347339e-05,
+      "loss": 0.7471,
+      "step": 1151
+    },
+    {
+      "epoch": 1.4745599999999999,
+      "grad_norm": 0.6360135674476624,
+      "learning_rate": 4.771508603441377e-05,
+      "loss": 0.7069,
+      "step": 1152
+    },
+    {
+      "epoch": 1.47584,
+      "grad_norm": 0.6397899389266968,
+      "learning_rate": 4.771308523409364e-05,
+      "loss": 0.6937,
+      "step": 1153
+    },
+    {
+      "epoch": 1.47712,
+      "grad_norm": 0.6112055778503418,
+      "learning_rate": 4.7711084433773515e-05,
+      "loss": 0.7514,
+      "step": 1154
+    },
+    {
+      "epoch": 1.4784,
+      "grad_norm": 0.608342170715332,
+      "learning_rate": 4.770908363345338e-05,
+      "loss": 0.7085,
+      "step": 1155
+    },
+    {
+      "epoch": 1.47968,
+      "grad_norm": 0.6625191569328308,
+      "learning_rate": 4.770708283313325e-05,
+      "loss": 0.7201,
+      "step": 1156
+    },
+    {
+      "epoch": 1.48096,
+      "grad_norm": 0.6654885411262512,
+      "learning_rate": 4.7705082032813124e-05,
+      "loss": 0.7487,
+      "step": 1157
+    },
+    {
+      "epoch": 1.48224,
+      "grad_norm": 0.6362460851669312,
+      "learning_rate": 4.7703081232493e-05,
+      "loss": 0.7151,
+      "step": 1158
+    },
+    {
+      "epoch": 1.48352,
+      "grad_norm": 0.6087484359741211,
+      "learning_rate": 4.7701080432172874e-05,
+      "loss": 0.7062,
+      "step": 1159
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.6134206652641296,
+      "learning_rate": 4.7699079631852746e-05,
+      "loss": 0.7385,
+      "step": 1160
+    },
+    {
+      "epoch": 1.48608,
+      "grad_norm": 0.5883345603942871,
+      "learning_rate": 4.769707883153262e-05,
+      "loss": 0.7447,
+      "step": 1161
+    },
+    {
+      "epoch": 1.48736,
+      "grad_norm": 0.6443062424659729,
+      "learning_rate": 4.769507803121249e-05,
+      "loss": 0.6859,
+      "step": 1162
+    },
+    {
+      "epoch": 1.48864,
+      "grad_norm": 0.6045071482658386,
+      "learning_rate": 4.7693077230892355e-05,
+      "loss": 0.7062,
+      "step": 1163
+    },
+    {
+      "epoch": 1.4899200000000001,
+      "grad_norm": 0.6627300381660461,
+      "learning_rate": 4.769107643057223e-05,
+      "loss": 0.7499,
+      "step": 1164
+    },
+    {
+      "epoch": 1.4912,
+      "grad_norm": 0.6466721892356873,
+      "learning_rate": 4.7689075630252105e-05,
+      "loss": 0.7218,
+      "step": 1165
+    },
+    {
+      "epoch": 1.49248,
+      "grad_norm": 0.6242368817329407,
+      "learning_rate": 4.768707482993198e-05,
+      "loss": 0.7273,
+      "step": 1166
+    },
+    {
+      "epoch": 1.49376,
+      "grad_norm": 0.614842414855957,
+      "learning_rate": 4.768507402961185e-05,
+      "loss": 0.7216,
+      "step": 1167
+    },
+    {
+      "epoch": 1.49504,
+      "grad_norm": 0.6312155723571777,
+      "learning_rate": 4.768307322929172e-05,
+      "loss": 0.7566,
+      "step": 1168
+    },
+    {
+      "epoch": 1.4963199999999999,
+      "grad_norm": 0.6303775906562805,
+      "learning_rate": 4.768107242897159e-05,
+      "loss": 0.7076,
+      "step": 1169
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.6277086734771729,
+      "learning_rate": 4.7679071628651465e-05,
+      "loss": 0.6755,
+      "step": 1170
+    },
+    {
+      "epoch": 1.49888,
+      "grad_norm": 0.6203471422195435,
+      "learning_rate": 4.767707082833133e-05,
+      "loss": 0.716,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5001600000000002,
+      "grad_norm": 0.6107301115989685,
+      "learning_rate": 4.767507002801121e-05,
+      "loss": 0.7121,
+      "step": 1172
+    },
+    {
+      "epoch": 1.50144,
+      "grad_norm": 0.6433287858963013,
+      "learning_rate": 4.767306922769108e-05,
+      "loss": 0.7116,
+      "step": 1173
+    },
+    {
+      "epoch": 1.50272,
+      "grad_norm": 0.6727626919746399,
+      "learning_rate": 4.767106842737095e-05,
+      "loss": 0.7444,
+      "step": 1174
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 0.6506801247596741,
+      "learning_rate": 4.7669067627050824e-05,
+      "loss": 0.6786,
+      "step": 1175
+    },
+    {
+      "epoch": 1.50528,
+      "grad_norm": 0.642342746257782,
+      "learning_rate": 4.7667066826730696e-05,
+      "loss": 0.7178,
+      "step": 1176
+    },
+    {
+      "epoch": 1.50656,
+      "grad_norm": 0.6692764759063721,
+      "learning_rate": 4.766506602641057e-05,
+      "loss": 0.7793,
+      "step": 1177
+    },
+    {
+      "epoch": 1.5078399999999998,
+      "grad_norm": 0.6183244585990906,
+      "learning_rate": 4.766306522609044e-05,
+      "loss": 0.7034,
+      "step": 1178
+    },
+    {
+      "epoch": 1.50912,
+      "grad_norm": 0.6591295003890991,
+      "learning_rate": 4.766106442577031e-05,
+      "loss": 0.7507,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.6517340540885925,
+      "learning_rate": 4.765906362545018e-05,
+      "loss": 0.7276,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5116800000000001,
+      "grad_norm": 0.6169537901878357,
+      "learning_rate": 4.7657062825130055e-05,
+      "loss": 0.7647,
+      "step": 1181
+    },
+    {
+      "epoch": 1.51296,
+      "grad_norm": 0.6543555855751038,
+      "learning_rate": 4.765506202480993e-05,
+      "loss": 0.7227,
+      "step": 1182
+    },
+    {
+      "epoch": 1.51424,
+      "grad_norm": 0.5778194665908813,
+      "learning_rate": 4.76530612244898e-05,
+      "loss": 0.6666,
+      "step": 1183
+    },
+    {
+      "epoch": 1.51552,
+      "grad_norm": 0.6411781907081604,
+      "learning_rate": 4.765106042416967e-05,
+      "loss": 0.7447,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5168,
+      "grad_norm": 0.6526811718940735,
+      "learning_rate": 4.764905962384954e-05,
+      "loss": 0.6795,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5180799999999999,
+      "grad_norm": 0.6295871734619141,
+      "learning_rate": 4.7647058823529414e-05,
+      "loss": 0.7025,
+      "step": 1186
+    },
+    {
+      "epoch": 1.51936,
+      "grad_norm": 0.6581454277038574,
+      "learning_rate": 4.7645058023209286e-05,
+      "loss": 0.7639,
+      "step": 1187
+    },
+    {
+      "epoch": 1.52064,
+      "grad_norm": 0.6672399640083313,
+      "learning_rate": 4.764305722288916e-05,
+      "loss": 0.7384,
+      "step": 1188
+    },
+    {
+      "epoch": 1.5219200000000002,
+      "grad_norm": 0.6672317385673523,
+      "learning_rate": 4.764105642256903e-05,
+      "loss": 0.7389,
+      "step": 1189
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.6444031000137329,
+      "learning_rate": 4.76390556222489e-05,
+      "loss": 0.7833,
+      "step": 1190
+    },
+    {
+      "epoch": 1.52448,
+      "grad_norm": 0.6317417621612549,
+      "learning_rate": 4.7637054821928774e-05,
+      "loss": 0.7112,
+      "step": 1191
+    },
+    {
+      "epoch": 1.52576,
+      "grad_norm": 0.627162516117096,
+      "learning_rate": 4.7635054021608646e-05,
+      "loss": 0.6628,
+      "step": 1192
+    },
+    {
+      "epoch": 1.52704,
+      "grad_norm": 0.5882164239883423,
+      "learning_rate": 4.763305322128852e-05,
+      "loss": 0.7103,
+      "step": 1193
+    },
+    {
+      "epoch": 1.52832,
+      "grad_norm": 0.6305367350578308,
+      "learning_rate": 4.763105242096839e-05,
+      "loss": 0.7103,
+      "step": 1194
+    },
+    {
+      "epoch": 1.5295999999999998,
+      "grad_norm": 0.6403663158416748,
+      "learning_rate": 4.762905162064826e-05,
+      "loss": 0.6768,
+      "step": 1195
+    },
+    {
+      "epoch": 1.53088,
+      "grad_norm": 0.6659502983093262,
+      "learning_rate": 4.762705082032813e-05,
+      "loss": 0.7166,
+      "step": 1196
+    },
+    {
+      "epoch": 1.53216,
+      "grad_norm": 0.6176123023033142,
+      "learning_rate": 4.7625050020008005e-05,
+      "loss": 0.6794,
+      "step": 1197
+    },
+    {
+      "epoch": 1.5334400000000001,
+      "grad_norm": 0.6616097688674927,
+      "learning_rate": 4.762304921968788e-05,
+      "loss": 0.7685,
+      "step": 1198
+    },
+    {
+      "epoch": 1.53472,
+      "grad_norm": 0.6266618967056274,
+      "learning_rate": 4.762104841936775e-05,
+      "loss": 0.6784,
+      "step": 1199
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.6575819849967957,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.7432,
+      "step": 1200
+    },
+    {
+      "epoch": 1.53728,
+      "grad_norm": 0.6002248525619507,
+      "learning_rate": 4.761704681872749e-05,
+      "loss": 0.6946,
+      "step": 1201
+    },
+    {
+      "epoch": 1.53856,
+      "grad_norm": 0.6573600172996521,
+      "learning_rate": 4.7615046018407364e-05,
+      "loss": 0.7185,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5398399999999999,
+      "grad_norm": 0.6227654218673706,
+      "learning_rate": 4.7613045218087236e-05,
+      "loss": 0.7381,
+      "step": 1203
+    },
+    {
+      "epoch": 1.54112,
+      "grad_norm": 0.6044524908065796,
+      "learning_rate": 4.761104441776711e-05,
+      "loss": 0.704,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5424,
+      "grad_norm": 0.6178364753723145,
+      "learning_rate": 4.760904361744698e-05,
+      "loss": 0.6916,
+      "step": 1205
+    },
+    {
+      "epoch": 1.5436800000000002,
+      "grad_norm": 0.6074690222740173,
+      "learning_rate": 4.760704281712685e-05,
+      "loss": 0.7256,
+      "step": 1206
+    },
+    {
+      "epoch": 1.54496,
+      "grad_norm": 0.6533793807029724,
+      "learning_rate": 4.760504201680672e-05,
+      "loss": 0.6969,
+      "step": 1207
+    },
+    {
+      "epoch": 1.54624,
+      "grad_norm": 0.6194373965263367,
+      "learning_rate": 4.76030412164866e-05,
+      "loss": 0.7138,
+      "step": 1208
+    },
+    {
+      "epoch": 1.54752,
+      "grad_norm": 0.6618860363960266,
+      "learning_rate": 4.760104041616647e-05,
+      "loss": 0.7754,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.6461208462715149,
+      "learning_rate": 4.759903961584634e-05,
+      "loss": 0.7114,
+      "step": 1210
+    },
+    {
+      "epoch": 1.55008,
+      "grad_norm": 0.6476582884788513,
+      "learning_rate": 4.759703881552621e-05,
+      "loss": 0.6973,
+      "step": 1211
+    },
+    {
+      "epoch": 1.5513599999999999,
+      "grad_norm": 0.6671127080917358,
+      "learning_rate": 4.759503801520608e-05,
+      "loss": 0.7009,
+      "step": 1212
+    },
+    {
+      "epoch": 1.55264,
+      "grad_norm": 0.6456823348999023,
+      "learning_rate": 4.7593037214885954e-05,
+      "loss": 0.7214,
+      "step": 1213
+    },
+    {
+      "epoch": 1.55392,
+      "grad_norm": 0.6629720330238342,
+      "learning_rate": 4.7591036414565826e-05,
+      "loss": 0.7583,
+      "step": 1214
+    },
+    {
+      "epoch": 1.5552000000000001,
+      "grad_norm": 0.649437427520752,
+      "learning_rate": 4.7589035614245705e-05,
+      "loss": 0.7532,
+      "step": 1215
+    },
+    {
+      "epoch": 1.55648,
+      "grad_norm": 0.6041565537452698,
+      "learning_rate": 4.758703481392558e-05,
+      "loss": 0.6884,
+      "step": 1216
+    },
+    {
+      "epoch": 1.55776,
+      "grad_norm": 0.6424383521080017,
+      "learning_rate": 4.758503401360544e-05,
+      "loss": 0.7084,
+      "step": 1217
+    },
+    {
+      "epoch": 1.55904,
+      "grad_norm": 0.6945663690567017,
+      "learning_rate": 4.7583033213285314e-05,
+      "loss": 0.75,
+      "step": 1218
+    },
+    {
+      "epoch": 1.56032,
+      "grad_norm": 0.6336300373077393,
+      "learning_rate": 4.7581032412965186e-05,
+      "loss": 0.7537,
+      "step": 1219
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.6776460409164429,
+      "learning_rate": 4.757903161264506e-05,
+      "loss": 0.7617,
+      "step": 1220
+    },
+    {
+      "epoch": 1.56288,
+      "grad_norm": 0.6466001272201538,
+      "learning_rate": 4.757703081232493e-05,
+      "loss": 0.7482,
+      "step": 1221
+    },
+    {
+      "epoch": 1.56416,
+      "grad_norm": 0.630434513092041,
+      "learning_rate": 4.757503001200481e-05,
+      "loss": 0.724,
+      "step": 1222
+    },
+    {
+      "epoch": 1.5654400000000002,
+      "grad_norm": 0.6976414322853088,
+      "learning_rate": 4.757302921168468e-05,
+      "loss": 0.7818,
+      "step": 1223
+    },
+    {
+      "epoch": 1.5667200000000001,
+      "grad_norm": 0.6347379684448242,
+      "learning_rate": 4.757102841136455e-05,
+      "loss": 0.7551,
+      "step": 1224
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 0.6358786225318909,
+      "learning_rate": 4.756902761104442e-05,
+      "loss": 0.6909,
+      "step": 1225
+    },
+    {
+      "epoch": 1.56928,
+      "grad_norm": 0.6460133194923401,
+      "learning_rate": 4.756702681072429e-05,
+      "loss": 0.7101,
+      "step": 1226
+    },
+    {
+      "epoch": 1.57056,
+      "grad_norm": 0.6669771671295166,
+      "learning_rate": 4.756502601040416e-05,
+      "loss": 0.7721,
+      "step": 1227
+    },
+    {
+      "epoch": 1.57184,
+      "grad_norm": 0.6039112210273743,
+      "learning_rate": 4.756302521008403e-05,
+      "loss": 0.712,
+      "step": 1228
+    },
+    {
+      "epoch": 1.5731199999999999,
+      "grad_norm": 0.6793951392173767,
+      "learning_rate": 4.756102440976391e-05,
+      "loss": 0.747,
+      "step": 1229
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.6434732675552368,
+      "learning_rate": 4.755902360944378e-05,
+      "loss": 0.6894,
+      "step": 1230
+    },
+    {
+      "epoch": 1.57568,
+      "grad_norm": 0.6476708650588989,
+      "learning_rate": 4.7557022809123655e-05,
+      "loss": 0.6805,
+      "step": 1231
+    },
+    {
+      "epoch": 1.5769600000000001,
+      "grad_norm": 0.6917094588279724,
+      "learning_rate": 4.7555022008803527e-05,
+      "loss": 0.6978,
+      "step": 1232
+    },
+    {
+      "epoch": 1.57824,
+      "grad_norm": 0.631247878074646,
+      "learning_rate": 4.755302120848339e-05,
+      "loss": 0.6858,
+      "step": 1233
+    },
+    {
+      "epoch": 1.57952,
+      "grad_norm": 0.6304442286491394,
+      "learning_rate": 4.7551020408163263e-05,
+      "loss": 0.6866,
+      "step": 1234
+    },
+    {
+      "epoch": 1.5808,
+      "grad_norm": 0.6345778703689575,
+      "learning_rate": 4.7549019607843135e-05,
+      "loss": 0.7146,
+      "step": 1235
+    },
+    {
+      "epoch": 1.58208,
+      "grad_norm": 0.639773964881897,
+      "learning_rate": 4.7547018807523014e-05,
+      "loss": 0.7268,
+      "step": 1236
+    },
+    {
+      "epoch": 1.5833599999999999,
+      "grad_norm": 0.6575612425804138,
+      "learning_rate": 4.7545018007202886e-05,
+      "loss": 0.7142,
+      "step": 1237
+    },
+    {
+      "epoch": 1.58464,
+      "grad_norm": 0.6063089966773987,
+      "learning_rate": 4.754301720688276e-05,
+      "loss": 0.7651,
+      "step": 1238
+    },
+    {
+      "epoch": 1.58592,
+      "grad_norm": 0.6471563577651978,
+      "learning_rate": 4.754101640656263e-05,
+      "loss": 0.7748,
+      "step": 1239
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.6223157048225403,
+      "learning_rate": 4.75390156062425e-05,
+      "loss": 0.7069,
+      "step": 1240
+    },
+    {
+      "epoch": 1.5884800000000001,
+      "grad_norm": 0.64573073387146,
+      "learning_rate": 4.7537014805922366e-05,
+      "loss": 0.766,
+      "step": 1241
+    },
+    {
+      "epoch": 1.58976,
+      "grad_norm": 0.6501642465591431,
+      "learning_rate": 4.753501400560224e-05,
+      "loss": 0.7455,
+      "step": 1242
+    },
+    {
+      "epoch": 1.59104,
+      "grad_norm": 0.6377203464508057,
+      "learning_rate": 4.753301320528212e-05,
+      "loss": 0.6988,
+      "step": 1243
+    },
+    {
+      "epoch": 1.59232,
+      "grad_norm": 0.6198776364326477,
+      "learning_rate": 4.753101240496199e-05,
+      "loss": 0.7555,
+      "step": 1244
+    },
+    {
+      "epoch": 1.5936,
+      "grad_norm": 0.6620365381240845,
+      "learning_rate": 4.752901160464186e-05,
+      "loss": 0.7119,
+      "step": 1245
+    },
+    {
+      "epoch": 1.5948799999999999,
+      "grad_norm": 0.6182314157485962,
+      "learning_rate": 4.752701080432173e-05,
+      "loss": 0.6488,
+      "step": 1246
+    },
+    {
+      "epoch": 1.59616,
+      "grad_norm": 0.6267403960227966,
+      "learning_rate": 4.7525010004001604e-05,
+      "loss": 0.7483,
+      "step": 1247
+    },
+    {
+      "epoch": 1.59744,
+      "grad_norm": 0.6369834542274475,
+      "learning_rate": 4.7523009203681476e-05,
+      "loss": 0.6785,
+      "step": 1248
+    },
+    {
+      "epoch": 1.5987200000000001,
+      "grad_norm": 0.6468684077262878,
+      "learning_rate": 4.752100840336134e-05,
+      "loss": 0.7245,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6526497006416321,
+      "learning_rate": 4.751900760304122e-05,
+      "loss": 0.7311,
+      "step": 1250
+    },
+    {
+      "epoch": 1.60128,
+      "grad_norm": 0.6682162284851074,
+      "learning_rate": 4.751700680272109e-05,
+      "loss": 0.7086,
+      "step": 1251
+    },
+    {
+      "epoch": 1.60256,
+      "grad_norm": 0.6937553286552429,
+      "learning_rate": 4.7515006002400964e-05,
+      "loss": 0.6991,
+      "step": 1252
+    },
+    {
+      "epoch": 1.60384,
+      "grad_norm": 0.6066246032714844,
+      "learning_rate": 4.7513005202080836e-05,
+      "loss": 0.6762,
+      "step": 1253
+    },
+    {
+      "epoch": 1.6051199999999999,
+      "grad_norm": 0.7019811272621155,
+      "learning_rate": 4.751100440176071e-05,
+      "loss": 0.7746,
+      "step": 1254
+    },
+    {
+      "epoch": 1.6064,
+      "grad_norm": 0.6307134032249451,
+      "learning_rate": 4.750900360144058e-05,
+      "loss": 0.7058,
+      "step": 1255
+    },
+    {
+      "epoch": 1.60768,
+      "grad_norm": 0.6239479780197144,
+      "learning_rate": 4.750700280112045e-05,
+      "loss": 0.6974,
+      "step": 1256
+    },
+    {
+      "epoch": 1.60896,
+      "grad_norm": 0.6378730535507202,
+      "learning_rate": 4.750500200080032e-05,
+      "loss": 0.7755,
+      "step": 1257
+    },
+    {
+      "epoch": 1.6102400000000001,
+      "grad_norm": 0.6123172640800476,
+      "learning_rate": 4.7503001200480195e-05,
+      "loss": 0.6952,
+      "step": 1258
+    },
+    {
+      "epoch": 1.61152,
+      "grad_norm": 0.6375351548194885,
+      "learning_rate": 4.750100040016007e-05,
+      "loss": 0.7457,
+      "step": 1259
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.6497233510017395,
+      "learning_rate": 4.749899959983994e-05,
+      "loss": 0.7198,
+      "step": 1260
+    },
+    {
+      "epoch": 1.61408,
+      "grad_norm": 0.6439145803451538,
+      "learning_rate": 4.749699879951981e-05,
+      "loss": 0.7273,
+      "step": 1261
+    },
+    {
+      "epoch": 1.61536,
+      "grad_norm": 0.590140700340271,
+      "learning_rate": 4.749499799919968e-05,
+      "loss": 0.6665,
+      "step": 1262
+    },
+    {
+      "epoch": 1.6166399999999999,
+      "grad_norm": 0.6305953860282898,
+      "learning_rate": 4.7492997198879554e-05,
+      "loss": 0.7422,
+      "step": 1263
+    },
+    {
+      "epoch": 1.61792,
+      "grad_norm": 0.6392373442649841,
+      "learning_rate": 4.7490996398559426e-05,
+      "loss": 0.7263,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6192,
+      "grad_norm": 0.6296020746231079,
+      "learning_rate": 4.74889955982393e-05,
+      "loss": 0.7455,
+      "step": 1265
+    },
+    {
+      "epoch": 1.6204800000000001,
+      "grad_norm": 0.6070896983146667,
+      "learning_rate": 4.748699479791917e-05,
+      "loss": 0.7014,
+      "step": 1266
+    },
+    {
+      "epoch": 1.62176,
+      "grad_norm": 0.6309097409248352,
+      "learning_rate": 4.748499399759904e-05,
+      "loss": 0.7737,
+      "step": 1267
+    },
+    {
+      "epoch": 1.62304,
+      "grad_norm": 0.591622531414032,
+      "learning_rate": 4.7482993197278913e-05,
+      "loss": 0.7032,
+      "step": 1268
+    },
+    {
+      "epoch": 1.62432,
+      "grad_norm": 0.5985342860221863,
+      "learning_rate": 4.7480992396958785e-05,
+      "loss": 0.7006,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.5956323146820068,
+      "learning_rate": 4.747899159663866e-05,
+      "loss": 0.6925,
+      "step": 1270
+    },
+    {
+      "epoch": 1.6268799999999999,
+      "grad_norm": 0.6757445931434631,
+      "learning_rate": 4.747699079631853e-05,
+      "loss": 0.7091,
+      "step": 1271
+    },
+    {
+      "epoch": 1.62816,
+      "grad_norm": 0.6045747399330139,
+      "learning_rate": 4.74749899959984e-05,
+      "loss": 0.6801,
+      "step": 1272
+    },
+    {
+      "epoch": 1.62944,
+      "grad_norm": 0.6032465696334839,
+      "learning_rate": 4.747298919567827e-05,
+      "loss": 0.6913,
+      "step": 1273
+    },
+    {
+      "epoch": 1.63072,
+      "grad_norm": 0.6105408072471619,
+      "learning_rate": 4.7470988395358145e-05,
+      "loss": 0.7102,
+      "step": 1274
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 0.6021878123283386,
+      "learning_rate": 4.7468987595038016e-05,
+      "loss": 0.66,
+      "step": 1275
+    },
+    {
+      "epoch": 1.63328,
+      "grad_norm": 0.6317958235740662,
+      "learning_rate": 4.746698679471789e-05,
+      "loss": 0.7622,
+      "step": 1276
+    },
+    {
+      "epoch": 1.63456,
+      "grad_norm": 0.6472170352935791,
+      "learning_rate": 4.746498599439776e-05,
+      "loss": 0.8176,
+      "step": 1277
+    },
+    {
+      "epoch": 1.63584,
+      "grad_norm": 0.6169067621231079,
+      "learning_rate": 4.746298519407764e-05,
+      "loss": 0.7191,
+      "step": 1278
+    },
+    {
+      "epoch": 1.63712,
+      "grad_norm": 0.6021884083747864,
+      "learning_rate": 4.7460984393757504e-05,
+      "loss": 0.7359,
+      "step": 1279
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.6332974433898926,
+      "learning_rate": 4.7458983593437376e-05,
+      "loss": 0.7494,
+      "step": 1280
+    },
+    {
+      "epoch": 1.63968,
+      "grad_norm": 0.6417760252952576,
+      "learning_rate": 4.745698279311725e-05,
+      "loss": 0.7384,
+      "step": 1281
+    },
+    {
+      "epoch": 1.64096,
+      "grad_norm": 0.6329305768013,
+      "learning_rate": 4.745498199279712e-05,
+      "loss": 0.7398,
+      "step": 1282
+    },
+    {
+      "epoch": 1.6422400000000001,
+      "grad_norm": 0.6128019690513611,
+      "learning_rate": 4.745298119247699e-05,
+      "loss": 0.6854,
+      "step": 1283
+    },
+    {
+      "epoch": 1.64352,
+      "grad_norm": 0.6532317399978638,
+      "learning_rate": 4.745098039215686e-05,
+      "loss": 0.6853,
+      "step": 1284
+    },
+    {
+      "epoch": 1.6448,
+      "grad_norm": 0.6134045720100403,
+      "learning_rate": 4.744897959183674e-05,
+      "loss": 0.6766,
+      "step": 1285
+    },
+    {
+      "epoch": 1.64608,
+      "grad_norm": 0.6327788829803467,
+      "learning_rate": 4.7446978791516614e-05,
+      "loss": 0.6836,
+      "step": 1286
+    },
+    {
+      "epoch": 1.64736,
+      "grad_norm": 0.669122040271759,
+      "learning_rate": 4.744497799119648e-05,
+      "loss": 0.7196,
+      "step": 1287
+    },
+    {
+      "epoch": 1.6486399999999999,
+      "grad_norm": 0.614807665348053,
+      "learning_rate": 4.744297719087635e-05,
+      "loss": 0.7626,
+      "step": 1288
+    },
+    {
+      "epoch": 1.64992,
+      "grad_norm": 0.6070460081100464,
+      "learning_rate": 4.744097639055622e-05,
+      "loss": 0.6698,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.6391851902008057,
+      "learning_rate": 4.7438975590236094e-05,
+      "loss": 0.7167,
+      "step": 1290
+    },
+    {
+      "epoch": 1.65248,
+      "grad_norm": 0.6460067629814148,
+      "learning_rate": 4.7436974789915966e-05,
+      "loss": 0.7165,
+      "step": 1291
+    },
+    {
+      "epoch": 1.6537600000000001,
+      "grad_norm": 0.6238282918930054,
+      "learning_rate": 4.7434973989595845e-05,
+      "loss": 0.7058,
+      "step": 1292
+    },
+    {
+      "epoch": 1.65504,
+      "grad_norm": 0.6249229311943054,
+      "learning_rate": 4.7432973189275717e-05,
+      "loss": 0.7196,
+      "step": 1293
+    },
+    {
+      "epoch": 1.65632,
+      "grad_norm": 0.6393983960151672,
+      "learning_rate": 4.743097238895559e-05,
+      "loss": 0.6811,
+      "step": 1294
+    },
+    {
+      "epoch": 1.6576,
+      "grad_norm": 0.6782253384590149,
+      "learning_rate": 4.7428971588635454e-05,
+      "loss": 0.7559,
+      "step": 1295
+    },
+    {
+      "epoch": 1.65888,
+      "grad_norm": 0.5967546701431274,
+      "learning_rate": 4.7426970788315325e-05,
+      "loss": 0.6603,
+      "step": 1296
+    },
+    {
+      "epoch": 1.6601599999999999,
+      "grad_norm": 0.6373478174209595,
+      "learning_rate": 4.74249699879952e-05,
+      "loss": 0.7279,
+      "step": 1297
+    },
+    {
+      "epoch": 1.66144,
+      "grad_norm": 0.6224157214164734,
+      "learning_rate": 4.742296918767507e-05,
+      "loss": 0.6636,
+      "step": 1298
+    },
+    {
+      "epoch": 1.66272,
+      "grad_norm": 0.6072138547897339,
+      "learning_rate": 4.742096838735494e-05,
+      "loss": 0.6899,
+      "step": 1299
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.6565014719963074,
+      "learning_rate": 4.741896758703482e-05,
+      "loss": 0.7157,
+      "step": 1300
+    },
+    {
+      "epoch": 1.66528,
+      "grad_norm": 0.653264045715332,
+      "learning_rate": 4.741696678671469e-05,
+      "loss": 0.7142,
+      "step": 1301
+    },
+    {
+      "epoch": 1.66656,
+      "grad_norm": 0.6316604018211365,
+      "learning_rate": 4.741496598639456e-05,
+      "loss": 0.6723,
+      "step": 1302
+    },
+    {
+      "epoch": 1.66784,
+      "grad_norm": 0.6184001564979553,
+      "learning_rate": 4.741296518607443e-05,
+      "loss": 0.7288,
+      "step": 1303
+    },
+    {
+      "epoch": 1.66912,
+      "grad_norm": 0.6457557082176208,
+      "learning_rate": 4.74109643857543e-05,
+      "loss": 0.7429,
+      "step": 1304
+    },
+    {
+      "epoch": 1.6703999999999999,
+      "grad_norm": 0.6284366250038147,
+      "learning_rate": 4.740896358543417e-05,
+      "loss": 0.6764,
+      "step": 1305
+    },
+    {
+      "epoch": 1.67168,
+      "grad_norm": 0.6020685434341431,
+      "learning_rate": 4.7406962785114044e-05,
+      "loss": 0.6978,
+      "step": 1306
+    },
+    {
+      "epoch": 1.67296,
+      "grad_norm": 0.6325390338897705,
+      "learning_rate": 4.740496198479392e-05,
+      "loss": 0.7271,
+      "step": 1307
+    },
+    {
+      "epoch": 1.67424,
+      "grad_norm": 0.6092963814735413,
+      "learning_rate": 4.7402961184473794e-05,
+      "loss": 0.7342,
+      "step": 1308
+    },
+    {
+      "epoch": 1.6755200000000001,
+      "grad_norm": 0.6341113448143005,
+      "learning_rate": 4.7400960384153666e-05,
+      "loss": 0.7043,
+      "step": 1309
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.6368173360824585,
+      "learning_rate": 4.739895958383354e-05,
+      "loss": 0.7099,
+      "step": 1310
+    },
+    {
+      "epoch": 1.67808,
+      "grad_norm": 0.6337018609046936,
+      "learning_rate": 4.73969587835134e-05,
+      "loss": 0.6765,
+      "step": 1311
+    },
+    {
+      "epoch": 1.67936,
+      "grad_norm": 0.7150176167488098,
+      "learning_rate": 4.7394957983193275e-05,
+      "loss": 0.7172,
+      "step": 1312
+    },
+    {
+      "epoch": 1.68064,
+      "grad_norm": 0.6691837906837463,
+      "learning_rate": 4.739295718287315e-05,
+      "loss": 0.7095,
+      "step": 1313
+    },
+    {
+      "epoch": 1.6819199999999999,
+      "grad_norm": 0.6361916661262512,
+      "learning_rate": 4.7390956382553026e-05,
+      "loss": 0.7367,
+      "step": 1314
+    },
+    {
+      "epoch": 1.6832,
+      "grad_norm": 0.6631631255149841,
+      "learning_rate": 4.73889555822329e-05,
+      "loss": 0.7083,
+      "step": 1315
+    },
+    {
+      "epoch": 1.68448,
+      "grad_norm": 0.6436387896537781,
+      "learning_rate": 4.738695478191277e-05,
+      "loss": 0.7754,
+      "step": 1316
+    },
+    {
+      "epoch": 1.6857600000000001,
+      "grad_norm": 0.644277811050415,
+      "learning_rate": 4.738495398159264e-05,
+      "loss": 0.7025,
+      "step": 1317
+    },
+    {
+      "epoch": 1.68704,
+      "grad_norm": 0.6064467430114746,
+      "learning_rate": 4.738295318127251e-05,
+      "loss": 0.7252,
+      "step": 1318
+    },
+    {
+      "epoch": 1.68832,
+      "grad_norm": 0.6227229237556458,
+      "learning_rate": 4.738095238095238e-05,
+      "loss": 0.6998,
+      "step": 1319
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.644025981426239,
+      "learning_rate": 4.737895158063225e-05,
+      "loss": 0.718,
+      "step": 1320
+    },
+    {
+      "epoch": 1.69088,
+      "grad_norm": 0.6950893998146057,
+      "learning_rate": 4.737695078031213e-05,
+      "loss": 0.7618,
+      "step": 1321
+    },
+    {
+      "epoch": 1.6921599999999999,
+      "grad_norm": 0.6096273064613342,
+      "learning_rate": 4.7374949979992e-05,
+      "loss": 0.7492,
+      "step": 1322
+    },
+    {
+      "epoch": 1.6934399999999998,
+      "grad_norm": 0.624535322189331,
+      "learning_rate": 4.737294917967187e-05,
+      "loss": 0.6749,
+      "step": 1323
+    },
+    {
+      "epoch": 1.69472,
+      "grad_norm": 0.7543303966522217,
+      "learning_rate": 4.7370948379351744e-05,
+      "loss": 0.7168,
+      "step": 1324
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 0.6681160926818848,
+      "learning_rate": 4.7368947579031616e-05,
+      "loss": 0.7414,
+      "step": 1325
+    },
+    {
+      "epoch": 1.6972800000000001,
+      "grad_norm": 0.6174480319023132,
+      "learning_rate": 4.736694677871149e-05,
+      "loss": 0.702,
+      "step": 1326
+    },
+    {
+      "epoch": 1.69856,
+      "grad_norm": 0.6031041145324707,
+      "learning_rate": 4.736494597839135e-05,
+      "loss": 0.6811,
+      "step": 1327
+    },
+    {
+      "epoch": 1.69984,
+      "grad_norm": 0.6663240194320679,
+      "learning_rate": 4.736294517807123e-05,
+      "loss": 0.7624,
+      "step": 1328
+    },
+    {
+      "epoch": 1.70112,
+      "grad_norm": 0.6095741391181946,
+      "learning_rate": 4.7360944377751103e-05,
+      "loss": 0.6708,
+      "step": 1329
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.6273066997528076,
+      "learning_rate": 4.7358943577430975e-05,
+      "loss": 0.7313,
+      "step": 1330
+    },
+    {
+      "epoch": 1.7036799999999999,
+      "grad_norm": 0.6379857063293457,
+      "learning_rate": 4.735694277711085e-05,
+      "loss": 0.7075,
+      "step": 1331
+    },
+    {
+      "epoch": 1.70496,
+      "grad_norm": 0.606400728225708,
+      "learning_rate": 4.735494197679072e-05,
+      "loss": 0.7267,
+      "step": 1332
+    },
+    {
+      "epoch": 1.70624,
+      "grad_norm": 0.6663306355476379,
+      "learning_rate": 4.735294117647059e-05,
+      "loss": 0.7154,
+      "step": 1333
+    },
+    {
+      "epoch": 1.7075200000000001,
+      "grad_norm": 0.6471573114395142,
+      "learning_rate": 4.735094037615046e-05,
+      "loss": 0.7464,
+      "step": 1334
+    },
+    {
+      "epoch": 1.7088,
+      "grad_norm": 0.6170628070831299,
+      "learning_rate": 4.7348939575830335e-05,
+      "loss": 0.7251,
+      "step": 1335
+    },
+    {
+      "epoch": 1.71008,
+      "grad_norm": 0.6125094890594482,
+      "learning_rate": 4.7346938775510206e-05,
+      "loss": 0.7548,
+      "step": 1336
+    },
+    {
+      "epoch": 1.71136,
+      "grad_norm": 0.6116237044334412,
+      "learning_rate": 4.734493797519008e-05,
+      "loss": 0.7347,
+      "step": 1337
+    },
+    {
+      "epoch": 1.71264,
+      "grad_norm": 0.6676164269447327,
+      "learning_rate": 4.734293717486995e-05,
+      "loss": 0.7439,
+      "step": 1338
+    },
+    {
+      "epoch": 1.7139199999999999,
+      "grad_norm": 0.6193406581878662,
+      "learning_rate": 4.734093637454982e-05,
+      "loss": 0.7186,
+      "step": 1339
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.5905833840370178,
+      "learning_rate": 4.7338935574229694e-05,
+      "loss": 0.6775,
+      "step": 1340
+    },
+    {
+      "epoch": 1.71648,
+      "grad_norm": 0.6464902758598328,
+      "learning_rate": 4.7336934773909566e-05,
+      "loss": 0.7586,
+      "step": 1341
+    },
+    {
+      "epoch": 1.71776,
+      "grad_norm": 0.6692549586296082,
+      "learning_rate": 4.733493397358944e-05,
+      "loss": 0.7933,
+      "step": 1342
+    },
+    {
+      "epoch": 1.7190400000000001,
+      "grad_norm": 0.6424185037612915,
+      "learning_rate": 4.733293317326931e-05,
+      "loss": 0.7422,
+      "step": 1343
+    },
+    {
+      "epoch": 1.72032,
+      "grad_norm": 0.6273303031921387,
+      "learning_rate": 4.733093237294918e-05,
+      "loss": 0.6837,
+      "step": 1344
+    },
+    {
+      "epoch": 1.7216,
+      "grad_norm": 0.6037375330924988,
+      "learning_rate": 4.732893157262905e-05,
+      "loss": 0.7306,
+      "step": 1345
+    },
+    {
+      "epoch": 1.72288,
+      "grad_norm": 0.6088179349899292,
+      "learning_rate": 4.7326930772308925e-05,
+      "loss": 0.7008,
+      "step": 1346
+    },
+    {
+      "epoch": 1.72416,
+      "grad_norm": 0.6654016375541687,
+      "learning_rate": 4.73249299719888e-05,
+      "loss": 0.7958,
+      "step": 1347
+    },
+    {
+      "epoch": 1.7254399999999999,
+      "grad_norm": 0.6818312406539917,
+      "learning_rate": 4.732292917166867e-05,
+      "loss": 0.7136,
+      "step": 1348
+    },
+    {
+      "epoch": 1.72672,
+      "grad_norm": 0.5944163799285889,
+      "learning_rate": 4.732092837134855e-05,
+      "loss": 0.7097,
+      "step": 1349
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.6351943016052246,
+      "learning_rate": 4.731892757102841e-05,
+      "loss": 0.7255,
+      "step": 1350
+    },
+    {
+      "epoch": 1.7292800000000002,
+      "grad_norm": 0.6025944948196411,
+      "learning_rate": 4.7316926770708284e-05,
+      "loss": 0.6275,
+      "step": 1351
+    },
+    {
+      "epoch": 1.73056,
+      "grad_norm": 0.6137747764587402,
+      "learning_rate": 4.7314925970388156e-05,
+      "loss": 0.722,
+      "step": 1352
+    },
+    {
+      "epoch": 1.73184,
+      "grad_norm": 0.6036489605903625,
+      "learning_rate": 4.731292517006803e-05,
+      "loss": 0.7132,
+      "step": 1353
+    },
+    {
+      "epoch": 1.73312,
+      "grad_norm": 0.6218430399894714,
+      "learning_rate": 4.73109243697479e-05,
+      "loss": 0.6794,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7344,
+      "grad_norm": 0.582331120967865,
+      "learning_rate": 4.730892356942777e-05,
+      "loss": 0.7159,
+      "step": 1355
+    },
+    {
+      "epoch": 1.73568,
+      "grad_norm": 0.626875638961792,
+      "learning_rate": 4.730692276910765e-05,
+      "loss": 0.7335,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7369599999999998,
+      "grad_norm": 0.6015035510063171,
+      "learning_rate": 4.730492196878752e-05,
+      "loss": 0.6889,
+      "step": 1357
+    },
+    {
+      "epoch": 1.73824,
+      "grad_norm": 0.6084362268447876,
+      "learning_rate": 4.730292116846739e-05,
+      "loss": 0.7034,
+      "step": 1358
+    },
+    {
+      "epoch": 1.73952,
+      "grad_norm": 0.6344590187072754,
+      "learning_rate": 4.730092036814726e-05,
+      "loss": 0.6738,
+      "step": 1359
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.6038119792938232,
+      "learning_rate": 4.729891956782713e-05,
+      "loss": 0.7495,
+      "step": 1360
+    },
+    {
+      "epoch": 1.74208,
+      "grad_norm": 0.5990369319915771,
+      "learning_rate": 4.7296918767507e-05,
+      "loss": 0.7376,
+      "step": 1361
+    },
+    {
+      "epoch": 1.74336,
+      "grad_norm": 0.6152486205101013,
+      "learning_rate": 4.7294917967186875e-05,
+      "loss": 0.6552,
+      "step": 1362
+    },
+    {
+      "epoch": 1.74464,
+      "grad_norm": 0.6658695936203003,
+      "learning_rate": 4.729291716686675e-05,
+      "loss": 0.7151,
+      "step": 1363
+    },
+    {
+      "epoch": 1.74592,
+      "grad_norm": 0.6392641067504883,
+      "learning_rate": 4.7290916366546625e-05,
+      "loss": 0.7298,
+      "step": 1364
+    },
+    {
+      "epoch": 1.7471999999999999,
+      "grad_norm": 0.6059170961380005,
+      "learning_rate": 4.72889155662265e-05,
+      "loss": 0.7054,
+      "step": 1365
+    },
+    {
+      "epoch": 1.74848,
+      "grad_norm": 0.5970711708068848,
+      "learning_rate": 4.728691476590636e-05,
+      "loss": 0.7038,
+      "step": 1366
+    },
+    {
+      "epoch": 1.74976,
+      "grad_norm": 0.6231157183647156,
+      "learning_rate": 4.7284913965586234e-05,
+      "loss": 0.6903,
+      "step": 1367
+    },
+    {
+      "epoch": 1.7510400000000002,
+      "grad_norm": 0.6200019121170044,
+      "learning_rate": 4.7282913165266106e-05,
+      "loss": 0.6916,
+      "step": 1368
+    },
+    {
+      "epoch": 1.75232,
+      "grad_norm": 0.62948077917099,
+      "learning_rate": 4.728091236494598e-05,
+      "loss": 0.7023,
+      "step": 1369
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.6121958494186401,
+      "learning_rate": 4.7278911564625856e-05,
+      "loss": 0.6912,
+      "step": 1370
+    },
+    {
+      "epoch": 1.75488,
+      "grad_norm": 0.6384387612342834,
+      "learning_rate": 4.727691076430573e-05,
+      "loss": 0.7216,
+      "step": 1371
+    },
+    {
+      "epoch": 1.75616,
+      "grad_norm": 0.6484329700469971,
+      "learning_rate": 4.72749099639856e-05,
+      "loss": 0.6696,
+      "step": 1372
+    },
+    {
+      "epoch": 1.75744,
+      "grad_norm": 0.6550239324569702,
+      "learning_rate": 4.727290916366547e-05,
+      "loss": 0.7213,
+      "step": 1373
+    },
+    {
+      "epoch": 1.7587199999999998,
+      "grad_norm": 0.6201356649398804,
+      "learning_rate": 4.727090836334534e-05,
+      "loss": 0.7322,
+      "step": 1374
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6567116379737854,
+      "learning_rate": 4.726890756302521e-05,
+      "loss": 0.7732,
+      "step": 1375
+    },
+    {
+      "epoch": 1.76128,
+      "grad_norm": 0.6301355957984924,
+      "learning_rate": 4.726690676270508e-05,
+      "loss": 0.7397,
+      "step": 1376
+    },
+    {
+      "epoch": 1.7625600000000001,
+      "grad_norm": 0.6590511202812195,
+      "learning_rate": 4.726490596238496e-05,
+      "loss": 0.7581,
+      "step": 1377
+    },
+    {
+      "epoch": 1.76384,
+      "grad_norm": 0.6398580074310303,
+      "learning_rate": 4.726290516206483e-05,
+      "loss": 0.7062,
+      "step": 1378
+    },
+    {
+      "epoch": 1.76512,
+      "grad_norm": 0.6533066630363464,
+      "learning_rate": 4.72609043617447e-05,
+      "loss": 0.7078,
+      "step": 1379
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.6254117488861084,
+      "learning_rate": 4.7258903561424575e-05,
+      "loss": 0.6917,
+      "step": 1380
+    },
+    {
+      "epoch": 1.76768,
+      "grad_norm": 0.6385900378227234,
+      "learning_rate": 4.725690276110445e-05,
+      "loss": 0.7951,
+      "step": 1381
+    },
+    {
+      "epoch": 1.7689599999999999,
+      "grad_norm": 0.6786746978759766,
+      "learning_rate": 4.725490196078431e-05,
+      "loss": 0.7645,
+      "step": 1382
+    },
+    {
+      "epoch": 1.77024,
+      "grad_norm": 0.6253941059112549,
+      "learning_rate": 4.7252901160464184e-05,
+      "loss": 0.7064,
+      "step": 1383
+    },
+    {
+      "epoch": 1.77152,
+      "grad_norm": 0.6229246258735657,
+      "learning_rate": 4.725090036014406e-05,
+      "loss": 0.7386,
+      "step": 1384
+    },
+    {
+      "epoch": 1.7728000000000002,
+      "grad_norm": 0.6111705303192139,
+      "learning_rate": 4.7248899559823934e-05,
+      "loss": 0.6646,
+      "step": 1385
+    },
+    {
+      "epoch": 1.77408,
+      "grad_norm": 0.6427714228630066,
+      "learning_rate": 4.7246898759503806e-05,
+      "loss": 0.7058,
+      "step": 1386
+    },
+    {
+      "epoch": 1.77536,
+      "grad_norm": 0.635543704032898,
+      "learning_rate": 4.724489795918368e-05,
+      "loss": 0.7016,
+      "step": 1387
+    },
+    {
+      "epoch": 1.77664,
+      "grad_norm": 0.6031673550605774,
+      "learning_rate": 4.724289715886355e-05,
+      "loss": 0.6427,
+      "step": 1388
+    },
+    {
+      "epoch": 1.77792,
+      "grad_norm": 0.6220544576644897,
+      "learning_rate": 4.724089635854342e-05,
+      "loss": 0.731,
+      "step": 1389
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.6029757261276245,
+      "learning_rate": 4.723889555822329e-05,
+      "loss": 0.7165,
+      "step": 1390
+    },
+    {
+      "epoch": 1.7804799999999998,
+      "grad_norm": 0.6523854732513428,
+      "learning_rate": 4.7236894757903165e-05,
+      "loss": 0.7179,
+      "step": 1391
+    },
+    {
+      "epoch": 1.78176,
+      "grad_norm": 0.6103372573852539,
+      "learning_rate": 4.723489395758304e-05,
+      "loss": 0.6943,
+      "step": 1392
+    },
+    {
+      "epoch": 1.78304,
+      "grad_norm": 0.631454348564148,
+      "learning_rate": 4.723289315726291e-05,
+      "loss": 0.7071,
+      "step": 1393
+    },
+    {
+      "epoch": 1.7843200000000001,
+      "grad_norm": 0.6054040193557739,
+      "learning_rate": 4.723089235694278e-05,
+      "loss": 0.6621,
+      "step": 1394
+    },
+    {
+      "epoch": 1.7856,
+      "grad_norm": 0.6552069783210754,
+      "learning_rate": 4.722889155662265e-05,
+      "loss": 0.7767,
+      "step": 1395
+    },
+    {
+      "epoch": 1.78688,
+      "grad_norm": 0.6483544111251831,
+      "learning_rate": 4.7226890756302525e-05,
+      "loss": 0.759,
+      "step": 1396
+    },
+    {
+      "epoch": 1.78816,
+      "grad_norm": 0.6473653316497803,
+      "learning_rate": 4.7224889955982396e-05,
+      "loss": 0.6879,
+      "step": 1397
+    },
+    {
+      "epoch": 1.78944,
+      "grad_norm": 0.6504471898078918,
+      "learning_rate": 4.722288915566227e-05,
+      "loss": 0.7329,
+      "step": 1398
+    },
+    {
+      "epoch": 1.7907199999999999,
+      "grad_norm": 0.6064606308937073,
+      "learning_rate": 4.722088835534214e-05,
+      "loss": 0.7472,
+      "step": 1399
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.6191816926002502,
+      "learning_rate": 4.721888755502201e-05,
+      "loss": 0.7031,
+      "step": 1400
+    },
+    {
+      "epoch": 1.79328,
+      "grad_norm": 0.6464985013008118,
+      "learning_rate": 4.7216886754701884e-05,
+      "loss": 0.7053,
+      "step": 1401
+    },
+    {
+      "epoch": 1.7945600000000002,
+      "grad_norm": 0.6391028165817261,
+      "learning_rate": 4.7214885954381756e-05,
+      "loss": 0.6984,
+      "step": 1402
+    },
+    {
+      "epoch": 1.79584,
+      "grad_norm": 0.6238522529602051,
+      "learning_rate": 4.721288515406163e-05,
+      "loss": 0.7134,
+      "step": 1403
+    },
+    {
+      "epoch": 1.79712,
+      "grad_norm": 0.6372814774513245,
+      "learning_rate": 4.72108843537415e-05,
+      "loss": 0.741,
+      "step": 1404
+    },
+    {
+      "epoch": 1.7984,
+      "grad_norm": 0.6125516295433044,
+      "learning_rate": 4.720888355342137e-05,
+      "loss": 0.7364,
+      "step": 1405
+    },
+    {
+      "epoch": 1.79968,
+      "grad_norm": 0.6260775923728943,
+      "learning_rate": 4.720688275310124e-05,
+      "loss": 0.7158,
+      "step": 1406
+    },
+    {
+      "epoch": 1.80096,
+      "grad_norm": 0.6567710041999817,
+      "learning_rate": 4.7204881952781115e-05,
+      "loss": 0.7745,
+      "step": 1407
+    },
+    {
+      "epoch": 1.8022399999999998,
+      "grad_norm": 0.6455616354942322,
+      "learning_rate": 4.720288115246099e-05,
+      "loss": 0.7038,
+      "step": 1408
+    },
+    {
+      "epoch": 1.80352,
+      "grad_norm": 0.6271963715553284,
+      "learning_rate": 4.720088035214086e-05,
+      "loss": 0.7339,
+      "step": 1409
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.6165766716003418,
+      "learning_rate": 4.719887955182073e-05,
+      "loss": 0.8227,
+      "step": 1410
+    },
+    {
+      "epoch": 1.8060800000000001,
+      "grad_norm": 0.6498475074768066,
+      "learning_rate": 4.71968787515006e-05,
+      "loss": 0.7552,
+      "step": 1411
+    },
+    {
+      "epoch": 1.80736,
+      "grad_norm": 0.6534706950187683,
+      "learning_rate": 4.7194877951180474e-05,
+      "loss": 0.7461,
+      "step": 1412
+    },
+    {
+      "epoch": 1.80864,
+      "grad_norm": 0.6396170854568481,
+      "learning_rate": 4.7192877150860346e-05,
+      "loss": 0.7051,
+      "step": 1413
+    },
+    {
+      "epoch": 1.80992,
+      "grad_norm": 0.6736233234405518,
+      "learning_rate": 4.719087635054022e-05,
+      "loss": 0.7181,
+      "step": 1414
+    },
+    {
+      "epoch": 1.8112,
+      "grad_norm": 0.7057044506072998,
+      "learning_rate": 4.718887555022009e-05,
+      "loss": 0.7447,
+      "step": 1415
+    },
+    {
+      "epoch": 1.8124799999999999,
+      "grad_norm": 0.639115571975708,
+      "learning_rate": 4.718687474989996e-05,
+      "loss": 0.702,
+      "step": 1416
+    },
+    {
+      "epoch": 1.81376,
+      "grad_norm": 0.6411137580871582,
+      "learning_rate": 4.7184873949579834e-05,
+      "loss": 0.7378,
+      "step": 1417
+    },
+    {
+      "epoch": 1.81504,
+      "grad_norm": 0.618817925453186,
+      "learning_rate": 4.7182873149259705e-05,
+      "loss": 0.6965,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8163200000000002,
+      "grad_norm": 0.6112127304077148,
+      "learning_rate": 4.718087234893958e-05,
+      "loss": 0.7387,
+      "step": 1419
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.6073436737060547,
+      "learning_rate": 4.717887154861945e-05,
+      "loss": 0.6605,
+      "step": 1420
+    },
+    {
+      "epoch": 1.81888,
+      "grad_norm": 0.5831905007362366,
+      "learning_rate": 4.717687074829932e-05,
+      "loss": 0.6903,
+      "step": 1421
+    },
+    {
+      "epoch": 1.82016,
+      "grad_norm": 0.6208034753799438,
+      "learning_rate": 4.717486994797919e-05,
+      "loss": 0.7061,
+      "step": 1422
+    },
+    {
+      "epoch": 1.82144,
+      "grad_norm": 0.6343234181404114,
+      "learning_rate": 4.7172869147659065e-05,
+      "loss": 0.7128,
+      "step": 1423
+    },
+    {
+      "epoch": 1.82272,
+      "grad_norm": 0.6611493229866028,
+      "learning_rate": 4.7170868347338937e-05,
+      "loss": 0.7567,
+      "step": 1424
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 0.6101629734039307,
+      "learning_rate": 4.716886754701881e-05,
+      "loss": 0.7273,
+      "step": 1425
+    },
+    {
+      "epoch": 1.82528,
+      "grad_norm": 0.5913015604019165,
+      "learning_rate": 4.716686674669868e-05,
+      "loss": 0.6477,
+      "step": 1426
+    },
+    {
+      "epoch": 1.82656,
+      "grad_norm": 0.5725111365318298,
+      "learning_rate": 4.716486594637856e-05,
+      "loss": 0.7042,
+      "step": 1427
+    },
+    {
+      "epoch": 1.8278400000000001,
+      "grad_norm": 0.5874539017677307,
+      "learning_rate": 4.7162865146058424e-05,
+      "loss": 0.6511,
+      "step": 1428
+    },
+    {
+      "epoch": 1.82912,
+      "grad_norm": 0.6199377179145813,
+      "learning_rate": 4.7160864345738296e-05,
+      "loss": 0.6984,
+      "step": 1429
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.6231164932250977,
+      "learning_rate": 4.715886354541817e-05,
+      "loss": 0.6653,
+      "step": 1430
+    },
+    {
+      "epoch": 1.83168,
+      "grad_norm": 0.6409709453582764,
+      "learning_rate": 4.715686274509804e-05,
+      "loss": 0.7573,
+      "step": 1431
+    },
+    {
+      "epoch": 1.83296,
+      "grad_norm": 0.5936411023139954,
+      "learning_rate": 4.715486194477791e-05,
+      "loss": 0.7043,
+      "step": 1432
+    },
+    {
+      "epoch": 1.8342399999999999,
+      "grad_norm": 0.6537723541259766,
+      "learning_rate": 4.715286114445778e-05,
+      "loss": 0.7084,
+      "step": 1433
+    },
+    {
+      "epoch": 1.83552,
+      "grad_norm": 0.5906792283058167,
+      "learning_rate": 4.715086034413766e-05,
+      "loss": 0.6799,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8368,
+      "grad_norm": 0.5938474535942078,
+      "learning_rate": 4.7148859543817534e-05,
+      "loss": 0.6764,
+      "step": 1435
+    },
+    {
+      "epoch": 1.8380800000000002,
+      "grad_norm": 0.6261917948722839,
+      "learning_rate": 4.71468587434974e-05,
+      "loss": 0.7256,
+      "step": 1436
+    },
+    {
+      "epoch": 1.83936,
+      "grad_norm": 0.6289463043212891,
+      "learning_rate": 4.714485794317727e-05,
+      "loss": 0.7277,
+      "step": 1437
+    },
+    {
+      "epoch": 1.84064,
+      "grad_norm": 0.606020987033844,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 0.6845,
+      "step": 1438
+    },
+    {
+      "epoch": 1.84192,
+      "grad_norm": 0.6375628709793091,
+      "learning_rate": 4.7140856342537014e-05,
+      "loss": 0.7181,
+      "step": 1439
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.6486073136329651,
+      "learning_rate": 4.7138855542216886e-05,
+      "loss": 0.7067,
+      "step": 1440
+    },
+    {
+      "epoch": 1.84448,
+      "grad_norm": 0.6745935082435608,
+      "learning_rate": 4.7136854741896765e-05,
+      "loss": 0.7309,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8457599999999998,
+      "grad_norm": 0.604806125164032,
+      "learning_rate": 4.713485394157664e-05,
+      "loss": 0.7268,
+      "step": 1442
+    },
+    {
+      "epoch": 1.84704,
+      "grad_norm": 0.6913792490959167,
+      "learning_rate": 4.713285314125651e-05,
+      "loss": 0.7569,
+      "step": 1443
+    },
+    {
+      "epoch": 1.84832,
+      "grad_norm": 0.5856841802597046,
+      "learning_rate": 4.7130852340936374e-05,
+      "loss": 0.6421,
+      "step": 1444
+    },
+    {
+      "epoch": 1.8496000000000001,
+      "grad_norm": 0.6320177316665649,
+      "learning_rate": 4.7128851540616246e-05,
+      "loss": 0.7385,
+      "step": 1445
+    },
+    {
+      "epoch": 1.85088,
+      "grad_norm": 0.6420151591300964,
+      "learning_rate": 4.712685074029612e-05,
+      "loss": 0.6881,
+      "step": 1446
+    },
+    {
+      "epoch": 1.85216,
+      "grad_norm": 0.6870813965797424,
+      "learning_rate": 4.712484993997599e-05,
+      "loss": 0.7299,
+      "step": 1447
+    },
+    {
+      "epoch": 1.85344,
+      "grad_norm": 0.6391651034355164,
+      "learning_rate": 4.712284913965587e-05,
+      "loss": 0.68,
+      "step": 1448
+    },
+    {
+      "epoch": 1.85472,
+      "grad_norm": 0.681445300579071,
+      "learning_rate": 4.712084833933574e-05,
+      "loss": 0.7795,
+      "step": 1449
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.6424078941345215,
+      "learning_rate": 4.711884753901561e-05,
+      "loss": 0.7234,
+      "step": 1450
+    },
+    {
+      "epoch": 1.85728,
+      "grad_norm": 0.625423014163971,
+      "learning_rate": 4.7116846738695484e-05,
+      "loss": 0.6631,
+      "step": 1451
+    },
+    {
+      "epoch": 1.85856,
+      "grad_norm": 0.611112117767334,
+      "learning_rate": 4.711484593837535e-05,
+      "loss": 0.7227,
+      "step": 1452
+    },
+    {
+      "epoch": 1.8598400000000002,
+      "grad_norm": 0.6501821279525757,
+      "learning_rate": 4.711284513805522e-05,
+      "loss": 0.7494,
+      "step": 1453
+    },
+    {
+      "epoch": 1.86112,
+      "grad_norm": 0.6695547699928284,
+      "learning_rate": 4.711084433773509e-05,
+      "loss": 0.7335,
+      "step": 1454
+    },
+    {
+      "epoch": 1.8624,
+      "grad_norm": 0.6903998851776123,
+      "learning_rate": 4.710884353741497e-05,
+      "loss": 0.7594,
+      "step": 1455
+    },
+    {
+      "epoch": 1.86368,
+      "grad_norm": 0.6672253608703613,
+      "learning_rate": 4.710684273709484e-05,
+      "loss": 0.7182,
+      "step": 1456
+    },
+    {
+      "epoch": 1.86496,
+      "grad_norm": 0.6630045175552368,
+      "learning_rate": 4.7104841936774715e-05,
+      "loss": 0.7251,
+      "step": 1457
+    },
+    {
+      "epoch": 1.86624,
+      "grad_norm": 0.6247894167900085,
+      "learning_rate": 4.7102841136454587e-05,
+      "loss": 0.7358,
+      "step": 1458
+    },
+    {
+      "epoch": 1.8675199999999998,
+      "grad_norm": 0.595182478427887,
+      "learning_rate": 4.710084033613446e-05,
+      "loss": 0.6497,
+      "step": 1459
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.6355342864990234,
+      "learning_rate": 4.7098839535814323e-05,
+      "loss": 0.6925,
+      "step": 1460
+    },
+    {
+      "epoch": 1.87008,
+      "grad_norm": 0.6582742929458618,
+      "learning_rate": 4.7096838735494195e-05,
+      "loss": 0.7062,
+      "step": 1461
+    },
+    {
+      "epoch": 1.8713600000000001,
+      "grad_norm": 0.5886150598526001,
+      "learning_rate": 4.7094837935174074e-05,
+      "loss": 0.6701,
+      "step": 1462
+    },
+    {
+      "epoch": 1.87264,
+      "grad_norm": 0.6048754453659058,
+      "learning_rate": 4.7092837134853946e-05,
+      "loss": 0.6841,
+      "step": 1463
+    },
+    {
+      "epoch": 1.87392,
+      "grad_norm": 0.6366539597511292,
+      "learning_rate": 4.709083633453382e-05,
+      "loss": 0.6991,
+      "step": 1464
+    },
+    {
+      "epoch": 1.8752,
+      "grad_norm": 0.5844667553901672,
+      "learning_rate": 4.708883553421369e-05,
+      "loss": 0.6269,
+      "step": 1465
+    },
+    {
+      "epoch": 1.87648,
+      "grad_norm": 0.5977579951286316,
+      "learning_rate": 4.708683473389356e-05,
+      "loss": 0.6561,
+      "step": 1466
+    },
+    {
+      "epoch": 1.8777599999999999,
+      "grad_norm": 0.6138444542884827,
+      "learning_rate": 4.708483393357343e-05,
+      "loss": 0.6814,
+      "step": 1467
+    },
+    {
+      "epoch": 1.87904,
+      "grad_norm": 0.6392626762390137,
+      "learning_rate": 4.70828331332533e-05,
+      "loss": 0.6931,
+      "step": 1468
+    },
+    {
+      "epoch": 1.88032,
+      "grad_norm": 0.666708767414093,
+      "learning_rate": 4.708083233293318e-05,
+      "loss": 0.7474,
+      "step": 1469
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.6352314949035645,
+      "learning_rate": 4.707883153261305e-05,
+      "loss": 0.7012,
+      "step": 1470
+    },
+    {
+      "epoch": 1.88288,
+      "grad_norm": 0.6222551465034485,
+      "learning_rate": 4.707683073229292e-05,
+      "loss": 0.7133,
+      "step": 1471
+    },
+    {
+      "epoch": 1.88416,
+      "grad_norm": 0.643947958946228,
+      "learning_rate": 4.707482993197279e-05,
+      "loss": 0.7556,
+      "step": 1472
+    },
+    {
+      "epoch": 1.88544,
+      "grad_norm": 0.5932632684707642,
+      "learning_rate": 4.7072829131652664e-05,
+      "loss": 0.6847,
+      "step": 1473
+    },
+    {
+      "epoch": 1.88672,
+      "grad_norm": 0.6231780648231506,
+      "learning_rate": 4.7070828331332536e-05,
+      "loss": 0.6971,
+      "step": 1474
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 0.6272242665290833,
+      "learning_rate": 4.706882753101241e-05,
+      "loss": 0.7149,
+      "step": 1475
+    },
+    {
+      "epoch": 1.8892799999999998,
+      "grad_norm": 0.6251195073127747,
+      "learning_rate": 4.706682673069228e-05,
+      "loss": 0.685,
+      "step": 1476
+    },
+    {
+      "epoch": 1.89056,
+      "grad_norm": 0.6322864890098572,
+      "learning_rate": 4.706482593037215e-05,
+      "loss": 0.7115,
+      "step": 1477
+    },
+    {
+      "epoch": 1.89184,
+      "grad_norm": 0.6262922286987305,
+      "learning_rate": 4.7062825130052024e-05,
+      "loss": 0.7272,
+      "step": 1478
+    },
+    {
+      "epoch": 1.8931200000000001,
+      "grad_norm": 0.6692927479743958,
+      "learning_rate": 4.7060824329731896e-05,
+      "loss": 0.7201,
+      "step": 1479
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.58826744556427,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.6663,
+      "step": 1480
+    },
+    {
+      "epoch": 1.89568,
+      "grad_norm": 0.6134018301963806,
+      "learning_rate": 4.705682272909164e-05,
+      "loss": 0.7319,
+      "step": 1481
+    },
+    {
+      "epoch": 1.89696,
+      "grad_norm": 0.6586698889732361,
+      "learning_rate": 4.705482192877151e-05,
+      "loss": 0.7324,
+      "step": 1482
+    },
+    {
+      "epoch": 1.89824,
+      "grad_norm": 0.6096124649047852,
+      "learning_rate": 4.705282112845138e-05,
+      "loss": 0.7266,
+      "step": 1483
+    },
+    {
+      "epoch": 1.8995199999999999,
+      "grad_norm": 0.6256430149078369,
+      "learning_rate": 4.7050820328131255e-05,
+      "loss": 0.6926,
+      "step": 1484
+    },
+    {
+      "epoch": 1.9008,
+      "grad_norm": 0.5955308675765991,
+      "learning_rate": 4.704881952781113e-05,
+      "loss": 0.7232,
+      "step": 1485
+    },
+    {
+      "epoch": 1.90208,
+      "grad_norm": 0.6191076040267944,
+      "learning_rate": 4.7046818727491e-05,
+      "loss": 0.6879,
+      "step": 1486
+    },
+    {
+      "epoch": 1.9033600000000002,
+      "grad_norm": 0.6182354688644409,
+      "learning_rate": 4.704481792717087e-05,
+      "loss": 0.7449,
+      "step": 1487
+    },
+    {
+      "epoch": 1.90464,
+      "grad_norm": 0.6782128214836121,
+      "learning_rate": 4.704281712685074e-05,
+      "loss": 0.7688,
+      "step": 1488
+    },
+    {
+      "epoch": 1.90592,
+      "grad_norm": 0.6402320265769958,
+      "learning_rate": 4.7040816326530614e-05,
+      "loss": 0.7376,
+      "step": 1489
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.6744869351387024,
+      "learning_rate": 4.7038815526210486e-05,
+      "loss": 0.8047,
+      "step": 1490
+    },
+    {
+      "epoch": 1.90848,
+      "grad_norm": 0.6262235045433044,
+      "learning_rate": 4.703681472589036e-05,
+      "loss": 0.6589,
+      "step": 1491
+    },
+    {
+      "epoch": 1.90976,
+      "grad_norm": 0.6338122487068176,
+      "learning_rate": 4.703481392557023e-05,
+      "loss": 0.7067,
+      "step": 1492
+    },
+    {
+      "epoch": 1.9110399999999998,
+      "grad_norm": 0.6535120010375977,
+      "learning_rate": 4.70328131252501e-05,
+      "loss": 0.6962,
+      "step": 1493
+    },
+    {
+      "epoch": 1.91232,
+      "grad_norm": 0.6537796854972839,
+      "learning_rate": 4.703081232492997e-05,
+      "loss": 0.7182,
+      "step": 1494
+    },
+    {
+      "epoch": 1.9136,
+      "grad_norm": 0.6153785586357117,
+      "learning_rate": 4.7028811524609845e-05,
+      "loss": 0.6872,
+      "step": 1495
+    },
+    {
+      "epoch": 1.9148800000000001,
+      "grad_norm": 0.6132374405860901,
+      "learning_rate": 4.702681072428972e-05,
+      "loss": 0.6838,
+      "step": 1496
+    },
+    {
+      "epoch": 1.91616,
+      "grad_norm": 0.6539681553840637,
+      "learning_rate": 4.7024809923969596e-05,
+      "loss": 0.7467,
+      "step": 1497
+    },
+    {
+      "epoch": 1.91744,
+      "grad_norm": 0.6296167373657227,
+      "learning_rate": 4.702280912364946e-05,
+      "loss": 0.6825,
+      "step": 1498
+    },
+    {
+      "epoch": 1.91872,
+      "grad_norm": 0.6461489796638489,
+      "learning_rate": 4.702080832332933e-05,
+      "loss": 0.7336,
+      "step": 1499
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6219279170036316,
+      "learning_rate": 4.7018807523009204e-05,
+      "loss": 0.7172,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9212799999999999,
+      "grad_norm": 0.6014849543571472,
+      "learning_rate": 4.7016806722689076e-05,
+      "loss": 0.7528,
+      "step": 1501
+    },
+    {
+      "epoch": 1.92256,
+      "grad_norm": 0.6287034749984741,
+      "learning_rate": 4.701480592236895e-05,
+      "loss": 0.7724,
+      "step": 1502
+    },
+    {
+      "epoch": 1.92384,
+      "grad_norm": 0.6444774270057678,
+      "learning_rate": 4.701280512204882e-05,
+      "loss": 0.7873,
+      "step": 1503
+    },
+    {
+      "epoch": 1.9251200000000002,
+      "grad_norm": 0.631040096282959,
+      "learning_rate": 4.70108043217287e-05,
+      "loss": 0.7204,
+      "step": 1504
+    },
+    {
+      "epoch": 1.9264000000000001,
+      "grad_norm": 0.6389703750610352,
+      "learning_rate": 4.700880352140857e-05,
+      "loss": 0.7623,
+      "step": 1505
+    },
+    {
+      "epoch": 1.92768,
+      "grad_norm": 0.6128464341163635,
+      "learning_rate": 4.7006802721088436e-05,
+      "loss": 0.7311,
+      "step": 1506
+    },
+    {
+      "epoch": 1.92896,
+      "grad_norm": 0.6340726613998413,
+      "learning_rate": 4.700480192076831e-05,
+      "loss": 0.7277,
+      "step": 1507
+    },
+    {
+      "epoch": 1.93024,
+      "grad_norm": 0.6153057813644409,
+      "learning_rate": 4.700280112044818e-05,
+      "loss": 0.7087,
+      "step": 1508
+    },
+    {
+      "epoch": 1.93152,
+      "grad_norm": 0.6262139081954956,
+      "learning_rate": 4.700080032012805e-05,
+      "loss": 0.6813,
+      "step": 1509
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.6372281908988953,
+      "learning_rate": 4.699879951980792e-05,
+      "loss": 0.6867,
+      "step": 1510
+    },
+    {
+      "epoch": 1.93408,
+      "grad_norm": 0.6328516602516174,
+      "learning_rate": 4.69967987194878e-05,
+      "loss": 0.7376,
+      "step": 1511
+    },
+    {
+      "epoch": 1.93536,
+      "grad_norm": 0.6271541118621826,
+      "learning_rate": 4.6994797919167674e-05,
+      "loss": 0.7694,
+      "step": 1512
+    },
+    {
+      "epoch": 1.9366400000000001,
+      "grad_norm": 0.6466159224510193,
+      "learning_rate": 4.6992797118847545e-05,
+      "loss": 0.7072,
+      "step": 1513
+    },
+    {
+      "epoch": 1.93792,
+      "grad_norm": 0.6909953355789185,
+      "learning_rate": 4.699079631852741e-05,
+      "loss": 0.7634,
+      "step": 1514
+    },
+    {
+      "epoch": 1.9392,
+      "grad_norm": 0.6303381323814392,
+      "learning_rate": 4.698879551820728e-05,
+      "loss": 0.802,
+      "step": 1515
+    },
+    {
+      "epoch": 1.94048,
+      "grad_norm": 0.6296682953834534,
+      "learning_rate": 4.6986794717887154e-05,
+      "loss": 0.7861,
+      "step": 1516
+    },
+    {
+      "epoch": 1.94176,
+      "grad_norm": 0.6130115985870361,
+      "learning_rate": 4.6984793917567026e-05,
+      "loss": 0.697,
+      "step": 1517
+    },
+    {
+      "epoch": 1.9430399999999999,
+      "grad_norm": 0.6313830614089966,
+      "learning_rate": 4.69827931172469e-05,
+      "loss": 0.7331,
+      "step": 1518
+    },
+    {
+      "epoch": 1.94432,
+      "grad_norm": 0.5878955125808716,
+      "learning_rate": 4.6980792316926777e-05,
+      "loss": 0.6528,
+      "step": 1519
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.6016858220100403,
+      "learning_rate": 4.697879151660665e-05,
+      "loss": 0.687,
+      "step": 1520
+    },
+    {
+      "epoch": 1.9468800000000002,
+      "grad_norm": 0.6477680802345276,
+      "learning_rate": 4.697679071628652e-05,
+      "loss": 0.7101,
+      "step": 1521
+    },
+    {
+      "epoch": 1.9481600000000001,
+      "grad_norm": 0.6151396632194519,
+      "learning_rate": 4.6974789915966385e-05,
+      "loss": 0.7067,
+      "step": 1522
+    },
+    {
+      "epoch": 1.94944,
+      "grad_norm": 0.6041195392608643,
+      "learning_rate": 4.697278911564626e-05,
+      "loss": 0.7499,
+      "step": 1523
+    },
+    {
+      "epoch": 1.95072,
+      "grad_norm": 0.6384760737419128,
+      "learning_rate": 4.697078831532613e-05,
+      "loss": 0.7452,
+      "step": 1524
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 0.6926515698432922,
+      "learning_rate": 4.6968787515006e-05,
+      "loss": 0.7656,
+      "step": 1525
+    },
+    {
+      "epoch": 1.95328,
+      "grad_norm": 0.652237594127655,
+      "learning_rate": 4.696678671468588e-05,
+      "loss": 0.7503,
+      "step": 1526
+    },
+    {
+      "epoch": 1.9545599999999999,
+      "grad_norm": 0.6000808477401733,
+      "learning_rate": 4.696478591436575e-05,
+      "loss": 0.6663,
+      "step": 1527
+    },
+    {
+      "epoch": 1.95584,
+      "grad_norm": 0.6068828105926514,
+      "learning_rate": 4.696278511404562e-05,
+      "loss": 0.8046,
+      "step": 1528
+    },
+    {
+      "epoch": 1.95712,
+      "grad_norm": 0.6311643123626709,
+      "learning_rate": 4.6960784313725495e-05,
+      "loss": 0.7473,
+      "step": 1529
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.6560335755348206,
+      "learning_rate": 4.695878351340536e-05,
+      "loss": 0.7188,
+      "step": 1530
+    },
+    {
+      "epoch": 1.95968,
+      "grad_norm": 0.6460652351379395,
+      "learning_rate": 4.695678271308523e-05,
+      "loss": 0.6784,
+      "step": 1531
+    },
+    {
+      "epoch": 1.96096,
+      "grad_norm": 0.6522865295410156,
+      "learning_rate": 4.6954781912765104e-05,
+      "loss": 0.7078,
+      "step": 1532
+    },
+    {
+      "epoch": 1.96224,
+      "grad_norm": 0.6220058798789978,
+      "learning_rate": 4.695278111244498e-05,
+      "loss": 0.7121,
+      "step": 1533
+    },
+    {
+      "epoch": 1.96352,
+      "grad_norm": 0.6365997791290283,
+      "learning_rate": 4.6950780312124854e-05,
+      "loss": 0.6876,
+      "step": 1534
+    },
+    {
+      "epoch": 1.9647999999999999,
+      "grad_norm": 0.6620326042175293,
+      "learning_rate": 4.6948779511804726e-05,
+      "loss": 0.7311,
+      "step": 1535
+    },
+    {
+      "epoch": 1.96608,
+      "grad_norm": 0.6124232411384583,
+      "learning_rate": 4.69467787114846e-05,
+      "loss": 0.7134,
+      "step": 1536
+    },
+    {
+      "epoch": 1.96736,
+      "grad_norm": 0.6828178763389587,
+      "learning_rate": 4.694477791116447e-05,
+      "loss": 0.7439,
+      "step": 1537
+    },
+    {
+      "epoch": 1.96864,
+      "grad_norm": 0.6395293474197388,
+      "learning_rate": 4.6942777110844335e-05,
+      "loss": 0.7836,
+      "step": 1538
+    },
+    {
+      "epoch": 1.9699200000000001,
+      "grad_norm": 0.6207759976387024,
+      "learning_rate": 4.694077631052421e-05,
+      "loss": 0.6739,
+      "step": 1539
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.6403487324714661,
+      "learning_rate": 4.6938775510204086e-05,
+      "loss": 0.7665,
+      "step": 1540
+    },
+    {
+      "epoch": 1.97248,
+      "grad_norm": 0.6810711026191711,
+      "learning_rate": 4.693677470988396e-05,
+      "loss": 0.7825,
+      "step": 1541
+    },
+    {
+      "epoch": 1.97376,
+      "grad_norm": 0.6638922095298767,
+      "learning_rate": 4.693477390956383e-05,
+      "loss": 0.7517,
+      "step": 1542
+    },
+    {
+      "epoch": 1.97504,
+      "grad_norm": 0.5913483500480652,
+      "learning_rate": 4.69327731092437e-05,
+      "loss": 0.6869,
+      "step": 1543
+    },
+    {
+      "epoch": 1.9763199999999999,
+      "grad_norm": 0.6187199354171753,
+      "learning_rate": 4.693077230892357e-05,
+      "loss": 0.7321,
+      "step": 1544
+    },
+    {
+      "epoch": 1.9776,
+      "grad_norm": 0.6598728895187378,
+      "learning_rate": 4.6928771508603445e-05,
+      "loss": 0.7358,
+      "step": 1545
+    },
+    {
+      "epoch": 1.97888,
+      "grad_norm": 0.6453779935836792,
+      "learning_rate": 4.692677070828331e-05,
+      "loss": 0.7885,
+      "step": 1546
+    },
+    {
+      "epoch": 1.9801600000000001,
+      "grad_norm": 0.62159663438797,
+      "learning_rate": 4.692476990796319e-05,
+      "loss": 0.6755,
+      "step": 1547
+    },
+    {
+      "epoch": 1.98144,
+      "grad_norm": 0.5968536734580994,
+      "learning_rate": 4.692276910764306e-05,
+      "loss": 0.6611,
+      "step": 1548
+    },
+    {
+      "epoch": 1.98272,
+      "grad_norm": 0.6375271081924438,
+      "learning_rate": 4.692076830732293e-05,
+      "loss": 0.6993,
+      "step": 1549
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.6619069576263428,
+      "learning_rate": 4.6918767507002804e-05,
+      "loss": 0.6928,
+      "step": 1550
+    },
+    {
+      "epoch": 1.98528,
+      "grad_norm": 0.6253162026405334,
+      "learning_rate": 4.6916766706682676e-05,
+      "loss": 0.66,
+      "step": 1551
+    },
+    {
+      "epoch": 1.9865599999999999,
+      "grad_norm": 0.6479495167732239,
+      "learning_rate": 4.691476590636255e-05,
+      "loss": 0.6365,
+      "step": 1552
+    },
+    {
+      "epoch": 1.98784,
+      "grad_norm": 0.6379725337028503,
+      "learning_rate": 4.691276510604242e-05,
+      "loss": 0.7077,
+      "step": 1553
+    },
+    {
+      "epoch": 1.98912,
+      "grad_norm": 0.6354682445526123,
+      "learning_rate": 4.691076430572229e-05,
+      "loss": 0.7052,
+      "step": 1554
+    },
+    {
+      "epoch": 1.9904,
+      "grad_norm": 0.6414213180541992,
+      "learning_rate": 4.6908763505402163e-05,
+      "loss": 0.6586,
+      "step": 1555
+    },
+    {
+      "epoch": 1.9916800000000001,
+      "grad_norm": 0.6093000173568726,
+      "learning_rate": 4.6906762705082035e-05,
+      "loss": 0.7352,
+      "step": 1556
+    },
+    {
+      "epoch": 1.99296,
+      "grad_norm": 0.6520658135414124,
+      "learning_rate": 4.690476190476191e-05,
+      "loss": 0.7425,
+      "step": 1557
+    },
+    {
+      "epoch": 1.99424,
+      "grad_norm": 0.635176420211792,
+      "learning_rate": 4.690276110444178e-05,
+      "loss": 0.7179,
+      "step": 1558
+    },
+    {
+      "epoch": 1.99552,
+      "grad_norm": 0.6376418471336365,
+      "learning_rate": 4.690076030412165e-05,
+      "loss": 0.7567,
+      "step": 1559
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.6329599022865295,
+      "learning_rate": 4.689875950380152e-05,
+      "loss": 0.7107,
+      "step": 1560
+    },
+    {
+      "epoch": 1.9980799999999999,
+      "grad_norm": 0.6018700003623962,
+      "learning_rate": 4.6896758703481395e-05,
+      "loss": 0.6841,
+      "step": 1561
+    },
+    {
+      "epoch": 1.99936,
+      "grad_norm": 0.6212313771247864,
+      "learning_rate": 4.6894757903161266e-05,
+      "loss": 0.7403,
+      "step": 1562
+    },
+    {
+      "epoch": 2.00064,
+      "grad_norm": 1.2921080589294434,
+      "learning_rate": 4.689275710284114e-05,
+      "loss": 1.147,
+      "step": 1563
+    },
+    {
+      "epoch": 2.00192,
+      "grad_norm": 0.6272991299629211,
+      "learning_rate": 4.689075630252101e-05,
+      "loss": 0.6922,
+      "step": 1564
+    },
+    {
+      "epoch": 2.0032,
+      "grad_norm": 0.6323875188827515,
+      "learning_rate": 4.688875550220088e-05,
+      "loss": 0.721,
+      "step": 1565
+    },
+    {
+      "epoch": 2.00448,
+      "grad_norm": 0.6615299582481384,
+      "learning_rate": 4.6886754701880754e-05,
+      "loss": 0.7782,
+      "step": 1566
+    },
+    {
+      "epoch": 2.00576,
+      "grad_norm": 0.6855239868164062,
+      "learning_rate": 4.6884753901560626e-05,
+      "loss": 0.7006,
+      "step": 1567
+    },
+    {
+      "epoch": 2.00704,
+      "grad_norm": 0.5969696640968323,
+      "learning_rate": 4.68827531012405e-05,
+      "loss": 0.6765,
+      "step": 1568
+    },
+    {
+      "epoch": 2.00832,
+      "grad_norm": 0.6360341310501099,
+      "learning_rate": 4.688075230092037e-05,
+      "loss": 0.7551,
+      "step": 1569
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 0.6344780325889587,
+      "learning_rate": 4.687875150060024e-05,
+      "loss": 0.7285,
+      "step": 1570
+    },
+    {
+      "epoch": 2.0108800000000002,
+      "grad_norm": 0.6265085935592651,
+      "learning_rate": 4.687675070028011e-05,
+      "loss": 0.693,
+      "step": 1571
+    },
+    {
+      "epoch": 2.01216,
+      "grad_norm": 0.6241620182991028,
+      "learning_rate": 4.6874749899959985e-05,
+      "loss": 0.6636,
+      "step": 1572
+    },
+    {
+      "epoch": 2.01344,
+      "grad_norm": 0.6464859843254089,
+      "learning_rate": 4.687274909963986e-05,
+      "loss": 0.7299,
+      "step": 1573
+    },
+    {
+      "epoch": 2.01472,
+      "grad_norm": 0.6327393651008606,
+      "learning_rate": 4.687074829931973e-05,
+      "loss": 0.6387,
+      "step": 1574
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.6345120072364807,
+      "learning_rate": 4.686874749899961e-05,
+      "loss": 0.6943,
+      "step": 1575
+    },
+    {
+      "epoch": 2.01728,
+      "grad_norm": 0.6209046244621277,
+      "learning_rate": 4.686674669867947e-05,
+      "loss": 0.6712,
+      "step": 1576
+    },
+    {
+      "epoch": 2.01856,
+      "grad_norm": 0.6531869173049927,
+      "learning_rate": 4.6864745898359344e-05,
+      "loss": 0.7195,
+      "step": 1577
+    },
+    {
+      "epoch": 2.01984,
+      "grad_norm": 0.6831356287002563,
+      "learning_rate": 4.6862745098039216e-05,
+      "loss": 0.7457,
+      "step": 1578
+    },
+    {
+      "epoch": 2.02112,
+      "grad_norm": 0.6502741575241089,
+      "learning_rate": 4.686074429771909e-05,
+      "loss": 0.7308,
+      "step": 1579
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.661711573600769,
+      "learning_rate": 4.685874349739896e-05,
+      "loss": 0.7113,
+      "step": 1580
+    },
+    {
+      "epoch": 2.02368,
+      "grad_norm": 0.6132927536964417,
+      "learning_rate": 4.685674269707883e-05,
+      "loss": 0.6587,
+      "step": 1581
+    },
+    {
+      "epoch": 2.02496,
+      "grad_norm": 0.6607900261878967,
+      "learning_rate": 4.685474189675871e-05,
+      "loss": 0.6695,
+      "step": 1582
+    },
+    {
+      "epoch": 2.02624,
+      "grad_norm": 0.6305644512176514,
+      "learning_rate": 4.685274109643858e-05,
+      "loss": 0.6834,
+      "step": 1583
+    },
+    {
+      "epoch": 2.02752,
+      "grad_norm": 0.6461015939712524,
+      "learning_rate": 4.685074029611845e-05,
+      "loss": 0.7033,
+      "step": 1584
+    },
+    {
+      "epoch": 2.0288,
+      "grad_norm": 0.6151067614555359,
+      "learning_rate": 4.684873949579832e-05,
+      "loss": 0.7138,
+      "step": 1585
+    },
+    {
+      "epoch": 2.03008,
+      "grad_norm": 0.6099941730499268,
+      "learning_rate": 4.684673869547819e-05,
+      "loss": 0.6364,
+      "step": 1586
+    },
+    {
+      "epoch": 2.03136,
+      "grad_norm": 0.6969642639160156,
+      "learning_rate": 4.684473789515806e-05,
+      "loss": 0.755,
+      "step": 1587
+    },
+    {
+      "epoch": 2.03264,
+      "grad_norm": 0.6614975929260254,
+      "learning_rate": 4.6842737094837935e-05,
+      "loss": 0.7028,
+      "step": 1588
+    },
+    {
+      "epoch": 2.03392,
+      "grad_norm": 0.6838937401771545,
+      "learning_rate": 4.684073629451781e-05,
+      "loss": 0.8004,
+      "step": 1589
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.6104751229286194,
+      "learning_rate": 4.6838735494197685e-05,
+      "loss": 0.7308,
+      "step": 1590
+    },
+    {
+      "epoch": 2.03648,
+      "grad_norm": 0.6918980479240417,
+      "learning_rate": 4.683673469387756e-05,
+      "loss": 0.7753,
+      "step": 1591
+    },
+    {
+      "epoch": 2.03776,
+      "grad_norm": 0.6084235906600952,
+      "learning_rate": 4.683473389355742e-05,
+      "loss": 0.6957,
+      "step": 1592
+    },
+    {
+      "epoch": 2.03904,
+      "grad_norm": 0.5943331718444824,
+      "learning_rate": 4.6832733093237294e-05,
+      "loss": 0.6609,
+      "step": 1593
+    },
+    {
+      "epoch": 2.04032,
+      "grad_norm": 0.6553683876991272,
+      "learning_rate": 4.6830732292917166e-05,
+      "loss": 0.7259,
+      "step": 1594
+    },
+    {
+      "epoch": 2.0416,
+      "grad_norm": 0.6380487680435181,
+      "learning_rate": 4.682873149259704e-05,
+      "loss": 0.7648,
+      "step": 1595
+    },
+    {
+      "epoch": 2.04288,
+      "grad_norm": 0.6057306528091431,
+      "learning_rate": 4.6826730692276916e-05,
+      "loss": 0.6415,
+      "step": 1596
+    },
+    {
+      "epoch": 2.04416,
+      "grad_norm": 0.6594085693359375,
+      "learning_rate": 4.682472989195679e-05,
+      "loss": 0.7322,
+      "step": 1597
+    },
+    {
+      "epoch": 2.04544,
+      "grad_norm": 0.6331745982170105,
+      "learning_rate": 4.682272909163666e-05,
+      "loss": 0.7065,
+      "step": 1598
+    },
+    {
+      "epoch": 2.04672,
+      "grad_norm": 0.7005310654640198,
+      "learning_rate": 4.682072829131653e-05,
+      "loss": 0.7582,
+      "step": 1599
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.6412195563316345,
+      "learning_rate": 4.68187274909964e-05,
+      "loss": 0.7186,
+      "step": 1600
+    },
+    {
+      "epoch": 2.04928,
+      "grad_norm": 0.6990569233894348,
+      "learning_rate": 4.681672669067627e-05,
+      "loss": 0.7195,
+      "step": 1601
+    },
+    {
+      "epoch": 2.05056,
+      "grad_norm": 0.6446850895881653,
+      "learning_rate": 4.681472589035614e-05,
+      "loss": 0.6461,
+      "step": 1602
+    },
+    {
+      "epoch": 2.05184,
+      "grad_norm": 0.6356726288795471,
+      "learning_rate": 4.681272509003602e-05,
+      "loss": 0.6688,
+      "step": 1603
+    },
+    {
+      "epoch": 2.05312,
+      "grad_norm": 0.6811710000038147,
+      "learning_rate": 4.681072428971589e-05,
+      "loss": 0.7049,
+      "step": 1604
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 0.6494273543357849,
+      "learning_rate": 4.680872348939576e-05,
+      "loss": 0.6843,
+      "step": 1605
+    },
+    {
+      "epoch": 2.05568,
+      "grad_norm": 0.6489901542663574,
+      "learning_rate": 4.6806722689075635e-05,
+      "loss": 0.6958,
+      "step": 1606
+    },
+    {
+      "epoch": 2.05696,
+      "grad_norm": 0.6661378145217896,
+      "learning_rate": 4.680472188875551e-05,
+      "loss": 0.7385,
+      "step": 1607
+    },
+    {
+      "epoch": 2.05824,
+      "grad_norm": 0.6342524290084839,
+      "learning_rate": 4.680272108843537e-05,
+      "loss": 0.7179,
+      "step": 1608
+    },
+    {
+      "epoch": 2.05952,
+      "grad_norm": 0.6314605474472046,
+      "learning_rate": 4.6800720288115244e-05,
+      "loss": 0.7109,
+      "step": 1609
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 0.665981650352478,
+      "learning_rate": 4.679871948779512e-05,
+      "loss": 0.7259,
+      "step": 1610
+    },
+    {
+      "epoch": 2.06208,
+      "grad_norm": 0.6300519704818726,
+      "learning_rate": 4.6796718687474994e-05,
+      "loss": 0.7014,
+      "step": 1611
+    },
+    {
+      "epoch": 2.06336,
+      "grad_norm": 0.6580458879470825,
+      "learning_rate": 4.6794717887154866e-05,
+      "loss": 0.6672,
+      "step": 1612
+    },
+    {
+      "epoch": 2.06464,
+      "grad_norm": 0.669180154800415,
+      "learning_rate": 4.679271708683474e-05,
+      "loss": 0.6437,
+      "step": 1613
+    },
+    {
+      "epoch": 2.06592,
+      "grad_norm": 0.690592348575592,
+      "learning_rate": 4.679071628651461e-05,
+      "loss": 0.7331,
+      "step": 1614
+    },
+    {
+      "epoch": 2.0672,
+      "grad_norm": 0.6528162360191345,
+      "learning_rate": 4.678871548619448e-05,
+      "loss": 0.6798,
+      "step": 1615
+    },
+    {
+      "epoch": 2.06848,
+      "grad_norm": 0.619225800037384,
+      "learning_rate": 4.678671468587435e-05,
+      "loss": 0.7003,
+      "step": 1616
+    },
+    {
+      "epoch": 2.06976,
+      "grad_norm": 0.6643122434616089,
+      "learning_rate": 4.6784713885554225e-05,
+      "loss": 0.764,
+      "step": 1617
+    },
+    {
+      "epoch": 2.07104,
+      "grad_norm": 0.6456024050712585,
+      "learning_rate": 4.67827130852341e-05,
+      "loss": 0.6792,
+      "step": 1618
+    },
+    {
+      "epoch": 2.07232,
+      "grad_norm": 0.6370171308517456,
+      "learning_rate": 4.678071228491397e-05,
+      "loss": 0.6821,
+      "step": 1619
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.646872341632843,
+      "learning_rate": 4.677871148459384e-05,
+      "loss": 0.6942,
+      "step": 1620
+    },
+    {
+      "epoch": 2.07488,
+      "grad_norm": 0.6307012438774109,
+      "learning_rate": 4.677671068427371e-05,
+      "loss": 0.6781,
+      "step": 1621
+    },
+    {
+      "epoch": 2.07616,
+      "grad_norm": 0.6719081401824951,
+      "learning_rate": 4.6774709883953585e-05,
+      "loss": 0.7221,
+      "step": 1622
+    },
+    {
+      "epoch": 2.07744,
+      "grad_norm": 0.6334235072135925,
+      "learning_rate": 4.6772709083633456e-05,
+      "loss": 0.6652,
+      "step": 1623
+    },
+    {
+      "epoch": 2.07872,
+      "grad_norm": 0.6129449009895325,
+      "learning_rate": 4.677070828331333e-05,
+      "loss": 0.6324,
+      "step": 1624
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.617857813835144,
+      "learning_rate": 4.67687074829932e-05,
+      "loss": 0.6846,
+      "step": 1625
+    },
+    {
+      "epoch": 2.08128,
+      "grad_norm": 0.6237433552742004,
+      "learning_rate": 4.676670668267307e-05,
+      "loss": 0.6622,
+      "step": 1626
+    },
+    {
+      "epoch": 2.08256,
+      "grad_norm": 0.606203019618988,
+      "learning_rate": 4.6764705882352944e-05,
+      "loss": 0.6895,
+      "step": 1627
+    },
+    {
+      "epoch": 2.08384,
+      "grad_norm": 0.6327537894248962,
+      "learning_rate": 4.6762705082032816e-05,
+      "loss": 0.6087,
+      "step": 1628
+    },
+    {
+      "epoch": 2.08512,
+      "grad_norm": 0.6112367510795593,
+      "learning_rate": 4.676070428171269e-05,
+      "loss": 0.6854,
+      "step": 1629
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.5821278691291809,
+      "learning_rate": 4.675870348139256e-05,
+      "loss": 0.6405,
+      "step": 1630
+    },
+    {
+      "epoch": 2.08768,
+      "grad_norm": 0.6095936894416809,
+      "learning_rate": 4.675670268107243e-05,
+      "loss": 0.6359,
+      "step": 1631
+    },
+    {
+      "epoch": 2.08896,
+      "grad_norm": 0.6587210893630981,
+      "learning_rate": 4.67547018807523e-05,
+      "loss": 0.7537,
+      "step": 1632
+    },
+    {
+      "epoch": 2.09024,
+      "grad_norm": 0.6856895089149475,
+      "learning_rate": 4.6752701080432175e-05,
+      "loss": 0.7386,
+      "step": 1633
+    },
+    {
+      "epoch": 2.09152,
+      "grad_norm": 0.6704199910163879,
+      "learning_rate": 4.675070028011205e-05,
+      "loss": 0.7063,
+      "step": 1634
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 0.6308407187461853,
+      "learning_rate": 4.674869947979192e-05,
+      "loss": 0.6713,
+      "step": 1635
+    },
+    {
+      "epoch": 2.09408,
+      "grad_norm": 0.6690980195999146,
+      "learning_rate": 4.674669867947179e-05,
+      "loss": 0.6961,
+      "step": 1636
+    },
+    {
+      "epoch": 2.09536,
+      "grad_norm": 0.6405916213989258,
+      "learning_rate": 4.674469787915166e-05,
+      "loss": 0.6758,
+      "step": 1637
+    },
+    {
+      "epoch": 2.09664,
+      "grad_norm": 0.638323962688446,
+      "learning_rate": 4.6742697078831534e-05,
+      "loss": 0.7383,
+      "step": 1638
+    },
+    {
+      "epoch": 2.09792,
+      "grad_norm": 0.6798389554023743,
+      "learning_rate": 4.6740696278511406e-05,
+      "loss": 0.7503,
+      "step": 1639
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.6478346586227417,
+      "learning_rate": 4.673869547819128e-05,
+      "loss": 0.6709,
+      "step": 1640
+    },
+    {
+      "epoch": 2.10048,
+      "grad_norm": 0.6550549864768982,
+      "learning_rate": 4.673669467787115e-05,
+      "loss": 0.7377,
+      "step": 1641
+    },
+    {
+      "epoch": 2.10176,
+      "grad_norm": 0.611666738986969,
+      "learning_rate": 4.673469387755102e-05,
+      "loss": 0.6414,
+      "step": 1642
+    },
+    {
+      "epoch": 2.10304,
+      "grad_norm": 0.6572237610816956,
+      "learning_rate": 4.6732693077230894e-05,
+      "loss": 0.7387,
+      "step": 1643
+    },
+    {
+      "epoch": 2.10432,
+      "grad_norm": 0.6161869764328003,
+      "learning_rate": 4.6730692276910765e-05,
+      "loss": 0.6336,
+      "step": 1644
+    },
+    {
+      "epoch": 2.1056,
+      "grad_norm": 0.6543852090835571,
+      "learning_rate": 4.672869147659064e-05,
+      "loss": 0.7487,
+      "step": 1645
+    },
+    {
+      "epoch": 2.10688,
+      "grad_norm": 0.657584011554718,
+      "learning_rate": 4.6726690676270516e-05,
+      "loss": 0.7025,
+      "step": 1646
+    },
+    {
+      "epoch": 2.10816,
+      "grad_norm": 0.6363831162452698,
+      "learning_rate": 4.672468987595038e-05,
+      "loss": 0.7098,
+      "step": 1647
+    },
+    {
+      "epoch": 2.10944,
+      "grad_norm": 0.6271912455558777,
+      "learning_rate": 4.672268907563025e-05,
+      "loss": 0.6601,
+      "step": 1648
+    },
+    {
+      "epoch": 2.11072,
+      "grad_norm": 0.6789674162864685,
+      "learning_rate": 4.6720688275310125e-05,
+      "loss": 0.739,
+      "step": 1649
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.6508240699768066,
+      "learning_rate": 4.6718687474989997e-05,
+      "loss": 0.7454,
+      "step": 1650
+    },
+    {
+      "epoch": 2.11328,
+      "grad_norm": 0.6586835980415344,
+      "learning_rate": 4.671668667466987e-05,
+      "loss": 0.6957,
+      "step": 1651
+    },
+    {
+      "epoch": 2.11456,
+      "grad_norm": 0.6778126955032349,
+      "learning_rate": 4.671468587434974e-05,
+      "loss": 0.7017,
+      "step": 1652
+    },
+    {
+      "epoch": 2.11584,
+      "grad_norm": 0.6158040165901184,
+      "learning_rate": 4.671268507402962e-05,
+      "loss": 0.6696,
+      "step": 1653
+    },
+    {
+      "epoch": 2.11712,
+      "grad_norm": 0.6658079624176025,
+      "learning_rate": 4.671068427370949e-05,
+      "loss": 0.735,
+      "step": 1654
+    },
+    {
+      "epoch": 2.1184,
+      "grad_norm": 0.6727830171585083,
+      "learning_rate": 4.6708683473389356e-05,
+      "loss": 0.685,
+      "step": 1655
+    },
+    {
+      "epoch": 2.11968,
+      "grad_norm": 0.614812970161438,
+      "learning_rate": 4.670668267306923e-05,
+      "loss": 0.6622,
+      "step": 1656
+    },
+    {
+      "epoch": 2.12096,
+      "grad_norm": 0.6530376076698303,
+      "learning_rate": 4.67046818727491e-05,
+      "loss": 0.7208,
+      "step": 1657
+    },
+    {
+      "epoch": 2.12224,
+      "grad_norm": 0.6509586572647095,
+      "learning_rate": 4.670268107242897e-05,
+      "loss": 0.7607,
+      "step": 1658
+    },
+    {
+      "epoch": 2.12352,
+      "grad_norm": 0.6249855756759644,
+      "learning_rate": 4.670068027210884e-05,
+      "loss": 0.6638,
+      "step": 1659
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.6153408885002136,
+      "learning_rate": 4.669867947178872e-05,
+      "loss": 0.699,
+      "step": 1660
+    },
+    {
+      "epoch": 2.12608,
+      "grad_norm": 0.6448095440864563,
+      "learning_rate": 4.6696678671468594e-05,
+      "loss": 0.7032,
+      "step": 1661
+    },
+    {
+      "epoch": 2.12736,
+      "grad_norm": 0.616129457950592,
+      "learning_rate": 4.6694677871148466e-05,
+      "loss": 0.68,
+      "step": 1662
+    },
+    {
+      "epoch": 2.12864,
+      "grad_norm": 0.6616100072860718,
+      "learning_rate": 4.669267707082833e-05,
+      "loss": 0.6944,
+      "step": 1663
+    },
+    {
+      "epoch": 2.12992,
+      "grad_norm": 0.6487681865692139,
+      "learning_rate": 4.66906762705082e-05,
+      "loss": 0.651,
+      "step": 1664
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 0.6391069889068604,
+      "learning_rate": 4.6688675470188074e-05,
+      "loss": 0.6815,
+      "step": 1665
+    },
+    {
+      "epoch": 2.13248,
+      "grad_norm": 0.6539847254753113,
+      "learning_rate": 4.6686674669867946e-05,
+      "loss": 0.7119,
+      "step": 1666
+    },
+    {
+      "epoch": 2.13376,
+      "grad_norm": 0.6501457691192627,
+      "learning_rate": 4.6684673869547825e-05,
+      "loss": 0.6649,
+      "step": 1667
+    },
+    {
+      "epoch": 2.13504,
+      "grad_norm": 0.6687941551208496,
+      "learning_rate": 4.66826730692277e-05,
+      "loss": 0.6558,
+      "step": 1668
+    },
+    {
+      "epoch": 2.13632,
+      "grad_norm": 0.690743625164032,
+      "learning_rate": 4.668067226890757e-05,
+      "loss": 0.6878,
+      "step": 1669
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.6361063122749329,
+      "learning_rate": 4.667867146858744e-05,
+      "loss": 0.7181,
+      "step": 1670
+    },
+    {
+      "epoch": 2.13888,
+      "grad_norm": 0.6546103954315186,
+      "learning_rate": 4.6676670668267306e-05,
+      "loss": 0.7105,
+      "step": 1671
+    },
+    {
+      "epoch": 2.14016,
+      "grad_norm": 0.6859259605407715,
+      "learning_rate": 4.667466986794718e-05,
+      "loss": 0.7206,
+      "step": 1672
+    },
+    {
+      "epoch": 2.1414400000000002,
+      "grad_norm": 0.6436448097229004,
+      "learning_rate": 4.667266906762705e-05,
+      "loss": 0.6466,
+      "step": 1673
+    },
+    {
+      "epoch": 2.14272,
+      "grad_norm": 0.6501901745796204,
+      "learning_rate": 4.667066826730693e-05,
+      "loss": 0.6934,
+      "step": 1674
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 0.6573314070701599,
+      "learning_rate": 4.66686674669868e-05,
+      "loss": 0.6902,
+      "step": 1675
+    },
+    {
+      "epoch": 2.14528,
+      "grad_norm": 0.6532529592514038,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.6551,
+      "step": 1676
+    },
+    {
+      "epoch": 2.14656,
+      "grad_norm": 0.6756119132041931,
+      "learning_rate": 4.6664665866346543e-05,
+      "loss": 0.7094,
+      "step": 1677
+    },
+    {
+      "epoch": 2.14784,
+      "grad_norm": 0.6881611943244934,
+      "learning_rate": 4.6662665066026415e-05,
+      "loss": 0.7356,
+      "step": 1678
+    },
+    {
+      "epoch": 2.14912,
+      "grad_norm": 0.6921321749687195,
+      "learning_rate": 4.666066426570628e-05,
+      "loss": 0.704,
+      "step": 1679
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.6233019828796387,
+      "learning_rate": 4.665866346538615e-05,
+      "loss": 0.6581,
+      "step": 1680
+    },
+    {
+      "epoch": 2.15168,
+      "grad_norm": 0.6320359110832214,
+      "learning_rate": 4.665666266506603e-05,
+      "loss": 0.7222,
+      "step": 1681
+    },
+    {
+      "epoch": 2.15296,
+      "grad_norm": 0.6233236789703369,
+      "learning_rate": 4.66546618647459e-05,
+      "loss": 0.6272,
+      "step": 1682
+    },
+    {
+      "epoch": 2.15424,
+      "grad_norm": 0.6322048306465149,
+      "learning_rate": 4.6652661064425775e-05,
+      "loss": 0.6732,
+      "step": 1683
+    },
+    {
+      "epoch": 2.15552,
+      "grad_norm": 0.6570281386375427,
+      "learning_rate": 4.6650660264105646e-05,
+      "loss": 0.6651,
+      "step": 1684
+    },
+    {
+      "epoch": 2.1568,
+      "grad_norm": 0.6593979597091675,
+      "learning_rate": 4.664865946378552e-05,
+      "loss": 0.7143,
+      "step": 1685
+    },
+    {
+      "epoch": 2.15808,
+      "grad_norm": 0.6577056050300598,
+      "learning_rate": 4.664665866346539e-05,
+      "loss": 0.6509,
+      "step": 1686
+    },
+    {
+      "epoch": 2.15936,
+      "grad_norm": 0.6036557555198669,
+      "learning_rate": 4.6644657863145255e-05,
+      "loss": 0.6528,
+      "step": 1687
+    },
+    {
+      "epoch": 2.16064,
+      "grad_norm": 0.6405543684959412,
+      "learning_rate": 4.6642657062825134e-05,
+      "loss": 0.6667,
+      "step": 1688
+    },
+    {
+      "epoch": 2.16192,
+      "grad_norm": 0.7007610201835632,
+      "learning_rate": 4.6640656262505006e-05,
+      "loss": 0.7203,
+      "step": 1689
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.6296299695968628,
+      "learning_rate": 4.663865546218488e-05,
+      "loss": 0.6863,
+      "step": 1690
+    },
+    {
+      "epoch": 2.16448,
+      "grad_norm": 0.6606785655021667,
+      "learning_rate": 4.663665466186475e-05,
+      "loss": 0.7131,
+      "step": 1691
+    },
+    {
+      "epoch": 2.16576,
+      "grad_norm": 0.6546313166618347,
+      "learning_rate": 4.663465386154462e-05,
+      "loss": 0.6784,
+      "step": 1692
+    },
+    {
+      "epoch": 2.16704,
+      "grad_norm": 0.6348084807395935,
+      "learning_rate": 4.663265306122449e-05,
+      "loss": 0.7113,
+      "step": 1693
+    },
+    {
+      "epoch": 2.16832,
+      "grad_norm": 0.6655004024505615,
+      "learning_rate": 4.6630652260904365e-05,
+      "loss": 0.6833,
+      "step": 1694
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 0.6627703905105591,
+      "learning_rate": 4.662865146058424e-05,
+      "loss": 0.6661,
+      "step": 1695
+    },
+    {
+      "epoch": 2.17088,
+      "grad_norm": 0.6439046263694763,
+      "learning_rate": 4.662665066026411e-05,
+      "loss": 0.7075,
+      "step": 1696
+    },
+    {
+      "epoch": 2.17216,
+      "grad_norm": 0.6709191203117371,
+      "learning_rate": 4.662464985994398e-05,
+      "loss": 0.7076,
+      "step": 1697
+    },
+    {
+      "epoch": 2.17344,
+      "grad_norm": 0.6734464168548584,
+      "learning_rate": 4.662264905962385e-05,
+      "loss": 0.7296,
+      "step": 1698
+    },
+    {
+      "epoch": 2.1747199999999998,
+      "grad_norm": 0.663478434085846,
+      "learning_rate": 4.6620648259303724e-05,
+      "loss": 0.6552,
+      "step": 1699
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.654140055179596,
+      "learning_rate": 4.6618647458983596e-05,
+      "loss": 0.7028,
+      "step": 1700
+    },
+    {
+      "epoch": 2.17728,
+      "grad_norm": 0.6052113175392151,
+      "learning_rate": 4.661664665866347e-05,
+      "loss": 0.6056,
+      "step": 1701
+    },
+    {
+      "epoch": 2.17856,
+      "grad_norm": 0.6727505922317505,
+      "learning_rate": 4.661464585834334e-05,
+      "loss": 0.6829,
+      "step": 1702
+    },
+    {
+      "epoch": 2.17984,
+      "grad_norm": 0.6504889130592346,
+      "learning_rate": 4.661264505802321e-05,
+      "loss": 0.6782,
+      "step": 1703
+    },
+    {
+      "epoch": 2.18112,
+      "grad_norm": 0.6764382123947144,
+      "learning_rate": 4.6610644257703084e-05,
+      "loss": 0.7543,
+      "step": 1704
+    },
+    {
+      "epoch": 2.1824,
+      "grad_norm": 0.6401752829551697,
+      "learning_rate": 4.6608643457382955e-05,
+      "loss": 0.6898,
+      "step": 1705
+    },
+    {
+      "epoch": 2.18368,
+      "grad_norm": 0.6742962002754211,
+      "learning_rate": 4.660664265706283e-05,
+      "loss": 0.6884,
+      "step": 1706
+    },
+    {
+      "epoch": 2.1849600000000002,
+      "grad_norm": 0.6535323262214661,
+      "learning_rate": 4.66046418567427e-05,
+      "loss": 0.6647,
+      "step": 1707
+    },
+    {
+      "epoch": 2.18624,
+      "grad_norm": 0.6376174092292786,
+      "learning_rate": 4.660264105642257e-05,
+      "loss": 0.6964,
+      "step": 1708
+    },
+    {
+      "epoch": 2.18752,
+      "grad_norm": 0.6661503314971924,
+      "learning_rate": 4.660064025610244e-05,
+      "loss": 0.6898,
+      "step": 1709
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.6385532021522522,
+      "learning_rate": 4.6598639455782315e-05,
+      "loss": 0.6951,
+      "step": 1710
+    },
+    {
+      "epoch": 2.19008,
+      "grad_norm": 0.6412584185600281,
+      "learning_rate": 4.6596638655462187e-05,
+      "loss": 0.7281,
+      "step": 1711
+    },
+    {
+      "epoch": 2.19136,
+      "grad_norm": 0.7024900317192078,
+      "learning_rate": 4.659463785514206e-05,
+      "loss": 0.7168,
+      "step": 1712
+    },
+    {
+      "epoch": 2.19264,
+      "grad_norm": 0.6655505895614624,
+      "learning_rate": 4.659263705482193e-05,
+      "loss": 0.6731,
+      "step": 1713
+    },
+    {
+      "epoch": 2.19392,
+      "grad_norm": 0.6673122048377991,
+      "learning_rate": 4.65906362545018e-05,
+      "loss": 0.757,
+      "step": 1714
+    },
+    {
+      "epoch": 2.1952,
+      "grad_norm": 0.6695178151130676,
+      "learning_rate": 4.6588635454181674e-05,
+      "loss": 0.7752,
+      "step": 1715
+    },
+    {
+      "epoch": 2.19648,
+      "grad_norm": 0.6591713428497314,
+      "learning_rate": 4.658663465386155e-05,
+      "loss": 0.6715,
+      "step": 1716
+    },
+    {
+      "epoch": 2.19776,
+      "grad_norm": 0.6705875396728516,
+      "learning_rate": 4.658463385354142e-05,
+      "loss": 0.6875,
+      "step": 1717
+    },
+    {
+      "epoch": 2.19904,
+      "grad_norm": 0.6780663728713989,
+      "learning_rate": 4.658263305322129e-05,
+      "loss": 0.697,
+      "step": 1718
+    },
+    {
+      "epoch": 2.20032,
+      "grad_norm": 0.6735832691192627,
+      "learning_rate": 4.658063225290116e-05,
+      "loss": 0.6991,
+      "step": 1719
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.640477180480957,
+      "learning_rate": 4.657863145258103e-05,
+      "loss": 0.694,
+      "step": 1720
+    },
+    {
+      "epoch": 2.20288,
+      "grad_norm": 0.6748470067977905,
+      "learning_rate": 4.6576630652260905e-05,
+      "loss": 0.6939,
+      "step": 1721
+    },
+    {
+      "epoch": 2.20416,
+      "grad_norm": 0.607291579246521,
+      "learning_rate": 4.657462985194078e-05,
+      "loss": 0.651,
+      "step": 1722
+    },
+    {
+      "epoch": 2.20544,
+      "grad_norm": 0.6633931398391724,
+      "learning_rate": 4.6572629051620656e-05,
+      "loss": 0.7131,
+      "step": 1723
+    },
+    {
+      "epoch": 2.20672,
+      "grad_norm": 0.7354316711425781,
+      "learning_rate": 4.657062825130053e-05,
+      "loss": 0.6963,
+      "step": 1724
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.708829402923584,
+      "learning_rate": 4.656862745098039e-05,
+      "loss": 0.736,
+      "step": 1725
+    },
+    {
+      "epoch": 2.20928,
+      "grad_norm": 0.6382274031639099,
+      "learning_rate": 4.6566626650660264e-05,
+      "loss": 0.663,
+      "step": 1726
+    },
+    {
+      "epoch": 2.21056,
+      "grad_norm": 0.6313372850418091,
+      "learning_rate": 4.6564625850340136e-05,
+      "loss": 0.6861,
+      "step": 1727
+    },
+    {
+      "epoch": 2.21184,
+      "grad_norm": 0.6234421730041504,
+      "learning_rate": 4.656262505002001e-05,
+      "loss": 0.6846,
+      "step": 1728
+    },
+    {
+      "epoch": 2.21312,
+      "grad_norm": 0.6155267953872681,
+      "learning_rate": 4.656062424969988e-05,
+      "loss": 0.6289,
+      "step": 1729
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.6332612633705139,
+      "learning_rate": 4.655862344937976e-05,
+      "loss": 0.6752,
+      "step": 1730
+    },
+    {
+      "epoch": 2.21568,
+      "grad_norm": 0.6155014634132385,
+      "learning_rate": 4.655662264905963e-05,
+      "loss": 0.6088,
+      "step": 1731
+    },
+    {
+      "epoch": 2.21696,
+      "grad_norm": 0.6483979821205139,
+      "learning_rate": 4.65546218487395e-05,
+      "loss": 0.688,
+      "step": 1732
+    },
+    {
+      "epoch": 2.2182399999999998,
+      "grad_norm": 0.6692848801612854,
+      "learning_rate": 4.655262104841937e-05,
+      "loss": 0.6918,
+      "step": 1733
+    },
+    {
+      "epoch": 2.21952,
+      "grad_norm": 0.6489692330360413,
+      "learning_rate": 4.655062024809924e-05,
+      "loss": 0.7016,
+      "step": 1734
+    },
+    {
+      "epoch": 2.2208,
+      "grad_norm": 0.6827375888824463,
+      "learning_rate": 4.654861944777911e-05,
+      "loss": 0.7165,
+      "step": 1735
+    },
+    {
+      "epoch": 2.22208,
+      "grad_norm": 0.652456521987915,
+      "learning_rate": 4.654661864745898e-05,
+      "loss": 0.7366,
+      "step": 1736
+    },
+    {
+      "epoch": 2.22336,
+      "grad_norm": 0.6942132711410522,
+      "learning_rate": 4.654461784713886e-05,
+      "loss": 0.674,
+      "step": 1737
+    },
+    {
+      "epoch": 2.22464,
+      "grad_norm": 0.6576859354972839,
+      "learning_rate": 4.6542617046818734e-05,
+      "loss": 0.7573,
+      "step": 1738
+    },
+    {
+      "epoch": 2.22592,
+      "grad_norm": 0.6581286191940308,
+      "learning_rate": 4.6540616246498605e-05,
+      "loss": 0.6484,
+      "step": 1739
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.6722337007522583,
+      "learning_rate": 4.653861544617848e-05,
+      "loss": 0.7328,
+      "step": 1740
+    },
+    {
+      "epoch": 2.22848,
+      "grad_norm": 0.6128942966461182,
+      "learning_rate": 4.653661464585834e-05,
+      "loss": 0.6935,
+      "step": 1741
+    },
+    {
+      "epoch": 2.22976,
+      "grad_norm": 0.632098913192749,
+      "learning_rate": 4.6534613845538214e-05,
+      "loss": 0.6824,
+      "step": 1742
+    },
+    {
+      "epoch": 2.23104,
+      "grad_norm": 0.6399368047714233,
+      "learning_rate": 4.6532613045218086e-05,
+      "loss": 0.6962,
+      "step": 1743
+    },
+    {
+      "epoch": 2.23232,
+      "grad_norm": 0.6291220188140869,
+      "learning_rate": 4.653061224489796e-05,
+      "loss": 0.6935,
+      "step": 1744
+    },
+    {
+      "epoch": 2.2336,
+      "grad_norm": 0.6679767966270447,
+      "learning_rate": 4.6528611444577837e-05,
+      "loss": 0.685,
+      "step": 1745
+    },
+    {
+      "epoch": 2.23488,
+      "grad_norm": 0.652439534664154,
+      "learning_rate": 4.652661064425771e-05,
+      "loss": 0.731,
+      "step": 1746
+    },
+    {
+      "epoch": 2.23616,
+      "grad_norm": 0.6402837038040161,
+      "learning_rate": 4.652460984393758e-05,
+      "loss": 0.6515,
+      "step": 1747
+    },
+    {
+      "epoch": 2.23744,
+      "grad_norm": 0.6584830284118652,
+      "learning_rate": 4.652260904361745e-05,
+      "loss": 0.7117,
+      "step": 1748
+    },
+    {
+      "epoch": 2.23872,
+      "grad_norm": 0.6643098592758179,
+      "learning_rate": 4.652060824329732e-05,
+      "loss": 0.6786,
+      "step": 1749
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6529515385627747,
+      "learning_rate": 4.651860744297719e-05,
+      "loss": 0.6985,
+      "step": 1750
+    },
+    {
+      "epoch": 2.24128,
+      "grad_norm": 0.6570754051208496,
+      "learning_rate": 4.651660664265706e-05,
+      "loss": 0.6892,
+      "step": 1751
+    },
+    {
+      "epoch": 2.24256,
+      "grad_norm": 0.6258246898651123,
+      "learning_rate": 4.651460584233694e-05,
+      "loss": 0.6781,
+      "step": 1752
+    },
+    {
+      "epoch": 2.24384,
+      "grad_norm": 0.6045702695846558,
+      "learning_rate": 4.651260504201681e-05,
+      "loss": 0.6712,
+      "step": 1753
+    },
+    {
+      "epoch": 2.24512,
+      "grad_norm": 0.6391083002090454,
+      "learning_rate": 4.651060424169668e-05,
+      "loss": 0.6869,
+      "step": 1754
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 0.6669121980667114,
+      "learning_rate": 4.6508603441376555e-05,
+      "loss": 0.7874,
+      "step": 1755
+    },
+    {
+      "epoch": 2.24768,
+      "grad_norm": 0.6193825006484985,
+      "learning_rate": 4.650660264105643e-05,
+      "loss": 0.6625,
+      "step": 1756
+    },
+    {
+      "epoch": 2.24896,
+      "grad_norm": 0.6318017840385437,
+      "learning_rate": 4.650460184073629e-05,
+      "loss": 0.6807,
+      "step": 1757
+    },
+    {
+      "epoch": 2.25024,
+      "grad_norm": 0.6676175594329834,
+      "learning_rate": 4.6502601040416164e-05,
+      "loss": 0.6965,
+      "step": 1758
+    },
+    {
+      "epoch": 2.25152,
+      "grad_norm": 0.652651846408844,
+      "learning_rate": 4.650060024009604e-05,
+      "loss": 0.6623,
+      "step": 1759
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.6952486038208008,
+      "learning_rate": 4.6498599439775914e-05,
+      "loss": 0.7059,
+      "step": 1760
+    },
+    {
+      "epoch": 2.25408,
+      "grad_norm": 0.713154673576355,
+      "learning_rate": 4.6496598639455786e-05,
+      "loss": 0.7397,
+      "step": 1761
+    },
+    {
+      "epoch": 2.25536,
+      "grad_norm": 0.636633574962616,
+      "learning_rate": 4.649459783913566e-05,
+      "loss": 0.6193,
+      "step": 1762
+    },
+    {
+      "epoch": 2.25664,
+      "grad_norm": 0.6551604866981506,
+      "learning_rate": 4.649259703881553e-05,
+      "loss": 0.7048,
+      "step": 1763
+    },
+    {
+      "epoch": 2.25792,
+      "grad_norm": 0.6308698654174805,
+      "learning_rate": 4.64905962384954e-05,
+      "loss": 0.6646,
+      "step": 1764
+    },
+    {
+      "epoch": 2.2592,
+      "grad_norm": 0.6496629118919373,
+      "learning_rate": 4.648859543817527e-05,
+      "loss": 0.6879,
+      "step": 1765
+    },
+    {
+      "epoch": 2.26048,
+      "grad_norm": 0.6222250461578369,
+      "learning_rate": 4.6486594637855145e-05,
+      "loss": 0.6635,
+      "step": 1766
+    },
+    {
+      "epoch": 2.2617599999999998,
+      "grad_norm": 0.6518458724021912,
+      "learning_rate": 4.648459383753502e-05,
+      "loss": 0.6757,
+      "step": 1767
+    },
+    {
+      "epoch": 2.26304,
+      "grad_norm": 0.656060516834259,
+      "learning_rate": 4.648259303721489e-05,
+      "loss": 0.6818,
+      "step": 1768
+    },
+    {
+      "epoch": 2.26432,
+      "grad_norm": 0.6683648228645325,
+      "learning_rate": 4.648059223689476e-05,
+      "loss": 0.7028,
+      "step": 1769
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.6672934889793396,
+      "learning_rate": 4.647859143657463e-05,
+      "loss": 0.6928,
+      "step": 1770
+    },
+    {
+      "epoch": 2.26688,
+      "grad_norm": 0.6802359223365784,
+      "learning_rate": 4.6476590636254505e-05,
+      "loss": 0.6616,
+      "step": 1771
+    },
+    {
+      "epoch": 2.26816,
+      "grad_norm": 0.6378535628318787,
+      "learning_rate": 4.647458983593438e-05,
+      "loss": 0.689,
+      "step": 1772
+    },
+    {
+      "epoch": 2.26944,
+      "grad_norm": 0.677901566028595,
+      "learning_rate": 4.647258903561425e-05,
+      "loss": 0.7436,
+      "step": 1773
+    },
+    {
+      "epoch": 2.27072,
+      "grad_norm": 0.6340510249137878,
+      "learning_rate": 4.647058823529412e-05,
+      "loss": 0.7241,
+      "step": 1774
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 0.6160155534744263,
+      "learning_rate": 4.646858743497399e-05,
+      "loss": 0.635,
+      "step": 1775
+    },
+    {
+      "epoch": 2.27328,
+      "grad_norm": 0.6366074085235596,
+      "learning_rate": 4.6466586634653864e-05,
+      "loss": 0.6915,
+      "step": 1776
+    },
+    {
+      "epoch": 2.27456,
+      "grad_norm": 0.6163462996482849,
+      "learning_rate": 4.6464585834333736e-05,
+      "loss": 0.6227,
+      "step": 1777
+    },
+    {
+      "epoch": 2.27584,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 4.646258503401361e-05,
+      "loss": 0.753,
+      "step": 1778
+    },
+    {
+      "epoch": 2.27712,
+      "grad_norm": 0.6583823561668396,
+      "learning_rate": 4.646058423369348e-05,
+      "loss": 0.7791,
+      "step": 1779
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.6398965120315552,
+      "learning_rate": 4.645858343337335e-05,
+      "loss": 0.6842,
+      "step": 1780
+    },
+    {
+      "epoch": 2.27968,
+      "grad_norm": 0.6243705153465271,
+      "learning_rate": 4.645658263305322e-05,
+      "loss": 0.6953,
+      "step": 1781
+    },
+    {
+      "epoch": 2.28096,
+      "grad_norm": 0.6466498374938965,
+      "learning_rate": 4.6454581832733095e-05,
+      "loss": 0.7053,
+      "step": 1782
+    },
+    {
+      "epoch": 2.28224,
+      "grad_norm": 0.6308572292327881,
+      "learning_rate": 4.645258103241297e-05,
+      "loss": 0.6739,
+      "step": 1783
+    },
+    {
+      "epoch": 2.28352,
+      "grad_norm": 0.6494999527931213,
+      "learning_rate": 4.645058023209284e-05,
+      "loss": 0.7472,
+      "step": 1784
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 0.633786678314209,
+      "learning_rate": 4.644857943177271e-05,
+      "loss": 0.6696,
+      "step": 1785
+    },
+    {
+      "epoch": 2.28608,
+      "grad_norm": 0.63746577501297,
+      "learning_rate": 4.644657863145258e-05,
+      "loss": 0.6897,
+      "step": 1786
+    },
+    {
+      "epoch": 2.28736,
+      "grad_norm": 0.6449893116950989,
+      "learning_rate": 4.6444577831132454e-05,
+      "loss": 0.657,
+      "step": 1787
+    },
+    {
+      "epoch": 2.28864,
+      "grad_norm": 0.6666371822357178,
+      "learning_rate": 4.6442577030812326e-05,
+      "loss": 0.7013,
+      "step": 1788
+    },
+    {
+      "epoch": 2.28992,
+      "grad_norm": 0.6510623693466187,
+      "learning_rate": 4.64405762304922e-05,
+      "loss": 0.7175,
+      "step": 1789
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.6616090536117554,
+      "learning_rate": 4.643857543017207e-05,
+      "loss": 0.6673,
+      "step": 1790
+    },
+    {
+      "epoch": 2.29248,
+      "grad_norm": 0.7210375070571899,
+      "learning_rate": 4.643657462985194e-05,
+      "loss": 0.7318,
+      "step": 1791
+    },
+    {
+      "epoch": 2.29376,
+      "grad_norm": 0.6955354809761047,
+      "learning_rate": 4.6434573829531814e-05,
+      "loss": 0.7548,
+      "step": 1792
+    },
+    {
+      "epoch": 2.29504,
+      "grad_norm": 0.6595535278320312,
+      "learning_rate": 4.6432573029211686e-05,
+      "loss": 0.6672,
+      "step": 1793
+    },
+    {
+      "epoch": 2.29632,
+      "grad_norm": 0.6795318126678467,
+      "learning_rate": 4.6430572228891564e-05,
+      "loss": 0.6465,
+      "step": 1794
+    },
+    {
+      "epoch": 2.2976,
+      "grad_norm": 0.700689971446991,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 0.7189,
+      "step": 1795
+    },
+    {
+      "epoch": 2.29888,
+      "grad_norm": 0.7070762515068054,
+      "learning_rate": 4.64265706282513e-05,
+      "loss": 0.7249,
+      "step": 1796
+    },
+    {
+      "epoch": 2.30016,
+      "grad_norm": 0.6750733256340027,
+      "learning_rate": 4.642456982793117e-05,
+      "loss": 0.7291,
+      "step": 1797
+    },
+    {
+      "epoch": 2.30144,
+      "grad_norm": 0.6518023610115051,
+      "learning_rate": 4.6422569027611045e-05,
+      "loss": 0.7376,
+      "step": 1798
+    },
+    {
+      "epoch": 2.30272,
+      "grad_norm": 0.658316433429718,
+      "learning_rate": 4.642056822729092e-05,
+      "loss": 0.7014,
+      "step": 1799
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.6573535799980164,
+      "learning_rate": 4.641856742697079e-05,
+      "loss": 0.673,
+      "step": 1800
+    },
+    {
+      "epoch": 2.3052799999999998,
+      "grad_norm": 0.6256746649742126,
+      "learning_rate": 4.641656662665067e-05,
+      "loss": 0.6777,
+      "step": 1801
+    },
+    {
+      "epoch": 2.30656,
+      "grad_norm": 0.6481615900993347,
+      "learning_rate": 4.641456582633054e-05,
+      "loss": 0.6787,
+      "step": 1802
+    },
+    {
+      "epoch": 2.30784,
+      "grad_norm": 0.6594406366348267,
+      "learning_rate": 4.6412565026010404e-05,
+      "loss": 0.6894,
+      "step": 1803
+    },
+    {
+      "epoch": 2.30912,
+      "grad_norm": 0.6397424936294556,
+      "learning_rate": 4.6410564225690276e-05,
+      "loss": 0.7325,
+      "step": 1804
+    },
+    {
+      "epoch": 2.3104,
+      "grad_norm": 0.6274937987327576,
+      "learning_rate": 4.640856342537015e-05,
+      "loss": 0.6842,
+      "step": 1805
+    },
+    {
+      "epoch": 2.31168,
+      "grad_norm": 0.7071935534477234,
+      "learning_rate": 4.640656262505002e-05,
+      "loss": 0.7394,
+      "step": 1806
+    },
+    {
+      "epoch": 2.31296,
+      "grad_norm": 0.6791349649429321,
+      "learning_rate": 4.640456182472989e-05,
+      "loss": 0.6907,
+      "step": 1807
+    },
+    {
+      "epoch": 2.31424,
+      "grad_norm": 0.6307082772254944,
+      "learning_rate": 4.640256102440977e-05,
+      "loss": 0.6735,
+      "step": 1808
+    },
+    {
+      "epoch": 2.3155200000000002,
+      "grad_norm": 0.6786848306655884,
+      "learning_rate": 4.640056022408964e-05,
+      "loss": 0.6944,
+      "step": 1809
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.6838906407356262,
+      "learning_rate": 4.6398559423769514e-05,
+      "loss": 0.7096,
+      "step": 1810
+    },
+    {
+      "epoch": 2.31808,
+      "grad_norm": 0.6599327921867371,
+      "learning_rate": 4.639655862344938e-05,
+      "loss": 0.6573,
+      "step": 1811
+    },
+    {
+      "epoch": 2.31936,
+      "grad_norm": 0.6728073358535767,
+      "learning_rate": 4.639455782312925e-05,
+      "loss": 0.7074,
+      "step": 1812
+    },
+    {
+      "epoch": 2.32064,
+      "grad_norm": 0.6599488258361816,
+      "learning_rate": 4.639255702280912e-05,
+      "loss": 0.6743,
+      "step": 1813
+    },
+    {
+      "epoch": 2.32192,
+      "grad_norm": 0.7124109864234924,
+      "learning_rate": 4.6390556222488995e-05,
+      "loss": 0.8051,
+      "step": 1814
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 0.6490312814712524,
+      "learning_rate": 4.638855542216887e-05,
+      "loss": 0.6868,
+      "step": 1815
+    },
+    {
+      "epoch": 2.32448,
+      "grad_norm": 0.6607807278633118,
+      "learning_rate": 4.6386554621848745e-05,
+      "loss": 0.7233,
+      "step": 1816
+    },
+    {
+      "epoch": 2.32576,
+      "grad_norm": 0.6332591772079468,
+      "learning_rate": 4.638455382152862e-05,
+      "loss": 0.6772,
+      "step": 1817
+    },
+    {
+      "epoch": 2.32704,
+      "grad_norm": 0.6462684273719788,
+      "learning_rate": 4.638255302120849e-05,
+      "loss": 0.6809,
+      "step": 1818
+    },
+    {
+      "epoch": 2.32832,
+      "grad_norm": 0.685043454170227,
+      "learning_rate": 4.6380552220888354e-05,
+      "loss": 0.672,
+      "step": 1819
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.6536259651184082,
+      "learning_rate": 4.6378551420568226e-05,
+      "loss": 0.6845,
+      "step": 1820
+    },
+    {
+      "epoch": 2.33088,
+      "grad_norm": 0.6470760703086853,
+      "learning_rate": 4.63765506202481e-05,
+      "loss": 0.6574,
+      "step": 1821
+    },
+    {
+      "epoch": 2.33216,
+      "grad_norm": 0.6793741583824158,
+      "learning_rate": 4.6374549819927976e-05,
+      "loss": 0.7286,
+      "step": 1822
+    },
+    {
+      "epoch": 2.33344,
+      "grad_norm": 0.633575975894928,
+      "learning_rate": 4.637254901960785e-05,
+      "loss": 0.689,
+      "step": 1823
+    },
+    {
+      "epoch": 2.33472,
+      "grad_norm": 0.6770448088645935,
+      "learning_rate": 4.637054821928772e-05,
+      "loss": 0.6672,
+      "step": 1824
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 0.6367180943489075,
+      "learning_rate": 4.636854741896759e-05,
+      "loss": 0.712,
+      "step": 1825
+    },
+    {
+      "epoch": 2.33728,
+      "grad_norm": 0.6444728970527649,
+      "learning_rate": 4.6366546618647464e-05,
+      "loss": 0.78,
+      "step": 1826
+    },
+    {
+      "epoch": 2.33856,
+      "grad_norm": 0.6529805064201355,
+      "learning_rate": 4.636454581832733e-05,
+      "loss": 0.6594,
+      "step": 1827
+    },
+    {
+      "epoch": 2.33984,
+      "grad_norm": 0.6504684090614319,
+      "learning_rate": 4.63625450180072e-05,
+      "loss": 0.6725,
+      "step": 1828
+    },
+    {
+      "epoch": 2.34112,
+      "grad_norm": 0.645531177520752,
+      "learning_rate": 4.636054421768708e-05,
+      "loss": 0.6721,
+      "step": 1829
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.6722829341888428,
+      "learning_rate": 4.635854341736695e-05,
+      "loss": 0.772,
+      "step": 1830
+    },
+    {
+      "epoch": 2.34368,
+      "grad_norm": 0.6697701215744019,
+      "learning_rate": 4.635654261704682e-05,
+      "loss": 0.7057,
+      "step": 1831
+    },
+    {
+      "epoch": 2.34496,
+      "grad_norm": 0.6601827144622803,
+      "learning_rate": 4.6354541816726695e-05,
+      "loss": 0.644,
+      "step": 1832
+    },
+    {
+      "epoch": 2.34624,
+      "grad_norm": 0.6771488189697266,
+      "learning_rate": 4.635254101640657e-05,
+      "loss": 0.7449,
+      "step": 1833
+    },
+    {
+      "epoch": 2.34752,
+      "grad_norm": 0.6421104073524475,
+      "learning_rate": 4.635054021608644e-05,
+      "loss": 0.7342,
+      "step": 1834
+    },
+    {
+      "epoch": 2.3487999999999998,
+      "grad_norm": 0.6297141909599304,
+      "learning_rate": 4.6348539415766304e-05,
+      "loss": 0.706,
+      "step": 1835
+    },
+    {
+      "epoch": 2.35008,
+      "grad_norm": 0.6440832018852234,
+      "learning_rate": 4.634653861544618e-05,
+      "loss": 0.7282,
+      "step": 1836
+    },
+    {
+      "epoch": 2.35136,
+      "grad_norm": 0.6416788101196289,
+      "learning_rate": 4.6344537815126054e-05,
+      "loss": 0.6494,
+      "step": 1837
+    },
+    {
+      "epoch": 2.35264,
+      "grad_norm": 0.6922294497489929,
+      "learning_rate": 4.6342537014805926e-05,
+      "loss": 0.7179,
+      "step": 1838
+    },
+    {
+      "epoch": 2.35392,
+      "grad_norm": 0.674335241317749,
+      "learning_rate": 4.63405362144858e-05,
+      "loss": 0.6737,
+      "step": 1839
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.6704332232475281,
+      "learning_rate": 4.633853541416567e-05,
+      "loss": 0.7289,
+      "step": 1840
+    },
+    {
+      "epoch": 2.35648,
+      "grad_norm": 0.6815084218978882,
+      "learning_rate": 4.633653461384554e-05,
+      "loss": 0.6504,
+      "step": 1841
+    },
+    {
+      "epoch": 2.35776,
+      "grad_norm": 0.6917728781700134,
+      "learning_rate": 4.6334533813525413e-05,
+      "loss": 0.7133,
+      "step": 1842
+    },
+    {
+      "epoch": 2.3590400000000002,
+      "grad_norm": 0.6595756411552429,
+      "learning_rate": 4.6332533013205285e-05,
+      "loss": 0.6622,
+      "step": 1843
+    },
+    {
+      "epoch": 2.3603199999999998,
+      "grad_norm": 0.6047163009643555,
+      "learning_rate": 4.633053221288516e-05,
+      "loss": 0.682,
+      "step": 1844
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 0.6585202217102051,
+      "learning_rate": 4.632853141256503e-05,
+      "loss": 0.7206,
+      "step": 1845
+    },
+    {
+      "epoch": 2.36288,
+      "grad_norm": 0.6378026604652405,
+      "learning_rate": 4.63265306122449e-05,
+      "loss": 0.6693,
+      "step": 1846
+    },
+    {
+      "epoch": 2.36416,
+      "grad_norm": 0.6626023054122925,
+      "learning_rate": 4.632452981192477e-05,
+      "loss": 0.7427,
+      "step": 1847
+    },
+    {
+      "epoch": 2.36544,
+      "grad_norm": 0.652944803237915,
+      "learning_rate": 4.6322529011604645e-05,
+      "loss": 0.6706,
+      "step": 1848
+    },
+    {
+      "epoch": 2.36672,
+      "grad_norm": 0.6104753613471985,
+      "learning_rate": 4.6320528211284516e-05,
+      "loss": 0.6231,
+      "step": 1849
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.5891187191009521,
+      "learning_rate": 4.631852741096439e-05,
+      "loss": 0.6612,
+      "step": 1850
+    },
+    {
+      "epoch": 2.36928,
+      "grad_norm": 0.6433348059654236,
+      "learning_rate": 4.631652661064426e-05,
+      "loss": 0.6903,
+      "step": 1851
+    },
+    {
+      "epoch": 2.3705600000000002,
+      "grad_norm": 0.6553032994270325,
+      "learning_rate": 4.631452581032413e-05,
+      "loss": 0.6837,
+      "step": 1852
+    },
+    {
+      "epoch": 2.37184,
+      "grad_norm": 0.6701604723930359,
+      "learning_rate": 4.6312525010004004e-05,
+      "loss": 0.7339,
+      "step": 1853
+    },
+    {
+      "epoch": 2.37312,
+      "grad_norm": 0.6654971837997437,
+      "learning_rate": 4.6310524209683876e-05,
+      "loss": 0.7244,
+      "step": 1854
+    },
+    {
+      "epoch": 2.3744,
+      "grad_norm": 0.6845753788948059,
+      "learning_rate": 4.630852340936375e-05,
+      "loss": 0.742,
+      "step": 1855
+    },
+    {
+      "epoch": 2.37568,
+      "grad_norm": 0.651664137840271,
+      "learning_rate": 4.630652260904362e-05,
+      "loss": 0.6837,
+      "step": 1856
+    },
+    {
+      "epoch": 2.37696,
+      "grad_norm": 0.7059352993965149,
+      "learning_rate": 4.630452180872349e-05,
+      "loss": 0.7274,
+      "step": 1857
+    },
+    {
+      "epoch": 2.37824,
+      "grad_norm": 0.6303350329399109,
+      "learning_rate": 4.630252100840336e-05,
+      "loss": 0.7218,
+      "step": 1858
+    },
+    {
+      "epoch": 2.37952,
+      "grad_norm": 0.6301616430282593,
+      "learning_rate": 4.6300520208083235e-05,
+      "loss": 0.6528,
+      "step": 1859
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.6489315032958984,
+      "learning_rate": 4.629851940776311e-05,
+      "loss": 0.6807,
+      "step": 1860
+    },
+    {
+      "epoch": 2.38208,
+      "grad_norm": 0.6146230697631836,
+      "learning_rate": 4.629651860744298e-05,
+      "loss": 0.671,
+      "step": 1861
+    },
+    {
+      "epoch": 2.38336,
+      "grad_norm": 0.6300916075706482,
+      "learning_rate": 4.629451780712285e-05,
+      "loss": 0.6797,
+      "step": 1862
+    },
+    {
+      "epoch": 2.38464,
+      "grad_norm": 0.6397101283073425,
+      "learning_rate": 4.629251700680272e-05,
+      "loss": 0.6652,
+      "step": 1863
+    },
+    {
+      "epoch": 2.38592,
+      "grad_norm": 0.6682645678520203,
+      "learning_rate": 4.6290516206482594e-05,
+      "loss": 0.7132,
+      "step": 1864
+    },
+    {
+      "epoch": 2.3872,
+      "grad_norm": 0.6869545578956604,
+      "learning_rate": 4.6288515406162466e-05,
+      "loss": 0.6986,
+      "step": 1865
+    },
+    {
+      "epoch": 2.38848,
+      "grad_norm": 0.6472262144088745,
+      "learning_rate": 4.628651460584234e-05,
+      "loss": 0.6647,
+      "step": 1866
+    },
+    {
+      "epoch": 2.38976,
+      "grad_norm": 0.6385058164596558,
+      "learning_rate": 4.628451380552221e-05,
+      "loss": 0.6489,
+      "step": 1867
+    },
+    {
+      "epoch": 2.39104,
+      "grad_norm": 0.6617813110351562,
+      "learning_rate": 4.628251300520208e-05,
+      "loss": 0.715,
+      "step": 1868
+    },
+    {
+      "epoch": 2.39232,
+      "grad_norm": 0.6766805648803711,
+      "learning_rate": 4.6280512204881954e-05,
+      "loss": 0.6902,
+      "step": 1869
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.6299700736999512,
+      "learning_rate": 4.6278511404561825e-05,
+      "loss": 0.6678,
+      "step": 1870
+    },
+    {
+      "epoch": 2.39488,
+      "grad_norm": 0.7133849263191223,
+      "learning_rate": 4.62765106042417e-05,
+      "loss": 0.7164,
+      "step": 1871
+    },
+    {
+      "epoch": 2.39616,
+      "grad_norm": 0.6625071167945862,
+      "learning_rate": 4.6274509803921576e-05,
+      "loss": 0.7179,
+      "step": 1872
+    },
+    {
+      "epoch": 2.39744,
+      "grad_norm": 0.6794329881668091,
+      "learning_rate": 4.627250900360144e-05,
+      "loss": 0.7199,
+      "step": 1873
+    },
+    {
+      "epoch": 2.39872,
+      "grad_norm": 0.630755603313446,
+      "learning_rate": 4.627050820328131e-05,
+      "loss": 0.7723,
+      "step": 1874
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6161158680915833,
+      "learning_rate": 4.6268507402961185e-05,
+      "loss": 0.5934,
+      "step": 1875
+    },
+    {
+      "epoch": 2.40128,
+      "grad_norm": 0.6303229928016663,
+      "learning_rate": 4.6266506602641057e-05,
+      "loss": 0.689,
+      "step": 1876
+    },
+    {
+      "epoch": 2.40256,
+      "grad_norm": 0.6833181977272034,
+      "learning_rate": 4.626450580232093e-05,
+      "loss": 0.7554,
+      "step": 1877
+    },
+    {
+      "epoch": 2.4038399999999998,
+      "grad_norm": 0.6531204581260681,
+      "learning_rate": 4.62625050020008e-05,
+      "loss": 0.6816,
+      "step": 1878
+    },
+    {
+      "epoch": 2.40512,
+      "grad_norm": 0.7215322256088257,
+      "learning_rate": 4.626050420168068e-05,
+      "loss": 0.7196,
+      "step": 1879
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.6577185988426208,
+      "learning_rate": 4.625850340136055e-05,
+      "loss": 0.6797,
+      "step": 1880
+    },
+    {
+      "epoch": 2.40768,
+      "grad_norm": 0.6590116620063782,
+      "learning_rate": 4.6256502601040416e-05,
+      "loss": 0.6712,
+      "step": 1881
+    },
+    {
+      "epoch": 2.40896,
+      "grad_norm": 0.715766191482544,
+      "learning_rate": 4.625450180072029e-05,
+      "loss": 0.7324,
+      "step": 1882
+    },
+    {
+      "epoch": 2.41024,
+      "grad_norm": 0.6765993237495422,
+      "learning_rate": 4.625250100040016e-05,
+      "loss": 0.726,
+      "step": 1883
+    },
+    {
+      "epoch": 2.41152,
+      "grad_norm": 0.6403140425682068,
+      "learning_rate": 4.625050020008003e-05,
+      "loss": 0.6877,
+      "step": 1884
+    },
+    {
+      "epoch": 2.4128,
+      "grad_norm": 0.6255505681037903,
+      "learning_rate": 4.62484993997599e-05,
+      "loss": 0.6142,
+      "step": 1885
+    },
+    {
+      "epoch": 2.4140800000000002,
+      "grad_norm": 0.6832484006881714,
+      "learning_rate": 4.624649859943978e-05,
+      "loss": 0.6964,
+      "step": 1886
+    },
+    {
+      "epoch": 2.41536,
+      "grad_norm": 0.6387737989425659,
+      "learning_rate": 4.6244497799119654e-05,
+      "loss": 0.7023,
+      "step": 1887
+    },
+    {
+      "epoch": 2.41664,
+      "grad_norm": 0.6823368072509766,
+      "learning_rate": 4.6242496998799526e-05,
+      "loss": 0.7162,
+      "step": 1888
+    },
+    {
+      "epoch": 2.41792,
+      "grad_norm": 0.6266238689422607,
+      "learning_rate": 4.624049619847939e-05,
+      "loss": 0.6378,
+      "step": 1889
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.705518901348114,
+      "learning_rate": 4.623849539815926e-05,
+      "loss": 0.7405,
+      "step": 1890
+    },
+    {
+      "epoch": 2.42048,
+      "grad_norm": 0.6303145885467529,
+      "learning_rate": 4.6236494597839134e-05,
+      "loss": 0.6536,
+      "step": 1891
+    },
+    {
+      "epoch": 2.42176,
+      "grad_norm": 0.6812713146209717,
+      "learning_rate": 4.6234493797519006e-05,
+      "loss": 0.6988,
+      "step": 1892
+    },
+    {
+      "epoch": 2.42304,
+      "grad_norm": 0.6497737169265747,
+      "learning_rate": 4.6232492997198885e-05,
+      "loss": 0.6926,
+      "step": 1893
+    },
+    {
+      "epoch": 2.42432,
+      "grad_norm": 0.669630765914917,
+      "learning_rate": 4.623049219687876e-05,
+      "loss": 0.7162,
+      "step": 1894
+    },
+    {
+      "epoch": 2.4256,
+      "grad_norm": 0.6586965918540955,
+      "learning_rate": 4.622849139655863e-05,
+      "loss": 0.6695,
+      "step": 1895
+    },
+    {
+      "epoch": 2.42688,
+      "grad_norm": 0.6347070336341858,
+      "learning_rate": 4.62264905962385e-05,
+      "loss": 0.6818,
+      "step": 1896
+    },
+    {
+      "epoch": 2.42816,
+      "grad_norm": 0.6584689617156982,
+      "learning_rate": 4.6224489795918366e-05,
+      "loss": 0.733,
+      "step": 1897
+    },
+    {
+      "epoch": 2.42944,
+      "grad_norm": 0.6446535587310791,
+      "learning_rate": 4.622248899559824e-05,
+      "loss": 0.6938,
+      "step": 1898
+    },
+    {
+      "epoch": 2.43072,
+      "grad_norm": 0.6775670051574707,
+      "learning_rate": 4.622048819527811e-05,
+      "loss": 0.7274,
+      "step": 1899
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.636137068271637,
+      "learning_rate": 4.621848739495799e-05,
+      "loss": 0.6261,
+      "step": 1900
+    },
+    {
+      "epoch": 2.43328,
+      "grad_norm": 0.6404497623443604,
+      "learning_rate": 4.621648659463786e-05,
+      "loss": 0.6962,
+      "step": 1901
+    },
+    {
+      "epoch": 2.43456,
+      "grad_norm": 0.661217451095581,
+      "learning_rate": 4.621448579431773e-05,
+      "loss": 0.7293,
+      "step": 1902
+    },
+    {
+      "epoch": 2.43584,
+      "grad_norm": 0.6509891748428345,
+      "learning_rate": 4.6212484993997603e-05,
+      "loss": 0.7254,
+      "step": 1903
+    },
+    {
+      "epoch": 2.43712,
+      "grad_norm": 0.644158661365509,
+      "learning_rate": 4.6210484193677475e-05,
+      "loss": 0.6958,
+      "step": 1904
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 0.6828984022140503,
+      "learning_rate": 4.620848339335734e-05,
+      "loss": 0.6665,
+      "step": 1905
+    },
+    {
+      "epoch": 2.43968,
+      "grad_norm": 0.6128406524658203,
+      "learning_rate": 4.620648259303721e-05,
+      "loss": 0.6624,
+      "step": 1906
+    },
+    {
+      "epoch": 2.44096,
+      "grad_norm": 0.7112566232681274,
+      "learning_rate": 4.620448179271709e-05,
+      "loss": 0.7079,
+      "step": 1907
+    },
+    {
+      "epoch": 2.44224,
+      "grad_norm": 0.6818531155586243,
+      "learning_rate": 4.620248099239696e-05,
+      "loss": 0.7034,
+      "step": 1908
+    },
+    {
+      "epoch": 2.44352,
+      "grad_norm": 0.6345267295837402,
+      "learning_rate": 4.6200480192076835e-05,
+      "loss": 0.6841,
+      "step": 1909
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.7214930057525635,
+      "learning_rate": 4.6198479391756706e-05,
+      "loss": 0.7026,
+      "step": 1910
+    },
+    {
+      "epoch": 2.44608,
+      "grad_norm": 0.6622048616409302,
+      "learning_rate": 4.619647859143658e-05,
+      "loss": 0.6964,
+      "step": 1911
+    },
+    {
+      "epoch": 2.4473599999999998,
+      "grad_norm": 0.6323049664497375,
+      "learning_rate": 4.619447779111645e-05,
+      "loss": 0.7194,
+      "step": 1912
+    },
+    {
+      "epoch": 2.44864,
+      "grad_norm": 0.6646015644073486,
+      "learning_rate": 4.6192476990796315e-05,
+      "loss": 0.6881,
+      "step": 1913
+    },
+    {
+      "epoch": 2.44992,
+      "grad_norm": 0.6655915975570679,
+      "learning_rate": 4.6190476190476194e-05,
+      "loss": 0.6974,
+      "step": 1914
+    },
+    {
+      "epoch": 2.4512,
+      "grad_norm": 0.6498141288757324,
+      "learning_rate": 4.6188475390156066e-05,
+      "loss": 0.6626,
+      "step": 1915
+    },
+    {
+      "epoch": 2.45248,
+      "grad_norm": 0.6257582306861877,
+      "learning_rate": 4.618647458983594e-05,
+      "loss": 0.6945,
+      "step": 1916
+    },
+    {
+      "epoch": 2.45376,
+      "grad_norm": 0.6620806455612183,
+      "learning_rate": 4.618447378951581e-05,
+      "loss": 0.6976,
+      "step": 1917
+    },
+    {
+      "epoch": 2.45504,
+      "grad_norm": 0.6166614890098572,
+      "learning_rate": 4.618247298919568e-05,
+      "loss": 0.7004,
+      "step": 1918
+    },
+    {
+      "epoch": 2.45632,
+      "grad_norm": 0.6367224454879761,
+      "learning_rate": 4.618047218887555e-05,
+      "loss": 0.7098,
+      "step": 1919
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.6246047019958496,
+      "learning_rate": 4.6178471388555425e-05,
+      "loss": 0.6933,
+      "step": 1920
+    },
+    {
+      "epoch": 2.45888,
+      "grad_norm": 0.6195082664489746,
+      "learning_rate": 4.61764705882353e-05,
+      "loss": 0.6455,
+      "step": 1921
+    },
+    {
+      "epoch": 2.46016,
+      "grad_norm": 0.6568025350570679,
+      "learning_rate": 4.617446978791517e-05,
+      "loss": 0.7072,
+      "step": 1922
+    },
+    {
+      "epoch": 2.46144,
+      "grad_norm": 0.6700795292854309,
+      "learning_rate": 4.617246898759504e-05,
+      "loss": 0.7353,
+      "step": 1923
+    },
+    {
+      "epoch": 2.46272,
+      "grad_norm": 0.6941995620727539,
+      "learning_rate": 4.617046818727491e-05,
+      "loss": 0.7014,
+      "step": 1924
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 0.6903062462806702,
+      "learning_rate": 4.6168467386954784e-05,
+      "loss": 0.7458,
+      "step": 1925
+    },
+    {
+      "epoch": 2.46528,
+      "grad_norm": 0.6764419078826904,
+      "learning_rate": 4.6166466586634656e-05,
+      "loss": 0.6934,
+      "step": 1926
+    },
+    {
+      "epoch": 2.46656,
+      "grad_norm": 0.6628318428993225,
+      "learning_rate": 4.616446578631453e-05,
+      "loss": 0.6735,
+      "step": 1927
+    },
+    {
+      "epoch": 2.46784,
+      "grad_norm": 0.6923201084136963,
+      "learning_rate": 4.61624649859944e-05,
+      "loss": 0.7516,
+      "step": 1928
+    },
+    {
+      "epoch": 2.46912,
+      "grad_norm": 0.6832184791564941,
+      "learning_rate": 4.616046418567427e-05,
+      "loss": 0.7106,
+      "step": 1929
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.6620609164237976,
+      "learning_rate": 4.6158463385354144e-05,
+      "loss": 0.6961,
+      "step": 1930
+    },
+    {
+      "epoch": 2.47168,
+      "grad_norm": 0.7005398273468018,
+      "learning_rate": 4.6156462585034015e-05,
+      "loss": 0.7223,
+      "step": 1931
+    },
+    {
+      "epoch": 2.47296,
+      "grad_norm": 0.6928220987319946,
+      "learning_rate": 4.615446178471389e-05,
+      "loss": 0.7758,
+      "step": 1932
+    },
+    {
+      "epoch": 2.47424,
+      "grad_norm": 0.6193888783454895,
+      "learning_rate": 4.615246098439376e-05,
+      "loss": 0.6345,
+      "step": 1933
+    },
+    {
+      "epoch": 2.47552,
+      "grad_norm": 0.6694657802581787,
+      "learning_rate": 4.615046018407363e-05,
+      "loss": 0.6957,
+      "step": 1934
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 0.6846767663955688,
+      "learning_rate": 4.61484593837535e-05,
+      "loss": 0.6667,
+      "step": 1935
+    },
+    {
+      "epoch": 2.47808,
+      "grad_norm": 0.6446174383163452,
+      "learning_rate": 4.6146458583433375e-05,
+      "loss": 0.7216,
+      "step": 1936
+    },
+    {
+      "epoch": 2.47936,
+      "grad_norm": 0.6601637601852417,
+      "learning_rate": 4.6144457783113247e-05,
+      "loss": 0.7262,
+      "step": 1937
+    },
+    {
+      "epoch": 2.48064,
+      "grad_norm": 0.6476382613182068,
+      "learning_rate": 4.614245698279312e-05,
+      "loss": 0.7142,
+      "step": 1938
+    },
+    {
+      "epoch": 2.48192,
+      "grad_norm": 0.6557601094245911,
+      "learning_rate": 4.614045618247299e-05,
+      "loss": 0.7386,
+      "step": 1939
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.6669868230819702,
+      "learning_rate": 4.613845538215286e-05,
+      "loss": 0.7643,
+      "step": 1940
+    },
+    {
+      "epoch": 2.48448,
+      "grad_norm": 0.6425213813781738,
+      "learning_rate": 4.6136454581832734e-05,
+      "loss": 0.643,
+      "step": 1941
+    },
+    {
+      "epoch": 2.48576,
+      "grad_norm": 0.6689165234565735,
+      "learning_rate": 4.613445378151261e-05,
+      "loss": 0.7025,
+      "step": 1942
+    },
+    {
+      "epoch": 2.48704,
+      "grad_norm": 0.6275350451469421,
+      "learning_rate": 4.613245298119248e-05,
+      "loss": 0.6482,
+      "step": 1943
+    },
+    {
+      "epoch": 2.48832,
+      "grad_norm": 0.6056965589523315,
+      "learning_rate": 4.613045218087235e-05,
+      "loss": 0.6807,
+      "step": 1944
+    },
+    {
+      "epoch": 2.4896,
+      "grad_norm": 0.6282046437263489,
+      "learning_rate": 4.612845138055222e-05,
+      "loss": 0.6564,
+      "step": 1945
+    },
+    {
+      "epoch": 2.4908799999999998,
+      "grad_norm": 0.640059769153595,
+      "learning_rate": 4.612645058023209e-05,
+      "loss": 0.6873,
+      "step": 1946
+    },
+    {
+      "epoch": 2.49216,
+      "grad_norm": 0.6487705707550049,
+      "learning_rate": 4.6124449779911965e-05,
+      "loss": 0.6516,
+      "step": 1947
+    },
+    {
+      "epoch": 2.49344,
+      "grad_norm": 0.6214262843132019,
+      "learning_rate": 4.612244897959184e-05,
+      "loss": 0.699,
+      "step": 1948
+    },
+    {
+      "epoch": 2.49472,
+      "grad_norm": 0.6714819073677063,
+      "learning_rate": 4.6120448179271716e-05,
+      "loss": 0.6946,
+      "step": 1949
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.6522096991539001,
+      "learning_rate": 4.611844737895159e-05,
+      "loss": 0.7193,
+      "step": 1950
+    },
+    {
+      "epoch": 2.49728,
+      "grad_norm": 0.6565654277801514,
+      "learning_rate": 4.611644657863145e-05,
+      "loss": 0.7604,
+      "step": 1951
+    },
+    {
+      "epoch": 2.49856,
+      "grad_norm": 0.602493166923523,
+      "learning_rate": 4.6114445778311324e-05,
+      "loss": 0.6623,
+      "step": 1952
+    },
+    {
+      "epoch": 2.49984,
+      "grad_norm": 0.6453181505203247,
+      "learning_rate": 4.6112444977991196e-05,
+      "loss": 0.6349,
+      "step": 1953
+    },
+    {
+      "epoch": 2.5011200000000002,
+      "grad_norm": 0.6643373966217041,
+      "learning_rate": 4.611044417767107e-05,
+      "loss": 0.7352,
+      "step": 1954
+    },
+    {
+      "epoch": 2.5023999999999997,
+      "grad_norm": 0.6704357862472534,
+      "learning_rate": 4.610844337735094e-05,
+      "loss": 0.744,
+      "step": 1955
+    },
+    {
+      "epoch": 2.50368,
+      "grad_norm": 0.6634871363639832,
+      "learning_rate": 4.610644257703082e-05,
+      "loss": 0.6748,
+      "step": 1956
+    },
+    {
+      "epoch": 2.50496,
+      "grad_norm": 0.6585037708282471,
+      "learning_rate": 4.610444177671069e-05,
+      "loss": 0.6893,
+      "step": 1957
+    },
+    {
+      "epoch": 2.50624,
+      "grad_norm": 0.6182425022125244,
+      "learning_rate": 4.610244097639056e-05,
+      "loss": 0.7226,
+      "step": 1958
+    },
+    {
+      "epoch": 2.50752,
+      "grad_norm": 0.6366656422615051,
+      "learning_rate": 4.610044017607043e-05,
+      "loss": 0.6535,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.6848028302192688,
+      "learning_rate": 4.60984393757503e-05,
+      "loss": 0.7085,
+      "step": 1960
+    },
+    {
+      "epoch": 2.51008,
+      "grad_norm": 0.6243534684181213,
+      "learning_rate": 4.609643857543017e-05,
+      "loss": 0.6508,
+      "step": 1961
+    },
+    {
+      "epoch": 2.51136,
+      "grad_norm": 0.6199184656143188,
+      "learning_rate": 4.609443777511004e-05,
+      "loss": 0.6997,
+      "step": 1962
+    },
+    {
+      "epoch": 2.51264,
+      "grad_norm": 0.6194076538085938,
+      "learning_rate": 4.6092436974789915e-05,
+      "loss": 0.6587,
+      "step": 1963
+    },
+    {
+      "epoch": 2.51392,
+      "grad_norm": 0.6427433490753174,
+      "learning_rate": 4.6090436174469793e-05,
+      "loss": 0.6716,
+      "step": 1964
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 0.6448578238487244,
+      "learning_rate": 4.6088435374149665e-05,
+      "loss": 0.7232,
+      "step": 1965
+    },
+    {
+      "epoch": 2.51648,
+      "grad_norm": 0.6704204678535461,
+      "learning_rate": 4.608643457382954e-05,
+      "loss": 0.7486,
+      "step": 1966
+    },
+    {
+      "epoch": 2.51776,
+      "grad_norm": 0.6715102195739746,
+      "learning_rate": 4.60844337735094e-05,
+      "loss": 0.6453,
+      "step": 1967
+    },
+    {
+      "epoch": 2.51904,
+      "grad_norm": 0.7025008201599121,
+      "learning_rate": 4.6082432973189274e-05,
+      "loss": 0.7043,
+      "step": 1968
+    },
+    {
+      "epoch": 2.52032,
+      "grad_norm": 0.689388632774353,
+      "learning_rate": 4.6080432172869146e-05,
+      "loss": 0.686,
+      "step": 1969
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.6979616284370422,
+      "learning_rate": 4.607843137254902e-05,
+      "loss": 0.7104,
+      "step": 1970
+    },
+    {
+      "epoch": 2.52288,
+      "grad_norm": 0.6773265600204468,
+      "learning_rate": 4.6076430572228896e-05,
+      "loss": 0.7321,
+      "step": 1971
+    },
+    {
+      "epoch": 2.52416,
+      "grad_norm": 0.6521515250205994,
+      "learning_rate": 4.607442977190877e-05,
+      "loss": 0.6446,
+      "step": 1972
+    },
+    {
+      "epoch": 2.52544,
+      "grad_norm": 0.6768288612365723,
+      "learning_rate": 4.607242897158864e-05,
+      "loss": 0.672,
+      "step": 1973
+    },
+    {
+      "epoch": 2.52672,
+      "grad_norm": 0.6806971430778503,
+      "learning_rate": 4.607042817126851e-05,
+      "loss": 0.7247,
+      "step": 1974
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 0.6461851596832275,
+      "learning_rate": 4.606842737094838e-05,
+      "loss": 0.6563,
+      "step": 1975
+    },
+    {
+      "epoch": 2.52928,
+      "grad_norm": 0.6352160573005676,
+      "learning_rate": 4.606642657062825e-05,
+      "loss": 0.6648,
+      "step": 1976
+    },
+    {
+      "epoch": 2.53056,
+      "grad_norm": 0.6496264338493347,
+      "learning_rate": 4.606442577030812e-05,
+      "loss": 0.7253,
+      "step": 1977
+    },
+    {
+      "epoch": 2.53184,
+      "grad_norm": 0.6366921067237854,
+      "learning_rate": 4.6062424969988e-05,
+      "loss": 0.6862,
+      "step": 1978
+    },
+    {
+      "epoch": 2.5331200000000003,
+      "grad_norm": 0.6239694952964783,
+      "learning_rate": 4.606042416966787e-05,
+      "loss": 0.6554,
+      "step": 1979
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.6517849564552307,
+      "learning_rate": 4.605842336934774e-05,
+      "loss": 0.7166,
+      "step": 1980
+    },
+    {
+      "epoch": 2.53568,
+      "grad_norm": 0.6458687782287598,
+      "learning_rate": 4.6056422569027615e-05,
+      "loss": 0.7159,
+      "step": 1981
+    },
+    {
+      "epoch": 2.53696,
+      "grad_norm": 0.6582731008529663,
+      "learning_rate": 4.605442176870749e-05,
+      "loss": 0.6434,
+      "step": 1982
+    },
+    {
+      "epoch": 2.53824,
+      "grad_norm": 0.6515088081359863,
+      "learning_rate": 4.605242096838735e-05,
+      "loss": 0.7313,
+      "step": 1983
+    },
+    {
+      "epoch": 2.53952,
+      "grad_norm": 0.65621417760849,
+      "learning_rate": 4.6050420168067224e-05,
+      "loss": 0.657,
+      "step": 1984
+    },
+    {
+      "epoch": 2.5408,
+      "grad_norm": 0.6879109740257263,
+      "learning_rate": 4.60484193677471e-05,
+      "loss": 0.6906,
+      "step": 1985
+    },
+    {
+      "epoch": 2.54208,
+      "grad_norm": 0.6400864124298096,
+      "learning_rate": 4.6046418567426974e-05,
+      "loss": 0.703,
+      "step": 1986
+    },
+    {
+      "epoch": 2.54336,
+      "grad_norm": 0.6810969114303589,
+      "learning_rate": 4.6044417767106846e-05,
+      "loss": 0.712,
+      "step": 1987
+    },
+    {
+      "epoch": 2.5446400000000002,
+      "grad_norm": 0.6492802500724792,
+      "learning_rate": 4.604241696678672e-05,
+      "loss": 0.6362,
+      "step": 1988
+    },
+    {
+      "epoch": 2.5459199999999997,
+      "grad_norm": 0.6736019253730774,
+      "learning_rate": 4.604041616646659e-05,
+      "loss": 0.7507,
+      "step": 1989
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.6537976861000061,
+      "learning_rate": 4.603841536614646e-05,
+      "loss": 0.6868,
+      "step": 1990
+    },
+    {
+      "epoch": 2.54848,
+      "grad_norm": 0.6979777812957764,
+      "learning_rate": 4.603641456582633e-05,
+      "loss": 0.7275,
+      "step": 1991
+    },
+    {
+      "epoch": 2.54976,
+      "grad_norm": 0.6660913825035095,
+      "learning_rate": 4.6034413765506205e-05,
+      "loss": 0.6566,
+      "step": 1992
+    },
+    {
+      "epoch": 2.55104,
+      "grad_norm": 0.6853949427604675,
+      "learning_rate": 4.603241296518608e-05,
+      "loss": 0.7594,
+      "step": 1993
+    },
+    {
+      "epoch": 2.55232,
+      "grad_norm": 0.6438176035881042,
+      "learning_rate": 4.603041216486595e-05,
+      "loss": 0.6594,
+      "step": 1994
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 0.6791808009147644,
+      "learning_rate": 4.602841136454582e-05,
+      "loss": 0.6871,
+      "step": 1995
+    },
+    {
+      "epoch": 2.55488,
+      "grad_norm": 0.6433961391448975,
+      "learning_rate": 4.602641056422569e-05,
+      "loss": 0.6616,
+      "step": 1996
+    },
+    {
+      "epoch": 2.55616,
+      "grad_norm": 0.6358078122138977,
+      "learning_rate": 4.6024409763905565e-05,
+      "loss": 0.6676,
+      "step": 1997
+    },
+    {
+      "epoch": 2.55744,
+      "grad_norm": 0.6184052228927612,
+      "learning_rate": 4.6022408963585437e-05,
+      "loss": 0.7061,
+      "step": 1998
+    },
+    {
+      "epoch": 2.55872,
+      "grad_norm": 0.6975054740905762,
+      "learning_rate": 4.602040816326531e-05,
+      "loss": 0.7147,
+      "step": 1999
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7169183492660522,
+      "learning_rate": 4.601840736294518e-05,
+      "loss": 0.7385,
+      "step": 2000
+    },
+    {
+      "epoch": 2.56128,
+      "grad_norm": 0.6839573383331299,
+      "learning_rate": 4.601640656262505e-05,
+      "loss": 0.7901,
+      "step": 2001
+    },
+    {
+      "epoch": 2.56256,
+      "grad_norm": 0.6511650681495667,
+      "learning_rate": 4.6014405762304924e-05,
+      "loss": 0.7281,
+      "step": 2002
+    },
+    {
+      "epoch": 2.56384,
+      "grad_norm": 0.6317888498306274,
+      "learning_rate": 4.6012404961984796e-05,
+      "loss": 0.6668,
+      "step": 2003
+    },
+    {
+      "epoch": 2.56512,
+      "grad_norm": 0.6695280075073242,
+      "learning_rate": 4.601040416166467e-05,
+      "loss": 0.6941,
+      "step": 2004
+    },
+    {
+      "epoch": 2.5664,
+      "grad_norm": 0.6781874895095825,
+      "learning_rate": 4.600840336134454e-05,
+      "loss": 0.7154,
+      "step": 2005
+    },
+    {
+      "epoch": 2.56768,
+      "grad_norm": 0.6626363396644592,
+      "learning_rate": 4.600640256102441e-05,
+      "loss": 0.6992,
+      "step": 2006
+    },
+    {
+      "epoch": 2.56896,
+      "grad_norm": 0.702985405921936,
+      "learning_rate": 4.600440176070428e-05,
+      "loss": 0.6923,
+      "step": 2007
+    },
+    {
+      "epoch": 2.57024,
+      "grad_norm": 0.6602315306663513,
+      "learning_rate": 4.6002400960384155e-05,
+      "loss": 0.6645,
+      "step": 2008
+    },
+    {
+      "epoch": 2.57152,
+      "grad_norm": 0.6675458550453186,
+      "learning_rate": 4.600040016006403e-05,
+      "loss": 0.6893,
+      "step": 2009
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.7174723744392395,
+      "learning_rate": 4.59983993597439e-05,
+      "loss": 0.7271,
+      "step": 2010
+    },
+    {
+      "epoch": 2.57408,
+      "grad_norm": 0.7206019163131714,
+      "learning_rate": 4.599639855942377e-05,
+      "loss": 0.711,
+      "step": 2011
+    },
+    {
+      "epoch": 2.57536,
+      "grad_norm": 0.6554808616638184,
+      "learning_rate": 4.599439775910364e-05,
+      "loss": 0.6322,
+      "step": 2012
+    },
+    {
+      "epoch": 2.5766400000000003,
+      "grad_norm": 0.6562361717224121,
+      "learning_rate": 4.599239695878352e-05,
+      "loss": 0.7233,
+      "step": 2013
+    },
+    {
+      "epoch": 2.5779199999999998,
+      "grad_norm": 0.6553013920783997,
+      "learning_rate": 4.5990396158463386e-05,
+      "loss": 0.6693,
+      "step": 2014
+    },
+    {
+      "epoch": 2.5792,
+      "grad_norm": 0.651155948638916,
+      "learning_rate": 4.598839535814326e-05,
+      "loss": 0.7005,
+      "step": 2015
+    },
+    {
+      "epoch": 2.58048,
+      "grad_norm": 0.6626810431480408,
+      "learning_rate": 4.598639455782313e-05,
+      "loss": 0.6962,
+      "step": 2016
+    },
+    {
+      "epoch": 2.58176,
+      "grad_norm": 0.6690220832824707,
+      "learning_rate": 4.5984393757503e-05,
+      "loss": 0.6897,
+      "step": 2017
+    },
+    {
+      "epoch": 2.58304,
+      "grad_norm": 0.6718343496322632,
+      "learning_rate": 4.5982392957182874e-05,
+      "loss": 0.7071,
+      "step": 2018
+    },
+    {
+      "epoch": 2.58432,
+      "grad_norm": 0.6581987142562866,
+      "learning_rate": 4.5980392156862746e-05,
+      "loss": 0.682,
+      "step": 2019
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.6765472888946533,
+      "learning_rate": 4.5978391356542624e-05,
+      "loss": 0.7696,
+      "step": 2020
+    },
+    {
+      "epoch": 2.58688,
+      "grad_norm": 0.630276083946228,
+      "learning_rate": 4.5976390556222496e-05,
+      "loss": 0.6681,
+      "step": 2021
+    },
+    {
+      "epoch": 2.5881600000000002,
+      "grad_norm": 0.6415339112281799,
+      "learning_rate": 4.597438975590236e-05,
+      "loss": 0.6739,
+      "step": 2022
+    },
+    {
+      "epoch": 2.5894399999999997,
+      "grad_norm": 0.6534903645515442,
+      "learning_rate": 4.597238895558223e-05,
+      "loss": 0.6743,
+      "step": 2023
+    },
+    {
+      "epoch": 2.59072,
+      "grad_norm": 0.704707145690918,
+      "learning_rate": 4.5970388155262105e-05,
+      "loss": 0.749,
+      "step": 2024
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.6433389782905579,
+      "learning_rate": 4.596838735494198e-05,
+      "loss": 0.6899,
+      "step": 2025
+    },
+    {
+      "epoch": 2.59328,
+      "grad_norm": 0.6151252388954163,
+      "learning_rate": 4.596638655462185e-05,
+      "loss": 0.6721,
+      "step": 2026
+    },
+    {
+      "epoch": 2.59456,
+      "grad_norm": 0.6191518902778625,
+      "learning_rate": 4.596438575430173e-05,
+      "loss": 0.7129,
+      "step": 2027
+    },
+    {
+      "epoch": 2.59584,
+      "grad_norm": 0.6529395580291748,
+      "learning_rate": 4.59623849539816e-05,
+      "loss": 0.6866,
+      "step": 2028
+    },
+    {
+      "epoch": 2.59712,
+      "grad_norm": 0.6465093493461609,
+      "learning_rate": 4.596038415366147e-05,
+      "loss": 0.6547,
+      "step": 2029
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.6575112342834473,
+      "learning_rate": 4.5958383353341336e-05,
+      "loss": 0.6982,
+      "step": 2030
+    },
+    {
+      "epoch": 2.59968,
+      "grad_norm": 0.6782650947570801,
+      "learning_rate": 4.595638255302121e-05,
+      "loss": 0.7034,
+      "step": 2031
+    },
+    {
+      "epoch": 2.60096,
+      "grad_norm": 0.7053045034408569,
+      "learning_rate": 4.595438175270108e-05,
+      "loss": 0.7387,
+      "step": 2032
+    },
+    {
+      "epoch": 2.60224,
+      "grad_norm": 0.6682645678520203,
+      "learning_rate": 4.595238095238095e-05,
+      "loss": 0.6812,
+      "step": 2033
+    },
+    {
+      "epoch": 2.60352,
+      "grad_norm": 0.6382338404655457,
+      "learning_rate": 4.595038015206083e-05,
+      "loss": 0.6417,
+      "step": 2034
+    },
+    {
+      "epoch": 2.6048,
+      "grad_norm": 0.6588988900184631,
+      "learning_rate": 4.59483793517407e-05,
+      "loss": 0.7028,
+      "step": 2035
+    },
+    {
+      "epoch": 2.60608,
+      "grad_norm": 0.6959824562072754,
+      "learning_rate": 4.5946378551420574e-05,
+      "loss": 0.7239,
+      "step": 2036
+    },
+    {
+      "epoch": 2.60736,
+      "grad_norm": 0.6393638253211975,
+      "learning_rate": 4.5944377751100446e-05,
+      "loss": 0.7437,
+      "step": 2037
+    },
+    {
+      "epoch": 2.60864,
+      "grad_norm": 0.6316011548042297,
+      "learning_rate": 4.594237695078031e-05,
+      "loss": 0.7059,
+      "step": 2038
+    },
+    {
+      "epoch": 2.60992,
+      "grad_norm": 0.6398905515670776,
+      "learning_rate": 4.594037615046018e-05,
+      "loss": 0.692,
+      "step": 2039
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.6583664417266846,
+      "learning_rate": 4.5938375350140055e-05,
+      "loss": 0.6969,
+      "step": 2040
+    },
+    {
+      "epoch": 2.61248,
+      "grad_norm": 0.6351532340049744,
+      "learning_rate": 4.593637454981993e-05,
+      "loss": 0.6436,
+      "step": 2041
+    },
+    {
+      "epoch": 2.61376,
+      "grad_norm": 0.6414986252784729,
+      "learning_rate": 4.5934373749499805e-05,
+      "loss": 0.6872,
+      "step": 2042
+    },
+    {
+      "epoch": 2.61504,
+      "grad_norm": 0.632988452911377,
+      "learning_rate": 4.593237294917968e-05,
+      "loss": 0.7222,
+      "step": 2043
+    },
+    {
+      "epoch": 2.61632,
+      "grad_norm": 0.6204145550727844,
+      "learning_rate": 4.593037214885955e-05,
+      "loss": 0.653,
+      "step": 2044
+    },
+    {
+      "epoch": 2.6176,
+      "grad_norm": 0.6726123094558716,
+      "learning_rate": 4.592837134853942e-05,
+      "loss": 0.7185,
+      "step": 2045
+    },
+    {
+      "epoch": 2.61888,
+      "grad_norm": 0.6624434590339661,
+      "learning_rate": 4.5926370548219286e-05,
+      "loss": 0.6756,
+      "step": 2046
+    },
+    {
+      "epoch": 2.6201600000000003,
+      "grad_norm": 0.66437166929245,
+      "learning_rate": 4.592436974789916e-05,
+      "loss": 0.6542,
+      "step": 2047
+    },
+    {
+      "epoch": 2.6214399999999998,
+      "grad_norm": 0.6689863801002502,
+      "learning_rate": 4.5922368947579036e-05,
+      "loss": 0.6601,
+      "step": 2048
+    },
+    {
+      "epoch": 2.62272,
+      "grad_norm": 0.6922819018363953,
+      "learning_rate": 4.592036814725891e-05,
+      "loss": 0.7033,
+      "step": 2049
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.6346952319145203,
+      "learning_rate": 4.591836734693878e-05,
+      "loss": 0.6276,
+      "step": 2050
+    },
+    {
+      "epoch": 2.62528,
+      "grad_norm": 0.6946719884872437,
+      "learning_rate": 4.591636654661865e-05,
+      "loss": 0.6687,
+      "step": 2051
+    },
+    {
+      "epoch": 2.62656,
+      "grad_norm": 0.6633927226066589,
+      "learning_rate": 4.5914365746298524e-05,
+      "loss": 0.6556,
+      "step": 2052
+    },
+    {
+      "epoch": 2.62784,
+      "grad_norm": 0.6315405964851379,
+      "learning_rate": 4.5912364945978395e-05,
+      "loss": 0.6286,
+      "step": 2053
+    },
+    {
+      "epoch": 2.62912,
+      "grad_norm": 0.6600960493087769,
+      "learning_rate": 4.591036414565826e-05,
+      "loss": 0.7148,
+      "step": 2054
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 0.6884539723396301,
+      "learning_rate": 4.590836334533814e-05,
+      "loss": 0.7345,
+      "step": 2055
+    },
+    {
+      "epoch": 2.6316800000000002,
+      "grad_norm": 0.6405014395713806,
+      "learning_rate": 4.590636254501801e-05,
+      "loss": 0.6786,
+      "step": 2056
+    },
+    {
+      "epoch": 2.6329599999999997,
+      "grad_norm": 0.6440674662590027,
+      "learning_rate": 4.590436174469788e-05,
+      "loss": 0.7232,
+      "step": 2057
+    },
+    {
+      "epoch": 2.63424,
+      "grad_norm": 0.6079061627388,
+      "learning_rate": 4.5902360944377755e-05,
+      "loss": 0.6926,
+      "step": 2058
+    },
+    {
+      "epoch": 2.63552,
+      "grad_norm": 0.6500112414360046,
+      "learning_rate": 4.590036014405763e-05,
+      "loss": 0.6942,
+      "step": 2059
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.6738446950912476,
+      "learning_rate": 4.58983593437375e-05,
+      "loss": 0.7175,
+      "step": 2060
+    },
+    {
+      "epoch": 2.63808,
+      "grad_norm": 0.6589712500572205,
+      "learning_rate": 4.589635854341737e-05,
+      "loss": 0.6648,
+      "step": 2061
+    },
+    {
+      "epoch": 2.63936,
+      "grad_norm": 0.6929957866668701,
+      "learning_rate": 4.589435774309724e-05,
+      "loss": 0.7793,
+      "step": 2062
+    },
+    {
+      "epoch": 2.64064,
+      "grad_norm": 0.6504958271980286,
+      "learning_rate": 4.5892356942777114e-05,
+      "loss": 0.6443,
+      "step": 2063
+    },
+    {
+      "epoch": 2.64192,
+      "grad_norm": 0.6380278468132019,
+      "learning_rate": 4.5890356142456986e-05,
+      "loss": 0.7045,
+      "step": 2064
+    },
+    {
+      "epoch": 2.6432,
+      "grad_norm": 0.6529109477996826,
+      "learning_rate": 4.588835534213686e-05,
+      "loss": 0.6324,
+      "step": 2065
+    },
+    {
+      "epoch": 2.64448,
+      "grad_norm": 0.647123396396637,
+      "learning_rate": 4.588635454181673e-05,
+      "loss": 0.6828,
+      "step": 2066
+    },
+    {
+      "epoch": 2.64576,
+      "grad_norm": 0.6754167079925537,
+      "learning_rate": 4.58843537414966e-05,
+      "loss": 0.6967,
+      "step": 2067
+    },
+    {
+      "epoch": 2.64704,
+      "grad_norm": 0.7286065220832825,
+      "learning_rate": 4.588235294117647e-05,
+      "loss": 0.7515,
+      "step": 2068
+    },
+    {
+      "epoch": 2.64832,
+      "grad_norm": 0.6146422624588013,
+      "learning_rate": 4.5880352140856345e-05,
+      "loss": 0.6442,
+      "step": 2069
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.6109498143196106,
+      "learning_rate": 4.587835134053622e-05,
+      "loss": 0.6513,
+      "step": 2070
+    },
+    {
+      "epoch": 2.65088,
+      "grad_norm": 0.6009126305580139,
+      "learning_rate": 4.587635054021609e-05,
+      "loss": 0.6343,
+      "step": 2071
+    },
+    {
+      "epoch": 2.65216,
+      "grad_norm": 0.6297327280044556,
+      "learning_rate": 4.587434973989596e-05,
+      "loss": 0.6925,
+      "step": 2072
+    },
+    {
+      "epoch": 2.65344,
+      "grad_norm": 0.6399157047271729,
+      "learning_rate": 4.587234893957583e-05,
+      "loss": 0.6359,
+      "step": 2073
+    },
+    {
+      "epoch": 2.65472,
+      "grad_norm": 0.626350998878479,
+      "learning_rate": 4.5870348139255704e-05,
+      "loss": 0.6937,
+      "step": 2074
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 0.6313692927360535,
+      "learning_rate": 4.5868347338935576e-05,
+      "loss": 0.644,
+      "step": 2075
+    },
+    {
+      "epoch": 2.65728,
+      "grad_norm": 0.671847403049469,
+      "learning_rate": 4.586634653861545e-05,
+      "loss": 0.7126,
+      "step": 2076
+    },
+    {
+      "epoch": 2.65856,
+      "grad_norm": 0.688568651676178,
+      "learning_rate": 4.586434573829532e-05,
+      "loss": 0.6857,
+      "step": 2077
+    },
+    {
+      "epoch": 2.65984,
+      "grad_norm": 0.6881417036056519,
+      "learning_rate": 4.586234493797519e-05,
+      "loss": 0.7783,
+      "step": 2078
+    },
+    {
+      "epoch": 2.66112,
+      "grad_norm": 0.648615300655365,
+      "learning_rate": 4.5860344137655064e-05,
+      "loss": 0.6637,
+      "step": 2079
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.6464314460754395,
+      "learning_rate": 4.5858343337334936e-05,
+      "loss": 0.6959,
+      "step": 2080
+    },
+    {
+      "epoch": 2.6636800000000003,
+      "grad_norm": 0.6369978189468384,
+      "learning_rate": 4.585634253701481e-05,
+      "loss": 0.6529,
+      "step": 2081
+    },
+    {
+      "epoch": 2.6649599999999998,
+      "grad_norm": 0.6303538084030151,
+      "learning_rate": 4.585434173669468e-05,
+      "loss": 0.6863,
+      "step": 2082
+    },
+    {
+      "epoch": 2.66624,
+      "grad_norm": 0.6283568143844604,
+      "learning_rate": 4.585234093637455e-05,
+      "loss": 0.6879,
+      "step": 2083
+    },
+    {
+      "epoch": 2.66752,
+      "grad_norm": 0.6396946310997009,
+      "learning_rate": 4.585034013605442e-05,
+      "loss": 0.6658,
+      "step": 2084
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 0.6821319460868835,
+      "learning_rate": 4.5848339335734295e-05,
+      "loss": 0.7968,
+      "step": 2085
+    },
+    {
+      "epoch": 2.67008,
+      "grad_norm": 0.6359503865242004,
+      "learning_rate": 4.584633853541417e-05,
+      "loss": 0.6889,
+      "step": 2086
+    },
+    {
+      "epoch": 2.67136,
+      "grad_norm": 0.6165139079093933,
+      "learning_rate": 4.584433773509404e-05,
+      "loss": 0.641,
+      "step": 2087
+    },
+    {
+      "epoch": 2.67264,
+      "grad_norm": 0.6269519329071045,
+      "learning_rate": 4.584233693477391e-05,
+      "loss": 0.6457,
+      "step": 2088
+    },
+    {
+      "epoch": 2.67392,
+      "grad_norm": 0.6549934148788452,
+      "learning_rate": 4.584033613445378e-05,
+      "loss": 0.7314,
+      "step": 2089
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.6697399616241455,
+      "learning_rate": 4.5838335334133654e-05,
+      "loss": 0.701,
+      "step": 2090
+    },
+    {
+      "epoch": 2.6764799999999997,
+      "grad_norm": 0.6521748304367065,
+      "learning_rate": 4.583633453381353e-05,
+      "loss": 0.6884,
+      "step": 2091
+    },
+    {
+      "epoch": 2.67776,
+      "grad_norm": 0.6441870331764221,
+      "learning_rate": 4.58343337334934e-05,
+      "loss": 0.6971,
+      "step": 2092
+    },
+    {
+      "epoch": 2.67904,
+      "grad_norm": 0.679710328578949,
+      "learning_rate": 4.583233293317327e-05,
+      "loss": 0.7295,
+      "step": 2093
+    },
+    {
+      "epoch": 2.68032,
+      "grad_norm": 0.6788901686668396,
+      "learning_rate": 4.583033213285314e-05,
+      "loss": 0.712,
+      "step": 2094
+    },
+    {
+      "epoch": 2.6816,
+      "grad_norm": 0.6429007053375244,
+      "learning_rate": 4.5828331332533013e-05,
+      "loss": 0.7209,
+      "step": 2095
+    },
+    {
+      "epoch": 2.68288,
+      "grad_norm": 0.6366958022117615,
+      "learning_rate": 4.5826330532212885e-05,
+      "loss": 0.6718,
+      "step": 2096
+    },
+    {
+      "epoch": 2.68416,
+      "grad_norm": 0.6946751475334167,
+      "learning_rate": 4.582432973189276e-05,
+      "loss": 0.7622,
+      "step": 2097
+    },
+    {
+      "epoch": 2.68544,
+      "grad_norm": 0.6607045531272888,
+      "learning_rate": 4.5822328931572636e-05,
+      "loss": 0.697,
+      "step": 2098
+    },
+    {
+      "epoch": 2.68672,
+      "grad_norm": 0.6583904027938843,
+      "learning_rate": 4.582032813125251e-05,
+      "loss": 0.7422,
+      "step": 2099
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.6572896838188171,
+      "learning_rate": 4.581832733093237e-05,
+      "loss": 0.7195,
+      "step": 2100
+    },
+    {
+      "epoch": 2.68928,
+      "grad_norm": 0.6300676465034485,
+      "learning_rate": 4.5816326530612245e-05,
+      "loss": 0.6429,
+      "step": 2101
+    },
+    {
+      "epoch": 2.69056,
+      "grad_norm": 0.6641280651092529,
+      "learning_rate": 4.5814325730292116e-05,
+      "loss": 0.7224,
+      "step": 2102
+    },
+    {
+      "epoch": 2.69184,
+      "grad_norm": 0.6447017788887024,
+      "learning_rate": 4.581232492997199e-05,
+      "loss": 0.7085,
+      "step": 2103
+    },
+    {
+      "epoch": 2.69312,
+      "grad_norm": 0.683560848236084,
+      "learning_rate": 4.581032412965186e-05,
+      "loss": 0.7218,
+      "step": 2104
+    },
+    {
+      "epoch": 2.6944,
+      "grad_norm": 0.6947901248931885,
+      "learning_rate": 4.580832332933174e-05,
+      "loss": 0.6343,
+      "step": 2105
+    },
+    {
+      "epoch": 2.69568,
+      "grad_norm": 0.6173972487449646,
+      "learning_rate": 4.580632252901161e-05,
+      "loss": 0.7048,
+      "step": 2106
+    },
+    {
+      "epoch": 2.69696,
+      "grad_norm": 0.6312255263328552,
+      "learning_rate": 4.580432172869148e-05,
+      "loss": 0.6644,
+      "step": 2107
+    },
+    {
+      "epoch": 2.69824,
+      "grad_norm": 0.6200210452079773,
+      "learning_rate": 4.580232092837135e-05,
+      "loss": 0.688,
+      "step": 2108
+    },
+    {
+      "epoch": 2.69952,
+      "grad_norm": 0.6290738582611084,
+      "learning_rate": 4.580032012805122e-05,
+      "loss": 0.6392,
+      "step": 2109
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.6984908580780029,
+      "learning_rate": 4.579831932773109e-05,
+      "loss": 0.7356,
+      "step": 2110
+    },
+    {
+      "epoch": 2.70208,
+      "grad_norm": 0.6521112322807312,
+      "learning_rate": 4.579631852741096e-05,
+      "loss": 0.6927,
+      "step": 2111
+    },
+    {
+      "epoch": 2.70336,
+      "grad_norm": 0.622561514377594,
+      "learning_rate": 4.579431772709084e-05,
+      "loss": 0.6233,
+      "step": 2112
+    },
+    {
+      "epoch": 2.70464,
+      "grad_norm": 0.6508548259735107,
+      "learning_rate": 4.5792316926770714e-05,
+      "loss": 0.678,
+      "step": 2113
+    },
+    {
+      "epoch": 2.70592,
+      "grad_norm": 0.6437860131263733,
+      "learning_rate": 4.5790316126450586e-05,
+      "loss": 0.6495,
+      "step": 2114
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 0.6867465376853943,
+      "learning_rate": 4.578831532613046e-05,
+      "loss": 0.6763,
+      "step": 2115
+    },
+    {
+      "epoch": 2.7084799999999998,
+      "grad_norm": 0.6844937801361084,
+      "learning_rate": 4.578631452581032e-05,
+      "loss": 0.7097,
+      "step": 2116
+    },
+    {
+      "epoch": 2.70976,
+      "grad_norm": 0.647807240486145,
+      "learning_rate": 4.5784313725490194e-05,
+      "loss": 0.7127,
+      "step": 2117
+    },
+    {
+      "epoch": 2.71104,
+      "grad_norm": 0.6669386625289917,
+      "learning_rate": 4.5782312925170066e-05,
+      "loss": 0.6953,
+      "step": 2118
+    },
+    {
+      "epoch": 2.71232,
+      "grad_norm": 0.6261999607086182,
+      "learning_rate": 4.5780312124849945e-05,
+      "loss": 0.623,
+      "step": 2119
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.655799925327301,
+      "learning_rate": 4.577831132452982e-05,
+      "loss": 0.6649,
+      "step": 2120
+    },
+    {
+      "epoch": 2.71488,
+      "grad_norm": 0.5952524542808533,
+      "learning_rate": 4.577631052420969e-05,
+      "loss": 0.6339,
+      "step": 2121
+    },
+    {
+      "epoch": 2.71616,
+      "grad_norm": 0.6749306321144104,
+      "learning_rate": 4.577430972388956e-05,
+      "loss": 0.7158,
+      "step": 2122
+    },
+    {
+      "epoch": 2.71744,
+      "grad_norm": 0.6410848498344421,
+      "learning_rate": 4.577230892356943e-05,
+      "loss": 0.6914,
+      "step": 2123
+    },
+    {
+      "epoch": 2.7187200000000002,
+      "grad_norm": 0.63245689868927,
+      "learning_rate": 4.57703081232493e-05,
+      "loss": 0.7241,
+      "step": 2124
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.6515913009643555,
+      "learning_rate": 4.576830732292917e-05,
+      "loss": 0.7299,
+      "step": 2125
+    },
+    {
+      "epoch": 2.72128,
+      "grad_norm": 0.6468390226364136,
+      "learning_rate": 4.576630652260905e-05,
+      "loss": 0.6929,
+      "step": 2126
+    },
+    {
+      "epoch": 2.72256,
+      "grad_norm": 0.6881574988365173,
+      "learning_rate": 4.576430572228892e-05,
+      "loss": 0.7325,
+      "step": 2127
+    },
+    {
+      "epoch": 2.72384,
+      "grad_norm": 0.6494055986404419,
+      "learning_rate": 4.576230492196879e-05,
+      "loss": 0.6923,
+      "step": 2128
+    },
+    {
+      "epoch": 2.72512,
+      "grad_norm": 0.6755759119987488,
+      "learning_rate": 4.576030412164866e-05,
+      "loss": 0.6909,
+      "step": 2129
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.6174530982971191,
+      "learning_rate": 4.5758303321328535e-05,
+      "loss": 0.6562,
+      "step": 2130
+    },
+    {
+      "epoch": 2.72768,
+      "grad_norm": 0.7030787467956543,
+      "learning_rate": 4.575630252100841e-05,
+      "loss": 0.6916,
+      "step": 2131
+    },
+    {
+      "epoch": 2.72896,
+      "grad_norm": 0.6742822527885437,
+      "learning_rate": 4.575430172068827e-05,
+      "loss": 0.6501,
+      "step": 2132
+    },
+    {
+      "epoch": 2.7302400000000002,
+      "grad_norm": 0.6416477560997009,
+      "learning_rate": 4.575230092036815e-05,
+      "loss": 0.7267,
+      "step": 2133
+    },
+    {
+      "epoch": 2.7315199999999997,
+      "grad_norm": 0.664212703704834,
+      "learning_rate": 4.575030012004802e-05,
+      "loss": 0.7052,
+      "step": 2134
+    },
+    {
+      "epoch": 2.7328,
+      "grad_norm": 0.6820107102394104,
+      "learning_rate": 4.5748299319727895e-05,
+      "loss": 0.7232,
+      "step": 2135
+    },
+    {
+      "epoch": 2.73408,
+      "grad_norm": 0.6734192967414856,
+      "learning_rate": 4.5746298519407766e-05,
+      "loss": 0.7513,
+      "step": 2136
+    },
+    {
+      "epoch": 2.73536,
+      "grad_norm": 0.6329904794692993,
+      "learning_rate": 4.574429771908764e-05,
+      "loss": 0.5988,
+      "step": 2137
+    },
+    {
+      "epoch": 2.73664,
+      "grad_norm": 0.6895211338996887,
+      "learning_rate": 4.574229691876751e-05,
+      "loss": 0.7194,
+      "step": 2138
+    },
+    {
+      "epoch": 2.73792,
+      "grad_norm": 0.6408979892730713,
+      "learning_rate": 4.574029611844738e-05,
+      "loss": 0.6905,
+      "step": 2139
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.6711249947547913,
+      "learning_rate": 4.5738295318127254e-05,
+      "loss": 0.7265,
+      "step": 2140
+    },
+    {
+      "epoch": 2.74048,
+      "grad_norm": 0.6658822894096375,
+      "learning_rate": 4.5736294517807126e-05,
+      "loss": 0.6549,
+      "step": 2141
+    },
+    {
+      "epoch": 2.74176,
+      "grad_norm": 0.6828049421310425,
+      "learning_rate": 4.5734293717487e-05,
+      "loss": 0.6709,
+      "step": 2142
+    },
+    {
+      "epoch": 2.74304,
+      "grad_norm": 0.6909437775611877,
+      "learning_rate": 4.573229291716687e-05,
+      "loss": 0.7724,
+      "step": 2143
+    },
+    {
+      "epoch": 2.74432,
+      "grad_norm": 0.639430582523346,
+      "learning_rate": 4.573029211684674e-05,
+      "loss": 0.6728,
+      "step": 2144
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 0.6498085856437683,
+      "learning_rate": 4.572829131652661e-05,
+      "loss": 0.7088,
+      "step": 2145
+    },
+    {
+      "epoch": 2.74688,
+      "grad_norm": 0.6685836911201477,
+      "learning_rate": 4.5726290516206485e-05,
+      "loss": 0.6646,
+      "step": 2146
+    },
+    {
+      "epoch": 2.74816,
+      "grad_norm": 0.6966959834098816,
+      "learning_rate": 4.572428971588636e-05,
+      "loss": 0.7059,
+      "step": 2147
+    },
+    {
+      "epoch": 2.74944,
+      "grad_norm": 0.6679326891899109,
+      "learning_rate": 4.572228891556623e-05,
+      "loss": 0.6814,
+      "step": 2148
+    },
+    {
+      "epoch": 2.7507200000000003,
+      "grad_norm": 0.675516664981842,
+      "learning_rate": 4.57202881152461e-05,
+      "loss": 0.8011,
+      "step": 2149
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.661491334438324,
+      "learning_rate": 4.571828731492597e-05,
+      "loss": 0.6971,
+      "step": 2150
+    },
+    {
+      "epoch": 2.75328,
+      "grad_norm": 0.6797171831130981,
+      "learning_rate": 4.5716286514605844e-05,
+      "loss": 0.6434,
+      "step": 2151
+    },
+    {
+      "epoch": 2.75456,
+      "grad_norm": 0.6701538562774658,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.6716,
+      "step": 2152
+    },
+    {
+      "epoch": 2.75584,
+      "grad_norm": 0.651821494102478,
+      "learning_rate": 4.571228491396559e-05,
+      "loss": 0.6881,
+      "step": 2153
+    },
+    {
+      "epoch": 2.75712,
+      "grad_norm": 0.6361088156700134,
+      "learning_rate": 4.571028411364546e-05,
+      "loss": 0.6827,
+      "step": 2154
+    },
+    {
+      "epoch": 2.7584,
+      "grad_norm": 0.6350401043891907,
+      "learning_rate": 4.570828331332533e-05,
+      "loss": 0.7337,
+      "step": 2155
+    },
+    {
+      "epoch": 2.75968,
+      "grad_norm": 0.7136991024017334,
+      "learning_rate": 4.5706282513005204e-05,
+      "loss": 0.6458,
+      "step": 2156
+    },
+    {
+      "epoch": 2.76096,
+      "grad_norm": 0.6868817210197449,
+      "learning_rate": 4.5704281712685075e-05,
+      "loss": 0.7136,
+      "step": 2157
+    },
+    {
+      "epoch": 2.7622400000000003,
+      "grad_norm": 0.676253616809845,
+      "learning_rate": 4.570228091236495e-05,
+      "loss": 0.7317,
+      "step": 2158
+    },
+    {
+      "epoch": 2.7635199999999998,
+      "grad_norm": 0.6780885457992554,
+      "learning_rate": 4.570028011204482e-05,
+      "loss": 0.7259,
+      "step": 2159
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.6793497204780579,
+      "learning_rate": 4.569827931172469e-05,
+      "loss": 0.7271,
+      "step": 2160
+    },
+    {
+      "epoch": 2.76608,
+      "grad_norm": 0.6723445653915405,
+      "learning_rate": 4.569627851140457e-05,
+      "loss": 0.6884,
+      "step": 2161
+    },
+    {
+      "epoch": 2.76736,
+      "grad_norm": 0.634547233581543,
+      "learning_rate": 4.5694277711084435e-05,
+      "loss": 0.6513,
+      "step": 2162
+    },
+    {
+      "epoch": 2.76864,
+      "grad_norm": 0.6849969625473022,
+      "learning_rate": 4.5692276910764307e-05,
+      "loss": 0.742,
+      "step": 2163
+    },
+    {
+      "epoch": 2.76992,
+      "grad_norm": 0.678898274898529,
+      "learning_rate": 4.569027611044418e-05,
+      "loss": 0.6751,
+      "step": 2164
+    },
+    {
+      "epoch": 2.7712,
+      "grad_norm": 0.6287670135498047,
+      "learning_rate": 4.568827531012405e-05,
+      "loss": 0.6337,
+      "step": 2165
+    },
+    {
+      "epoch": 2.77248,
+      "grad_norm": 0.6324494481086731,
+      "learning_rate": 4.568627450980392e-05,
+      "loss": 0.6311,
+      "step": 2166
+    },
+    {
+      "epoch": 2.7737600000000002,
+      "grad_norm": 0.6523123383522034,
+      "learning_rate": 4.5684273709483794e-05,
+      "loss": 0.6919,
+      "step": 2167
+    },
+    {
+      "epoch": 2.7750399999999997,
+      "grad_norm": 0.6603489518165588,
+      "learning_rate": 4.568227290916367e-05,
+      "loss": 0.6948,
+      "step": 2168
+    },
+    {
+      "epoch": 2.77632,
+      "grad_norm": 0.6603338122367859,
+      "learning_rate": 4.5680272108843544e-05,
+      "loss": 0.6322,
+      "step": 2169
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.6867691874504089,
+      "learning_rate": 4.567827130852341e-05,
+      "loss": 0.6886,
+      "step": 2170
+    },
+    {
+      "epoch": 2.77888,
+      "grad_norm": 0.6577850580215454,
+      "learning_rate": 4.567627050820328e-05,
+      "loss": 0.6591,
+      "step": 2171
+    },
+    {
+      "epoch": 2.78016,
+      "grad_norm": 0.6284435987472534,
+      "learning_rate": 4.567426970788315e-05,
+      "loss": 0.6754,
+      "step": 2172
+    },
+    {
+      "epoch": 2.78144,
+      "grad_norm": 0.6540303826332092,
+      "learning_rate": 4.5672268907563025e-05,
+      "loss": 0.6868,
+      "step": 2173
+    },
+    {
+      "epoch": 2.78272,
+      "grad_norm": 0.6498759984970093,
+      "learning_rate": 4.56702681072429e-05,
+      "loss": 0.7164,
+      "step": 2174
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.686326265335083,
+      "learning_rate": 4.5668267306922776e-05,
+      "loss": 0.7155,
+      "step": 2175
+    },
+    {
+      "epoch": 2.78528,
+      "grad_norm": 0.6437170505523682,
+      "learning_rate": 4.566626650660265e-05,
+      "loss": 0.6718,
+      "step": 2176
+    },
+    {
+      "epoch": 2.78656,
+      "grad_norm": 0.6241775155067444,
+      "learning_rate": 4.566426570628252e-05,
+      "loss": 0.6818,
+      "step": 2177
+    },
+    {
+      "epoch": 2.78784,
+      "grad_norm": 0.6604918837547302,
+      "learning_rate": 4.5662264905962384e-05,
+      "loss": 0.6968,
+      "step": 2178
+    },
+    {
+      "epoch": 2.78912,
+      "grad_norm": 0.6711975932121277,
+      "learning_rate": 4.5660264105642256e-05,
+      "loss": 0.6325,
+      "step": 2179
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.6874549984931946,
+      "learning_rate": 4.565826330532213e-05,
+      "loss": 0.7189,
+      "step": 2180
+    },
+    {
+      "epoch": 2.79168,
+      "grad_norm": 0.6761820316314697,
+      "learning_rate": 4.5656262505002e-05,
+      "loss": 0.6852,
+      "step": 2181
+    },
+    {
+      "epoch": 2.79296,
+      "grad_norm": 0.6783404350280762,
+      "learning_rate": 4.565426170468188e-05,
+      "loss": 0.6999,
+      "step": 2182
+    },
+    {
+      "epoch": 2.79424,
+      "grad_norm": 0.692419171333313,
+      "learning_rate": 4.565226090436175e-05,
+      "loss": 0.6883,
+      "step": 2183
+    },
+    {
+      "epoch": 2.79552,
+      "grad_norm": 0.6205313205718994,
+      "learning_rate": 4.565026010404162e-05,
+      "loss": 0.6789,
+      "step": 2184
+    },
+    {
+      "epoch": 2.7968,
+      "grad_norm": 0.6563310027122498,
+      "learning_rate": 4.5648259303721494e-05,
+      "loss": 0.6559,
+      "step": 2185
+    },
+    {
+      "epoch": 2.79808,
+      "grad_norm": 0.6110013127326965,
+      "learning_rate": 4.564625850340136e-05,
+      "loss": 0.6487,
+      "step": 2186
+    },
+    {
+      "epoch": 2.79936,
+      "grad_norm": 0.6310566067695618,
+      "learning_rate": 4.564425770308123e-05,
+      "loss": 0.6614,
+      "step": 2187
+    },
+    {
+      "epoch": 2.80064,
+      "grad_norm": 0.6590198278427124,
+      "learning_rate": 4.56422569027611e-05,
+      "loss": 0.6885,
+      "step": 2188
+    },
+    {
+      "epoch": 2.80192,
+      "grad_norm": 0.6192741394042969,
+      "learning_rate": 4.5640256102440975e-05,
+      "loss": 0.668,
+      "step": 2189
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.6905452013015747,
+      "learning_rate": 4.5638255302120853e-05,
+      "loss": 0.7039,
+      "step": 2190
+    },
+    {
+      "epoch": 2.80448,
+      "grad_norm": 0.7045333385467529,
+      "learning_rate": 4.5636254501800725e-05,
+      "loss": 0.7192,
+      "step": 2191
+    },
+    {
+      "epoch": 2.8057600000000003,
+      "grad_norm": 0.6958116888999939,
+      "learning_rate": 4.56342537014806e-05,
+      "loss": 0.6825,
+      "step": 2192
+    },
+    {
+      "epoch": 2.8070399999999998,
+      "grad_norm": 0.6681202054023743,
+      "learning_rate": 4.563225290116047e-05,
+      "loss": 0.6615,
+      "step": 2193
+    },
+    {
+      "epoch": 2.80832,
+      "grad_norm": 0.6813097596168518,
+      "learning_rate": 4.5630252100840334e-05,
+      "loss": 0.7027,
+      "step": 2194
+    },
+    {
+      "epoch": 2.8096,
+      "grad_norm": 0.6501699090003967,
+      "learning_rate": 4.5628251300520206e-05,
+      "loss": 0.6703,
+      "step": 2195
+    },
+    {
+      "epoch": 2.81088,
+      "grad_norm": 0.6794613599777222,
+      "learning_rate": 4.562625050020008e-05,
+      "loss": 0.681,
+      "step": 2196
+    },
+    {
+      "epoch": 2.81216,
+      "grad_norm": 0.6721237301826477,
+      "learning_rate": 4.5624249699879956e-05,
+      "loss": 0.6923,
+      "step": 2197
+    },
+    {
+      "epoch": 2.81344,
+      "grad_norm": 0.6417657732963562,
+      "learning_rate": 4.562224889955983e-05,
+      "loss": 0.6285,
+      "step": 2198
+    },
+    {
+      "epoch": 2.81472,
+      "grad_norm": 0.6990517973899841,
+      "learning_rate": 4.56202480992397e-05,
+      "loss": 0.6986,
+      "step": 2199
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.6576136350631714,
+      "learning_rate": 4.561824729891957e-05,
+      "loss": 0.7217,
+      "step": 2200
+    },
+    {
+      "epoch": 2.8172800000000002,
+      "grad_norm": 0.6665319204330444,
+      "learning_rate": 4.5616246498599444e-05,
+      "loss": 0.6833,
+      "step": 2201
+    },
+    {
+      "epoch": 2.8185599999999997,
+      "grad_norm": 0.6458298563957214,
+      "learning_rate": 4.561424569827931e-05,
+      "loss": 0.7176,
+      "step": 2202
+    },
+    {
+      "epoch": 2.81984,
+      "grad_norm": 0.624575674533844,
+      "learning_rate": 4.561224489795918e-05,
+      "loss": 0.6285,
+      "step": 2203
+    },
+    {
+      "epoch": 2.82112,
+      "grad_norm": 0.644964337348938,
+      "learning_rate": 4.561024409763906e-05,
+      "loss": 0.7044,
+      "step": 2204
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 0.6390849947929382,
+      "learning_rate": 4.560824329731893e-05,
+      "loss": 0.7089,
+      "step": 2205
+    },
+    {
+      "epoch": 2.82368,
+      "grad_norm": 0.6157050132751465,
+      "learning_rate": 4.56062424969988e-05,
+      "loss": 0.6307,
+      "step": 2206
+    },
+    {
+      "epoch": 2.82496,
+      "grad_norm": 0.6521828770637512,
+      "learning_rate": 4.5604241696678675e-05,
+      "loss": 0.6278,
+      "step": 2207
+    },
+    {
+      "epoch": 2.82624,
+      "grad_norm": 0.6566979289054871,
+      "learning_rate": 4.560224089635855e-05,
+      "loss": 0.698,
+      "step": 2208
+    },
+    {
+      "epoch": 2.82752,
+      "grad_norm": 0.6705225110054016,
+      "learning_rate": 4.560024009603842e-05,
+      "loss": 0.7031,
+      "step": 2209
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.6485233306884766,
+      "learning_rate": 4.5598239295718284e-05,
+      "loss": 0.6328,
+      "step": 2210
+    },
+    {
+      "epoch": 2.83008,
+      "grad_norm": 0.6989656686782837,
+      "learning_rate": 4.559623849539816e-05,
+      "loss": 0.7367,
+      "step": 2211
+    },
+    {
+      "epoch": 2.83136,
+      "grad_norm": 0.705863356590271,
+      "learning_rate": 4.5594237695078034e-05,
+      "loss": 0.6894,
+      "step": 2212
+    },
+    {
+      "epoch": 2.83264,
+      "grad_norm": 0.65732741355896,
+      "learning_rate": 4.5592236894757906e-05,
+      "loss": 0.7177,
+      "step": 2213
+    },
+    {
+      "epoch": 2.83392,
+      "grad_norm": 0.6545193195343018,
+      "learning_rate": 4.559023609443778e-05,
+      "loss": 0.7393,
+      "step": 2214
+    },
+    {
+      "epoch": 2.8352,
+      "grad_norm": 0.6775088906288147,
+      "learning_rate": 4.558823529411765e-05,
+      "loss": 0.7286,
+      "step": 2215
+    },
+    {
+      "epoch": 2.83648,
+      "grad_norm": 0.7127577662467957,
+      "learning_rate": 4.558623449379752e-05,
+      "loss": 0.6682,
+      "step": 2216
+    },
+    {
+      "epoch": 2.83776,
+      "grad_norm": 0.6672844886779785,
+      "learning_rate": 4.5584233693477394e-05,
+      "loss": 0.6333,
+      "step": 2217
+    },
+    {
+      "epoch": 2.83904,
+      "grad_norm": 0.6709384322166443,
+      "learning_rate": 4.5582232893157265e-05,
+      "loss": 0.6946,
+      "step": 2218
+    },
+    {
+      "epoch": 2.84032,
+      "grad_norm": 0.6919859647750854,
+      "learning_rate": 4.558023209283714e-05,
+      "loss": 0.7289,
+      "step": 2219
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.663409411907196,
+      "learning_rate": 4.557823129251701e-05,
+      "loss": 0.6832,
+      "step": 2220
+    },
+    {
+      "epoch": 2.84288,
+      "grad_norm": 0.7317196130752563,
+      "learning_rate": 4.557623049219688e-05,
+      "loss": 0.7423,
+      "step": 2221
+    },
+    {
+      "epoch": 2.84416,
+      "grad_norm": 0.6240825057029724,
+      "learning_rate": 4.557422969187675e-05,
+      "loss": 0.6365,
+      "step": 2222
+    },
+    {
+      "epoch": 2.84544,
+      "grad_norm": 0.6449801921844482,
+      "learning_rate": 4.5572228891556625e-05,
+      "loss": 0.7191,
+      "step": 2223
+    },
+    {
+      "epoch": 2.84672,
+      "grad_norm": 0.6569290161132812,
+      "learning_rate": 4.5570228091236497e-05,
+      "loss": 0.6668,
+      "step": 2224
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 0.6520342826843262,
+      "learning_rate": 4.556822729091637e-05,
+      "loss": 0.6671,
+      "step": 2225
+    },
+    {
+      "epoch": 2.8492800000000003,
+      "grad_norm": 0.708814263343811,
+      "learning_rate": 4.556622649059624e-05,
+      "loss": 0.7664,
+      "step": 2226
+    },
+    {
+      "epoch": 2.8505599999999998,
+      "grad_norm": 0.6431106925010681,
+      "learning_rate": 4.556422569027611e-05,
+      "loss": 0.6481,
+      "step": 2227
+    },
+    {
+      "epoch": 2.85184,
+      "grad_norm": 0.6694732904434204,
+      "learning_rate": 4.5562224889955984e-05,
+      "loss": 0.6742,
+      "step": 2228
+    },
+    {
+      "epoch": 2.85312,
+      "grad_norm": 0.6356359124183655,
+      "learning_rate": 4.5560224089635856e-05,
+      "loss": 0.6854,
+      "step": 2229
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.6448719501495361,
+      "learning_rate": 4.555822328931573e-05,
+      "loss": 0.7208,
+      "step": 2230
+    },
+    {
+      "epoch": 2.85568,
+      "grad_norm": 0.6238778233528137,
+      "learning_rate": 4.55562224889956e-05,
+      "loss": 0.6774,
+      "step": 2231
+    },
+    {
+      "epoch": 2.85696,
+      "grad_norm": 0.6807205677032471,
+      "learning_rate": 4.555422168867547e-05,
+      "loss": 0.6773,
+      "step": 2232
+    },
+    {
+      "epoch": 2.85824,
+      "grad_norm": 0.6966913938522339,
+      "learning_rate": 4.555222088835534e-05,
+      "loss": 0.7078,
+      "step": 2233
+    },
+    {
+      "epoch": 2.85952,
+      "grad_norm": 0.666556715965271,
+      "learning_rate": 4.5550220088035215e-05,
+      "loss": 0.714,
+      "step": 2234
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 0.6896193027496338,
+      "learning_rate": 4.554821928771509e-05,
+      "loss": 0.7264,
+      "step": 2235
+    },
+    {
+      "epoch": 2.8620799999999997,
+      "grad_norm": 0.6884152293205261,
+      "learning_rate": 4.554621848739496e-05,
+      "loss": 0.7435,
+      "step": 2236
+    },
+    {
+      "epoch": 2.86336,
+      "grad_norm": 0.7019590735435486,
+      "learning_rate": 4.554421768707483e-05,
+      "loss": 0.7466,
+      "step": 2237
+    },
+    {
+      "epoch": 2.86464,
+      "grad_norm": 0.6901242733001709,
+      "learning_rate": 4.55422168867547e-05,
+      "loss": 0.6879,
+      "step": 2238
+    },
+    {
+      "epoch": 2.86592,
+      "grad_norm": 0.6424590349197388,
+      "learning_rate": 4.554021608643458e-05,
+      "loss": 0.6982,
+      "step": 2239
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.6135256290435791,
+      "learning_rate": 4.5538215286114446e-05,
+      "loss": 0.6547,
+      "step": 2240
+    },
+    {
+      "epoch": 2.86848,
+      "grad_norm": 0.6518721580505371,
+      "learning_rate": 4.553621448579432e-05,
+      "loss": 0.6951,
+      "step": 2241
+    },
+    {
+      "epoch": 2.86976,
+      "grad_norm": 0.6382459402084351,
+      "learning_rate": 4.553421368547419e-05,
+      "loss": 0.6453,
+      "step": 2242
+    },
+    {
+      "epoch": 2.87104,
+      "grad_norm": 0.6710212230682373,
+      "learning_rate": 4.553221288515406e-05,
+      "loss": 0.718,
+      "step": 2243
+    },
+    {
+      "epoch": 2.87232,
+      "grad_norm": 0.6573660373687744,
+      "learning_rate": 4.5530212084833934e-05,
+      "loss": 0.6752,
+      "step": 2244
+    },
+    {
+      "epoch": 2.8736,
+      "grad_norm": 0.6997604966163635,
+      "learning_rate": 4.5528211284513806e-05,
+      "loss": 0.7185,
+      "step": 2245
+    },
+    {
+      "epoch": 2.87488,
+      "grad_norm": 0.6322104930877686,
+      "learning_rate": 4.5526210484193684e-05,
+      "loss": 0.6109,
+      "step": 2246
+    },
+    {
+      "epoch": 2.87616,
+      "grad_norm": 0.6722203493118286,
+      "learning_rate": 4.5524209683873556e-05,
+      "loss": 0.702,
+      "step": 2247
+    },
+    {
+      "epoch": 2.87744,
+      "grad_norm": 0.6825948357582092,
+      "learning_rate": 4.552220888355342e-05,
+      "loss": 0.6802,
+      "step": 2248
+    },
+    {
+      "epoch": 2.87872,
+      "grad_norm": 0.6763216853141785,
+      "learning_rate": 4.552020808323329e-05,
+      "loss": 0.7054,
+      "step": 2249
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6940529346466064,
+      "learning_rate": 4.5518207282913165e-05,
+      "loss": 0.7395,
+      "step": 2250
+    },
+    {
+      "epoch": 2.88128,
+      "grad_norm": 0.6334916949272156,
+      "learning_rate": 4.551620648259304e-05,
+      "loss": 0.6658,
+      "step": 2251
+    },
+    {
+      "epoch": 2.88256,
+      "grad_norm": 0.6614584922790527,
+      "learning_rate": 4.551420568227291e-05,
+      "loss": 0.7377,
+      "step": 2252
+    },
+    {
+      "epoch": 2.88384,
+      "grad_norm": 0.6642113924026489,
+      "learning_rate": 4.551220488195279e-05,
+      "loss": 0.6981,
+      "step": 2253
+    },
+    {
+      "epoch": 2.88512,
+      "grad_norm": 0.624686598777771,
+      "learning_rate": 4.551020408163266e-05,
+      "loss": 0.7133,
+      "step": 2254
+    },
+    {
+      "epoch": 2.8864,
+      "grad_norm": 0.67439204454422,
+      "learning_rate": 4.550820328131253e-05,
+      "loss": 0.7079,
+      "step": 2255
+    },
+    {
+      "epoch": 2.88768,
+      "grad_norm": 0.6298664212226868,
+      "learning_rate": 4.5506202480992396e-05,
+      "loss": 0.6848,
+      "step": 2256
+    },
+    {
+      "epoch": 2.88896,
+      "grad_norm": 0.604558527469635,
+      "learning_rate": 4.550420168067227e-05,
+      "loss": 0.6442,
+      "step": 2257
+    },
+    {
+      "epoch": 2.89024,
+      "grad_norm": 0.6485775113105774,
+      "learning_rate": 4.550220088035214e-05,
+      "loss": 0.6867,
+      "step": 2258
+    },
+    {
+      "epoch": 2.89152,
+      "grad_norm": 0.6755024194717407,
+      "learning_rate": 4.550020008003201e-05,
+      "loss": 0.705,
+      "step": 2259
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.664548397064209,
+      "learning_rate": 4.549819927971189e-05,
+      "loss": 0.657,
+      "step": 2260
+    },
+    {
+      "epoch": 2.8940799999999998,
+      "grad_norm": 0.6621407866477966,
+      "learning_rate": 4.549619847939176e-05,
+      "loss": 0.7003,
+      "step": 2261
+    },
+    {
+      "epoch": 2.89536,
+      "grad_norm": 0.6824017763137817,
+      "learning_rate": 4.5494197679071634e-05,
+      "loss": 0.7026,
+      "step": 2262
+    },
+    {
+      "epoch": 2.89664,
+      "grad_norm": 0.6626001000404358,
+      "learning_rate": 4.5492196878751506e-05,
+      "loss": 0.6712,
+      "step": 2263
+    },
+    {
+      "epoch": 2.89792,
+      "grad_norm": 0.6786776781082153,
+      "learning_rate": 4.549019607843137e-05,
+      "loss": 0.6515,
+      "step": 2264
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 0.677880048751831,
+      "learning_rate": 4.548819527811124e-05,
+      "loss": 0.6956,
+      "step": 2265
+    },
+    {
+      "epoch": 2.90048,
+      "grad_norm": 0.6466307640075684,
+      "learning_rate": 4.5486194477791115e-05,
+      "loss": 0.7,
+      "step": 2266
+    },
+    {
+      "epoch": 2.90176,
+      "grad_norm": 0.6700591444969177,
+      "learning_rate": 4.548419367747099e-05,
+      "loss": 0.708,
+      "step": 2267
+    },
+    {
+      "epoch": 2.90304,
+      "grad_norm": 0.6479355096817017,
+      "learning_rate": 4.5482192877150865e-05,
+      "loss": 0.6921,
+      "step": 2268
+    },
+    {
+      "epoch": 2.9043200000000002,
+      "grad_norm": 0.6609050631523132,
+      "learning_rate": 4.548019207683074e-05,
+      "loss": 0.7045,
+      "step": 2269
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.6681724190711975,
+      "learning_rate": 4.547819127651061e-05,
+      "loss": 0.6764,
+      "step": 2270
+    },
+    {
+      "epoch": 2.90688,
+      "grad_norm": 0.6616440415382385,
+      "learning_rate": 4.547619047619048e-05,
+      "loss": 0.6934,
+      "step": 2271
+    },
+    {
+      "epoch": 2.90816,
+      "grad_norm": 0.6577678322792053,
+      "learning_rate": 4.5474189675870346e-05,
+      "loss": 0.7208,
+      "step": 2272
+    },
+    {
+      "epoch": 2.90944,
+      "grad_norm": 0.6378509402275085,
+      "learning_rate": 4.547218887555022e-05,
+      "loss": 0.6296,
+      "step": 2273
+    },
+    {
+      "epoch": 2.91072,
+      "grad_norm": 0.6398113965988159,
+      "learning_rate": 4.5470188075230096e-05,
+      "loss": 0.631,
+      "step": 2274
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 0.6845940947532654,
+      "learning_rate": 4.546818727490997e-05,
+      "loss": 0.7193,
+      "step": 2275
+    },
+    {
+      "epoch": 2.91328,
+      "grad_norm": 0.6516909003257751,
+      "learning_rate": 4.546618647458984e-05,
+      "loss": 0.7033,
+      "step": 2276
+    },
+    {
+      "epoch": 2.91456,
+      "grad_norm": 0.6223090887069702,
+      "learning_rate": 4.546418567426971e-05,
+      "loss": 0.6473,
+      "step": 2277
+    },
+    {
+      "epoch": 2.91584,
+      "grad_norm": 0.6670078635215759,
+      "learning_rate": 4.5462184873949584e-05,
+      "loss": 0.7206,
+      "step": 2278
+    },
+    {
+      "epoch": 2.91712,
+      "grad_norm": 0.6714773774147034,
+      "learning_rate": 4.5460184073629455e-05,
+      "loss": 0.6349,
+      "step": 2279
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.6824267506599426,
+      "learning_rate": 4.545818327330932e-05,
+      "loss": 0.6335,
+      "step": 2280
+    },
+    {
+      "epoch": 2.91968,
+      "grad_norm": 0.7514560222625732,
+      "learning_rate": 4.54561824729892e-05,
+      "loss": 0.7474,
+      "step": 2281
+    },
+    {
+      "epoch": 2.92096,
+      "grad_norm": 0.6569649577140808,
+      "learning_rate": 4.545418167266907e-05,
+      "loss": 0.7029,
+      "step": 2282
+    },
+    {
+      "epoch": 2.92224,
+      "grad_norm": 0.700855553150177,
+      "learning_rate": 4.545218087234894e-05,
+      "loss": 0.7706,
+      "step": 2283
+    },
+    {
+      "epoch": 2.92352,
+      "grad_norm": 0.6456267833709717,
+      "learning_rate": 4.5450180072028815e-05,
+      "loss": 0.7021,
+      "step": 2284
+    },
+    {
+      "epoch": 2.9248,
+      "grad_norm": 0.6533935070037842,
+      "learning_rate": 4.5448179271708687e-05,
+      "loss": 0.6852,
+      "step": 2285
+    },
+    {
+      "epoch": 2.92608,
+      "grad_norm": 0.6358088850975037,
+      "learning_rate": 4.544617847138856e-05,
+      "loss": 0.7173,
+      "step": 2286
+    },
+    {
+      "epoch": 2.92736,
+      "grad_norm": 0.6510772109031677,
+      "learning_rate": 4.544417767106843e-05,
+      "loss": 0.6866,
+      "step": 2287
+    },
+    {
+      "epoch": 2.92864,
+      "grad_norm": 0.642501950263977,
+      "learning_rate": 4.54421768707483e-05,
+      "loss": 0.7078,
+      "step": 2288
+    },
+    {
+      "epoch": 2.92992,
+      "grad_norm": 0.6543188095092773,
+      "learning_rate": 4.5440176070428174e-05,
+      "loss": 0.6925,
+      "step": 2289
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.7026666402816772,
+      "learning_rate": 4.5438175270108046e-05,
+      "loss": 0.7661,
+      "step": 2290
+    },
+    {
+      "epoch": 2.93248,
+      "grad_norm": 0.6860962510108948,
+      "learning_rate": 4.543617446978792e-05,
+      "loss": 0.7165,
+      "step": 2291
+    },
+    {
+      "epoch": 2.93376,
+      "grad_norm": 0.6254853010177612,
+      "learning_rate": 4.543417366946779e-05,
+      "loss": 0.6598,
+      "step": 2292
+    },
+    {
+      "epoch": 2.93504,
+      "grad_norm": 0.6507871150970459,
+      "learning_rate": 4.543217286914766e-05,
+      "loss": 0.6557,
+      "step": 2293
+    },
+    {
+      "epoch": 2.9363200000000003,
+      "grad_norm": 0.6633324027061462,
+      "learning_rate": 4.543017206882753e-05,
+      "loss": 0.6782,
+      "step": 2294
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 0.6420159935951233,
+      "learning_rate": 4.5428171268507405e-05,
+      "loss": 0.7025,
+      "step": 2295
+    },
+    {
+      "epoch": 2.93888,
+      "grad_norm": 0.6702166795730591,
+      "learning_rate": 4.542617046818728e-05,
+      "loss": 0.7024,
+      "step": 2296
+    },
+    {
+      "epoch": 2.94016,
+      "grad_norm": 0.6596609950065613,
+      "learning_rate": 4.542416966786715e-05,
+      "loss": 0.6926,
+      "step": 2297
+    },
+    {
+      "epoch": 2.94144,
+      "grad_norm": 0.6574816703796387,
+      "learning_rate": 4.542216886754702e-05,
+      "loss": 0.6624,
+      "step": 2298
+    },
+    {
+      "epoch": 2.94272,
+      "grad_norm": 0.6869375109672546,
+      "learning_rate": 4.542016806722689e-05,
+      "loss": 0.6966,
+      "step": 2299
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.659354031085968,
+      "learning_rate": 4.5418167266906764e-05,
+      "loss": 0.6866,
+      "step": 2300
+    },
+    {
+      "epoch": 2.94528,
+      "grad_norm": 0.6519255042076111,
+      "learning_rate": 4.5416166466586636e-05,
+      "loss": 0.6968,
+      "step": 2301
+    },
+    {
+      "epoch": 2.94656,
+      "grad_norm": 0.6179057955741882,
+      "learning_rate": 4.541416566626651e-05,
+      "loss": 0.6456,
+      "step": 2302
+    },
+    {
+      "epoch": 2.9478400000000002,
+      "grad_norm": 0.6320080161094666,
+      "learning_rate": 4.541216486594638e-05,
+      "loss": 0.6268,
+      "step": 2303
+    },
+    {
+      "epoch": 2.9491199999999997,
+      "grad_norm": 0.6388428807258606,
+      "learning_rate": 4.541016406562625e-05,
+      "loss": 0.7033,
+      "step": 2304
+    },
+    {
+      "epoch": 2.9504,
+      "grad_norm": 0.6056840419769287,
+      "learning_rate": 4.5408163265306124e-05,
+      "loss": 0.6709,
+      "step": 2305
+    },
+    {
+      "epoch": 2.95168,
+      "grad_norm": 0.6787373423576355,
+      "learning_rate": 4.5406162464985996e-05,
+      "loss": 0.6955,
+      "step": 2306
+    },
+    {
+      "epoch": 2.95296,
+      "grad_norm": 0.6520240306854248,
+      "learning_rate": 4.540416166466587e-05,
+      "loss": 0.645,
+      "step": 2307
+    },
+    {
+      "epoch": 2.95424,
+      "grad_norm": 0.6489017605781555,
+      "learning_rate": 4.540216086434574e-05,
+      "loss": 0.6428,
+      "step": 2308
+    },
+    {
+      "epoch": 2.95552,
+      "grad_norm": 0.6715541481971741,
+      "learning_rate": 4.540016006402561e-05,
+      "loss": 0.6948,
+      "step": 2309
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.7156697511672974,
+      "learning_rate": 4.539815926370549e-05,
+      "loss": 0.696,
+      "step": 2310
+    },
+    {
+      "epoch": 2.95808,
+      "grad_norm": 0.6907376050949097,
+      "learning_rate": 4.5396158463385355e-05,
+      "loss": 0.7406,
+      "step": 2311
+    },
+    {
+      "epoch": 2.95936,
+      "grad_norm": 0.6593761444091797,
+      "learning_rate": 4.539415766306523e-05,
+      "loss": 0.6731,
+      "step": 2312
+    },
+    {
+      "epoch": 2.96064,
+      "grad_norm": 0.6654943227767944,
+      "learning_rate": 4.53921568627451e-05,
+      "loss": 0.6773,
+      "step": 2313
+    },
+    {
+      "epoch": 2.96192,
+      "grad_norm": 0.6800094246864319,
+      "learning_rate": 4.539015606242497e-05,
+      "loss": 0.7691,
+      "step": 2314
+    },
+    {
+      "epoch": 2.9632,
+      "grad_norm": 0.6977217793464661,
+      "learning_rate": 4.538815526210484e-05,
+      "loss": 0.6491,
+      "step": 2315
+    },
+    {
+      "epoch": 2.96448,
+      "grad_norm": 0.6390252113342285,
+      "learning_rate": 4.5386154461784714e-05,
+      "loss": 0.6644,
+      "step": 2316
+    },
+    {
+      "epoch": 2.96576,
+      "grad_norm": 0.6670240163803101,
+      "learning_rate": 4.538415366146459e-05,
+      "loss": 0.6976,
+      "step": 2317
+    },
+    {
+      "epoch": 2.96704,
+      "grad_norm": 0.6732922196388245,
+      "learning_rate": 4.5382152861144465e-05,
+      "loss": 0.6924,
+      "step": 2318
+    },
+    {
+      "epoch": 2.96832,
+      "grad_norm": 0.6866724491119385,
+      "learning_rate": 4.538015206082433e-05,
+      "loss": 0.6729,
+      "step": 2319
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.6516912579536438,
+      "learning_rate": 4.53781512605042e-05,
+      "loss": 0.7145,
+      "step": 2320
+    },
+    {
+      "epoch": 2.97088,
+      "grad_norm": 0.6390223503112793,
+      "learning_rate": 4.5376150460184073e-05,
+      "loss": 0.6462,
+      "step": 2321
+    },
+    {
+      "epoch": 2.97216,
+      "grad_norm": 0.6120672821998596,
+      "learning_rate": 4.5374149659863945e-05,
+      "loss": 0.6401,
+      "step": 2322
+    },
+    {
+      "epoch": 2.97344,
+      "grad_norm": 0.6073777675628662,
+      "learning_rate": 4.537214885954382e-05,
+      "loss": 0.6114,
+      "step": 2323
+    },
+    {
+      "epoch": 2.97472,
+      "grad_norm": 0.698907732963562,
+      "learning_rate": 4.5370148059223696e-05,
+      "loss": 0.7557,
+      "step": 2324
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 0.7005800008773804,
+      "learning_rate": 4.536814725890357e-05,
+      "loss": 0.7338,
+      "step": 2325
+    },
+    {
+      "epoch": 2.97728,
+      "grad_norm": 0.673547089099884,
+      "learning_rate": 4.536614645858344e-05,
+      "loss": 0.6493,
+      "step": 2326
+    },
+    {
+      "epoch": 2.97856,
+      "grad_norm": 0.6851591467857361,
+      "learning_rate": 4.5364145658263305e-05,
+      "loss": 0.7215,
+      "step": 2327
+    },
+    {
+      "epoch": 2.9798400000000003,
+      "grad_norm": 0.5817590355873108,
+      "learning_rate": 4.5362144857943176e-05,
+      "loss": 0.5828,
+      "step": 2328
+    },
+    {
+      "epoch": 2.9811199999999998,
+      "grad_norm": 0.6559504866600037,
+      "learning_rate": 4.536014405762305e-05,
+      "loss": 0.6832,
+      "step": 2329
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.6387779712677002,
+      "learning_rate": 4.535814325730292e-05,
+      "loss": 0.6812,
+      "step": 2330
+    },
+    {
+      "epoch": 2.98368,
+      "grad_norm": 0.6166095733642578,
+      "learning_rate": 4.53561424569828e-05,
+      "loss": 0.6638,
+      "step": 2331
+    },
+    {
+      "epoch": 2.98496,
+      "grad_norm": 0.6481328010559082,
+      "learning_rate": 4.535414165666267e-05,
+      "loss": 0.6361,
+      "step": 2332
+    },
+    {
+      "epoch": 2.98624,
+      "grad_norm": 0.6827143430709839,
+      "learning_rate": 4.535214085634254e-05,
+      "loss": 0.687,
+      "step": 2333
+    },
+    {
+      "epoch": 2.98752,
+      "grad_norm": 0.6266081929206848,
+      "learning_rate": 4.5350140056022414e-05,
+      "loss": 0.6573,
+      "step": 2334
+    },
+    {
+      "epoch": 2.9888,
+      "grad_norm": 0.6480961441993713,
+      "learning_rate": 4.534813925570228e-05,
+      "loss": 0.6742,
+      "step": 2335
+    },
+    {
+      "epoch": 2.99008,
+      "grad_norm": 0.6894843578338623,
+      "learning_rate": 4.534613845538215e-05,
+      "loss": 0.672,
+      "step": 2336
+    },
+    {
+      "epoch": 2.9913600000000002,
+      "grad_norm": 0.6723330020904541,
+      "learning_rate": 4.534413765506202e-05,
+      "loss": 0.6999,
+      "step": 2337
+    },
+    {
+      "epoch": 2.9926399999999997,
+      "grad_norm": 0.7179523706436157,
+      "learning_rate": 4.53421368547419e-05,
+      "loss": 0.7847,
+      "step": 2338
+    },
+    {
+      "epoch": 2.99392,
+      "grad_norm": 0.648439884185791,
+      "learning_rate": 4.5340136054421774e-05,
+      "loss": 0.6769,
+      "step": 2339
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.6338523030281067,
+      "learning_rate": 4.5338135254101645e-05,
+      "loss": 0.6469,
+      "step": 2340
+    },
+    {
+      "epoch": 2.99648,
+      "grad_norm": 0.6572247743606567,
+      "learning_rate": 4.533613445378152e-05,
+      "loss": 0.7077,
+      "step": 2341
+    },
+    {
+      "epoch": 2.99776,
+      "grad_norm": 0.6434239149093628,
+      "learning_rate": 4.533413365346139e-05,
+      "loss": 0.6655,
+      "step": 2342
+    },
+    {
+      "epoch": 2.99904,
+      "grad_norm": 0.67351895570755,
+      "learning_rate": 4.5332132853141254e-05,
+      "loss": 0.7103,
+      "step": 2343
+    },
+    {
+      "epoch": 3.00032,
+      "grad_norm": 1.486992359161377,
+      "learning_rate": 4.5330132052821126e-05,
+      "loss": 1.2061,
+      "step": 2344
+    },
+    {
+      "epoch": 3.0016,
+      "grad_norm": 0.6494001150131226,
+      "learning_rate": 4.5328131252501005e-05,
+      "loss": 0.7025,
+      "step": 2345
+    },
+    {
+      "epoch": 3.00288,
+      "grad_norm": 0.6204409003257751,
+      "learning_rate": 4.532613045218088e-05,
+      "loss": 0.665,
+      "step": 2346
+    },
+    {
+      "epoch": 3.00416,
+      "grad_norm": 0.6547411680221558,
+      "learning_rate": 4.532412965186075e-05,
+      "loss": 0.6653,
+      "step": 2347
+    },
+    {
+      "epoch": 3.00544,
+      "grad_norm": 0.6656926274299622,
+      "learning_rate": 4.532212885154062e-05,
+      "loss": 0.6842,
+      "step": 2348
+    },
+    {
+      "epoch": 3.00672,
+      "grad_norm": 0.7050084471702576,
+      "learning_rate": 4.532012805122049e-05,
+      "loss": 0.6961,
+      "step": 2349
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 0.6964038610458374,
+      "learning_rate": 4.5318127250900364e-05,
+      "loss": 0.6978,
+      "step": 2350
+    },
+    {
+      "epoch": 3.00928,
+      "grad_norm": 0.6723408699035645,
+      "learning_rate": 4.531612645058023e-05,
+      "loss": 0.7163,
+      "step": 2351
+    },
+    {
+      "epoch": 3.01056,
+      "grad_norm": 0.6292080879211426,
+      "learning_rate": 4.531412565026011e-05,
+      "loss": 0.6484,
+      "step": 2352
+    },
+    {
+      "epoch": 3.01184,
+      "grad_norm": 0.6262539625167847,
+      "learning_rate": 4.531212484993998e-05,
+      "loss": 0.6386,
+      "step": 2353
+    },
+    {
+      "epoch": 3.01312,
+      "grad_norm": 0.650016725063324,
+      "learning_rate": 4.531012404961985e-05,
+      "loss": 0.6554,
+      "step": 2354
+    },
+    {
+      "epoch": 3.0144,
+      "grad_norm": 0.6762639880180359,
+      "learning_rate": 4.530812324929972e-05,
+      "loss": 0.6389,
+      "step": 2355
+    },
+    {
+      "epoch": 3.01568,
+      "grad_norm": 0.7196240425109863,
+      "learning_rate": 4.5306122448979595e-05,
+      "loss": 0.6837,
+      "step": 2356
+    },
+    {
+      "epoch": 3.01696,
+      "grad_norm": 0.6458487510681152,
+      "learning_rate": 4.530412164865947e-05,
+      "loss": 0.6213,
+      "step": 2357
+    },
+    {
+      "epoch": 3.01824,
+      "grad_norm": 0.6251707077026367,
+      "learning_rate": 4.530212084833934e-05,
+      "loss": 0.612,
+      "step": 2358
+    },
+    {
+      "epoch": 3.01952,
+      "grad_norm": 0.6338829398155212,
+      "learning_rate": 4.530012004801921e-05,
+      "loss": 0.6327,
+      "step": 2359
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.6653558611869812,
+      "learning_rate": 4.529811924769908e-05,
+      "loss": 0.6595,
+      "step": 2360
+    },
+    {
+      "epoch": 3.02208,
+      "grad_norm": 0.677976131439209,
+      "learning_rate": 4.5296118447378954e-05,
+      "loss": 0.6325,
+      "step": 2361
+    },
+    {
+      "epoch": 3.02336,
+      "grad_norm": 0.6878952980041504,
+      "learning_rate": 4.5294117647058826e-05,
+      "loss": 0.7399,
+      "step": 2362
+    },
+    {
+      "epoch": 3.02464,
+      "grad_norm": 0.6632867455482483,
+      "learning_rate": 4.52921168467387e-05,
+      "loss": 0.7067,
+      "step": 2363
+    },
+    {
+      "epoch": 3.02592,
+      "grad_norm": 0.6564420461654663,
+      "learning_rate": 4.529011604641857e-05,
+      "loss": 0.6509,
+      "step": 2364
+    },
+    {
+      "epoch": 3.0272,
+      "grad_norm": 0.6492841243743896,
+      "learning_rate": 4.528811524609844e-05,
+      "loss": 0.6339,
+      "step": 2365
+    },
+    {
+      "epoch": 3.02848,
+      "grad_norm": 0.682312548160553,
+      "learning_rate": 4.5286114445778314e-05,
+      "loss": 0.6906,
+      "step": 2366
+    },
+    {
+      "epoch": 3.02976,
+      "grad_norm": 0.6361678242683411,
+      "learning_rate": 4.5284113645458186e-05,
+      "loss": 0.6743,
+      "step": 2367
+    },
+    {
+      "epoch": 3.03104,
+      "grad_norm": 0.6740372776985168,
+      "learning_rate": 4.528211284513806e-05,
+      "loss": 0.643,
+      "step": 2368
+    },
+    {
+      "epoch": 3.03232,
+      "grad_norm": 0.6516414880752563,
+      "learning_rate": 4.528011204481793e-05,
+      "loss": 0.6258,
+      "step": 2369
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.677844762802124,
+      "learning_rate": 4.52781112444978e-05,
+      "loss": 0.7175,
+      "step": 2370
+    },
+    {
+      "epoch": 3.03488,
+      "grad_norm": 0.6672567129135132,
+      "learning_rate": 4.527611044417767e-05,
+      "loss": 0.6363,
+      "step": 2371
+    },
+    {
+      "epoch": 3.03616,
+      "grad_norm": 0.6582069993019104,
+      "learning_rate": 4.5274109643857545e-05,
+      "loss": 0.6774,
+      "step": 2372
+    },
+    {
+      "epoch": 3.03744,
+      "grad_norm": 0.670803427696228,
+      "learning_rate": 4.527210884353742e-05,
+      "loss": 0.6809,
+      "step": 2373
+    },
+    {
+      "epoch": 3.03872,
+      "grad_norm": 0.6593008041381836,
+      "learning_rate": 4.527010804321729e-05,
+      "loss": 0.648,
+      "step": 2374
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.6505773067474365,
+      "learning_rate": 4.526810724289716e-05,
+      "loss": 0.6273,
+      "step": 2375
+    },
+    {
+      "epoch": 3.04128,
+      "grad_norm": 0.7069177031517029,
+      "learning_rate": 4.526610644257703e-05,
+      "loss": 0.7197,
+      "step": 2376
+    },
+    {
+      "epoch": 3.04256,
+      "grad_norm": 0.6976633667945862,
+      "learning_rate": 4.5264105642256904e-05,
+      "loss": 0.6741,
+      "step": 2377
+    },
+    {
+      "epoch": 3.04384,
+      "grad_norm": 0.7404224276542664,
+      "learning_rate": 4.5262104841936776e-05,
+      "loss": 0.7145,
+      "step": 2378
+    },
+    {
+      "epoch": 3.04512,
+      "grad_norm": 0.6723030209541321,
+      "learning_rate": 4.526010404161665e-05,
+      "loss": 0.6488,
+      "step": 2379
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.689201831817627,
+      "learning_rate": 4.5258103241296527e-05,
+      "loss": 0.7129,
+      "step": 2380
+    },
+    {
+      "epoch": 3.04768,
+      "grad_norm": 0.7270947098731995,
+      "learning_rate": 4.525610244097639e-05,
+      "loss": 0.6891,
+      "step": 2381
+    },
+    {
+      "epoch": 3.04896,
+      "grad_norm": 0.7034469246864319,
+      "learning_rate": 4.5254101640656263e-05,
+      "loss": 0.7154,
+      "step": 2382
+    },
+    {
+      "epoch": 3.05024,
+      "grad_norm": 0.7317506074905396,
+      "learning_rate": 4.5252100840336135e-05,
+      "loss": 0.6951,
+      "step": 2383
+    },
+    {
+      "epoch": 3.05152,
+      "grad_norm": 0.6207937598228455,
+      "learning_rate": 4.525010004001601e-05,
+      "loss": 0.6367,
+      "step": 2384
+    },
+    {
+      "epoch": 3.0528,
+      "grad_norm": 0.6545078158378601,
+      "learning_rate": 4.524809923969588e-05,
+      "loss": 0.6703,
+      "step": 2385
+    },
+    {
+      "epoch": 3.05408,
+      "grad_norm": 0.6788381934165955,
+      "learning_rate": 4.524609843937575e-05,
+      "loss": 0.6538,
+      "step": 2386
+    },
+    {
+      "epoch": 3.05536,
+      "grad_norm": 0.64985191822052,
+      "learning_rate": 4.524409763905563e-05,
+      "loss": 0.681,
+      "step": 2387
+    },
+    {
+      "epoch": 3.05664,
+      "grad_norm": 0.6559796333312988,
+      "learning_rate": 4.52420968387355e-05,
+      "loss": 0.7017,
+      "step": 2388
+    },
+    {
+      "epoch": 3.05792,
+      "grad_norm": 0.677920937538147,
+      "learning_rate": 4.5240096038415366e-05,
+      "loss": 0.6879,
+      "step": 2389
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.6159390211105347,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.6216,
+      "step": 2390
+    },
+    {
+      "epoch": 3.06048,
+      "grad_norm": 0.7163643836975098,
+      "learning_rate": 4.523609443777511e-05,
+      "loss": 0.7317,
+      "step": 2391
+    },
+    {
+      "epoch": 3.06176,
+      "grad_norm": 0.6356967091560364,
+      "learning_rate": 4.523409363745498e-05,
+      "loss": 0.6964,
+      "step": 2392
+    },
+    {
+      "epoch": 3.06304,
+      "grad_norm": 0.7133451700210571,
+      "learning_rate": 4.5232092837134854e-05,
+      "loss": 0.683,
+      "step": 2393
+    },
+    {
+      "epoch": 3.06432,
+      "grad_norm": 0.6646026372909546,
+      "learning_rate": 4.523009203681473e-05,
+      "loss": 0.6141,
+      "step": 2394
+    },
+    {
+      "epoch": 3.0656,
+      "grad_norm": 0.6448561549186707,
+      "learning_rate": 4.5228091236494604e-05,
+      "loss": 0.6628,
+      "step": 2395
+    },
+    {
+      "epoch": 3.06688,
+      "grad_norm": 0.6711647510528564,
+      "learning_rate": 4.5226090436174476e-05,
+      "loss": 0.6526,
+      "step": 2396
+    },
+    {
+      "epoch": 3.0681599999999998,
+      "grad_norm": 0.6663058996200562,
+      "learning_rate": 4.522408963585434e-05,
+      "loss": 0.6812,
+      "step": 2397
+    },
+    {
+      "epoch": 3.06944,
+      "grad_norm": 0.6636464595794678,
+      "learning_rate": 4.522208883553421e-05,
+      "loss": 0.6382,
+      "step": 2398
+    },
+    {
+      "epoch": 3.07072,
+      "grad_norm": 0.6658351421356201,
+      "learning_rate": 4.5220088035214085e-05,
+      "loss": 0.7137,
+      "step": 2399
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.6860455870628357,
+      "learning_rate": 4.521808723489396e-05,
+      "loss": 0.679,
+      "step": 2400
+    },
+    {
+      "epoch": 3.07328,
+      "grad_norm": 0.6328697800636292,
+      "learning_rate": 4.5216086434573836e-05,
+      "loss": 0.6579,
+      "step": 2401
+    },
+    {
+      "epoch": 3.07456,
+      "grad_norm": 0.665577232837677,
+      "learning_rate": 4.521408563425371e-05,
+      "loss": 0.6864,
+      "step": 2402
+    },
+    {
+      "epoch": 3.07584,
+      "grad_norm": 0.6703231334686279,
+      "learning_rate": 4.521208483393358e-05,
+      "loss": 0.6518,
+      "step": 2403
+    },
+    {
+      "epoch": 3.07712,
+      "grad_norm": 0.7464866042137146,
+      "learning_rate": 4.521008403361345e-05,
+      "loss": 0.7464,
+      "step": 2404
+    },
+    {
+      "epoch": 3.0784,
+      "grad_norm": 0.6916598677635193,
+      "learning_rate": 4.5208083233293316e-05,
+      "loss": 0.6668,
+      "step": 2405
+    },
+    {
+      "epoch": 3.07968,
+      "grad_norm": 0.7007398009300232,
+      "learning_rate": 4.520608243297319e-05,
+      "loss": 0.6506,
+      "step": 2406
+    },
+    {
+      "epoch": 3.08096,
+      "grad_norm": 0.6712905168533325,
+      "learning_rate": 4.520408163265306e-05,
+      "loss": 0.6568,
+      "step": 2407
+    },
+    {
+      "epoch": 3.08224,
+      "grad_norm": 0.687311053276062,
+      "learning_rate": 4.520208083233294e-05,
+      "loss": 0.6959,
+      "step": 2408
+    },
+    {
+      "epoch": 3.08352,
+      "grad_norm": 0.6788797378540039,
+      "learning_rate": 4.520008003201281e-05,
+      "loss": 0.6703,
+      "step": 2409
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.6620363593101501,
+      "learning_rate": 4.519807923169268e-05,
+      "loss": 0.6754,
+      "step": 2410
+    },
+    {
+      "epoch": 3.08608,
+      "grad_norm": 0.6966630220413208,
+      "learning_rate": 4.5196078431372554e-05,
+      "loss": 0.7254,
+      "step": 2411
+    },
+    {
+      "epoch": 3.08736,
+      "grad_norm": 0.6657626032829285,
+      "learning_rate": 4.5194077631052426e-05,
+      "loss": 0.6667,
+      "step": 2412
+    },
+    {
+      "epoch": 3.08864,
+      "grad_norm": 0.6981019377708435,
+      "learning_rate": 4.519207683073229e-05,
+      "loss": 0.6876,
+      "step": 2413
+    },
+    {
+      "epoch": 3.08992,
+      "grad_norm": 0.7262740731239319,
+      "learning_rate": 4.519007603041216e-05,
+      "loss": 0.7387,
+      "step": 2414
+    },
+    {
+      "epoch": 3.0912,
+      "grad_norm": 0.7206056714057922,
+      "learning_rate": 4.5188075230092035e-05,
+      "loss": 0.7254,
+      "step": 2415
+    },
+    {
+      "epoch": 3.09248,
+      "grad_norm": 0.700833261013031,
+      "learning_rate": 4.518607442977191e-05,
+      "loss": 0.6647,
+      "step": 2416
+    },
+    {
+      "epoch": 3.09376,
+      "grad_norm": 0.6799070239067078,
+      "learning_rate": 4.5184073629451785e-05,
+      "loss": 0.6117,
+      "step": 2417
+    },
+    {
+      "epoch": 3.09504,
+      "grad_norm": 0.7079071998596191,
+      "learning_rate": 4.518207282913166e-05,
+      "loss": 0.714,
+      "step": 2418
+    },
+    {
+      "epoch": 3.09632,
+      "grad_norm": 0.7541791200637817,
+      "learning_rate": 4.518007202881153e-05,
+      "loss": 0.7597,
+      "step": 2419
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.7132668495178223,
+      "learning_rate": 4.51780712284914e-05,
+      "loss": 0.6674,
+      "step": 2420
+    },
+    {
+      "epoch": 3.09888,
+      "grad_norm": 0.6781483292579651,
+      "learning_rate": 4.5176070428171266e-05,
+      "loss": 0.7066,
+      "step": 2421
+    },
+    {
+      "epoch": 3.10016,
+      "grad_norm": 0.6593953371047974,
+      "learning_rate": 4.517406962785114e-05,
+      "loss": 0.6443,
+      "step": 2422
+    },
+    {
+      "epoch": 3.10144,
+      "grad_norm": 0.6513901948928833,
+      "learning_rate": 4.5172068827531016e-05,
+      "loss": 0.6421,
+      "step": 2423
+    },
+    {
+      "epoch": 3.10272,
+      "grad_norm": 0.6750826239585876,
+      "learning_rate": 4.517006802721089e-05,
+      "loss": 0.6934,
+      "step": 2424
+    },
+    {
+      "epoch": 3.104,
+      "grad_norm": 0.6588460803031921,
+      "learning_rate": 4.516806722689076e-05,
+      "loss": 0.6222,
+      "step": 2425
+    },
+    {
+      "epoch": 3.10528,
+      "grad_norm": 0.6957754492759705,
+      "learning_rate": 4.516606642657063e-05,
+      "loss": 0.7149,
+      "step": 2426
+    },
+    {
+      "epoch": 3.10656,
+      "grad_norm": 0.644356906414032,
+      "learning_rate": 4.5164065626250504e-05,
+      "loss": 0.6543,
+      "step": 2427
+    },
+    {
+      "epoch": 3.10784,
+      "grad_norm": 0.65400230884552,
+      "learning_rate": 4.5162064825930376e-05,
+      "loss": 0.6291,
+      "step": 2428
+    },
+    {
+      "epoch": 3.10912,
+      "grad_norm": 0.6292628049850464,
+      "learning_rate": 4.516006402561024e-05,
+      "loss": 0.6608,
+      "step": 2429
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.6592010855674744,
+      "learning_rate": 4.515806322529012e-05,
+      "loss": 0.6669,
+      "step": 2430
+    },
+    {
+      "epoch": 3.11168,
+      "grad_norm": 0.6952319145202637,
+      "learning_rate": 4.515606242496999e-05,
+      "loss": 0.6414,
+      "step": 2431
+    },
+    {
+      "epoch": 3.11296,
+      "grad_norm": 0.679885745048523,
+      "learning_rate": 4.515406162464986e-05,
+      "loss": 0.6805,
+      "step": 2432
+    },
+    {
+      "epoch": 3.11424,
+      "grad_norm": 0.6636596918106079,
+      "learning_rate": 4.5152060824329735e-05,
+      "loss": 0.673,
+      "step": 2433
+    },
+    {
+      "epoch": 3.11552,
+      "grad_norm": 0.6836968660354614,
+      "learning_rate": 4.515006002400961e-05,
+      "loss": 0.6919,
+      "step": 2434
+    },
+    {
+      "epoch": 3.1168,
+      "grad_norm": 0.6769798994064331,
+      "learning_rate": 4.514805922368948e-05,
+      "loss": 0.6437,
+      "step": 2435
+    },
+    {
+      "epoch": 3.11808,
+      "grad_norm": 0.6906677484512329,
+      "learning_rate": 4.514605842336935e-05,
+      "loss": 0.6757,
+      "step": 2436
+    },
+    {
+      "epoch": 3.11936,
+      "grad_norm": 0.6719872951507568,
+      "learning_rate": 4.514405762304922e-05,
+      "loss": 0.7031,
+      "step": 2437
+    },
+    {
+      "epoch": 3.12064,
+      "grad_norm": 0.667978048324585,
+      "learning_rate": 4.5142056822729094e-05,
+      "loss": 0.6664,
+      "step": 2438
+    },
+    {
+      "epoch": 3.12192,
+      "grad_norm": 0.6768871545791626,
+      "learning_rate": 4.5140056022408966e-05,
+      "loss": 0.6375,
+      "step": 2439
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.6770617961883545,
+      "learning_rate": 4.513805522208884e-05,
+      "loss": 0.7146,
+      "step": 2440
+    },
+    {
+      "epoch": 3.12448,
+      "grad_norm": 0.7039771676063538,
+      "learning_rate": 4.513605442176871e-05,
+      "loss": 0.6729,
+      "step": 2441
+    },
+    {
+      "epoch": 3.12576,
+      "grad_norm": 0.7217344641685486,
+      "learning_rate": 4.513405362144858e-05,
+      "loss": 0.7069,
+      "step": 2442
+    },
+    {
+      "epoch": 3.12704,
+      "grad_norm": 0.6702495217323303,
+      "learning_rate": 4.5132052821128454e-05,
+      "loss": 0.707,
+      "step": 2443
+    },
+    {
+      "epoch": 3.12832,
+      "grad_norm": 0.6823554635047913,
+      "learning_rate": 4.5130052020808325e-05,
+      "loss": 0.6586,
+      "step": 2444
+    },
+    {
+      "epoch": 3.1296,
+      "grad_norm": 0.6816961765289307,
+      "learning_rate": 4.51280512204882e-05,
+      "loss": 0.716,
+      "step": 2445
+    },
+    {
+      "epoch": 3.13088,
+      "grad_norm": 0.6727588176727295,
+      "learning_rate": 4.512605042016807e-05,
+      "loss": 0.6628,
+      "step": 2446
+    },
+    {
+      "epoch": 3.13216,
+      "grad_norm": 0.6545873284339905,
+      "learning_rate": 4.512404961984794e-05,
+      "loss": 0.6654,
+      "step": 2447
+    },
+    {
+      "epoch": 3.1334400000000002,
+      "grad_norm": 0.6535388827323914,
+      "learning_rate": 4.512204881952781e-05,
+      "loss": 0.6513,
+      "step": 2448
+    },
+    {
+      "epoch": 3.13472,
+      "grad_norm": 0.6998023986816406,
+      "learning_rate": 4.5120048019207685e-05,
+      "loss": 0.6721,
+      "step": 2449
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.7023734450340271,
+      "learning_rate": 4.5118047218887557e-05,
+      "loss": 0.705,
+      "step": 2450
+    },
+    {
+      "epoch": 3.13728,
+      "grad_norm": 0.6585432291030884,
+      "learning_rate": 4.511604641856743e-05,
+      "loss": 0.6781,
+      "step": 2451
+    },
+    {
+      "epoch": 3.13856,
+      "grad_norm": 0.6865501403808594,
+      "learning_rate": 4.51140456182473e-05,
+      "loss": 0.6667,
+      "step": 2452
+    },
+    {
+      "epoch": 3.13984,
+      "grad_norm": 0.6533448100090027,
+      "learning_rate": 4.511204481792717e-05,
+      "loss": 0.6903,
+      "step": 2453
+    },
+    {
+      "epoch": 3.14112,
+      "grad_norm": 0.6679571270942688,
+      "learning_rate": 4.5110044017607044e-05,
+      "loss": 0.6596,
+      "step": 2454
+    },
+    {
+      "epoch": 3.1424,
+      "grad_norm": 0.6311309337615967,
+      "learning_rate": 4.5108043217286916e-05,
+      "loss": 0.6224,
+      "step": 2455
+    },
+    {
+      "epoch": 3.14368,
+      "grad_norm": 0.6664997935295105,
+      "learning_rate": 4.510604241696679e-05,
+      "loss": 0.6438,
+      "step": 2456
+    },
+    {
+      "epoch": 3.14496,
+      "grad_norm": 0.6539314389228821,
+      "learning_rate": 4.510404161664666e-05,
+      "loss": 0.6767,
+      "step": 2457
+    },
+    {
+      "epoch": 3.14624,
+      "grad_norm": 0.6571037769317627,
+      "learning_rate": 4.510204081632654e-05,
+      "loss": 0.6322,
+      "step": 2458
+    },
+    {
+      "epoch": 3.14752,
+      "grad_norm": 0.6855980753898621,
+      "learning_rate": 4.51000400160064e-05,
+      "loss": 0.7187,
+      "step": 2459
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.6939951181411743,
+      "learning_rate": 4.5098039215686275e-05,
+      "loss": 0.7095,
+      "step": 2460
+    },
+    {
+      "epoch": 3.15008,
+      "grad_norm": 0.6432140469551086,
+      "learning_rate": 4.509603841536615e-05,
+      "loss": 0.6465,
+      "step": 2461
+    },
+    {
+      "epoch": 3.15136,
+      "grad_norm": 0.670173704624176,
+      "learning_rate": 4.509403761504602e-05,
+      "loss": 0.6325,
+      "step": 2462
+    },
+    {
+      "epoch": 3.15264,
+      "grad_norm": 0.6773591041564941,
+      "learning_rate": 4.509203681472589e-05,
+      "loss": 0.6375,
+      "step": 2463
+    },
+    {
+      "epoch": 3.15392,
+      "grad_norm": 0.6705629229545593,
+      "learning_rate": 4.509003601440576e-05,
+      "loss": 0.6594,
+      "step": 2464
+    },
+    {
+      "epoch": 3.1552,
+      "grad_norm": 0.7062819004058838,
+      "learning_rate": 4.508803521408564e-05,
+      "loss": 0.6698,
+      "step": 2465
+    },
+    {
+      "epoch": 3.15648,
+      "grad_norm": 0.7161005139350891,
+      "learning_rate": 4.508603441376551e-05,
+      "loss": 0.643,
+      "step": 2466
+    },
+    {
+      "epoch": 3.15776,
+      "grad_norm": 0.6911036968231201,
+      "learning_rate": 4.508403361344538e-05,
+      "loss": 0.6432,
+      "step": 2467
+    },
+    {
+      "epoch": 3.15904,
+      "grad_norm": 0.6515189409255981,
+      "learning_rate": 4.508203281312525e-05,
+      "loss": 0.6367,
+      "step": 2468
+    },
+    {
+      "epoch": 3.16032,
+      "grad_norm": 0.6516139507293701,
+      "learning_rate": 4.508003201280512e-05,
+      "loss": 0.6113,
+      "step": 2469
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.687893807888031,
+      "learning_rate": 4.5078031212484994e-05,
+      "loss": 0.6723,
+      "step": 2470
+    },
+    {
+      "epoch": 3.16288,
+      "grad_norm": 0.6721757650375366,
+      "learning_rate": 4.5076030412164866e-05,
+      "loss": 0.6546,
+      "step": 2471
+    },
+    {
+      "epoch": 3.16416,
+      "grad_norm": 0.67435622215271,
+      "learning_rate": 4.5074029611844744e-05,
+      "loss": 0.6481,
+      "step": 2472
+    },
+    {
+      "epoch": 3.16544,
+      "grad_norm": 0.6942387223243713,
+      "learning_rate": 4.5072028811524616e-05,
+      "loss": 0.7223,
+      "step": 2473
+    },
+    {
+      "epoch": 3.16672,
+      "grad_norm": 0.717036247253418,
+      "learning_rate": 4.507002801120449e-05,
+      "loss": 0.6983,
+      "step": 2474
+    },
+    {
+      "epoch": 3.168,
+      "grad_norm": 0.6933386325836182,
+      "learning_rate": 4.506802721088435e-05,
+      "loss": 0.6512,
+      "step": 2475
+    },
+    {
+      "epoch": 3.16928,
+      "grad_norm": 0.6787186861038208,
+      "learning_rate": 4.5066026410564225e-05,
+      "loss": 0.7028,
+      "step": 2476
+    },
+    {
+      "epoch": 3.17056,
+      "grad_norm": 0.6807264089584351,
+      "learning_rate": 4.50640256102441e-05,
+      "loss": 0.6713,
+      "step": 2477
+    },
+    {
+      "epoch": 3.17184,
+      "grad_norm": 0.6492279767990112,
+      "learning_rate": 4.506202480992397e-05,
+      "loss": 0.6451,
+      "step": 2478
+    },
+    {
+      "epoch": 3.17312,
+      "grad_norm": 0.6500716209411621,
+      "learning_rate": 4.506002400960385e-05,
+      "loss": 0.6752,
+      "step": 2479
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.7048171162605286,
+      "learning_rate": 4.505802320928372e-05,
+      "loss": 0.73,
+      "step": 2480
+    },
+    {
+      "epoch": 3.17568,
+      "grad_norm": 0.709684431552887,
+      "learning_rate": 4.505602240896359e-05,
+      "loss": 0.6861,
+      "step": 2481
+    },
+    {
+      "epoch": 3.1769600000000002,
+      "grad_norm": 0.6857057213783264,
+      "learning_rate": 4.505402160864346e-05,
+      "loss": 0.6706,
+      "step": 2482
+    },
+    {
+      "epoch": 3.17824,
+      "grad_norm": 0.7002062797546387,
+      "learning_rate": 4.505202080832333e-05,
+      "loss": 0.735,
+      "step": 2483
+    },
+    {
+      "epoch": 3.17952,
+      "grad_norm": 0.6401265859603882,
+      "learning_rate": 4.50500200080032e-05,
+      "loss": 0.6662,
+      "step": 2484
+    },
+    {
+      "epoch": 3.1808,
+      "grad_norm": 0.6491225361824036,
+      "learning_rate": 4.504801920768307e-05,
+      "loss": 0.6084,
+      "step": 2485
+    },
+    {
+      "epoch": 3.18208,
+      "grad_norm": 0.7253592014312744,
+      "learning_rate": 4.504601840736295e-05,
+      "loss": 0.6837,
+      "step": 2486
+    },
+    {
+      "epoch": 3.18336,
+      "grad_norm": 0.6980118751525879,
+      "learning_rate": 4.504401760704282e-05,
+      "loss": 0.663,
+      "step": 2487
+    },
+    {
+      "epoch": 3.18464,
+      "grad_norm": 0.6534168124198914,
+      "learning_rate": 4.5042016806722694e-05,
+      "loss": 0.633,
+      "step": 2488
+    },
+    {
+      "epoch": 3.18592,
+      "grad_norm": 0.7110708355903625,
+      "learning_rate": 4.5040016006402566e-05,
+      "loss": 0.6896,
+      "step": 2489
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.7201905846595764,
+      "learning_rate": 4.503801520608244e-05,
+      "loss": 0.6583,
+      "step": 2490
+    },
+    {
+      "epoch": 3.18848,
+      "grad_norm": 0.7668898701667786,
+      "learning_rate": 4.50360144057623e-05,
+      "loss": 0.7062,
+      "step": 2491
+    },
+    {
+      "epoch": 3.18976,
+      "grad_norm": 0.6760601997375488,
+      "learning_rate": 4.5034013605442174e-05,
+      "loss": 0.6544,
+      "step": 2492
+    },
+    {
+      "epoch": 3.19104,
+      "grad_norm": 0.7329314351081848,
+      "learning_rate": 4.503201280512205e-05,
+      "loss": 0.6991,
+      "step": 2493
+    },
+    {
+      "epoch": 3.19232,
+      "grad_norm": 0.7096596956253052,
+      "learning_rate": 4.5030012004801925e-05,
+      "loss": 0.7134,
+      "step": 2494
+    },
+    {
+      "epoch": 3.1936,
+      "grad_norm": 0.688031017780304,
+      "learning_rate": 4.50280112044818e-05,
+      "loss": 0.6861,
+      "step": 2495
+    },
+    {
+      "epoch": 3.19488,
+      "grad_norm": 0.6957057118415833,
+      "learning_rate": 4.502601040416167e-05,
+      "loss": 0.6978,
+      "step": 2496
+    },
+    {
+      "epoch": 3.19616,
+      "grad_norm": 0.6752075552940369,
+      "learning_rate": 4.502400960384154e-05,
+      "loss": 0.6878,
+      "step": 2497
+    },
+    {
+      "epoch": 3.19744,
+      "grad_norm": 0.6884586811065674,
+      "learning_rate": 4.502200880352141e-05,
+      "loss": 0.6982,
+      "step": 2498
+    },
+    {
+      "epoch": 3.19872,
+      "grad_norm": 0.6521093249320984,
+      "learning_rate": 4.502000800320128e-05,
+      "loss": 0.6606,
+      "step": 2499
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.6644425392150879,
+      "learning_rate": 4.5018007202881156e-05,
+      "loss": 0.6753,
+      "step": 2500
+    },
+    {
+      "epoch": 3.20128,
+      "grad_norm": 0.6681901216506958,
+      "learning_rate": 4.501600640256103e-05,
+      "loss": 0.6476,
+      "step": 2501
+    },
+    {
+      "epoch": 3.20256,
+      "grad_norm": 0.6761760711669922,
+      "learning_rate": 4.50140056022409e-05,
+      "loss": 0.7179,
+      "step": 2502
+    },
+    {
+      "epoch": 3.20384,
+      "grad_norm": 0.6894738078117371,
+      "learning_rate": 4.501200480192077e-05,
+      "loss": 0.6737,
+      "step": 2503
+    },
+    {
+      "epoch": 3.20512,
+      "grad_norm": 0.714769184589386,
+      "learning_rate": 4.5010004001600644e-05,
+      "loss": 0.6867,
+      "step": 2504
+    },
+    {
+      "epoch": 3.2064,
+      "grad_norm": 0.6950749754905701,
+      "learning_rate": 4.5008003201280515e-05,
+      "loss": 0.6284,
+      "step": 2505
+    },
+    {
+      "epoch": 3.20768,
+      "grad_norm": 0.7155801057815552,
+      "learning_rate": 4.500600240096039e-05,
+      "loss": 0.7747,
+      "step": 2506
+    },
+    {
+      "epoch": 3.20896,
+      "grad_norm": 0.6814357042312622,
+      "learning_rate": 4.500400160064026e-05,
+      "loss": 0.6696,
+      "step": 2507
+    },
+    {
+      "epoch": 3.21024,
+      "grad_norm": 0.6706905364990234,
+      "learning_rate": 4.500200080032013e-05,
+      "loss": 0.6327,
+      "step": 2508
+    },
+    {
+      "epoch": 3.21152,
+      "grad_norm": 0.6843993067741394,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6887,
+      "step": 2509
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.690980076789856,
+      "learning_rate": 4.4997999199679875e-05,
+      "loss": 0.6454,
+      "step": 2510
+    },
+    {
+      "epoch": 3.21408,
+      "grad_norm": 0.6931962370872498,
+      "learning_rate": 4.4995998399359747e-05,
+      "loss": 0.7268,
+      "step": 2511
+    },
+    {
+      "epoch": 3.21536,
+      "grad_norm": 0.6837199926376343,
+      "learning_rate": 4.499399759903962e-05,
+      "loss": 0.6747,
+      "step": 2512
+    },
+    {
+      "epoch": 3.21664,
+      "grad_norm": 0.6582213640213013,
+      "learning_rate": 4.499199679871949e-05,
+      "loss": 0.6761,
+      "step": 2513
+    },
+    {
+      "epoch": 3.21792,
+      "grad_norm": 0.6401382088661194,
+      "learning_rate": 4.498999599839936e-05,
+      "loss": 0.6742,
+      "step": 2514
+    },
+    {
+      "epoch": 3.2192,
+      "grad_norm": 0.6773951053619385,
+      "learning_rate": 4.4987995198079234e-05,
+      "loss": 0.6328,
+      "step": 2515
+    },
+    {
+      "epoch": 3.2204800000000002,
+      "grad_norm": 0.7156820893287659,
+      "learning_rate": 4.4985994397759106e-05,
+      "loss": 0.7116,
+      "step": 2516
+    },
+    {
+      "epoch": 3.22176,
+      "grad_norm": 0.6766327619552612,
+      "learning_rate": 4.498399359743898e-05,
+      "loss": 0.66,
+      "step": 2517
+    },
+    {
+      "epoch": 3.22304,
+      "grad_norm": 0.6747710704803467,
+      "learning_rate": 4.498199279711885e-05,
+      "loss": 0.6555,
+      "step": 2518
+    },
+    {
+      "epoch": 3.22432,
+      "grad_norm": 0.6878660917282104,
+      "learning_rate": 4.497999199679872e-05,
+      "loss": 0.7085,
+      "step": 2519
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.7145971655845642,
+      "learning_rate": 4.497799119647859e-05,
+      "loss": 0.65,
+      "step": 2520
+    },
+    {
+      "epoch": 3.22688,
+      "grad_norm": 0.664560854434967,
+      "learning_rate": 4.4975990396158465e-05,
+      "loss": 0.692,
+      "step": 2521
+    },
+    {
+      "epoch": 3.22816,
+      "grad_norm": 0.6799471378326416,
+      "learning_rate": 4.497398959583834e-05,
+      "loss": 0.7305,
+      "step": 2522
+    },
+    {
+      "epoch": 3.22944,
+      "grad_norm": 0.7038140892982483,
+      "learning_rate": 4.497198879551821e-05,
+      "loss": 0.6561,
+      "step": 2523
+    },
+    {
+      "epoch": 3.23072,
+      "grad_norm": 0.6589545607566833,
+      "learning_rate": 4.496998799519808e-05,
+      "loss": 0.6906,
+      "step": 2524
+    },
+    {
+      "epoch": 3.232,
+      "grad_norm": 0.6808072924613953,
+      "learning_rate": 4.496798719487795e-05,
+      "loss": 0.652,
+      "step": 2525
+    },
+    {
+      "epoch": 3.23328,
+      "grad_norm": 0.6572796702384949,
+      "learning_rate": 4.4965986394557824e-05,
+      "loss": 0.6557,
+      "step": 2526
+    },
+    {
+      "epoch": 3.23456,
+      "grad_norm": 0.6141901016235352,
+      "learning_rate": 4.4963985594237696e-05,
+      "loss": 0.6165,
+      "step": 2527
+    },
+    {
+      "epoch": 3.23584,
+      "grad_norm": 0.6819775104522705,
+      "learning_rate": 4.496198479391757e-05,
+      "loss": 0.6397,
+      "step": 2528
+    },
+    {
+      "epoch": 3.23712,
+      "grad_norm": 0.6745738387107849,
+      "learning_rate": 4.495998399359744e-05,
+      "loss": 0.6836,
+      "step": 2529
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.7165648937225342,
+      "learning_rate": 4.495798319327731e-05,
+      "loss": 0.7437,
+      "step": 2530
+    },
+    {
+      "epoch": 3.23968,
+      "grad_norm": 0.6725075840950012,
+      "learning_rate": 4.4955982392957184e-05,
+      "loss": 0.6651,
+      "step": 2531
+    },
+    {
+      "epoch": 3.24096,
+      "grad_norm": 0.6513887643814087,
+      "learning_rate": 4.4953981592637056e-05,
+      "loss": 0.6218,
+      "step": 2532
+    },
+    {
+      "epoch": 3.24224,
+      "grad_norm": 0.7130393385887146,
+      "learning_rate": 4.495198079231693e-05,
+      "loss": 0.7008,
+      "step": 2533
+    },
+    {
+      "epoch": 3.24352,
+      "grad_norm": 0.7220726013183594,
+      "learning_rate": 4.49499799919968e-05,
+      "loss": 0.7437,
+      "step": 2534
+    },
+    {
+      "epoch": 3.2448,
+      "grad_norm": 0.6959324479103088,
+      "learning_rate": 4.494797919167667e-05,
+      "loss": 0.6412,
+      "step": 2535
+    },
+    {
+      "epoch": 3.24608,
+      "grad_norm": 0.6925476789474487,
+      "learning_rate": 4.494597839135655e-05,
+      "loss": 0.671,
+      "step": 2536
+    },
+    {
+      "epoch": 3.24736,
+      "grad_norm": 0.6907139420509338,
+      "learning_rate": 4.4943977591036415e-05,
+      "loss": 0.7017,
+      "step": 2537
+    },
+    {
+      "epoch": 3.24864,
+      "grad_norm": 0.7118213176727295,
+      "learning_rate": 4.494197679071629e-05,
+      "loss": 0.6902,
+      "step": 2538
+    },
+    {
+      "epoch": 3.24992,
+      "grad_norm": 0.6986705660820007,
+      "learning_rate": 4.493997599039616e-05,
+      "loss": 0.6936,
+      "step": 2539
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.6764949560165405,
+      "learning_rate": 4.493797519007603e-05,
+      "loss": 0.6506,
+      "step": 2540
+    },
+    {
+      "epoch": 3.25248,
+      "grad_norm": 0.6302717328071594,
+      "learning_rate": 4.49359743897559e-05,
+      "loss": 0.6249,
+      "step": 2541
+    },
+    {
+      "epoch": 3.2537599999999998,
+      "grad_norm": 0.6727375388145447,
+      "learning_rate": 4.4933973589435774e-05,
+      "loss": 0.6644,
+      "step": 2542
+    },
+    {
+      "epoch": 3.25504,
+      "grad_norm": 0.6621202826499939,
+      "learning_rate": 4.493197278911565e-05,
+      "loss": 0.6559,
+      "step": 2543
+    },
+    {
+      "epoch": 3.25632,
+      "grad_norm": 0.67811119556427,
+      "learning_rate": 4.4929971988795525e-05,
+      "loss": 0.6871,
+      "step": 2544
+    },
+    {
+      "epoch": 3.2576,
+      "grad_norm": 0.7005154490470886,
+      "learning_rate": 4.492797118847539e-05,
+      "loss": 0.6732,
+      "step": 2545
+    },
+    {
+      "epoch": 3.25888,
+      "grad_norm": 0.6961964964866638,
+      "learning_rate": 4.492597038815526e-05,
+      "loss": 0.6727,
+      "step": 2546
+    },
+    {
+      "epoch": 3.26016,
+      "grad_norm": 0.6468445062637329,
+      "learning_rate": 4.4923969587835133e-05,
+      "loss": 0.6521,
+      "step": 2547
+    },
+    {
+      "epoch": 3.26144,
+      "grad_norm": 0.6833783984184265,
+      "learning_rate": 4.4921968787515005e-05,
+      "loss": 0.6713,
+      "step": 2548
+    },
+    {
+      "epoch": 3.26272,
+      "grad_norm": 0.6983469724655151,
+      "learning_rate": 4.491996798719488e-05,
+      "loss": 0.7341,
+      "step": 2549
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.6976143717765808,
+      "learning_rate": 4.4917967186874756e-05,
+      "loss": 0.6902,
+      "step": 2550
+    },
+    {
+      "epoch": 3.26528,
+      "grad_norm": 0.6914013028144836,
+      "learning_rate": 4.491596638655463e-05,
+      "loss": 0.6953,
+      "step": 2551
+    },
+    {
+      "epoch": 3.26656,
+      "grad_norm": 0.6625963449478149,
+      "learning_rate": 4.49139655862345e-05,
+      "loss": 0.6624,
+      "step": 2552
+    },
+    {
+      "epoch": 3.26784,
+      "grad_norm": 0.6785646677017212,
+      "learning_rate": 4.4911964785914365e-05,
+      "loss": 0.6387,
+      "step": 2553
+    },
+    {
+      "epoch": 3.26912,
+      "grad_norm": 0.6610428094863892,
+      "learning_rate": 4.4909963985594236e-05,
+      "loss": 0.6896,
+      "step": 2554
+    },
+    {
+      "epoch": 3.2704,
+      "grad_norm": 0.7095894813537598,
+      "learning_rate": 4.490796318527411e-05,
+      "loss": 0.6717,
+      "step": 2555
+    },
+    {
+      "epoch": 3.27168,
+      "grad_norm": 0.6596770286560059,
+      "learning_rate": 4.490596238495398e-05,
+      "loss": 0.627,
+      "step": 2556
+    },
+    {
+      "epoch": 3.27296,
+      "grad_norm": 0.6432579755783081,
+      "learning_rate": 4.490396158463386e-05,
+      "loss": 0.6196,
+      "step": 2557
+    },
+    {
+      "epoch": 3.27424,
+      "grad_norm": 0.6961793303489685,
+      "learning_rate": 4.490196078431373e-05,
+      "loss": 0.6624,
+      "step": 2558
+    },
+    {
+      "epoch": 3.27552,
+      "grad_norm": 0.6829546093940735,
+      "learning_rate": 4.48999599839936e-05,
+      "loss": 0.6567,
+      "step": 2559
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.7708462476730347,
+      "learning_rate": 4.4897959183673474e-05,
+      "loss": 0.7334,
+      "step": 2560
+    },
+    {
+      "epoch": 3.27808,
+      "grad_norm": 0.6607860326766968,
+      "learning_rate": 4.489595838335334e-05,
+      "loss": 0.6581,
+      "step": 2561
+    },
+    {
+      "epoch": 3.27936,
+      "grad_norm": 0.6741865873336792,
+      "learning_rate": 4.489395758303321e-05,
+      "loss": 0.6928,
+      "step": 2562
+    },
+    {
+      "epoch": 3.28064,
+      "grad_norm": 0.6427492499351501,
+      "learning_rate": 4.489195678271308e-05,
+      "loss": 0.6294,
+      "step": 2563
+    },
+    {
+      "epoch": 3.28192,
+      "grad_norm": 0.6945633292198181,
+      "learning_rate": 4.488995598239296e-05,
+      "loss": 0.7027,
+      "step": 2564
+    },
+    {
+      "epoch": 3.2832,
+      "grad_norm": 0.6338651776313782,
+      "learning_rate": 4.4887955182072834e-05,
+      "loss": 0.6274,
+      "step": 2565
+    },
+    {
+      "epoch": 3.28448,
+      "grad_norm": 0.6843262314796448,
+      "learning_rate": 4.4885954381752705e-05,
+      "loss": 0.6814,
+      "step": 2566
+    },
+    {
+      "epoch": 3.28576,
+      "grad_norm": 0.6916020512580872,
+      "learning_rate": 4.488395358143258e-05,
+      "loss": 0.7082,
+      "step": 2567
+    },
+    {
+      "epoch": 3.28704,
+      "grad_norm": 0.6470593214035034,
+      "learning_rate": 4.488195278111245e-05,
+      "loss": 0.6058,
+      "step": 2568
+    },
+    {
+      "epoch": 3.28832,
+      "grad_norm": 0.7212356328964233,
+      "learning_rate": 4.4879951980792314e-05,
+      "loss": 0.6994,
+      "step": 2569
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.6888067722320557,
+      "learning_rate": 4.4877951180472186e-05,
+      "loss": 0.6621,
+      "step": 2570
+    },
+    {
+      "epoch": 3.29088,
+      "grad_norm": 0.7158266305923462,
+      "learning_rate": 4.4875950380152065e-05,
+      "loss": 0.6782,
+      "step": 2571
+    },
+    {
+      "epoch": 3.29216,
+      "grad_norm": 0.6973638534545898,
+      "learning_rate": 4.4873949579831937e-05,
+      "loss": 0.7213,
+      "step": 2572
+    },
+    {
+      "epoch": 3.29344,
+      "grad_norm": 0.7036058306694031,
+      "learning_rate": 4.487194877951181e-05,
+      "loss": 0.6762,
+      "step": 2573
+    },
+    {
+      "epoch": 3.29472,
+      "grad_norm": 0.6940634846687317,
+      "learning_rate": 4.486994797919168e-05,
+      "loss": 0.6541,
+      "step": 2574
+    },
+    {
+      "epoch": 3.296,
+      "grad_norm": 0.7082141041755676,
+      "learning_rate": 4.486794717887155e-05,
+      "loss": 0.6707,
+      "step": 2575
+    },
+    {
+      "epoch": 3.2972799999999998,
+      "grad_norm": 0.6922471523284912,
+      "learning_rate": 4.4865946378551424e-05,
+      "loss": 0.6843,
+      "step": 2576
+    },
+    {
+      "epoch": 3.29856,
+      "grad_norm": 0.6800956130027771,
+      "learning_rate": 4.486394557823129e-05,
+      "loss": 0.6028,
+      "step": 2577
+    },
+    {
+      "epoch": 3.29984,
+      "grad_norm": 0.6691498160362244,
+      "learning_rate": 4.486194477791117e-05,
+      "loss": 0.6706,
+      "step": 2578
+    },
+    {
+      "epoch": 3.30112,
+      "grad_norm": 0.6471302509307861,
+      "learning_rate": 4.485994397759104e-05,
+      "loss": 0.6753,
+      "step": 2579
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.6529038548469543,
+      "learning_rate": 4.485794317727091e-05,
+      "loss": 0.6006,
+      "step": 2580
+    },
+    {
+      "epoch": 3.30368,
+      "grad_norm": 0.6317946314811707,
+      "learning_rate": 4.485594237695078e-05,
+      "loss": 0.6842,
+      "step": 2581
+    },
+    {
+      "epoch": 3.30496,
+      "grad_norm": 0.6559657454490662,
+      "learning_rate": 4.4853941576630655e-05,
+      "loss": 0.6628,
+      "step": 2582
+    },
+    {
+      "epoch": 3.30624,
+      "grad_norm": 0.6486949324607849,
+      "learning_rate": 4.485194077631053e-05,
+      "loss": 0.6775,
+      "step": 2583
+    },
+    {
+      "epoch": 3.3075200000000002,
+      "grad_norm": 0.6692925691604614,
+      "learning_rate": 4.48499399759904e-05,
+      "loss": 0.6721,
+      "step": 2584
+    },
+    {
+      "epoch": 3.3088,
+      "grad_norm": 0.6672170162200928,
+      "learning_rate": 4.484793917567027e-05,
+      "loss": 0.6424,
+      "step": 2585
+    },
+    {
+      "epoch": 3.31008,
+      "grad_norm": 0.668408215045929,
+      "learning_rate": 4.484593837535014e-05,
+      "loss": 0.6445,
+      "step": 2586
+    },
+    {
+      "epoch": 3.31136,
+      "grad_norm": 0.6862883567810059,
+      "learning_rate": 4.4843937575030014e-05,
+      "loss": 0.6994,
+      "step": 2587
+    },
+    {
+      "epoch": 3.31264,
+      "grad_norm": 0.6536902189254761,
+      "learning_rate": 4.4841936774709886e-05,
+      "loss": 0.6149,
+      "step": 2588
+    },
+    {
+      "epoch": 3.31392,
+      "grad_norm": 0.6626643538475037,
+      "learning_rate": 4.483993597438976e-05,
+      "loss": 0.6631,
+      "step": 2589
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.6667909622192383,
+      "learning_rate": 4.483793517406963e-05,
+      "loss": 0.6781,
+      "step": 2590
+    },
+    {
+      "epoch": 3.31648,
+      "grad_norm": 0.6631030440330505,
+      "learning_rate": 4.48359343737495e-05,
+      "loss": 0.6586,
+      "step": 2591
+    },
+    {
+      "epoch": 3.31776,
+      "grad_norm": 0.6788957118988037,
+      "learning_rate": 4.4833933573429374e-05,
+      "loss": 0.6355,
+      "step": 2592
+    },
+    {
+      "epoch": 3.31904,
+      "grad_norm": 0.7192303538322449,
+      "learning_rate": 4.4831932773109246e-05,
+      "loss": 0.6597,
+      "step": 2593
+    },
+    {
+      "epoch": 3.32032,
+      "grad_norm": 0.7045040726661682,
+      "learning_rate": 4.482993197278912e-05,
+      "loss": 0.7322,
+      "step": 2594
+    },
+    {
+      "epoch": 3.3216,
+      "grad_norm": 0.6571533679962158,
+      "learning_rate": 4.482793117246899e-05,
+      "loss": 0.6443,
+      "step": 2595
+    },
+    {
+      "epoch": 3.32288,
+      "grad_norm": 0.6708471775054932,
+      "learning_rate": 4.482593037214886e-05,
+      "loss": 0.6391,
+      "step": 2596
+    },
+    {
+      "epoch": 3.32416,
+      "grad_norm": 0.6779941916465759,
+      "learning_rate": 4.482392957182873e-05,
+      "loss": 0.6226,
+      "step": 2597
+    },
+    {
+      "epoch": 3.32544,
+      "grad_norm": 0.7229152321815491,
+      "learning_rate": 4.4821928771508605e-05,
+      "loss": 0.6434,
+      "step": 2598
+    },
+    {
+      "epoch": 3.32672,
+      "grad_norm": 0.7014409303665161,
+      "learning_rate": 4.481992797118848e-05,
+      "loss": 0.7082,
+      "step": 2599
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.6607677340507507,
+      "learning_rate": 4.481792717086835e-05,
+      "loss": 0.6574,
+      "step": 2600
+    },
+    {
+      "epoch": 3.32928,
+      "grad_norm": 0.7238780856132507,
+      "learning_rate": 4.481592637054822e-05,
+      "loss": 0.6938,
+      "step": 2601
+    },
+    {
+      "epoch": 3.33056,
+      "grad_norm": 0.6686570048332214,
+      "learning_rate": 4.481392557022809e-05,
+      "loss": 0.6405,
+      "step": 2602
+    },
+    {
+      "epoch": 3.33184,
+      "grad_norm": 0.708084225654602,
+      "learning_rate": 4.4811924769907964e-05,
+      "loss": 0.6745,
+      "step": 2603
+    },
+    {
+      "epoch": 3.33312,
+      "grad_norm": 0.684820830821991,
+      "learning_rate": 4.4809923969587836e-05,
+      "loss": 0.6385,
+      "step": 2604
+    },
+    {
+      "epoch": 3.3344,
+      "grad_norm": 0.6804541945457458,
+      "learning_rate": 4.480792316926771e-05,
+      "loss": 0.677,
+      "step": 2605
+    },
+    {
+      "epoch": 3.33568,
+      "grad_norm": 0.6551414728164673,
+      "learning_rate": 4.4805922368947586e-05,
+      "loss": 0.6381,
+      "step": 2606
+    },
+    {
+      "epoch": 3.33696,
+      "grad_norm": 0.7107576131820679,
+      "learning_rate": 4.480392156862745e-05,
+      "loss": 0.7287,
+      "step": 2607
+    },
+    {
+      "epoch": 3.33824,
+      "grad_norm": 0.6644242405891418,
+      "learning_rate": 4.4801920768307323e-05,
+      "loss": 0.6999,
+      "step": 2608
+    },
+    {
+      "epoch": 3.33952,
+      "grad_norm": 0.6727672815322876,
+      "learning_rate": 4.4799919967987195e-05,
+      "loss": 0.6438,
+      "step": 2609
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.67037433385849,
+      "learning_rate": 4.479791916766707e-05,
+      "loss": 0.6282,
+      "step": 2610
+    },
+    {
+      "epoch": 3.34208,
+      "grad_norm": 0.650492250919342,
+      "learning_rate": 4.479591836734694e-05,
+      "loss": 0.6333,
+      "step": 2611
+    },
+    {
+      "epoch": 3.34336,
+      "grad_norm": 0.6534383893013,
+      "learning_rate": 4.479391756702681e-05,
+      "loss": 0.671,
+      "step": 2612
+    },
+    {
+      "epoch": 3.34464,
+      "grad_norm": 0.6915240287780762,
+      "learning_rate": 4.479191676670669e-05,
+      "loss": 0.6673,
+      "step": 2613
+    },
+    {
+      "epoch": 3.34592,
+      "grad_norm": 0.6916387677192688,
+      "learning_rate": 4.478991596638656e-05,
+      "loss": 0.6572,
+      "step": 2614
+    },
+    {
+      "epoch": 3.3472,
+      "grad_norm": 0.6631138324737549,
+      "learning_rate": 4.4787915166066426e-05,
+      "loss": 0.6512,
+      "step": 2615
+    },
+    {
+      "epoch": 3.34848,
+      "grad_norm": 0.6894577145576477,
+      "learning_rate": 4.47859143657463e-05,
+      "loss": 0.6438,
+      "step": 2616
+    },
+    {
+      "epoch": 3.34976,
+      "grad_norm": 0.6938015818595886,
+      "learning_rate": 4.478391356542617e-05,
+      "loss": 0.69,
+      "step": 2617
+    },
+    {
+      "epoch": 3.3510400000000002,
+      "grad_norm": 0.6625933647155762,
+      "learning_rate": 4.478191276510604e-05,
+      "loss": 0.6358,
+      "step": 2618
+    },
+    {
+      "epoch": 3.35232,
+      "grad_norm": 0.7050860524177551,
+      "learning_rate": 4.4779911964785914e-05,
+      "loss": 0.7098,
+      "step": 2619
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.7034277319908142,
+      "learning_rate": 4.477791116446579e-05,
+      "loss": 0.7057,
+      "step": 2620
+    },
+    {
+      "epoch": 3.35488,
+      "grad_norm": 0.7165651917457581,
+      "learning_rate": 4.4775910364145664e-05,
+      "loss": 0.7058,
+      "step": 2621
+    },
+    {
+      "epoch": 3.35616,
+      "grad_norm": 0.6885167956352234,
+      "learning_rate": 4.4773909563825536e-05,
+      "loss": 0.6334,
+      "step": 2622
+    },
+    {
+      "epoch": 3.35744,
+      "grad_norm": 0.6467083692550659,
+      "learning_rate": 4.47719087635054e-05,
+      "loss": 0.6097,
+      "step": 2623
+    },
+    {
+      "epoch": 3.35872,
+      "grad_norm": 0.7264028787612915,
+      "learning_rate": 4.476990796318527e-05,
+      "loss": 0.7396,
+      "step": 2624
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.7132958769798279,
+      "learning_rate": 4.4767907162865145e-05,
+      "loss": 0.6642,
+      "step": 2625
+    },
+    {
+      "epoch": 3.36128,
+      "grad_norm": 0.7386727333068848,
+      "learning_rate": 4.476590636254502e-05,
+      "loss": 0.726,
+      "step": 2626
+    },
+    {
+      "epoch": 3.36256,
+      "grad_norm": 0.6800748705863953,
+      "learning_rate": 4.4763905562224895e-05,
+      "loss": 0.6936,
+      "step": 2627
+    },
+    {
+      "epoch": 3.36384,
+      "grad_norm": 0.6800684332847595,
+      "learning_rate": 4.476190476190477e-05,
+      "loss": 0.6065,
+      "step": 2628
+    },
+    {
+      "epoch": 3.36512,
+      "grad_norm": 0.6520423293113708,
+      "learning_rate": 4.475990396158464e-05,
+      "loss": 0.6725,
+      "step": 2629
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.6812400221824646,
+      "learning_rate": 4.475790316126451e-05,
+      "loss": 0.6495,
+      "step": 2630
+    },
+    {
+      "epoch": 3.36768,
+      "grad_norm": 0.6640177965164185,
+      "learning_rate": 4.4755902360944376e-05,
+      "loss": 0.6885,
+      "step": 2631
+    },
+    {
+      "epoch": 3.36896,
+      "grad_norm": 0.6409942507743835,
+      "learning_rate": 4.475390156062425e-05,
+      "loss": 0.6343,
+      "step": 2632
+    },
+    {
+      "epoch": 3.37024,
+      "grad_norm": 0.6690536141395569,
+      "learning_rate": 4.475190076030412e-05,
+      "loss": 0.6792,
+      "step": 2633
+    },
+    {
+      "epoch": 3.37152,
+      "grad_norm": 0.7020560503005981,
+      "learning_rate": 4.474989995998399e-05,
+      "loss": 0.7442,
+      "step": 2634
+    },
+    {
+      "epoch": 3.3728,
+      "grad_norm": 0.698049783706665,
+      "learning_rate": 4.474789915966387e-05,
+      "loss": 0.6836,
+      "step": 2635
+    },
+    {
+      "epoch": 3.37408,
+      "grad_norm": 0.6824793815612793,
+      "learning_rate": 4.474589835934374e-05,
+      "loss": 0.6818,
+      "step": 2636
+    },
+    {
+      "epoch": 3.37536,
+      "grad_norm": 0.6641604900360107,
+      "learning_rate": 4.4743897559023614e-05,
+      "loss": 0.6497,
+      "step": 2637
+    },
+    {
+      "epoch": 3.37664,
+      "grad_norm": 0.6908559203147888,
+      "learning_rate": 4.4741896758703486e-05,
+      "loss": 0.6655,
+      "step": 2638
+    },
+    {
+      "epoch": 3.37792,
+      "grad_norm": 0.6793168783187866,
+      "learning_rate": 4.473989595838335e-05,
+      "loss": 0.6648,
+      "step": 2639
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.7036957144737244,
+      "learning_rate": 4.473789515806322e-05,
+      "loss": 0.6691,
+      "step": 2640
+    },
+    {
+      "epoch": 3.38048,
+      "grad_norm": 0.684795081615448,
+      "learning_rate": 4.4735894357743095e-05,
+      "loss": 0.6324,
+      "step": 2641
+    },
+    {
+      "epoch": 3.38176,
+      "grad_norm": 0.6578769683837891,
+      "learning_rate": 4.473389355742297e-05,
+      "loss": 0.6746,
+      "step": 2642
+    },
+    {
+      "epoch": 3.38304,
+      "grad_norm": 0.6963245272636414,
+      "learning_rate": 4.4731892757102845e-05,
+      "loss": 0.7146,
+      "step": 2643
+    },
+    {
+      "epoch": 3.3843199999999998,
+      "grad_norm": 0.6984001398086548,
+      "learning_rate": 4.472989195678272e-05,
+      "loss": 0.6317,
+      "step": 2644
+    },
+    {
+      "epoch": 3.3856,
+      "grad_norm": 0.7258270978927612,
+      "learning_rate": 4.472789115646259e-05,
+      "loss": 0.6904,
+      "step": 2645
+    },
+    {
+      "epoch": 3.38688,
+      "grad_norm": 0.6447382569313049,
+      "learning_rate": 4.472589035614246e-05,
+      "loss": 0.6477,
+      "step": 2646
+    },
+    {
+      "epoch": 3.38816,
+      "grad_norm": 0.6710345149040222,
+      "learning_rate": 4.4723889555822326e-05,
+      "loss": 0.6563,
+      "step": 2647
+    },
+    {
+      "epoch": 3.38944,
+      "grad_norm": 0.7310519218444824,
+      "learning_rate": 4.47218887555022e-05,
+      "loss": 0.6717,
+      "step": 2648
+    },
+    {
+      "epoch": 3.39072,
+      "grad_norm": 0.7194939851760864,
+      "learning_rate": 4.4719887955182076e-05,
+      "loss": 0.6474,
+      "step": 2649
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.6350496411323547,
+      "learning_rate": 4.471788715486195e-05,
+      "loss": 0.6131,
+      "step": 2650
+    },
+    {
+      "epoch": 3.39328,
+      "grad_norm": 0.6812024712562561,
+      "learning_rate": 4.471588635454182e-05,
+      "loss": 0.6778,
+      "step": 2651
+    },
+    {
+      "epoch": 3.3945600000000002,
+      "grad_norm": 0.6636951565742493,
+      "learning_rate": 4.471388555422169e-05,
+      "loss": 0.6575,
+      "step": 2652
+    },
+    {
+      "epoch": 3.39584,
+      "grad_norm": 0.6647194027900696,
+      "learning_rate": 4.4711884753901564e-05,
+      "loss": 0.6415,
+      "step": 2653
+    },
+    {
+      "epoch": 3.39712,
+      "grad_norm": 0.6898042559623718,
+      "learning_rate": 4.4709883953581436e-05,
+      "loss": 0.7263,
+      "step": 2654
+    },
+    {
+      "epoch": 3.3984,
+      "grad_norm": 0.7069958448410034,
+      "learning_rate": 4.47078831532613e-05,
+      "loss": 0.6351,
+      "step": 2655
+    },
+    {
+      "epoch": 3.39968,
+      "grad_norm": 0.6806748509407043,
+      "learning_rate": 4.470588235294118e-05,
+      "loss": 0.665,
+      "step": 2656
+    },
+    {
+      "epoch": 3.40096,
+      "grad_norm": 0.7422452569007874,
+      "learning_rate": 4.470388155262105e-05,
+      "loss": 0.73,
+      "step": 2657
+    },
+    {
+      "epoch": 3.40224,
+      "grad_norm": 0.7665037512779236,
+      "learning_rate": 4.470188075230092e-05,
+      "loss": 0.7148,
+      "step": 2658
+    },
+    {
+      "epoch": 3.40352,
+      "grad_norm": 0.7186295986175537,
+      "learning_rate": 4.4699879951980795e-05,
+      "loss": 0.6665,
+      "step": 2659
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.705100417137146,
+      "learning_rate": 4.469787915166067e-05,
+      "loss": 0.6724,
+      "step": 2660
+    },
+    {
+      "epoch": 3.40608,
+      "grad_norm": 0.6744673252105713,
+      "learning_rate": 4.469587835134054e-05,
+      "loss": 0.6658,
+      "step": 2661
+    },
+    {
+      "epoch": 3.40736,
+      "grad_norm": 0.6781793236732483,
+      "learning_rate": 4.469387755102041e-05,
+      "loss": 0.6664,
+      "step": 2662
+    },
+    {
+      "epoch": 3.40864,
+      "grad_norm": 0.6606245040893555,
+      "learning_rate": 4.469187675070028e-05,
+      "loss": 0.6189,
+      "step": 2663
+    },
+    {
+      "epoch": 3.40992,
+      "grad_norm": 0.6804131865501404,
+      "learning_rate": 4.4689875950380154e-05,
+      "loss": 0.7092,
+      "step": 2664
+    },
+    {
+      "epoch": 3.4112,
+      "grad_norm": 0.6703636050224304,
+      "learning_rate": 4.4687875150060026e-05,
+      "loss": 0.7,
+      "step": 2665
+    },
+    {
+      "epoch": 3.41248,
+      "grad_norm": 0.6748145818710327,
+      "learning_rate": 4.46858743497399e-05,
+      "loss": 0.6143,
+      "step": 2666
+    },
+    {
+      "epoch": 3.41376,
+      "grad_norm": 0.6967974305152893,
+      "learning_rate": 4.468387354941977e-05,
+      "loss": 0.6877,
+      "step": 2667
+    },
+    {
+      "epoch": 3.41504,
+      "grad_norm": 0.6868359446525574,
+      "learning_rate": 4.468187274909964e-05,
+      "loss": 0.6435,
+      "step": 2668
+    },
+    {
+      "epoch": 3.41632,
+      "grad_norm": 0.7249009013175964,
+      "learning_rate": 4.4679871948779513e-05,
+      "loss": 0.7003,
+      "step": 2669
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.6801961064338684,
+      "learning_rate": 4.4677871148459385e-05,
+      "loss": 0.6565,
+      "step": 2670
+    },
+    {
+      "epoch": 3.41888,
+      "grad_norm": 0.7236019372940063,
+      "learning_rate": 4.467587034813926e-05,
+      "loss": 0.6586,
+      "step": 2671
+    },
+    {
+      "epoch": 3.42016,
+      "grad_norm": 0.6557265520095825,
+      "learning_rate": 4.467386954781913e-05,
+      "loss": 0.6919,
+      "step": 2672
+    },
+    {
+      "epoch": 3.42144,
+      "grad_norm": 0.7524291276931763,
+      "learning_rate": 4.4671868747499e-05,
+      "loss": 0.7208,
+      "step": 2673
+    },
+    {
+      "epoch": 3.42272,
+      "grad_norm": 0.7023764848709106,
+      "learning_rate": 4.466986794717887e-05,
+      "loss": 0.7148,
+      "step": 2674
+    },
+    {
+      "epoch": 3.424,
+      "grad_norm": 0.7110322117805481,
+      "learning_rate": 4.4667867146858745e-05,
+      "loss": 0.7194,
+      "step": 2675
+    },
+    {
+      "epoch": 3.42528,
+      "grad_norm": 0.668428361415863,
+      "learning_rate": 4.4665866346538616e-05,
+      "loss": 0.663,
+      "step": 2676
+    },
+    {
+      "epoch": 3.42656,
+      "grad_norm": 0.6546775698661804,
+      "learning_rate": 4.4663865546218495e-05,
+      "loss": 0.5899,
+      "step": 2677
+    },
+    {
+      "epoch": 3.4278399999999998,
+      "grad_norm": 0.6761878132820129,
+      "learning_rate": 4.466186474589836e-05,
+      "loss": 0.6309,
+      "step": 2678
+    },
+    {
+      "epoch": 3.42912,
+      "grad_norm": 0.7158021330833435,
+      "learning_rate": 4.465986394557823e-05,
+      "loss": 0.737,
+      "step": 2679
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.7490652799606323,
+      "learning_rate": 4.4657863145258104e-05,
+      "loss": 0.6946,
+      "step": 2680
+    },
+    {
+      "epoch": 3.43168,
+      "grad_norm": 0.7451316118240356,
+      "learning_rate": 4.4655862344937976e-05,
+      "loss": 0.6851,
+      "step": 2681
+    },
+    {
+      "epoch": 3.43296,
+      "grad_norm": 0.6753928065299988,
+      "learning_rate": 4.465386154461785e-05,
+      "loss": 0.6876,
+      "step": 2682
+    },
+    {
+      "epoch": 3.43424,
+      "grad_norm": 0.6749107837677002,
+      "learning_rate": 4.465186074429772e-05,
+      "loss": 0.6256,
+      "step": 2683
+    },
+    {
+      "epoch": 3.43552,
+      "grad_norm": 0.6775442361831665,
+      "learning_rate": 4.46498599439776e-05,
+      "loss": 0.7097,
+      "step": 2684
+    },
+    {
+      "epoch": 3.4368,
+      "grad_norm": 0.7066680192947388,
+      "learning_rate": 4.464785914365747e-05,
+      "loss": 0.6892,
+      "step": 2685
+    },
+    {
+      "epoch": 3.4380800000000002,
+      "grad_norm": 0.7244362831115723,
+      "learning_rate": 4.4645858343337335e-05,
+      "loss": 0.6675,
+      "step": 2686
+    },
+    {
+      "epoch": 3.4393599999999998,
+      "grad_norm": 0.7037179470062256,
+      "learning_rate": 4.464385754301721e-05,
+      "loss": 0.7126,
+      "step": 2687
+    },
+    {
+      "epoch": 3.44064,
+      "grad_norm": 0.6872085332870483,
+      "learning_rate": 4.464185674269708e-05,
+      "loss": 0.6409,
+      "step": 2688
+    },
+    {
+      "epoch": 3.44192,
+      "grad_norm": 0.6594861149787903,
+      "learning_rate": 4.463985594237695e-05,
+      "loss": 0.6789,
+      "step": 2689
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.6587255001068115,
+      "learning_rate": 4.463785514205682e-05,
+      "loss": 0.6327,
+      "step": 2690
+    },
+    {
+      "epoch": 3.44448,
+      "grad_norm": 0.6550960540771484,
+      "learning_rate": 4.46358543417367e-05,
+      "loss": 0.5824,
+      "step": 2691
+    },
+    {
+      "epoch": 3.44576,
+      "grad_norm": 0.7198725342750549,
+      "learning_rate": 4.463385354141657e-05,
+      "loss": 0.6853,
+      "step": 2692
+    },
+    {
+      "epoch": 3.44704,
+      "grad_norm": 0.6870785355567932,
+      "learning_rate": 4.4631852741096445e-05,
+      "loss": 0.6366,
+      "step": 2693
+    },
+    {
+      "epoch": 3.44832,
+      "grad_norm": 0.6846055388450623,
+      "learning_rate": 4.462985194077631e-05,
+      "loss": 0.6755,
+      "step": 2694
+    },
+    {
+      "epoch": 3.4496,
+      "grad_norm": 0.6860753893852234,
+      "learning_rate": 4.462785114045618e-05,
+      "loss": 0.6605,
+      "step": 2695
+    },
+    {
+      "epoch": 3.45088,
+      "grad_norm": 0.7068583369255066,
+      "learning_rate": 4.4625850340136054e-05,
+      "loss": 0.6775,
+      "step": 2696
+    },
+    {
+      "epoch": 3.45216,
+      "grad_norm": 0.6669130921363831,
+      "learning_rate": 4.4623849539815925e-05,
+      "loss": 0.674,
+      "step": 2697
+    },
+    {
+      "epoch": 3.45344,
+      "grad_norm": 0.6609745025634766,
+      "learning_rate": 4.4621848739495804e-05,
+      "loss": 0.6717,
+      "step": 2698
+    },
+    {
+      "epoch": 3.45472,
+      "grad_norm": 0.6721084713935852,
+      "learning_rate": 4.4619847939175676e-05,
+      "loss": 0.6345,
+      "step": 2699
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.6484143733978271,
+      "learning_rate": 4.461784713885555e-05,
+      "loss": 0.6415,
+      "step": 2700
+    },
+    {
+      "epoch": 3.45728,
+      "grad_norm": 0.6830167174339294,
+      "learning_rate": 4.461584633853542e-05,
+      "loss": 0.613,
+      "step": 2701
+    },
+    {
+      "epoch": 3.45856,
+      "grad_norm": 0.7231410145759583,
+      "learning_rate": 4.4613845538215285e-05,
+      "loss": 0.7048,
+      "step": 2702
+    },
+    {
+      "epoch": 3.45984,
+      "grad_norm": 0.6800318956375122,
+      "learning_rate": 4.4611844737895157e-05,
+      "loss": 0.6584,
+      "step": 2703
+    },
+    {
+      "epoch": 3.46112,
+      "grad_norm": 0.6745948791503906,
+      "learning_rate": 4.460984393757503e-05,
+      "loss": 0.6266,
+      "step": 2704
+    },
+    {
+      "epoch": 3.4624,
+      "grad_norm": 0.6727662086486816,
+      "learning_rate": 4.460784313725491e-05,
+      "loss": 0.6853,
+      "step": 2705
+    },
+    {
+      "epoch": 3.46368,
+      "grad_norm": 0.7079063653945923,
+      "learning_rate": 4.460584233693478e-05,
+      "loss": 0.659,
+      "step": 2706
+    },
+    {
+      "epoch": 3.46496,
+      "grad_norm": 0.6946644186973572,
+      "learning_rate": 4.460384153661465e-05,
+      "loss": 0.6874,
+      "step": 2707
+    },
+    {
+      "epoch": 3.46624,
+      "grad_norm": 0.6658247113227844,
+      "learning_rate": 4.460184073629452e-05,
+      "loss": 0.6348,
+      "step": 2708
+    },
+    {
+      "epoch": 3.46752,
+      "grad_norm": 0.6603049635887146,
+      "learning_rate": 4.4599839935974395e-05,
+      "loss": 0.6797,
+      "step": 2709
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.6833632588386536,
+      "learning_rate": 4.459783913565426e-05,
+      "loss": 0.6698,
+      "step": 2710
+    },
+    {
+      "epoch": 3.47008,
+      "grad_norm": 0.6787217855453491,
+      "learning_rate": 4.459583833533413e-05,
+      "loss": 0.6639,
+      "step": 2711
+    },
+    {
+      "epoch": 3.47136,
+      "grad_norm": 0.6830318570137024,
+      "learning_rate": 4.459383753501401e-05,
+      "loss": 0.6569,
+      "step": 2712
+    },
+    {
+      "epoch": 3.47264,
+      "grad_norm": 0.7140050530433655,
+      "learning_rate": 4.459183673469388e-05,
+      "loss": 0.7267,
+      "step": 2713
+    },
+    {
+      "epoch": 3.47392,
+      "grad_norm": 0.7092685103416443,
+      "learning_rate": 4.4589835934373754e-05,
+      "loss": 0.6364,
+      "step": 2714
+    },
+    {
+      "epoch": 3.4752,
+      "grad_norm": 0.6709200143814087,
+      "learning_rate": 4.4587835134053626e-05,
+      "loss": 0.6491,
+      "step": 2715
+    },
+    {
+      "epoch": 3.47648,
+      "grad_norm": 0.6295605897903442,
+      "learning_rate": 4.45858343337335e-05,
+      "loss": 0.6316,
+      "step": 2716
+    },
+    {
+      "epoch": 3.47776,
+      "grad_norm": 0.6191185116767883,
+      "learning_rate": 4.458383353341337e-05,
+      "loss": 0.6765,
+      "step": 2717
+    },
+    {
+      "epoch": 3.47904,
+      "grad_norm": 0.6618558168411255,
+      "learning_rate": 4.4581832733093234e-05,
+      "loss": 0.6509,
+      "step": 2718
+    },
+    {
+      "epoch": 3.48032,
+      "grad_norm": 0.6785896420478821,
+      "learning_rate": 4.457983193277311e-05,
+      "loss": 0.6234,
+      "step": 2719
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.7022203803062439,
+      "learning_rate": 4.4577831132452985e-05,
+      "loss": 0.7037,
+      "step": 2720
+    },
+    {
+      "epoch": 3.4828799999999998,
+      "grad_norm": 0.6869513988494873,
+      "learning_rate": 4.457583033213286e-05,
+      "loss": 0.6943,
+      "step": 2721
+    },
+    {
+      "epoch": 3.48416,
+      "grad_norm": 0.7011018395423889,
+      "learning_rate": 4.457382953181273e-05,
+      "loss": 0.693,
+      "step": 2722
+    },
+    {
+      "epoch": 3.48544,
+      "grad_norm": 0.6646562814712524,
+      "learning_rate": 4.45718287314926e-05,
+      "loss": 0.6767,
+      "step": 2723
+    },
+    {
+      "epoch": 3.48672,
+      "grad_norm": 0.6229949593544006,
+      "learning_rate": 4.456982793117247e-05,
+      "loss": 0.6177,
+      "step": 2724
+    },
+    {
+      "epoch": 3.488,
+      "grad_norm": 0.6661481261253357,
+      "learning_rate": 4.4567827130852344e-05,
+      "loss": 0.6401,
+      "step": 2725
+    },
+    {
+      "epoch": 3.48928,
+      "grad_norm": 0.6885098218917847,
+      "learning_rate": 4.4565826330532216e-05,
+      "loss": 0.6593,
+      "step": 2726
+    },
+    {
+      "epoch": 3.49056,
+      "grad_norm": 0.6794683933258057,
+      "learning_rate": 4.456382553021209e-05,
+      "loss": 0.641,
+      "step": 2727
+    },
+    {
+      "epoch": 3.49184,
+      "grad_norm": 0.6937509775161743,
+      "learning_rate": 4.456182472989196e-05,
+      "loss": 0.7112,
+      "step": 2728
+    },
+    {
+      "epoch": 3.4931200000000002,
+      "grad_norm": 0.7058037519454956,
+      "learning_rate": 4.455982392957183e-05,
+      "loss": 0.644,
+      "step": 2729
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.7502039670944214,
+      "learning_rate": 4.4557823129251704e-05,
+      "loss": 0.6871,
+      "step": 2730
+    },
+    {
+      "epoch": 3.49568,
+      "grad_norm": 0.6988638639450073,
+      "learning_rate": 4.4555822328931575e-05,
+      "loss": 0.6788,
+      "step": 2731
+    },
+    {
+      "epoch": 3.49696,
+      "grad_norm": 0.7283390760421753,
+      "learning_rate": 4.455382152861145e-05,
+      "loss": 0.6979,
+      "step": 2732
+    },
+    {
+      "epoch": 3.49824,
+      "grad_norm": 0.7279435396194458,
+      "learning_rate": 4.455182072829132e-05,
+      "loss": 0.7258,
+      "step": 2733
+    },
+    {
+      "epoch": 3.49952,
+      "grad_norm": 0.64103102684021,
+      "learning_rate": 4.454981992797119e-05,
+      "loss": 0.6253,
+      "step": 2734
+    },
+    {
+      "epoch": 3.5008,
+      "grad_norm": 0.6883367896080017,
+      "learning_rate": 4.454781912765106e-05,
+      "loss": 0.6615,
+      "step": 2735
+    },
+    {
+      "epoch": 3.50208,
+      "grad_norm": 0.6586665511131287,
+      "learning_rate": 4.4545818327330935e-05,
+      "loss": 0.6741,
+      "step": 2736
+    },
+    {
+      "epoch": 3.50336,
+      "grad_norm": 0.6852685213088989,
+      "learning_rate": 4.4543817527010807e-05,
+      "loss": 0.7288,
+      "step": 2737
+    },
+    {
+      "epoch": 3.50464,
+      "grad_norm": 0.6741801500320435,
+      "learning_rate": 4.454181672669068e-05,
+      "loss": 0.7345,
+      "step": 2738
+    },
+    {
+      "epoch": 3.50592,
+      "grad_norm": 0.6715037822723389,
+      "learning_rate": 4.453981592637055e-05,
+      "loss": 0.6525,
+      "step": 2739
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.6797910332679749,
+      "learning_rate": 4.453781512605042e-05,
+      "loss": 0.668,
+      "step": 2740
+    },
+    {
+      "epoch": 3.50848,
+      "grad_norm": 0.7185656428337097,
+      "learning_rate": 4.4535814325730294e-05,
+      "loss": 0.7052,
+      "step": 2741
+    },
+    {
+      "epoch": 3.50976,
+      "grad_norm": 0.6579523086547852,
+      "learning_rate": 4.4533813525410166e-05,
+      "loss": 0.6825,
+      "step": 2742
+    },
+    {
+      "epoch": 3.51104,
+      "grad_norm": 0.7128697037696838,
+      "learning_rate": 4.453181272509004e-05,
+      "loss": 0.7095,
+      "step": 2743
+    },
+    {
+      "epoch": 3.51232,
+      "grad_norm": 0.6932650208473206,
+      "learning_rate": 4.452981192476991e-05,
+      "loss": 0.6745,
+      "step": 2744
+    },
+    {
+      "epoch": 3.5136,
+      "grad_norm": 0.7073693871498108,
+      "learning_rate": 4.452781112444978e-05,
+      "loss": 0.6845,
+      "step": 2745
+    },
+    {
+      "epoch": 3.51488,
+      "grad_norm": 0.6620619893074036,
+      "learning_rate": 4.452581032412965e-05,
+      "loss": 0.6135,
+      "step": 2746
+    },
+    {
+      "epoch": 3.51616,
+      "grad_norm": 0.695382297039032,
+      "learning_rate": 4.4523809523809525e-05,
+      "loss": 0.6533,
+      "step": 2747
+    },
+    {
+      "epoch": 3.51744,
+      "grad_norm": 0.6715102195739746,
+      "learning_rate": 4.45218087234894e-05,
+      "loss": 0.6409,
+      "step": 2748
+    },
+    {
+      "epoch": 3.51872,
+      "grad_norm": 0.6616403460502625,
+      "learning_rate": 4.451980792316927e-05,
+      "loss": 0.6637,
+      "step": 2749
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.6827410459518433,
+      "learning_rate": 4.451780712284914e-05,
+      "loss": 0.6183,
+      "step": 2750
+    },
+    {
+      "epoch": 3.52128,
+      "grad_norm": 0.7313747406005859,
+      "learning_rate": 4.451580632252901e-05,
+      "loss": 0.6647,
+      "step": 2751
+    },
+    {
+      "epoch": 3.52256,
+      "grad_norm": 0.6950268149375916,
+      "learning_rate": 4.4513805522208884e-05,
+      "loss": 0.6705,
+      "step": 2752
+    },
+    {
+      "epoch": 3.52384,
+      "grad_norm": 0.6453522443771362,
+      "learning_rate": 4.4511804721888756e-05,
+      "loss": 0.666,
+      "step": 2753
+    },
+    {
+      "epoch": 3.5251200000000003,
+      "grad_norm": 0.6691879034042358,
+      "learning_rate": 4.450980392156863e-05,
+      "loss": 0.7045,
+      "step": 2754
+    },
+    {
+      "epoch": 3.5263999999999998,
+      "grad_norm": 0.7268353700637817,
+      "learning_rate": 4.450780312124851e-05,
+      "loss": 0.6803,
+      "step": 2755
+    },
+    {
+      "epoch": 3.52768,
+      "grad_norm": 0.6457922458648682,
+      "learning_rate": 4.450580232092837e-05,
+      "loss": 0.6423,
+      "step": 2756
+    },
+    {
+      "epoch": 3.52896,
+      "grad_norm": 0.6515114307403564,
+      "learning_rate": 4.4503801520608244e-05,
+      "loss": 0.7078,
+      "step": 2757
+    },
+    {
+      "epoch": 3.53024,
+      "grad_norm": 0.7063488960266113,
+      "learning_rate": 4.4501800720288115e-05,
+      "loss": 0.6547,
+      "step": 2758
+    },
+    {
+      "epoch": 3.53152,
+      "grad_norm": 0.7043952345848083,
+      "learning_rate": 4.449979991996799e-05,
+      "loss": 0.6184,
+      "step": 2759
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.7184996008872986,
+      "learning_rate": 4.449779911964786e-05,
+      "loss": 0.7037,
+      "step": 2760
+    },
+    {
+      "epoch": 3.53408,
+      "grad_norm": 0.6916483640670776,
+      "learning_rate": 4.449579831932773e-05,
+      "loss": 0.6334,
+      "step": 2761
+    },
+    {
+      "epoch": 3.53536,
+      "grad_norm": 0.7161397337913513,
+      "learning_rate": 4.449379751900761e-05,
+      "loss": 0.654,
+      "step": 2762
+    },
+    {
+      "epoch": 3.5366400000000002,
+      "grad_norm": 0.7488949298858643,
+      "learning_rate": 4.449179671868748e-05,
+      "loss": 0.7023,
+      "step": 2763
+    },
+    {
+      "epoch": 3.5379199999999997,
+      "grad_norm": 0.7058433890342712,
+      "learning_rate": 4.448979591836735e-05,
+      "loss": 0.6553,
+      "step": 2764
+    },
+    {
+      "epoch": 3.5392,
+      "grad_norm": 0.6619923114776611,
+      "learning_rate": 4.448779511804722e-05,
+      "loss": 0.6527,
+      "step": 2765
+    },
+    {
+      "epoch": 3.54048,
+      "grad_norm": 0.6837087869644165,
+      "learning_rate": 4.448579431772709e-05,
+      "loss": 0.6352,
+      "step": 2766
+    },
+    {
+      "epoch": 3.54176,
+      "grad_norm": 0.7077759504318237,
+      "learning_rate": 4.448379351740696e-05,
+      "loss": 0.6671,
+      "step": 2767
+    },
+    {
+      "epoch": 3.54304,
+      "grad_norm": 0.6346209049224854,
+      "learning_rate": 4.4481792717086834e-05,
+      "loss": 0.612,
+      "step": 2768
+    },
+    {
+      "epoch": 3.54432,
+      "grad_norm": 0.6402051448822021,
+      "learning_rate": 4.447979191676671e-05,
+      "loss": 0.6399,
+      "step": 2769
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.6839762926101685,
+      "learning_rate": 4.4477791116446585e-05,
+      "loss": 0.7407,
+      "step": 2770
+    },
+    {
+      "epoch": 3.54688,
+      "grad_norm": 0.688944399356842,
+      "learning_rate": 4.4475790316126456e-05,
+      "loss": 0.6762,
+      "step": 2771
+    },
+    {
+      "epoch": 3.54816,
+      "grad_norm": 0.642185628414154,
+      "learning_rate": 4.447378951580632e-05,
+      "loss": 0.6375,
+      "step": 2772
+    },
+    {
+      "epoch": 3.54944,
+      "grad_norm": 0.6515256762504578,
+      "learning_rate": 4.447178871548619e-05,
+      "loss": 0.5938,
+      "step": 2773
+    },
+    {
+      "epoch": 3.55072,
+      "grad_norm": 0.6609624624252319,
+      "learning_rate": 4.4469787915166065e-05,
+      "loss": 0.7017,
+      "step": 2774
+    },
+    {
+      "epoch": 3.552,
+      "grad_norm": 0.6748828291893005,
+      "learning_rate": 4.446778711484594e-05,
+      "loss": 0.6911,
+      "step": 2775
+    },
+    {
+      "epoch": 3.55328,
+      "grad_norm": 0.7154529094696045,
+      "learning_rate": 4.4465786314525816e-05,
+      "loss": 0.6708,
+      "step": 2776
+    },
+    {
+      "epoch": 3.55456,
+      "grad_norm": 0.6632035374641418,
+      "learning_rate": 4.446378551420569e-05,
+      "loss": 0.6461,
+      "step": 2777
+    },
+    {
+      "epoch": 3.55584,
+      "grad_norm": 0.7373719811439514,
+      "learning_rate": 4.446178471388556e-05,
+      "loss": 0.7789,
+      "step": 2778
+    },
+    {
+      "epoch": 3.55712,
+      "grad_norm": 0.7006744146347046,
+      "learning_rate": 4.445978391356543e-05,
+      "loss": 0.6923,
+      "step": 2779
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.699441134929657,
+      "learning_rate": 4.4457783113245296e-05,
+      "loss": 0.7505,
+      "step": 2780
+    },
+    {
+      "epoch": 3.55968,
+      "grad_norm": 0.6537452936172485,
+      "learning_rate": 4.445578231292517e-05,
+      "loss": 0.671,
+      "step": 2781
+    },
+    {
+      "epoch": 3.56096,
+      "grad_norm": 0.6859011650085449,
+      "learning_rate": 4.445378151260504e-05,
+      "loss": 0.6701,
+      "step": 2782
+    },
+    {
+      "epoch": 3.56224,
+      "grad_norm": 0.6629458069801331,
+      "learning_rate": 4.445178071228492e-05,
+      "loss": 0.6463,
+      "step": 2783
+    },
+    {
+      "epoch": 3.56352,
+      "grad_norm": 0.6698895692825317,
+      "learning_rate": 4.444977991196479e-05,
+      "loss": 0.6575,
+      "step": 2784
+    },
+    {
+      "epoch": 3.5648,
+      "grad_norm": 0.7185133695602417,
+      "learning_rate": 4.444777911164466e-05,
+      "loss": 0.7278,
+      "step": 2785
+    },
+    {
+      "epoch": 3.56608,
+      "grad_norm": 0.699368417263031,
+      "learning_rate": 4.4445778311324534e-05,
+      "loss": 0.7271,
+      "step": 2786
+    },
+    {
+      "epoch": 3.56736,
+      "grad_norm": 0.6818522810935974,
+      "learning_rate": 4.4443777511004406e-05,
+      "loss": 0.6762,
+      "step": 2787
+    },
+    {
+      "epoch": 3.5686400000000003,
+      "grad_norm": 0.6821078062057495,
+      "learning_rate": 4.444177671068427e-05,
+      "loss": 0.6452,
+      "step": 2788
+    },
+    {
+      "epoch": 3.5699199999999998,
+      "grad_norm": 0.6925137042999268,
+      "learning_rate": 4.443977591036414e-05,
+      "loss": 0.6354,
+      "step": 2789
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.6931013464927673,
+      "learning_rate": 4.443777511004402e-05,
+      "loss": 0.6438,
+      "step": 2790
+    },
+    {
+      "epoch": 3.57248,
+      "grad_norm": 0.6680043935775757,
+      "learning_rate": 4.4435774309723894e-05,
+      "loss": 0.6732,
+      "step": 2791
+    },
+    {
+      "epoch": 3.57376,
+      "grad_norm": 0.7447218894958496,
+      "learning_rate": 4.4433773509403765e-05,
+      "loss": 0.7314,
+      "step": 2792
+    },
+    {
+      "epoch": 3.57504,
+      "grad_norm": 0.6944176554679871,
+      "learning_rate": 4.443177270908364e-05,
+      "loss": 0.6333,
+      "step": 2793
+    },
+    {
+      "epoch": 3.57632,
+      "grad_norm": 0.7107727527618408,
+      "learning_rate": 4.442977190876351e-05,
+      "loss": 0.673,
+      "step": 2794
+    },
+    {
+      "epoch": 3.5776,
+      "grad_norm": 0.6927473545074463,
+      "learning_rate": 4.442777110844338e-05,
+      "loss": 0.6667,
+      "step": 2795
+    },
+    {
+      "epoch": 3.57888,
+      "grad_norm": 0.7122114896774292,
+      "learning_rate": 4.4425770308123246e-05,
+      "loss": 0.6544,
+      "step": 2796
+    },
+    {
+      "epoch": 3.5801600000000002,
+      "grad_norm": 0.6794743537902832,
+      "learning_rate": 4.4423769507803125e-05,
+      "loss": 0.6743,
+      "step": 2797
+    },
+    {
+      "epoch": 3.5814399999999997,
+      "grad_norm": 0.6645591259002686,
+      "learning_rate": 4.4421768707482997e-05,
+      "loss": 0.7121,
+      "step": 2798
+    },
+    {
+      "epoch": 3.58272,
+      "grad_norm": 0.6730340123176575,
+      "learning_rate": 4.441976790716287e-05,
+      "loss": 0.6619,
+      "step": 2799
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.6491687297821045,
+      "learning_rate": 4.441776710684274e-05,
+      "loss": 0.6227,
+      "step": 2800
+    },
+    {
+      "epoch": 3.58528,
+      "grad_norm": 0.698142945766449,
+      "learning_rate": 4.441576630652261e-05,
+      "loss": 0.6708,
+      "step": 2801
+    },
+    {
+      "epoch": 3.58656,
+      "grad_norm": 0.7014725804328918,
+      "learning_rate": 4.4413765506202484e-05,
+      "loss": 0.6487,
+      "step": 2802
+    },
+    {
+      "epoch": 3.58784,
+      "grad_norm": 0.6868982315063477,
+      "learning_rate": 4.4411764705882356e-05,
+      "loss": 0.719,
+      "step": 2803
+    },
+    {
+      "epoch": 3.58912,
+      "grad_norm": 0.6748588681221008,
+      "learning_rate": 4.440976390556223e-05,
+      "loss": 0.6158,
+      "step": 2804
+    },
+    {
+      "epoch": 3.5904,
+      "grad_norm": 0.7001504898071289,
+      "learning_rate": 4.44077631052421e-05,
+      "loss": 0.6752,
+      "step": 2805
+    },
+    {
+      "epoch": 3.59168,
+      "grad_norm": 0.6581389904022217,
+      "learning_rate": 4.440576230492197e-05,
+      "loss": 0.6909,
+      "step": 2806
+    },
+    {
+      "epoch": 3.59296,
+      "grad_norm": 0.6843972206115723,
+      "learning_rate": 4.440376150460184e-05,
+      "loss": 0.7346,
+      "step": 2807
+    },
+    {
+      "epoch": 3.59424,
+      "grad_norm": 0.6761201620101929,
+      "learning_rate": 4.4401760704281715e-05,
+      "loss": 0.6827,
+      "step": 2808
+    },
+    {
+      "epoch": 3.59552,
+      "grad_norm": 0.7041255235671997,
+      "learning_rate": 4.439975990396159e-05,
+      "loss": 0.6768,
+      "step": 2809
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.650201141834259,
+      "learning_rate": 4.439775910364146e-05,
+      "loss": 0.6364,
+      "step": 2810
+    },
+    {
+      "epoch": 3.59808,
+      "grad_norm": 0.7019467353820801,
+      "learning_rate": 4.439575830332133e-05,
+      "loss": 0.6991,
+      "step": 2811
+    },
+    {
+      "epoch": 3.59936,
+      "grad_norm": 0.6987199783325195,
+      "learning_rate": 4.43937575030012e-05,
+      "loss": 0.6525,
+      "step": 2812
+    },
+    {
+      "epoch": 3.60064,
+      "grad_norm": 0.6764307618141174,
+      "learning_rate": 4.4391756702681074e-05,
+      "loss": 0.6271,
+      "step": 2813
+    },
+    {
+      "epoch": 3.60192,
+      "grad_norm": 0.6464497447013855,
+      "learning_rate": 4.4389755902360946e-05,
+      "loss": 0.6348,
+      "step": 2814
+    },
+    {
+      "epoch": 3.6032,
+      "grad_norm": 0.704531192779541,
+      "learning_rate": 4.438775510204082e-05,
+      "loss": 0.6944,
+      "step": 2815
+    },
+    {
+      "epoch": 3.60448,
+      "grad_norm": 0.7004350423812866,
+      "learning_rate": 4.438575430172069e-05,
+      "loss": 0.6299,
+      "step": 2816
+    },
+    {
+      "epoch": 3.60576,
+      "grad_norm": 0.6948818564414978,
+      "learning_rate": 4.438375350140056e-05,
+      "loss": 0.6662,
+      "step": 2817
+    },
+    {
+      "epoch": 3.60704,
+      "grad_norm": 0.668674647808075,
+      "learning_rate": 4.4381752701080434e-05,
+      "loss": 0.6203,
+      "step": 2818
+    },
+    {
+      "epoch": 3.60832,
+      "grad_norm": 0.6800994277000427,
+      "learning_rate": 4.4379751900760306e-05,
+      "loss": 0.6619,
+      "step": 2819
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.667832612991333,
+      "learning_rate": 4.437775110044018e-05,
+      "loss": 0.6441,
+      "step": 2820
+    },
+    {
+      "epoch": 3.61088,
+      "grad_norm": 0.695292592048645,
+      "learning_rate": 4.437575030012005e-05,
+      "loss": 0.7054,
+      "step": 2821
+    },
+    {
+      "epoch": 3.6121600000000003,
+      "grad_norm": 0.7079086303710938,
+      "learning_rate": 4.437374949979992e-05,
+      "loss": 0.6772,
+      "step": 2822
+    },
+    {
+      "epoch": 3.6134399999999998,
+      "grad_norm": 0.6839944124221802,
+      "learning_rate": 4.437174869947979e-05,
+      "loss": 0.6553,
+      "step": 2823
+    },
+    {
+      "epoch": 3.61472,
+      "grad_norm": 0.6628500819206238,
+      "learning_rate": 4.4369747899159665e-05,
+      "loss": 0.6787,
+      "step": 2824
+    },
+    {
+      "epoch": 3.616,
+      "grad_norm": 0.6501266360282898,
+      "learning_rate": 4.4367747098839543e-05,
+      "loss": 0.6575,
+      "step": 2825
+    },
+    {
+      "epoch": 3.61728,
+      "grad_norm": 0.6804488301277161,
+      "learning_rate": 4.436574629851941e-05,
+      "loss": 0.647,
+      "step": 2826
+    },
+    {
+      "epoch": 3.61856,
+      "grad_norm": 0.7033595442771912,
+      "learning_rate": 4.436374549819928e-05,
+      "loss": 0.6988,
+      "step": 2827
+    },
+    {
+      "epoch": 3.61984,
+      "grad_norm": 0.6809185147285461,
+      "learning_rate": 4.436174469787915e-05,
+      "loss": 0.6726,
+      "step": 2828
+    },
+    {
+      "epoch": 3.62112,
+      "grad_norm": 0.7250016927719116,
+      "learning_rate": 4.4359743897559024e-05,
+      "loss": 0.6378,
+      "step": 2829
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.6764475703239441,
+      "learning_rate": 4.4357743097238896e-05,
+      "loss": 0.6564,
+      "step": 2830
+    },
+    {
+      "epoch": 3.6236800000000002,
+      "grad_norm": 0.7175212502479553,
+      "learning_rate": 4.435574229691877e-05,
+      "loss": 0.7327,
+      "step": 2831
+    },
+    {
+      "epoch": 3.6249599999999997,
+      "grad_norm": 0.6657410264015198,
+      "learning_rate": 4.4353741496598646e-05,
+      "loss": 0.6475,
+      "step": 2832
+    },
+    {
+      "epoch": 3.62624,
+      "grad_norm": 0.6739627122879028,
+      "learning_rate": 4.435174069627852e-05,
+      "loss": 0.6444,
+      "step": 2833
+    },
+    {
+      "epoch": 3.62752,
+      "grad_norm": 0.7084604501724243,
+      "learning_rate": 4.434973989595838e-05,
+      "loss": 0.6923,
+      "step": 2834
+    },
+    {
+      "epoch": 3.6288,
+      "grad_norm": 0.7034947276115417,
+      "learning_rate": 4.4347739095638255e-05,
+      "loss": 0.696,
+      "step": 2835
+    },
+    {
+      "epoch": 3.63008,
+      "grad_norm": 0.6603802442550659,
+      "learning_rate": 4.434573829531813e-05,
+      "loss": 0.6625,
+      "step": 2836
+    },
+    {
+      "epoch": 3.63136,
+      "grad_norm": 0.6719616055488586,
+      "learning_rate": 4.4343737494998e-05,
+      "loss": 0.6623,
+      "step": 2837
+    },
+    {
+      "epoch": 3.63264,
+      "grad_norm": 0.6790746450424194,
+      "learning_rate": 4.434173669467787e-05,
+      "loss": 0.7016,
+      "step": 2838
+    },
+    {
+      "epoch": 3.63392,
+      "grad_norm": 0.6688959002494812,
+      "learning_rate": 4.433973589435775e-05,
+      "loss": 0.6411,
+      "step": 2839
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.7280130386352539,
+      "learning_rate": 4.433773509403762e-05,
+      "loss": 0.7061,
+      "step": 2840
+    },
+    {
+      "epoch": 3.63648,
+      "grad_norm": 0.6903562545776367,
+      "learning_rate": 4.433573429371749e-05,
+      "loss": 0.6433,
+      "step": 2841
+    },
+    {
+      "epoch": 3.63776,
+      "grad_norm": 0.713874101638794,
+      "learning_rate": 4.433373349339736e-05,
+      "loss": 0.6956,
+      "step": 2842
+    },
+    {
+      "epoch": 3.63904,
+      "grad_norm": 0.6836928129196167,
+      "learning_rate": 4.433173269307723e-05,
+      "loss": 0.6829,
+      "step": 2843
+    },
+    {
+      "epoch": 3.64032,
+      "grad_norm": 0.6967549920082092,
+      "learning_rate": 4.43297318927571e-05,
+      "loss": 0.6793,
+      "step": 2844
+    },
+    {
+      "epoch": 3.6416,
+      "grad_norm": 0.7121291160583496,
+      "learning_rate": 4.4327731092436974e-05,
+      "loss": 0.6855,
+      "step": 2845
+    },
+    {
+      "epoch": 3.64288,
+      "grad_norm": 0.6488426327705383,
+      "learning_rate": 4.432573029211685e-05,
+      "loss": 0.6063,
+      "step": 2846
+    },
+    {
+      "epoch": 3.64416,
+      "grad_norm": 0.669741153717041,
+      "learning_rate": 4.4323729491796724e-05,
+      "loss": 0.6383,
+      "step": 2847
+    },
+    {
+      "epoch": 3.64544,
+      "grad_norm": 0.6769450306892395,
+      "learning_rate": 4.4321728691476596e-05,
+      "loss": 0.6116,
+      "step": 2848
+    },
+    {
+      "epoch": 3.64672,
+      "grad_norm": 0.6796509623527527,
+      "learning_rate": 4.431972789115647e-05,
+      "loss": 0.6567,
+      "step": 2849
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.6543831825256348,
+      "learning_rate": 4.431772709083633e-05,
+      "loss": 0.6122,
+      "step": 2850
+    },
+    {
+      "epoch": 3.64928,
+      "grad_norm": 0.6729878187179565,
+      "learning_rate": 4.4315726290516205e-05,
+      "loss": 0.6636,
+      "step": 2851
+    },
+    {
+      "epoch": 3.65056,
+      "grad_norm": 0.7151674628257751,
+      "learning_rate": 4.431372549019608e-05,
+      "loss": 0.6863,
+      "step": 2852
+    },
+    {
+      "epoch": 3.65184,
+      "grad_norm": 0.6241359710693359,
+      "learning_rate": 4.4311724689875955e-05,
+      "loss": 0.6208,
+      "step": 2853
+    },
+    {
+      "epoch": 3.65312,
+      "grad_norm": 0.6886858940124512,
+      "learning_rate": 4.430972388955583e-05,
+      "loss": 0.679,
+      "step": 2854
+    },
+    {
+      "epoch": 3.6544,
+      "grad_norm": 0.7163287997245789,
+      "learning_rate": 4.43077230892357e-05,
+      "loss": 0.6783,
+      "step": 2855
+    },
+    {
+      "epoch": 3.6556800000000003,
+      "grad_norm": 0.698589026927948,
+      "learning_rate": 4.430572228891557e-05,
+      "loss": 0.6695,
+      "step": 2856
+    },
+    {
+      "epoch": 3.6569599999999998,
+      "grad_norm": 0.7296429872512817,
+      "learning_rate": 4.430372148859544e-05,
+      "loss": 0.7372,
+      "step": 2857
+    },
+    {
+      "epoch": 3.65824,
+      "grad_norm": 0.6619814038276672,
+      "learning_rate": 4.430172068827531e-05,
+      "loss": 0.6973,
+      "step": 2858
+    },
+    {
+      "epoch": 3.65952,
+      "grad_norm": 0.6731798052787781,
+      "learning_rate": 4.429971988795518e-05,
+      "loss": 0.652,
+      "step": 2859
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.7068068981170654,
+      "learning_rate": 4.429771908763505e-05,
+      "loss": 0.715,
+      "step": 2860
+    },
+    {
+      "epoch": 3.66208,
+      "grad_norm": 0.6874963641166687,
+      "learning_rate": 4.429571828731493e-05,
+      "loss": 0.7342,
+      "step": 2861
+    },
+    {
+      "epoch": 3.66336,
+      "grad_norm": 0.6647247672080994,
+      "learning_rate": 4.42937174869948e-05,
+      "loss": 0.6638,
+      "step": 2862
+    },
+    {
+      "epoch": 3.66464,
+      "grad_norm": 0.6708807349205017,
+      "learning_rate": 4.4291716686674674e-05,
+      "loss": 0.6494,
+      "step": 2863
+    },
+    {
+      "epoch": 3.66592,
+      "grad_norm": 0.6945130825042725,
+      "learning_rate": 4.4289715886354546e-05,
+      "loss": 0.6974,
+      "step": 2864
+    },
+    {
+      "epoch": 3.6672000000000002,
+      "grad_norm": 0.6519247889518738,
+      "learning_rate": 4.428771508603442e-05,
+      "loss": 0.5756,
+      "step": 2865
+    },
+    {
+      "epoch": 3.6684799999999997,
+      "grad_norm": 0.6893465518951416,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 0.6507,
+      "step": 2866
+    },
+    {
+      "epoch": 3.66976,
+      "grad_norm": 0.6804561018943787,
+      "learning_rate": 4.4283713485394155e-05,
+      "loss": 0.6864,
+      "step": 2867
+    },
+    {
+      "epoch": 3.67104,
+      "grad_norm": 0.7122829556465149,
+      "learning_rate": 4.428171268507403e-05,
+      "loss": 0.7419,
+      "step": 2868
+    },
+    {
+      "epoch": 3.67232,
+      "grad_norm": 0.64589524269104,
+      "learning_rate": 4.4279711884753905e-05,
+      "loss": 0.6927,
+      "step": 2869
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.6762197613716125,
+      "learning_rate": 4.427771108443378e-05,
+      "loss": 0.6725,
+      "step": 2870
+    },
+    {
+      "epoch": 3.67488,
+      "grad_norm": 0.6425807476043701,
+      "learning_rate": 4.427571028411365e-05,
+      "loss": 0.631,
+      "step": 2871
+    },
+    {
+      "epoch": 3.67616,
+      "grad_norm": 0.6912189722061157,
+      "learning_rate": 4.427370948379352e-05,
+      "loss": 0.6934,
+      "step": 2872
+    },
+    {
+      "epoch": 3.67744,
+      "grad_norm": 0.6705701947212219,
+      "learning_rate": 4.427170868347339e-05,
+      "loss": 0.6638,
+      "step": 2873
+    },
+    {
+      "epoch": 3.67872,
+      "grad_norm": 0.7085809111595154,
+      "learning_rate": 4.426970788315326e-05,
+      "loss": 0.6887,
+      "step": 2874
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.683226466178894,
+      "learning_rate": 4.4267707082833136e-05,
+      "loss": 0.658,
+      "step": 2875
+    },
+    {
+      "epoch": 3.68128,
+      "grad_norm": 0.6895299553871155,
+      "learning_rate": 4.426570628251301e-05,
+      "loss": 0.638,
+      "step": 2876
+    },
+    {
+      "epoch": 3.68256,
+      "grad_norm": 0.6925261616706848,
+      "learning_rate": 4.426370548219288e-05,
+      "loss": 0.6365,
+      "step": 2877
+    },
+    {
+      "epoch": 3.68384,
+      "grad_norm": 0.7004500031471252,
+      "learning_rate": 4.426170468187275e-05,
+      "loss": 0.6739,
+      "step": 2878
+    },
+    {
+      "epoch": 3.68512,
+      "grad_norm": 0.7290331125259399,
+      "learning_rate": 4.4259703881552624e-05,
+      "loss": 0.6495,
+      "step": 2879
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.6692587733268738,
+      "learning_rate": 4.4257703081232496e-05,
+      "loss": 0.6887,
+      "step": 2880
+    },
+    {
+      "epoch": 3.68768,
+      "grad_norm": 0.6933017373085022,
+      "learning_rate": 4.425570228091237e-05,
+      "loss": 0.6722,
+      "step": 2881
+    },
+    {
+      "epoch": 3.68896,
+      "grad_norm": 0.707172691822052,
+      "learning_rate": 4.425370148059224e-05,
+      "loss": 0.7151,
+      "step": 2882
+    },
+    {
+      "epoch": 3.69024,
+      "grad_norm": 0.6823884844779968,
+      "learning_rate": 4.425170068027211e-05,
+      "loss": 0.6213,
+      "step": 2883
+    },
+    {
+      "epoch": 3.69152,
+      "grad_norm": 0.7094587087631226,
+      "learning_rate": 4.424969987995198e-05,
+      "loss": 0.7033,
+      "step": 2884
+    },
+    {
+      "epoch": 3.6928,
+      "grad_norm": 0.6758087277412415,
+      "learning_rate": 4.4247699079631855e-05,
+      "loss": 0.687,
+      "step": 2885
+    },
+    {
+      "epoch": 3.69408,
+      "grad_norm": 0.6843044757843018,
+      "learning_rate": 4.424569827931173e-05,
+      "loss": 0.6536,
+      "step": 2886
+    },
+    {
+      "epoch": 3.69536,
+      "grad_norm": 0.7044915556907654,
+      "learning_rate": 4.42436974789916e-05,
+      "loss": 0.7335,
+      "step": 2887
+    },
+    {
+      "epoch": 3.69664,
+      "grad_norm": 0.6591890454292297,
+      "learning_rate": 4.424169667867147e-05,
+      "loss": 0.645,
+      "step": 2888
+    },
+    {
+      "epoch": 3.69792,
+      "grad_norm": 0.7025469541549683,
+      "learning_rate": 4.423969587835134e-05,
+      "loss": 0.6961,
+      "step": 2889
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.663060188293457,
+      "learning_rate": 4.4237695078031214e-05,
+      "loss": 0.6113,
+      "step": 2890
+    },
+    {
+      "epoch": 3.7004799999999998,
+      "grad_norm": 0.6972171068191528,
+      "learning_rate": 4.4235694277711086e-05,
+      "loss": 0.6683,
+      "step": 2891
+    },
+    {
+      "epoch": 3.70176,
+      "grad_norm": 0.7062972187995911,
+      "learning_rate": 4.423369347739096e-05,
+      "loss": 0.7317,
+      "step": 2892
+    },
+    {
+      "epoch": 3.70304,
+      "grad_norm": 0.6868116855621338,
+      "learning_rate": 4.423169267707083e-05,
+      "loss": 0.6834,
+      "step": 2893
+    },
+    {
+      "epoch": 3.70432,
+      "grad_norm": 0.6620085835456848,
+      "learning_rate": 4.42296918767507e-05,
+      "loss": 0.6806,
+      "step": 2894
+    },
+    {
+      "epoch": 3.7056,
+      "grad_norm": 0.6830658912658691,
+      "learning_rate": 4.4227691076430573e-05,
+      "loss": 0.6771,
+      "step": 2895
+    },
+    {
+      "epoch": 3.70688,
+      "grad_norm": 0.6770226359367371,
+      "learning_rate": 4.4225690276110445e-05,
+      "loss": 0.6476,
+      "step": 2896
+    },
+    {
+      "epoch": 3.70816,
+      "grad_norm": 0.6751776933670044,
+      "learning_rate": 4.422368947579032e-05,
+      "loss": 0.6402,
+      "step": 2897
+    },
+    {
+      "epoch": 3.70944,
+      "grad_norm": 0.7226163148880005,
+      "learning_rate": 4.422168867547019e-05,
+      "loss": 0.7128,
+      "step": 2898
+    },
+    {
+      "epoch": 3.7107200000000002,
+      "grad_norm": 0.6574375033378601,
+      "learning_rate": 4.421968787515006e-05,
+      "loss": 0.7031,
+      "step": 2899
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.6752791404724121,
+      "learning_rate": 4.421768707482993e-05,
+      "loss": 0.7141,
+      "step": 2900
+    },
+    {
+      "epoch": 3.71328,
+      "grad_norm": 0.6757639646530151,
+      "learning_rate": 4.4215686274509805e-05,
+      "loss": 0.707,
+      "step": 2901
+    },
+    {
+      "epoch": 3.71456,
+      "grad_norm": 0.6414822936058044,
+      "learning_rate": 4.4213685474189676e-05,
+      "loss": 0.6637,
+      "step": 2902
+    },
+    {
+      "epoch": 3.71584,
+      "grad_norm": 0.7052628397941589,
+      "learning_rate": 4.4211684673869555e-05,
+      "loss": 0.6547,
+      "step": 2903
+    },
+    {
+      "epoch": 3.71712,
+      "grad_norm": 0.6637598872184753,
+      "learning_rate": 4.420968387354942e-05,
+      "loss": 0.6694,
+      "step": 2904
+    },
+    {
+      "epoch": 3.7184,
+      "grad_norm": 0.684610903263092,
+      "learning_rate": 4.420768307322929e-05,
+      "loss": 0.6892,
+      "step": 2905
+    },
+    {
+      "epoch": 3.71968,
+      "grad_norm": 0.6572564244270325,
+      "learning_rate": 4.4205682272909164e-05,
+      "loss": 0.6579,
+      "step": 2906
+    },
+    {
+      "epoch": 3.72096,
+      "grad_norm": 0.6893813014030457,
+      "learning_rate": 4.4203681472589036e-05,
+      "loss": 0.6612,
+      "step": 2907
+    },
+    {
+      "epoch": 3.72224,
+      "grad_norm": 0.6515512466430664,
+      "learning_rate": 4.420168067226891e-05,
+      "loss": 0.6371,
+      "step": 2908
+    },
+    {
+      "epoch": 3.72352,
+      "grad_norm": 0.7329177260398865,
+      "learning_rate": 4.419967987194878e-05,
+      "loss": 0.7172,
+      "step": 2909
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.6419927477836609,
+      "learning_rate": 4.419767907162866e-05,
+      "loss": 0.6113,
+      "step": 2910
+    },
+    {
+      "epoch": 3.72608,
+      "grad_norm": 0.6344096064567566,
+      "learning_rate": 4.419567827130853e-05,
+      "loss": 0.6277,
+      "step": 2911
+    },
+    {
+      "epoch": 3.72736,
+      "grad_norm": 0.7253386974334717,
+      "learning_rate": 4.4193677470988395e-05,
+      "loss": 0.6749,
+      "step": 2912
+    },
+    {
+      "epoch": 3.72864,
+      "grad_norm": 0.6708924174308777,
+      "learning_rate": 4.419167667066827e-05,
+      "loss": 0.6019,
+      "step": 2913
+    },
+    {
+      "epoch": 3.72992,
+      "grad_norm": 0.6822198033332825,
+      "learning_rate": 4.418967587034814e-05,
+      "loss": 0.6393,
+      "step": 2914
+    },
+    {
+      "epoch": 3.7312,
+      "grad_norm": 0.6997161507606506,
+      "learning_rate": 4.418767507002801e-05,
+      "loss": 0.668,
+      "step": 2915
+    },
+    {
+      "epoch": 3.73248,
+      "grad_norm": 0.67588210105896,
+      "learning_rate": 4.418567426970788e-05,
+      "loss": 0.6117,
+      "step": 2916
+    },
+    {
+      "epoch": 3.73376,
+      "grad_norm": 0.7122563123703003,
+      "learning_rate": 4.418367346938776e-05,
+      "loss": 0.6749,
+      "step": 2917
+    },
+    {
+      "epoch": 3.73504,
+      "grad_norm": 0.681527316570282,
+      "learning_rate": 4.418167266906763e-05,
+      "loss": 0.682,
+      "step": 2918
+    },
+    {
+      "epoch": 3.73632,
+      "grad_norm": 0.6719216704368591,
+      "learning_rate": 4.4179671868747505e-05,
+      "loss": 0.684,
+      "step": 2919
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.6405019164085388,
+      "learning_rate": 4.417767106842737e-05,
+      "loss": 0.6527,
+      "step": 2920
+    },
+    {
+      "epoch": 3.73888,
+      "grad_norm": 0.6861347556114197,
+      "learning_rate": 4.417567026810724e-05,
+      "loss": 0.6983,
+      "step": 2921
+    },
+    {
+      "epoch": 3.74016,
+      "grad_norm": 0.6605308055877686,
+      "learning_rate": 4.4173669467787114e-05,
+      "loss": 0.6757,
+      "step": 2922
+    },
+    {
+      "epoch": 3.74144,
+      "grad_norm": 0.6555644869804382,
+      "learning_rate": 4.4171668667466985e-05,
+      "loss": 0.681,
+      "step": 2923
+    },
+    {
+      "epoch": 3.7427200000000003,
+      "grad_norm": 0.6855587363243103,
+      "learning_rate": 4.4169667867146864e-05,
+      "loss": 0.6807,
+      "step": 2924
+    },
+    {
+      "epoch": 3.7439999999999998,
+      "grad_norm": 0.7204174399375916,
+      "learning_rate": 4.4167667066826736e-05,
+      "loss": 0.7226,
+      "step": 2925
+    },
+    {
+      "epoch": 3.74528,
+      "grad_norm": 0.6882557272911072,
+      "learning_rate": 4.416566626650661e-05,
+      "loss": 0.7127,
+      "step": 2926
+    },
+    {
+      "epoch": 3.74656,
+      "grad_norm": 0.6996170878410339,
+      "learning_rate": 4.416366546618648e-05,
+      "loss": 0.6309,
+      "step": 2927
+    },
+    {
+      "epoch": 3.74784,
+      "grad_norm": 0.6991925239562988,
+      "learning_rate": 4.4161664665866345e-05,
+      "loss": 0.6571,
+      "step": 2928
+    },
+    {
+      "epoch": 3.74912,
+      "grad_norm": 0.7202059030532837,
+      "learning_rate": 4.4159663865546217e-05,
+      "loss": 0.6948,
+      "step": 2929
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.7513526678085327,
+      "learning_rate": 4.415766306522609e-05,
+      "loss": 0.7387,
+      "step": 2930
+    },
+    {
+      "epoch": 3.75168,
+      "grad_norm": 0.6940328478813171,
+      "learning_rate": 4.415566226490597e-05,
+      "loss": 0.7083,
+      "step": 2931
+    },
+    {
+      "epoch": 3.75296,
+      "grad_norm": 0.665745198726654,
+      "learning_rate": 4.415366146458584e-05,
+      "loss": 0.6208,
+      "step": 2932
+    },
+    {
+      "epoch": 3.7542400000000002,
+      "grad_norm": 0.7266408801078796,
+      "learning_rate": 4.415166066426571e-05,
+      "loss": 0.6952,
+      "step": 2933
+    },
+    {
+      "epoch": 3.7555199999999997,
+      "grad_norm": 0.7092468738555908,
+      "learning_rate": 4.414965986394558e-05,
+      "loss": 0.6157,
+      "step": 2934
+    },
+    {
+      "epoch": 3.7568,
+      "grad_norm": 0.6691955327987671,
+      "learning_rate": 4.4147659063625454e-05,
+      "loss": 0.644,
+      "step": 2935
+    },
+    {
+      "epoch": 3.75808,
+      "grad_norm": 0.679655909538269,
+      "learning_rate": 4.414565826330532e-05,
+      "loss": 0.6441,
+      "step": 2936
+    },
+    {
+      "epoch": 3.75936,
+      "grad_norm": 0.7097312211990356,
+      "learning_rate": 4.414365746298519e-05,
+      "loss": 0.7353,
+      "step": 2937
+    },
+    {
+      "epoch": 3.76064,
+      "grad_norm": 0.730174720287323,
+      "learning_rate": 4.414165666266507e-05,
+      "loss": 0.6899,
+      "step": 2938
+    },
+    {
+      "epoch": 3.76192,
+      "grad_norm": 0.6965400576591492,
+      "learning_rate": 4.413965586234494e-05,
+      "loss": 0.654,
+      "step": 2939
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.7156361937522888,
+      "learning_rate": 4.4137655062024814e-05,
+      "loss": 0.6471,
+      "step": 2940
+    },
+    {
+      "epoch": 3.76448,
+      "grad_norm": 0.7233887910842896,
+      "learning_rate": 4.4135654261704686e-05,
+      "loss": 0.6766,
+      "step": 2941
+    },
+    {
+      "epoch": 3.76576,
+      "grad_norm": 0.708551287651062,
+      "learning_rate": 4.413365346138456e-05,
+      "loss": 0.6497,
+      "step": 2942
+    },
+    {
+      "epoch": 3.76704,
+      "grad_norm": 0.6958677768707275,
+      "learning_rate": 4.413165266106443e-05,
+      "loss": 0.6522,
+      "step": 2943
+    },
+    {
+      "epoch": 3.76832,
+      "grad_norm": 0.693276047706604,
+      "learning_rate": 4.4129651860744294e-05,
+      "loss": 0.6597,
+      "step": 2944
+    },
+    {
+      "epoch": 3.7696,
+      "grad_norm": 0.7062802910804749,
+      "learning_rate": 4.412765106042417e-05,
+      "loss": 0.6558,
+      "step": 2945
+    },
+    {
+      "epoch": 3.77088,
+      "grad_norm": 0.7109881639480591,
+      "learning_rate": 4.4125650260104045e-05,
+      "loss": 0.6426,
+      "step": 2946
+    },
+    {
+      "epoch": 3.77216,
+      "grad_norm": 0.6618438959121704,
+      "learning_rate": 4.412364945978392e-05,
+      "loss": 0.6421,
+      "step": 2947
+    },
+    {
+      "epoch": 3.77344,
+      "grad_norm": 0.6410247683525085,
+      "learning_rate": 4.412164865946379e-05,
+      "loss": 0.644,
+      "step": 2948
+    },
+    {
+      "epoch": 3.77472,
+      "grad_norm": 0.705467939376831,
+      "learning_rate": 4.411964785914366e-05,
+      "loss": 0.6873,
+      "step": 2949
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.6557707190513611,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.6677,
+      "step": 2950
+    },
+    {
+      "epoch": 3.77728,
+      "grad_norm": 0.6946511268615723,
+      "learning_rate": 4.4115646258503404e-05,
+      "loss": 0.6773,
+      "step": 2951
+    },
+    {
+      "epoch": 3.77856,
+      "grad_norm": 0.6707186102867126,
+      "learning_rate": 4.4113645458183276e-05,
+      "loss": 0.6772,
+      "step": 2952
+    },
+    {
+      "epoch": 3.77984,
+      "grad_norm": 0.6657763123512268,
+      "learning_rate": 4.411164465786315e-05,
+      "loss": 0.6664,
+      "step": 2953
+    },
+    {
+      "epoch": 3.78112,
+      "grad_norm": 0.6746175289154053,
+      "learning_rate": 4.410964385754302e-05,
+      "loss": 0.6286,
+      "step": 2954
+    },
+    {
+      "epoch": 3.7824,
+      "grad_norm": 0.6620097756385803,
+      "learning_rate": 4.410764305722289e-05,
+      "loss": 0.6133,
+      "step": 2955
+    },
+    {
+      "epoch": 3.78368,
+      "grad_norm": 0.6786773204803467,
+      "learning_rate": 4.4105642256902763e-05,
+      "loss": 0.6386,
+      "step": 2956
+    },
+    {
+      "epoch": 3.78496,
+      "grad_norm": 0.6965882182121277,
+      "learning_rate": 4.4103641456582635e-05,
+      "loss": 0.6586,
+      "step": 2957
+    },
+    {
+      "epoch": 3.7862400000000003,
+      "grad_norm": 0.7074950337409973,
+      "learning_rate": 4.410164065626251e-05,
+      "loss": 0.6971,
+      "step": 2958
+    },
+    {
+      "epoch": 3.7875199999999998,
+      "grad_norm": 0.6736770272254944,
+      "learning_rate": 4.409963985594238e-05,
+      "loss": 0.6395,
+      "step": 2959
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.736041784286499,
+      "learning_rate": 4.409763905562225e-05,
+      "loss": 0.6891,
+      "step": 2960
+    },
+    {
+      "epoch": 3.79008,
+      "grad_norm": 0.6859921813011169,
+      "learning_rate": 4.409563825530212e-05,
+      "loss": 0.6503,
+      "step": 2961
+    },
+    {
+      "epoch": 3.79136,
+      "grad_norm": 0.7309426069259644,
+      "learning_rate": 4.4093637454981995e-05,
+      "loss": 0.7584,
+      "step": 2962
+    },
+    {
+      "epoch": 3.79264,
+      "grad_norm": 0.6685876846313477,
+      "learning_rate": 4.4091636654661866e-05,
+      "loss": 0.6737,
+      "step": 2963
+    },
+    {
+      "epoch": 3.79392,
+      "grad_norm": 0.6798075437545776,
+      "learning_rate": 4.408963585434174e-05,
+      "loss": 0.6157,
+      "step": 2964
+    },
+    {
+      "epoch": 3.7952,
+      "grad_norm": 0.677910327911377,
+      "learning_rate": 4.408763505402161e-05,
+      "loss": 0.6221,
+      "step": 2965
+    },
+    {
+      "epoch": 3.79648,
+      "grad_norm": 0.7607006430625916,
+      "learning_rate": 4.408563425370149e-05,
+      "loss": 0.7309,
+      "step": 2966
+    },
+    {
+      "epoch": 3.7977600000000002,
+      "grad_norm": 0.6692076921463013,
+      "learning_rate": 4.4083633453381354e-05,
+      "loss": 0.6585,
+      "step": 2967
+    },
+    {
+      "epoch": 3.7990399999999998,
+      "grad_norm": 0.6770291328430176,
+      "learning_rate": 4.4081632653061226e-05,
+      "loss": 0.6702,
+      "step": 2968
+    },
+    {
+      "epoch": 3.80032,
+      "grad_norm": 0.7067806720733643,
+      "learning_rate": 4.40796318527411e-05,
+      "loss": 0.6986,
+      "step": 2969
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.6886568069458008,
+      "learning_rate": 4.407763105242097e-05,
+      "loss": 0.7016,
+      "step": 2970
+    },
+    {
+      "epoch": 3.80288,
+      "grad_norm": 0.6451705694198608,
+      "learning_rate": 4.407563025210084e-05,
+      "loss": 0.6291,
+      "step": 2971
+    },
+    {
+      "epoch": 3.80416,
+      "grad_norm": 0.663578450679779,
+      "learning_rate": 4.407362945178071e-05,
+      "loss": 0.6653,
+      "step": 2972
+    },
+    {
+      "epoch": 3.80544,
+      "grad_norm": 0.71304851770401,
+      "learning_rate": 4.4071628651460585e-05,
+      "loss": 0.6694,
+      "step": 2973
+    },
+    {
+      "epoch": 3.80672,
+      "grad_norm": 0.6625712513923645,
+      "learning_rate": 4.4069627851140464e-05,
+      "loss": 0.6296,
+      "step": 2974
+    },
+    {
+      "epoch": 3.808,
+      "grad_norm": 0.6760493516921997,
+      "learning_rate": 4.406762705082033e-05,
+      "loss": 0.6438,
+      "step": 2975
+    },
+    {
+      "epoch": 3.80928,
+      "grad_norm": 0.6758971214294434,
+      "learning_rate": 4.40656262505002e-05,
+      "loss": 0.7146,
+      "step": 2976
+    },
+    {
+      "epoch": 3.8105599999999997,
+      "grad_norm": 0.6622211337089539,
+      "learning_rate": 4.406362545018007e-05,
+      "loss": 0.6595,
+      "step": 2977
+    },
+    {
+      "epoch": 3.81184,
+      "grad_norm": 0.6578704714775085,
+      "learning_rate": 4.4061624649859944e-05,
+      "loss": 0.6264,
+      "step": 2978
+    },
+    {
+      "epoch": 3.81312,
+      "grad_norm": 0.7323094606399536,
+      "learning_rate": 4.4059623849539816e-05,
+      "loss": 0.693,
+      "step": 2979
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.6730750799179077,
+      "learning_rate": 4.405762304921969e-05,
+      "loss": 0.6267,
+      "step": 2980
+    },
+    {
+      "epoch": 3.81568,
+      "grad_norm": 0.7026966214179993,
+      "learning_rate": 4.405562224889957e-05,
+      "loss": 0.6686,
+      "step": 2981
+    },
+    {
+      "epoch": 3.81696,
+      "grad_norm": 0.7159264087677002,
+      "learning_rate": 4.405362144857944e-05,
+      "loss": 0.6856,
+      "step": 2982
+    },
+    {
+      "epoch": 3.81824,
+      "grad_norm": 0.6785261034965515,
+      "learning_rate": 4.4051620648259304e-05,
+      "loss": 0.7115,
+      "step": 2983
+    },
+    {
+      "epoch": 3.81952,
+      "grad_norm": 0.6551268100738525,
+      "learning_rate": 4.4049619847939175e-05,
+      "loss": 0.6466,
+      "step": 2984
+    },
+    {
+      "epoch": 3.8208,
+      "grad_norm": 0.7004695534706116,
+      "learning_rate": 4.404761904761905e-05,
+      "loss": 0.6808,
+      "step": 2985
+    },
+    {
+      "epoch": 3.82208,
+      "grad_norm": 0.7341533303260803,
+      "learning_rate": 4.404561824729892e-05,
+      "loss": 0.7086,
+      "step": 2986
+    },
+    {
+      "epoch": 3.82336,
+      "grad_norm": 0.6831992268562317,
+      "learning_rate": 4.404361744697879e-05,
+      "loss": 0.6795,
+      "step": 2987
+    },
+    {
+      "epoch": 3.82464,
+      "grad_norm": 0.6428318023681641,
+      "learning_rate": 4.404161664665867e-05,
+      "loss": 0.6104,
+      "step": 2988
+    },
+    {
+      "epoch": 3.82592,
+      "grad_norm": 0.6942797899246216,
+      "learning_rate": 4.403961584633854e-05,
+      "loss": 0.7077,
+      "step": 2989
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.6718285083770752,
+      "learning_rate": 4.403761504601841e-05,
+      "loss": 0.6673,
+      "step": 2990
+    },
+    {
+      "epoch": 3.82848,
+      "grad_norm": 0.669506311416626,
+      "learning_rate": 4.403561424569828e-05,
+      "loss": 0.6671,
+      "step": 2991
+    },
+    {
+      "epoch": 3.8297600000000003,
+      "grad_norm": 0.6954408884048462,
+      "learning_rate": 4.403361344537815e-05,
+      "loss": 0.66,
+      "step": 2992
+    },
+    {
+      "epoch": 3.83104,
+      "grad_norm": 0.6513415575027466,
+      "learning_rate": 4.403161264505802e-05,
+      "loss": 0.6519,
+      "step": 2993
+    },
+    {
+      "epoch": 3.83232,
+      "grad_norm": 0.6512577533721924,
+      "learning_rate": 4.4029611844737894e-05,
+      "loss": 0.6356,
+      "step": 2994
+    },
+    {
+      "epoch": 3.8336,
+      "grad_norm": 0.69619220495224,
+      "learning_rate": 4.402761104441777e-05,
+      "loss": 0.6911,
+      "step": 2995
+    },
+    {
+      "epoch": 3.83488,
+      "grad_norm": 0.6886332035064697,
+      "learning_rate": 4.4025610244097645e-05,
+      "loss": 0.7061,
+      "step": 2996
+    },
+    {
+      "epoch": 3.83616,
+      "grad_norm": 0.7018471360206604,
+      "learning_rate": 4.4023609443777516e-05,
+      "loss": 0.6589,
+      "step": 2997
+    },
+    {
+      "epoch": 3.83744,
+      "grad_norm": 0.6925556659698486,
+      "learning_rate": 4.402160864345739e-05,
+      "loss": 0.7039,
+      "step": 2998
+    },
+    {
+      "epoch": 3.83872,
+      "grad_norm": 0.7044040560722351,
+      "learning_rate": 4.401960784313725e-05,
+      "loss": 0.7379,
+      "step": 2999
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.6619036197662354,
+      "learning_rate": 4.4017607042817125e-05,
+      "loss": 0.6497,
+      "step": 3000
+    },
+    {
+      "epoch": 3.8412800000000002,
+      "grad_norm": 0.677604615688324,
+      "learning_rate": 4.4015606242497e-05,
+      "loss": 0.6742,
+      "step": 3001
+    },
+    {
+      "epoch": 3.8425599999999998,
+      "grad_norm": 0.6716386079788208,
+      "learning_rate": 4.4013605442176876e-05,
+      "loss": 0.6129,
+      "step": 3002
+    },
+    {
+      "epoch": 3.84384,
+      "grad_norm": 0.6958229541778564,
+      "learning_rate": 4.401160464185675e-05,
+      "loss": 0.6419,
+      "step": 3003
+    },
+    {
+      "epoch": 3.84512,
+      "grad_norm": 0.6683501601219177,
+      "learning_rate": 4.400960384153662e-05,
+      "loss": 0.583,
+      "step": 3004
+    },
+    {
+      "epoch": 3.8464,
+      "grad_norm": 0.6913495063781738,
+      "learning_rate": 4.400760304121649e-05,
+      "loss": 0.6845,
+      "step": 3005
+    },
+    {
+      "epoch": 3.84768,
+      "grad_norm": 0.7633008360862732,
+      "learning_rate": 4.400560224089636e-05,
+      "loss": 0.7119,
+      "step": 3006
+    },
+    {
+      "epoch": 3.84896,
+      "grad_norm": 0.68817138671875,
+      "learning_rate": 4.400360144057623e-05,
+      "loss": 0.6531,
+      "step": 3007
+    },
+    {
+      "epoch": 3.85024,
+      "grad_norm": 0.6548225283622742,
+      "learning_rate": 4.40016006402561e-05,
+      "loss": 0.6511,
+      "step": 3008
+    },
+    {
+      "epoch": 3.85152,
+      "grad_norm": 0.7012094259262085,
+      "learning_rate": 4.399959983993598e-05,
+      "loss": 0.6826,
+      "step": 3009
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.7215191721916199,
+      "learning_rate": 4.399759903961585e-05,
+      "loss": 0.7206,
+      "step": 3010
+    },
+    {
+      "epoch": 3.8540799999999997,
+      "grad_norm": 0.7269889116287231,
+      "learning_rate": 4.399559823929572e-05,
+      "loss": 0.6886,
+      "step": 3011
+    },
+    {
+      "epoch": 3.85536,
+      "grad_norm": 0.6882228255271912,
+      "learning_rate": 4.3993597438975594e-05,
+      "loss": 0.6738,
+      "step": 3012
+    },
+    {
+      "epoch": 3.85664,
+      "grad_norm": 0.6564989686012268,
+      "learning_rate": 4.3991596638655466e-05,
+      "loss": 0.614,
+      "step": 3013
+    },
+    {
+      "epoch": 3.85792,
+      "grad_norm": 0.7501393556594849,
+      "learning_rate": 4.398959583833534e-05,
+      "loss": 0.6957,
+      "step": 3014
+    },
+    {
+      "epoch": 3.8592,
+      "grad_norm": 0.6881626844406128,
+      "learning_rate": 4.39875950380152e-05,
+      "loss": 0.6503,
+      "step": 3015
+    },
+    {
+      "epoch": 3.86048,
+      "grad_norm": 0.7230293154716492,
+      "learning_rate": 4.398559423769508e-05,
+      "loss": 0.6886,
+      "step": 3016
+    },
+    {
+      "epoch": 3.86176,
+      "grad_norm": 0.7025682330131531,
+      "learning_rate": 4.3983593437374954e-05,
+      "loss": 0.6752,
+      "step": 3017
+    },
+    {
+      "epoch": 3.86304,
+      "grad_norm": 0.7032774686813354,
+      "learning_rate": 4.3981592637054825e-05,
+      "loss": 0.6771,
+      "step": 3018
+    },
+    {
+      "epoch": 3.86432,
+      "grad_norm": 0.6938534379005432,
+      "learning_rate": 4.39795918367347e-05,
+      "loss": 0.6688,
+      "step": 3019
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.6492636203765869,
+      "learning_rate": 4.397759103641457e-05,
+      "loss": 0.6702,
+      "step": 3020
+    },
+    {
+      "epoch": 3.86688,
+      "grad_norm": 0.698173463344574,
+      "learning_rate": 4.397559023609444e-05,
+      "loss": 0.616,
+      "step": 3021
+    },
+    {
+      "epoch": 3.86816,
+      "grad_norm": 0.7170678377151489,
+      "learning_rate": 4.397358943577431e-05,
+      "loss": 0.6048,
+      "step": 3022
+    },
+    {
+      "epoch": 3.86944,
+      "grad_norm": 0.6741195917129517,
+      "learning_rate": 4.3971588635454185e-05,
+      "loss": 0.6607,
+      "step": 3023
+    },
+    {
+      "epoch": 3.87072,
+      "grad_norm": 0.7016419172286987,
+      "learning_rate": 4.3969587835134056e-05,
+      "loss": 0.7229,
+      "step": 3024
+    },
+    {
+      "epoch": 3.872,
+      "grad_norm": 0.6970353126525879,
+      "learning_rate": 4.396758703481393e-05,
+      "loss": 0.6658,
+      "step": 3025
+    },
+    {
+      "epoch": 3.87328,
+      "grad_norm": 0.6936695575714111,
+      "learning_rate": 4.39655862344938e-05,
+      "loss": 0.7113,
+      "step": 3026
+    },
+    {
+      "epoch": 3.87456,
+      "grad_norm": 0.6938142776489258,
+      "learning_rate": 4.396358543417367e-05,
+      "loss": 0.6535,
+      "step": 3027
+    },
+    {
+      "epoch": 3.87584,
+      "grad_norm": 0.6950214505195618,
+      "learning_rate": 4.3961584633853544e-05,
+      "loss": 0.701,
+      "step": 3028
+    },
+    {
+      "epoch": 3.87712,
+      "grad_norm": 0.6649338603019714,
+      "learning_rate": 4.3959583833533416e-05,
+      "loss": 0.635,
+      "step": 3029
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.7024021744728088,
+      "learning_rate": 4.395758303321329e-05,
+      "loss": 0.7062,
+      "step": 3030
+    },
+    {
+      "epoch": 3.87968,
+      "grad_norm": 0.6412244439125061,
+      "learning_rate": 4.395558223289316e-05,
+      "loss": 0.6081,
+      "step": 3031
+    },
+    {
+      "epoch": 3.88096,
+      "grad_norm": 0.6740350127220154,
+      "learning_rate": 4.395358143257303e-05,
+      "loss": 0.6404,
+      "step": 3032
+    },
+    {
+      "epoch": 3.88224,
+      "grad_norm": 0.647661566734314,
+      "learning_rate": 4.39515806322529e-05,
+      "loss": 0.6506,
+      "step": 3033
+    },
+    {
+      "epoch": 3.88352,
+      "grad_norm": 0.6819959878921509,
+      "learning_rate": 4.3949579831932775e-05,
+      "loss": 0.6258,
+      "step": 3034
+    },
+    {
+      "epoch": 3.8848000000000003,
+      "grad_norm": 0.7153730988502502,
+      "learning_rate": 4.394757903161265e-05,
+      "loss": 0.6579,
+      "step": 3035
+    },
+    {
+      "epoch": 3.8860799999999998,
+      "grad_norm": 0.6946179270744324,
+      "learning_rate": 4.394557823129252e-05,
+      "loss": 0.6526,
+      "step": 3036
+    },
+    {
+      "epoch": 3.88736,
+      "grad_norm": 0.6954408288002014,
+      "learning_rate": 4.394357743097239e-05,
+      "loss": 0.6596,
+      "step": 3037
+    },
+    {
+      "epoch": 3.88864,
+      "grad_norm": 0.6411071419715881,
+      "learning_rate": 4.394157663065226e-05,
+      "loss": 0.6556,
+      "step": 3038
+    },
+    {
+      "epoch": 3.88992,
+      "grad_norm": 0.6873469948768616,
+      "learning_rate": 4.3939575830332134e-05,
+      "loss": 0.6791,
+      "step": 3039
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.6711555123329163,
+      "learning_rate": 4.3937575030012006e-05,
+      "loss": 0.6624,
+      "step": 3040
+    },
+    {
+      "epoch": 3.89248,
+      "grad_norm": 0.6283385753631592,
+      "learning_rate": 4.393557422969188e-05,
+      "loss": 0.6152,
+      "step": 3041
+    },
+    {
+      "epoch": 3.89376,
+      "grad_norm": 0.6898487210273743,
+      "learning_rate": 4.393357342937175e-05,
+      "loss": 0.7011,
+      "step": 3042
+    },
+    {
+      "epoch": 3.89504,
+      "grad_norm": 0.6428130865097046,
+      "learning_rate": 4.393157262905162e-05,
+      "loss": 0.6157,
+      "step": 3043
+    },
+    {
+      "epoch": 3.8963200000000002,
+      "grad_norm": 0.6543678641319275,
+      "learning_rate": 4.39295718287315e-05,
+      "loss": 0.6196,
+      "step": 3044
+    },
+    {
+      "epoch": 3.8975999999999997,
+      "grad_norm": 0.6453118324279785,
+      "learning_rate": 4.3927571028411365e-05,
+      "loss": 0.6832,
+      "step": 3045
+    },
+    {
+      "epoch": 3.89888,
+      "grad_norm": 0.6822550892829895,
+      "learning_rate": 4.392557022809124e-05,
+      "loss": 0.6523,
+      "step": 3046
+    },
+    {
+      "epoch": 3.90016,
+      "grad_norm": 0.7187889814376831,
+      "learning_rate": 4.392356942777111e-05,
+      "loss": 0.7118,
+      "step": 3047
+    },
+    {
+      "epoch": 3.90144,
+      "grad_norm": 0.7157731056213379,
+      "learning_rate": 4.392156862745098e-05,
+      "loss": 0.7197,
+      "step": 3048
+    },
+    {
+      "epoch": 3.90272,
+      "grad_norm": 0.686944305896759,
+      "learning_rate": 4.391956782713085e-05,
+      "loss": 0.6051,
+      "step": 3049
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.7068070769309998,
+      "learning_rate": 4.3917567026810725e-05,
+      "loss": 0.7256,
+      "step": 3050
+    },
+    {
+      "epoch": 3.90528,
+      "grad_norm": 0.7173423767089844,
+      "learning_rate": 4.3915566226490603e-05,
+      "loss": 0.6701,
+      "step": 3051
+    },
+    {
+      "epoch": 3.90656,
+      "grad_norm": 0.6758041977882385,
+      "learning_rate": 4.3913565426170475e-05,
+      "loss": 0.6383,
+      "step": 3052
+    },
+    {
+      "epoch": 3.90784,
+      "grad_norm": 0.6902790665626526,
+      "learning_rate": 4.391156462585034e-05,
+      "loss": 0.6533,
+      "step": 3053
+    },
+    {
+      "epoch": 3.90912,
+      "grad_norm": 0.70408034324646,
+      "learning_rate": 4.390956382553021e-05,
+      "loss": 0.6577,
+      "step": 3054
+    },
+    {
+      "epoch": 3.9104,
+      "grad_norm": 0.6818796396255493,
+      "learning_rate": 4.3907563025210084e-05,
+      "loss": 0.6786,
+      "step": 3055
+    },
+    {
+      "epoch": 3.91168,
+      "grad_norm": 0.6927520036697388,
+      "learning_rate": 4.3905562224889956e-05,
+      "loss": 0.6924,
+      "step": 3056
+    },
+    {
+      "epoch": 3.91296,
+      "grad_norm": 0.7096193432807922,
+      "learning_rate": 4.390356142456983e-05,
+      "loss": 0.6852,
+      "step": 3057
+    },
+    {
+      "epoch": 3.91424,
+      "grad_norm": 0.6638479828834534,
+      "learning_rate": 4.3901560624249706e-05,
+      "loss": 0.6388,
+      "step": 3058
+    },
+    {
+      "epoch": 3.91552,
+      "grad_norm": 0.7214940190315247,
+      "learning_rate": 4.389955982392958e-05,
+      "loss": 0.7158,
+      "step": 3059
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.681658148765564,
+      "learning_rate": 4.389755902360945e-05,
+      "loss": 0.657,
+      "step": 3060
+    },
+    {
+      "epoch": 3.91808,
+      "grad_norm": 0.6954881548881531,
+      "learning_rate": 4.3895558223289315e-05,
+      "loss": 0.7083,
+      "step": 3061
+    },
+    {
+      "epoch": 3.91936,
+      "grad_norm": 0.681389331817627,
+      "learning_rate": 4.389355742296919e-05,
+      "loss": 0.6743,
+      "step": 3062
+    },
+    {
+      "epoch": 3.92064,
+      "grad_norm": 0.688248336315155,
+      "learning_rate": 4.389155662264906e-05,
+      "loss": 0.6632,
+      "step": 3063
+    },
+    {
+      "epoch": 3.92192,
+      "grad_norm": 0.7140293717384338,
+      "learning_rate": 4.388955582232893e-05,
+      "loss": 0.6964,
+      "step": 3064
+    },
+    {
+      "epoch": 3.9232,
+      "grad_norm": 0.692408561706543,
+      "learning_rate": 4.388755502200881e-05,
+      "loss": 0.6953,
+      "step": 3065
+    },
+    {
+      "epoch": 3.92448,
+      "grad_norm": 0.6348261833190918,
+      "learning_rate": 4.388555422168868e-05,
+      "loss": 0.6256,
+      "step": 3066
+    },
+    {
+      "epoch": 3.92576,
+      "grad_norm": 0.6396204233169556,
+      "learning_rate": 4.388355342136855e-05,
+      "loss": 0.6958,
+      "step": 3067
+    },
+    {
+      "epoch": 3.92704,
+      "grad_norm": 0.6573483347892761,
+      "learning_rate": 4.3881552621048425e-05,
+      "loss": 0.6314,
+      "step": 3068
+    },
+    {
+      "epoch": 3.9283200000000003,
+      "grad_norm": 0.6830142736434937,
+      "learning_rate": 4.387955182072829e-05,
+      "loss": 0.6347,
+      "step": 3069
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.6613914370536804,
+      "learning_rate": 4.387755102040816e-05,
+      "loss": 0.7057,
+      "step": 3070
+    },
+    {
+      "epoch": 3.93088,
+      "grad_norm": 0.7034798860549927,
+      "learning_rate": 4.3875550220088034e-05,
+      "loss": 0.6773,
+      "step": 3071
+    },
+    {
+      "epoch": 3.93216,
+      "grad_norm": 0.741765022277832,
+      "learning_rate": 4.387354941976791e-05,
+      "loss": 0.6941,
+      "step": 3072
+    },
+    {
+      "epoch": 3.93344,
+      "grad_norm": 0.7319201827049255,
+      "learning_rate": 4.3871548619447784e-05,
+      "loss": 0.7408,
+      "step": 3073
+    },
+    {
+      "epoch": 3.93472,
+      "grad_norm": 0.66302090883255,
+      "learning_rate": 4.3869547819127656e-05,
+      "loss": 0.624,
+      "step": 3074
+    },
+    {
+      "epoch": 3.936,
+      "grad_norm": 0.6886874437332153,
+      "learning_rate": 4.386754701880753e-05,
+      "loss": 0.6537,
+      "step": 3075
+    },
+    {
+      "epoch": 3.93728,
+      "grad_norm": 0.6793836951255798,
+      "learning_rate": 4.38655462184874e-05,
+      "loss": 0.6656,
+      "step": 3076
+    },
+    {
+      "epoch": 3.93856,
+      "grad_norm": 0.6942064762115479,
+      "learning_rate": 4.3863545418167265e-05,
+      "loss": 0.7116,
+      "step": 3077
+    },
+    {
+      "epoch": 3.9398400000000002,
+      "grad_norm": 0.6668751239776611,
+      "learning_rate": 4.386154461784714e-05,
+      "loss": 0.626,
+      "step": 3078
+    },
+    {
+      "epoch": 3.9411199999999997,
+      "grad_norm": 0.6711156964302063,
+      "learning_rate": 4.385954381752701e-05,
+      "loss": 0.6653,
+      "step": 3079
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.6742311716079712,
+      "learning_rate": 4.385754301720689e-05,
+      "loss": 0.7238,
+      "step": 3080
+    },
+    {
+      "epoch": 3.94368,
+      "grad_norm": 0.6560841202735901,
+      "learning_rate": 4.385554221688676e-05,
+      "loss": 0.6364,
+      "step": 3081
+    },
+    {
+      "epoch": 3.94496,
+      "grad_norm": 0.6376678347587585,
+      "learning_rate": 4.385354141656663e-05,
+      "loss": 0.6447,
+      "step": 3082
+    },
+    {
+      "epoch": 3.94624,
+      "grad_norm": 0.6691120862960815,
+      "learning_rate": 4.38515406162465e-05,
+      "loss": 0.6557,
+      "step": 3083
+    },
+    {
+      "epoch": 3.94752,
+      "grad_norm": 0.6742989420890808,
+      "learning_rate": 4.3849539815926375e-05,
+      "loss": 0.6029,
+      "step": 3084
+    },
+    {
+      "epoch": 3.9488,
+      "grad_norm": 0.7199971675872803,
+      "learning_rate": 4.384753901560624e-05,
+      "loss": 0.6881,
+      "step": 3085
+    },
+    {
+      "epoch": 3.95008,
+      "grad_norm": 0.7232701182365417,
+      "learning_rate": 4.384553821528611e-05,
+      "loss": 0.6548,
+      "step": 3086
+    },
+    {
+      "epoch": 3.95136,
+      "grad_norm": 0.7027313113212585,
+      "learning_rate": 4.384353741496599e-05,
+      "loss": 0.6444,
+      "step": 3087
+    },
+    {
+      "epoch": 3.95264,
+      "grad_norm": 0.6895410418510437,
+      "learning_rate": 4.384153661464586e-05,
+      "loss": 0.6935,
+      "step": 3088
+    },
+    {
+      "epoch": 3.95392,
+      "grad_norm": 0.6641124486923218,
+      "learning_rate": 4.3839535814325734e-05,
+      "loss": 0.6709,
+      "step": 3089
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.6645619869232178,
+      "learning_rate": 4.3837535014005606e-05,
+      "loss": 0.6736,
+      "step": 3090
+    },
+    {
+      "epoch": 3.95648,
+      "grad_norm": 0.680464506149292,
+      "learning_rate": 4.383553421368548e-05,
+      "loss": 0.6601,
+      "step": 3091
+    },
+    {
+      "epoch": 3.95776,
+      "grad_norm": 0.6810153126716614,
+      "learning_rate": 4.383353341336535e-05,
+      "loss": 0.6463,
+      "step": 3092
+    },
+    {
+      "epoch": 3.95904,
+      "grad_norm": 0.6904779076576233,
+      "learning_rate": 4.3831532613045215e-05,
+      "loss": 0.6375,
+      "step": 3093
+    },
+    {
+      "epoch": 3.96032,
+      "grad_norm": 0.7193115949630737,
+      "learning_rate": 4.382953181272509e-05,
+      "loss": 0.6649,
+      "step": 3094
+    },
+    {
+      "epoch": 3.9616,
+      "grad_norm": 0.6670559644699097,
+      "learning_rate": 4.3827531012404965e-05,
+      "loss": 0.6505,
+      "step": 3095
+    },
+    {
+      "epoch": 3.96288,
+      "grad_norm": 0.707496166229248,
+      "learning_rate": 4.382553021208484e-05,
+      "loss": 0.6726,
+      "step": 3096
+    },
+    {
+      "epoch": 3.96416,
+      "grad_norm": 0.6769973635673523,
+      "learning_rate": 4.382352941176471e-05,
+      "loss": 0.6529,
+      "step": 3097
+    },
+    {
+      "epoch": 3.96544,
+      "grad_norm": 0.6613832712173462,
+      "learning_rate": 4.382152861144458e-05,
+      "loss": 0.6218,
+      "step": 3098
+    },
+    {
+      "epoch": 3.96672,
+      "grad_norm": 0.6404051184654236,
+      "learning_rate": 4.381952781112445e-05,
+      "loss": 0.6685,
+      "step": 3099
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.7350611090660095,
+      "learning_rate": 4.3817527010804324e-05,
+      "loss": 0.7538,
+      "step": 3100
+    },
+    {
+      "epoch": 3.96928,
+      "grad_norm": 0.7339857220649719,
+      "learning_rate": 4.3815526210484196e-05,
+      "loss": 0.7104,
+      "step": 3101
+    },
+    {
+      "epoch": 3.97056,
+      "grad_norm": 0.6971475481987,
+      "learning_rate": 4.381352541016407e-05,
+      "loss": 0.6595,
+      "step": 3102
+    },
+    {
+      "epoch": 3.9718400000000003,
+      "grad_norm": 0.6795316934585571,
+      "learning_rate": 4.381152460984394e-05,
+      "loss": 0.63,
+      "step": 3103
+    },
+    {
+      "epoch": 3.9731199999999998,
+      "grad_norm": 0.6590287685394287,
+      "learning_rate": 4.380952380952381e-05,
+      "loss": 0.6013,
+      "step": 3104
+    },
+    {
+      "epoch": 3.9744,
+      "grad_norm": 0.6492599248886108,
+      "learning_rate": 4.3807523009203684e-05,
+      "loss": 0.6626,
+      "step": 3105
+    },
+    {
+      "epoch": 3.97568,
+      "grad_norm": 0.7017719149589539,
+      "learning_rate": 4.3805522208883556e-05,
+      "loss": 0.6685,
+      "step": 3106
+    },
+    {
+      "epoch": 3.97696,
+      "grad_norm": 0.6492130160331726,
+      "learning_rate": 4.380352140856343e-05,
+      "loss": 0.6498,
+      "step": 3107
+    },
+    {
+      "epoch": 3.97824,
+      "grad_norm": 0.6498051881790161,
+      "learning_rate": 4.38015206082433e-05,
+      "loss": 0.6152,
+      "step": 3108
+    },
+    {
+      "epoch": 3.97952,
+      "grad_norm": 0.710146963596344,
+      "learning_rate": 4.379951980792317e-05,
+      "loss": 0.7109,
+      "step": 3109
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.6580629348754883,
+      "learning_rate": 4.379751900760304e-05,
+      "loss": 0.6464,
+      "step": 3110
+    },
+    {
+      "epoch": 3.98208,
+      "grad_norm": 0.671927809715271,
+      "learning_rate": 4.3795518207282915e-05,
+      "loss": 0.7003,
+      "step": 3111
+    },
+    {
+      "epoch": 3.9833600000000002,
+      "grad_norm": 0.64473956823349,
+      "learning_rate": 4.379351740696279e-05,
+      "loss": 0.6789,
+      "step": 3112
+    },
+    {
+      "epoch": 3.9846399999999997,
+      "grad_norm": 0.6603186130523682,
+      "learning_rate": 4.379151660664266e-05,
+      "loss": 0.6558,
+      "step": 3113
+    },
+    {
+      "epoch": 3.98592,
+      "grad_norm": 0.7090316414833069,
+      "learning_rate": 4.378951580632253e-05,
+      "loss": 0.6923,
+      "step": 3114
+    },
+    {
+      "epoch": 3.9872,
+      "grad_norm": 0.6954898238182068,
+      "learning_rate": 4.37875150060024e-05,
+      "loss": 0.6341,
+      "step": 3115
+    },
+    {
+      "epoch": 3.98848,
+      "grad_norm": 0.6501321196556091,
+      "learning_rate": 4.3785514205682274e-05,
+      "loss": 0.6481,
+      "step": 3116
+    },
+    {
+      "epoch": 3.98976,
+      "grad_norm": 0.6654294729232788,
+      "learning_rate": 4.3783513405362146e-05,
+      "loss": 0.6379,
+      "step": 3117
+    },
+    {
+      "epoch": 3.99104,
+      "grad_norm": 0.698321521282196,
+      "learning_rate": 4.378151260504202e-05,
+      "loss": 0.6881,
+      "step": 3118
+    },
+    {
+      "epoch": 3.99232,
+      "grad_norm": 0.7432014346122742,
+      "learning_rate": 4.377951180472189e-05,
+      "loss": 0.6836,
+      "step": 3119
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.6992458701133728,
+      "learning_rate": 4.377751100440176e-05,
+      "loss": 0.7221,
+      "step": 3120
+    },
+    {
+      "epoch": 3.99488,
+      "grad_norm": 0.6965485215187073,
+      "learning_rate": 4.377551020408163e-05,
+      "loss": 0.604,
+      "step": 3121
+    },
+    {
+      "epoch": 3.99616,
+      "grad_norm": 0.7211214900016785,
+      "learning_rate": 4.377350940376151e-05,
+      "loss": 0.7074,
+      "step": 3122
+    },
+    {
+      "epoch": 3.99744,
+      "grad_norm": 0.6553539633750916,
+      "learning_rate": 4.377150860344138e-05,
+      "loss": 0.6204,
+      "step": 3123
+    },
+    {
+      "epoch": 3.99872,
+      "grad_norm": 0.6686910390853882,
+      "learning_rate": 4.376950780312125e-05,
+      "loss": 0.7271,
+      "step": 3124
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.5496490001678467,
+      "learning_rate": 4.376750700280112e-05,
+      "loss": 1.3281,
+      "step": 3125
+    },
+    {
+      "epoch": 4.00128,
+      "grad_norm": 0.6808484792709351,
+      "learning_rate": 4.376550620248099e-05,
+      "loss": 0.6388,
+      "step": 3126
+    },
+    {
+      "epoch": 4.00256,
+      "grad_norm": 0.6682015061378479,
+      "learning_rate": 4.3763505402160865e-05,
+      "loss": 0.6512,
+      "step": 3127
+    },
+    {
+      "epoch": 4.00384,
+      "grad_norm": 0.6814844608306885,
+      "learning_rate": 4.3761504601840736e-05,
+      "loss": 0.6496,
+      "step": 3128
+    },
+    {
+      "epoch": 4.00512,
+      "grad_norm": 0.698432981967926,
+      "learning_rate": 4.3759503801520615e-05,
+      "loss": 0.6272,
+      "step": 3129
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 0.6985964179039001,
+      "learning_rate": 4.375750300120049e-05,
+      "loss": 0.6447,
+      "step": 3130
+    },
+    {
+      "epoch": 4.00768,
+      "grad_norm": 0.648468554019928,
+      "learning_rate": 4.375550220088035e-05,
+      "loss": 0.6205,
+      "step": 3131
+    },
+    {
+      "epoch": 4.00896,
+      "grad_norm": 0.694003164768219,
+      "learning_rate": 4.3753501400560224e-05,
+      "loss": 0.6613,
+      "step": 3132
+    },
+    {
+      "epoch": 4.01024,
+      "grad_norm": 0.7265833616256714,
+      "learning_rate": 4.3751500600240096e-05,
+      "loss": 0.623,
+      "step": 3133
+    },
+    {
+      "epoch": 4.01152,
+      "grad_norm": 0.7631153464317322,
+      "learning_rate": 4.374949979991997e-05,
+      "loss": 0.6963,
+      "step": 3134
+    },
+    {
+      "epoch": 4.0128,
+      "grad_norm": 0.7075992822647095,
+      "learning_rate": 4.374749899959984e-05,
+      "loss": 0.6754,
+      "step": 3135
+    },
+    {
+      "epoch": 4.01408,
+      "grad_norm": 0.7054523229598999,
+      "learning_rate": 4.374549819927972e-05,
+      "loss": 0.6145,
+      "step": 3136
+    },
+    {
+      "epoch": 4.01536,
+      "grad_norm": 0.6870067119598389,
+      "learning_rate": 4.374349739895959e-05,
+      "loss": 0.663,
+      "step": 3137
+    },
+    {
+      "epoch": 4.01664,
+      "grad_norm": 0.6565841436386108,
+      "learning_rate": 4.374149659863946e-05,
+      "loss": 0.6111,
+      "step": 3138
+    },
+    {
+      "epoch": 4.01792,
+      "grad_norm": 0.6983011960983276,
+      "learning_rate": 4.373949579831933e-05,
+      "loss": 0.6528,
+      "step": 3139
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.6741187572479248,
+      "learning_rate": 4.37374949979992e-05,
+      "loss": 0.6625,
+      "step": 3140
+    },
+    {
+      "epoch": 4.02048,
+      "grad_norm": 0.6667806506156921,
+      "learning_rate": 4.373549419767907e-05,
+      "loss": 0.629,
+      "step": 3141
+    },
+    {
+      "epoch": 4.0217600000000004,
+      "grad_norm": 0.6826735734939575,
+      "learning_rate": 4.373349339735894e-05,
+      "loss": 0.6538,
+      "step": 3142
+    },
+    {
+      "epoch": 4.02304,
+      "grad_norm": 0.7120842933654785,
+      "learning_rate": 4.373149259703882e-05,
+      "loss": 0.6368,
+      "step": 3143
+    },
+    {
+      "epoch": 4.02432,
+      "grad_norm": 0.6919832825660706,
+      "learning_rate": 4.372949179671869e-05,
+      "loss": 0.6451,
+      "step": 3144
+    },
+    {
+      "epoch": 4.0256,
+      "grad_norm": 0.7019147276878357,
+      "learning_rate": 4.3727490996398565e-05,
+      "loss": 0.6605,
+      "step": 3145
+    },
+    {
+      "epoch": 4.02688,
+      "grad_norm": 0.7638674378395081,
+      "learning_rate": 4.3725490196078437e-05,
+      "loss": 0.6817,
+      "step": 3146
+    },
+    {
+      "epoch": 4.02816,
+      "grad_norm": 0.7084218263626099,
+      "learning_rate": 4.37234893957583e-05,
+      "loss": 0.6572,
+      "step": 3147
+    },
+    {
+      "epoch": 4.02944,
+      "grad_norm": 0.6604458093643188,
+      "learning_rate": 4.3721488595438174e-05,
+      "loss": 0.6048,
+      "step": 3148
+    },
+    {
+      "epoch": 4.03072,
+      "grad_norm": 0.6742260456085205,
+      "learning_rate": 4.3719487795118045e-05,
+      "loss": 0.6481,
+      "step": 3149
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.666289210319519,
+      "learning_rate": 4.3717486994797924e-05,
+      "loss": 0.6421,
+      "step": 3150
+    },
+    {
+      "epoch": 4.03328,
+      "grad_norm": 0.6697606444358826,
+      "learning_rate": 4.3715486194477796e-05,
+      "loss": 0.6479,
+      "step": 3151
+    },
+    {
+      "epoch": 4.03456,
+      "grad_norm": 0.7304966449737549,
+      "learning_rate": 4.371348539415767e-05,
+      "loss": 0.6792,
+      "step": 3152
+    },
+    {
+      "epoch": 4.03584,
+      "grad_norm": 0.6816575527191162,
+      "learning_rate": 4.371148459383754e-05,
+      "loss": 0.601,
+      "step": 3153
+    },
+    {
+      "epoch": 4.03712,
+      "grad_norm": 0.7284610271453857,
+      "learning_rate": 4.370948379351741e-05,
+      "loss": 0.64,
+      "step": 3154
+    },
+    {
+      "epoch": 4.0384,
+      "grad_norm": 0.7191203832626343,
+      "learning_rate": 4.3707482993197277e-05,
+      "loss": 0.64,
+      "step": 3155
+    },
+    {
+      "epoch": 4.03968,
+      "grad_norm": 0.7250388264656067,
+      "learning_rate": 4.370548219287715e-05,
+      "loss": 0.696,
+      "step": 3156
+    },
+    {
+      "epoch": 4.04096,
+      "grad_norm": 0.6809269785881042,
+      "learning_rate": 4.370348139255703e-05,
+      "loss": 0.6339,
+      "step": 3157
+    },
+    {
+      "epoch": 4.04224,
+      "grad_norm": 0.6679171323776245,
+      "learning_rate": 4.37014805922369e-05,
+      "loss": 0.6256,
+      "step": 3158
+    },
+    {
+      "epoch": 4.04352,
+      "grad_norm": 0.7094346880912781,
+      "learning_rate": 4.369947979191677e-05,
+      "loss": 0.6825,
+      "step": 3159
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.6463338732719421,
+      "learning_rate": 4.369747899159664e-05,
+      "loss": 0.6085,
+      "step": 3160
+    },
+    {
+      "epoch": 4.04608,
+      "grad_norm": 0.6958500146865845,
+      "learning_rate": 4.3695478191276514e-05,
+      "loss": 0.6654,
+      "step": 3161
+    },
+    {
+      "epoch": 4.04736,
+      "grad_norm": 0.7284070253372192,
+      "learning_rate": 4.3693477390956386e-05,
+      "loss": 0.6706,
+      "step": 3162
+    },
+    {
+      "epoch": 4.04864,
+      "grad_norm": 0.7025485038757324,
+      "learning_rate": 4.369147659063625e-05,
+      "loss": 0.677,
+      "step": 3163
+    },
+    {
+      "epoch": 4.04992,
+      "grad_norm": 0.7521453499794006,
+      "learning_rate": 4.368947579031613e-05,
+      "loss": 0.7294,
+      "step": 3164
+    },
+    {
+      "epoch": 4.0512,
+      "grad_norm": 0.6905098557472229,
+      "learning_rate": 4.3687474989996e-05,
+      "loss": 0.7215,
+      "step": 3165
+    },
+    {
+      "epoch": 4.05248,
+      "grad_norm": 0.7111669182777405,
+      "learning_rate": 4.3685474189675874e-05,
+      "loss": 0.6637,
+      "step": 3166
+    },
+    {
+      "epoch": 4.05376,
+      "grad_norm": 0.7028812766075134,
+      "learning_rate": 4.3683473389355746e-05,
+      "loss": 0.665,
+      "step": 3167
+    },
+    {
+      "epoch": 4.05504,
+      "grad_norm": 0.7170726656913757,
+      "learning_rate": 4.368147258903562e-05,
+      "loss": 0.6078,
+      "step": 3168
+    },
+    {
+      "epoch": 4.05632,
+      "grad_norm": 0.702850878238678,
+      "learning_rate": 4.367947178871549e-05,
+      "loss": 0.6381,
+      "step": 3169
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.7177186012268066,
+      "learning_rate": 4.367747098839536e-05,
+      "loss": 0.6339,
+      "step": 3170
+    },
+    {
+      "epoch": 4.05888,
+      "grad_norm": 0.7100415229797363,
+      "learning_rate": 4.367547018807523e-05,
+      "loss": 0.6966,
+      "step": 3171
+    },
+    {
+      "epoch": 4.06016,
+      "grad_norm": 0.6687876582145691,
+      "learning_rate": 4.3673469387755105e-05,
+      "loss": 0.6423,
+      "step": 3172
+    },
+    {
+      "epoch": 4.06144,
+      "grad_norm": 0.7261446714401245,
+      "learning_rate": 4.367146858743498e-05,
+      "loss": 0.6613,
+      "step": 3173
+    },
+    {
+      "epoch": 4.06272,
+      "grad_norm": 0.6871818900108337,
+      "learning_rate": 4.366946778711485e-05,
+      "loss": 0.6553,
+      "step": 3174
+    },
+    {
+      "epoch": 4.064,
+      "grad_norm": 0.6970064640045166,
+      "learning_rate": 4.366746698679472e-05,
+      "loss": 0.6334,
+      "step": 3175
+    },
+    {
+      "epoch": 4.06528,
+      "grad_norm": 0.6694562435150146,
+      "learning_rate": 4.366546618647459e-05,
+      "loss": 0.5959,
+      "step": 3176
+    },
+    {
+      "epoch": 4.06656,
+      "grad_norm": 0.7168928384780884,
+      "learning_rate": 4.3663465386154464e-05,
+      "loss": 0.6292,
+      "step": 3177
+    },
+    {
+      "epoch": 4.06784,
+      "grad_norm": 0.6959748864173889,
+      "learning_rate": 4.3661464585834336e-05,
+      "loss": 0.6437,
+      "step": 3178
+    },
+    {
+      "epoch": 4.06912,
+      "grad_norm": 0.6844011545181274,
+      "learning_rate": 4.365946378551421e-05,
+      "loss": 0.6478,
+      "step": 3179
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.7261403799057007,
+      "learning_rate": 4.365746298519408e-05,
+      "loss": 0.7068,
+      "step": 3180
+    },
+    {
+      "epoch": 4.07168,
+      "grad_norm": 0.7560855746269226,
+      "learning_rate": 4.365546218487395e-05,
+      "loss": 0.6783,
+      "step": 3181
+    },
+    {
+      "epoch": 4.07296,
+      "grad_norm": 0.6811922788619995,
+      "learning_rate": 4.3653461384553823e-05,
+      "loss": 0.6544,
+      "step": 3182
+    },
+    {
+      "epoch": 4.07424,
+      "grad_norm": 0.7000817060470581,
+      "learning_rate": 4.3651460584233695e-05,
+      "loss": 0.7211,
+      "step": 3183
+    },
+    {
+      "epoch": 4.07552,
+      "grad_norm": 0.7141358852386475,
+      "learning_rate": 4.364945978391357e-05,
+      "loss": 0.6627,
+      "step": 3184
+    },
+    {
+      "epoch": 4.0768,
+      "grad_norm": 0.7260948419570923,
+      "learning_rate": 4.364745898359344e-05,
+      "loss": 0.6885,
+      "step": 3185
+    },
+    {
+      "epoch": 4.07808,
+      "grad_norm": 0.7163599133491516,
+      "learning_rate": 4.364545818327331e-05,
+      "loss": 0.6581,
+      "step": 3186
+    },
+    {
+      "epoch": 4.07936,
+      "grad_norm": 0.6844797730445862,
+      "learning_rate": 4.364345738295318e-05,
+      "loss": 0.6654,
+      "step": 3187
+    },
+    {
+      "epoch": 4.08064,
+      "grad_norm": 0.6640191078186035,
+      "learning_rate": 4.3641456582633055e-05,
+      "loss": 0.5865,
+      "step": 3188
+    },
+    {
+      "epoch": 4.08192,
+      "grad_norm": 0.6174594163894653,
+      "learning_rate": 4.3639455782312926e-05,
+      "loss": 0.595,
+      "step": 3189
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.6691709160804749,
+      "learning_rate": 4.36374549819928e-05,
+      "loss": 0.6298,
+      "step": 3190
+    },
+    {
+      "epoch": 4.08448,
+      "grad_norm": 0.7091382741928101,
+      "learning_rate": 4.363545418167267e-05,
+      "loss": 0.6369,
+      "step": 3191
+    },
+    {
+      "epoch": 4.08576,
+      "grad_norm": 0.6721692681312561,
+      "learning_rate": 4.363345338135254e-05,
+      "loss": 0.6149,
+      "step": 3192
+    },
+    {
+      "epoch": 4.08704,
+      "grad_norm": 0.7344068288803101,
+      "learning_rate": 4.3631452581032414e-05,
+      "loss": 0.6966,
+      "step": 3193
+    },
+    {
+      "epoch": 4.08832,
+      "grad_norm": 0.7092353105545044,
+      "learning_rate": 4.3629451780712286e-05,
+      "loss": 0.6499,
+      "step": 3194
+    },
+    {
+      "epoch": 4.0896,
+      "grad_norm": 0.7213243246078491,
+      "learning_rate": 4.362745098039216e-05,
+      "loss": 0.6762,
+      "step": 3195
+    },
+    {
+      "epoch": 4.09088,
+      "grad_norm": 0.734781801700592,
+      "learning_rate": 4.362545018007203e-05,
+      "loss": 0.7091,
+      "step": 3196
+    },
+    {
+      "epoch": 4.09216,
+      "grad_norm": 0.7117027044296265,
+      "learning_rate": 4.36234493797519e-05,
+      "loss": 0.6111,
+      "step": 3197
+    },
+    {
+      "epoch": 4.09344,
+      "grad_norm": 0.7281467914581299,
+      "learning_rate": 4.362144857943177e-05,
+      "loss": 0.6682,
+      "step": 3198
+    },
+    {
+      "epoch": 4.09472,
+      "grad_norm": 0.7199225425720215,
+      "learning_rate": 4.3619447779111645e-05,
+      "loss": 0.6847,
+      "step": 3199
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.6403080224990845,
+      "learning_rate": 4.3617446978791524e-05,
+      "loss": 0.6134,
+      "step": 3200
+    },
+    {
+      "epoch": 4.09728,
+      "grad_norm": 0.7282823920249939,
+      "learning_rate": 4.361544617847139e-05,
+      "loss": 0.6369,
+      "step": 3201
+    },
+    {
+      "epoch": 4.09856,
+      "grad_norm": 0.673549234867096,
+      "learning_rate": 4.361344537815126e-05,
+      "loss": 0.6577,
+      "step": 3202
+    },
+    {
+      "epoch": 4.09984,
+      "grad_norm": 0.717682421207428,
+      "learning_rate": 4.361144457783113e-05,
+      "loss": 0.6413,
+      "step": 3203
+    },
+    {
+      "epoch": 4.10112,
+      "grad_norm": 0.6735298037528992,
+      "learning_rate": 4.3609443777511004e-05,
+      "loss": 0.6404,
+      "step": 3204
+    },
+    {
+      "epoch": 4.1024,
+      "grad_norm": 0.715994656085968,
+      "learning_rate": 4.3607442977190876e-05,
+      "loss": 0.677,
+      "step": 3205
+    },
+    {
+      "epoch": 4.10368,
+      "grad_norm": 0.7662949562072754,
+      "learning_rate": 4.360544217687075e-05,
+      "loss": 0.7697,
+      "step": 3206
+    },
+    {
+      "epoch": 4.10496,
+      "grad_norm": 0.7299420833587646,
+      "learning_rate": 4.360344137655063e-05,
+      "loss": 0.688,
+      "step": 3207
+    },
+    {
+      "epoch": 4.10624,
+      "grad_norm": 0.7071677446365356,
+      "learning_rate": 4.36014405762305e-05,
+      "loss": 0.6872,
+      "step": 3208
+    },
+    {
+      "epoch": 4.10752,
+      "grad_norm": 0.7498626708984375,
+      "learning_rate": 4.3599439775910364e-05,
+      "loss": 0.647,
+      "step": 3209
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.6997010707855225,
+      "learning_rate": 4.3597438975590235e-05,
+      "loss": 0.5932,
+      "step": 3210
+    },
+    {
+      "epoch": 4.11008,
+      "grad_norm": 0.6870605945587158,
+      "learning_rate": 4.359543817527011e-05,
+      "loss": 0.6438,
+      "step": 3211
+    },
+    {
+      "epoch": 4.11136,
+      "grad_norm": 0.6903749704360962,
+      "learning_rate": 4.359343737494998e-05,
+      "loss": 0.6307,
+      "step": 3212
+    },
+    {
+      "epoch": 4.11264,
+      "grad_norm": 0.7005348801612854,
+      "learning_rate": 4.359143657462985e-05,
+      "loss": 0.6214,
+      "step": 3213
+    },
+    {
+      "epoch": 4.11392,
+      "grad_norm": 0.6793337464332581,
+      "learning_rate": 4.358943577430973e-05,
+      "loss": 0.6668,
+      "step": 3214
+    },
+    {
+      "epoch": 4.1152,
+      "grad_norm": 0.7198171019554138,
+      "learning_rate": 4.35874349739896e-05,
+      "loss": 0.6384,
+      "step": 3215
+    },
+    {
+      "epoch": 4.11648,
+      "grad_norm": 0.7152752876281738,
+      "learning_rate": 4.358543417366947e-05,
+      "loss": 0.6446,
+      "step": 3216
+    },
+    {
+      "epoch": 4.11776,
+      "grad_norm": 0.7162675857543945,
+      "learning_rate": 4.358343337334934e-05,
+      "loss": 0.6621,
+      "step": 3217
+    },
+    {
+      "epoch": 4.11904,
+      "grad_norm": 0.72551029920578,
+      "learning_rate": 4.358143257302921e-05,
+      "loss": 0.6708,
+      "step": 3218
+    },
+    {
+      "epoch": 4.12032,
+      "grad_norm": 0.7436588406562805,
+      "learning_rate": 4.357943177270908e-05,
+      "loss": 0.6543,
+      "step": 3219
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.7259624600410461,
+      "learning_rate": 4.3577430972388954e-05,
+      "loss": 0.6642,
+      "step": 3220
+    },
+    {
+      "epoch": 4.12288,
+      "grad_norm": 0.6752135157585144,
+      "learning_rate": 4.357543017206883e-05,
+      "loss": 0.6495,
+      "step": 3221
+    },
+    {
+      "epoch": 4.12416,
+      "grad_norm": 0.698948323726654,
+      "learning_rate": 4.3573429371748704e-05,
+      "loss": 0.703,
+      "step": 3222
+    },
+    {
+      "epoch": 4.12544,
+      "grad_norm": 0.7195350527763367,
+      "learning_rate": 4.3571428571428576e-05,
+      "loss": 0.6924,
+      "step": 3223
+    },
+    {
+      "epoch": 4.12672,
+      "grad_norm": 0.6823775172233582,
+      "learning_rate": 4.356942777110845e-05,
+      "loss": 0.6378,
+      "step": 3224
+    },
+    {
+      "epoch": 4.128,
+      "grad_norm": 0.719937801361084,
+      "learning_rate": 4.356742697078831e-05,
+      "loss": 0.6959,
+      "step": 3225
+    },
+    {
+      "epoch": 4.12928,
+      "grad_norm": 0.6940419673919678,
+      "learning_rate": 4.3565426170468185e-05,
+      "loss": 0.6788,
+      "step": 3226
+    },
+    {
+      "epoch": 4.13056,
+      "grad_norm": 0.7241349220275879,
+      "learning_rate": 4.356342537014806e-05,
+      "loss": 0.6565,
+      "step": 3227
+    },
+    {
+      "epoch": 4.13184,
+      "grad_norm": 0.7489647269248962,
+      "learning_rate": 4.3561424569827936e-05,
+      "loss": 0.6415,
+      "step": 3228
+    },
+    {
+      "epoch": 4.13312,
+      "grad_norm": 0.7111489176750183,
+      "learning_rate": 4.355942376950781e-05,
+      "loss": 0.6258,
+      "step": 3229
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.6695640087127686,
+      "learning_rate": 4.355742296918768e-05,
+      "loss": 0.612,
+      "step": 3230
+    },
+    {
+      "epoch": 4.13568,
+      "grad_norm": 0.6810483336448669,
+      "learning_rate": 4.355542216886755e-05,
+      "loss": 0.6202,
+      "step": 3231
+    },
+    {
+      "epoch": 4.13696,
+      "grad_norm": 0.7654765844345093,
+      "learning_rate": 4.355342136854742e-05,
+      "loss": 0.6865,
+      "step": 3232
+    },
+    {
+      "epoch": 4.13824,
+      "grad_norm": 0.7125704288482666,
+      "learning_rate": 4.355142056822729e-05,
+      "loss": 0.6864,
+      "step": 3233
+    },
+    {
+      "epoch": 4.13952,
+      "grad_norm": 0.7397105097770691,
+      "learning_rate": 4.354941976790716e-05,
+      "loss": 0.6494,
+      "step": 3234
+    },
+    {
+      "epoch": 4.1408,
+      "grad_norm": 0.6764322519302368,
+      "learning_rate": 4.354741896758704e-05,
+      "loss": 0.6203,
+      "step": 3235
+    },
+    {
+      "epoch": 4.14208,
+      "grad_norm": 0.6737890839576721,
+      "learning_rate": 4.354541816726691e-05,
+      "loss": 0.6531,
+      "step": 3236
+    },
+    {
+      "epoch": 4.14336,
+      "grad_norm": 0.6234689950942993,
+      "learning_rate": 4.354341736694678e-05,
+      "loss": 0.5897,
+      "step": 3237
+    },
+    {
+      "epoch": 4.14464,
+      "grad_norm": 0.7278350591659546,
+      "learning_rate": 4.3541416566626654e-05,
+      "loss": 0.7032,
+      "step": 3238
+    },
+    {
+      "epoch": 4.14592,
+      "grad_norm": 0.645071804523468,
+      "learning_rate": 4.3539415766306526e-05,
+      "loss": 0.6347,
+      "step": 3239
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.6949378848075867,
+      "learning_rate": 4.35374149659864e-05,
+      "loss": 0.624,
+      "step": 3240
+    },
+    {
+      "epoch": 4.14848,
+      "grad_norm": 0.6974506974220276,
+      "learning_rate": 4.353541416566626e-05,
+      "loss": 0.6416,
+      "step": 3241
+    },
+    {
+      "epoch": 4.14976,
+      "grad_norm": 0.7012295722961426,
+      "learning_rate": 4.353341336534614e-05,
+      "loss": 0.6061,
+      "step": 3242
+    },
+    {
+      "epoch": 4.15104,
+      "grad_norm": 0.6732747554779053,
+      "learning_rate": 4.3531412565026013e-05,
+      "loss": 0.6482,
+      "step": 3243
+    },
+    {
+      "epoch": 4.15232,
+      "grad_norm": 0.6656915545463562,
+      "learning_rate": 4.3529411764705885e-05,
+      "loss": 0.6304,
+      "step": 3244
+    },
+    {
+      "epoch": 4.1536,
+      "grad_norm": 0.6815019845962524,
+      "learning_rate": 4.352741096438576e-05,
+      "loss": 0.5652,
+      "step": 3245
+    },
+    {
+      "epoch": 4.15488,
+      "grad_norm": 0.739524245262146,
+      "learning_rate": 4.352541016406563e-05,
+      "loss": 0.6389,
+      "step": 3246
+    },
+    {
+      "epoch": 4.15616,
+      "grad_norm": 0.641751766204834,
+      "learning_rate": 4.35234093637455e-05,
+      "loss": 0.6259,
+      "step": 3247
+    },
+    {
+      "epoch": 4.15744,
+      "grad_norm": 0.6507452130317688,
+      "learning_rate": 4.352140856342537e-05,
+      "loss": 0.5613,
+      "step": 3248
+    },
+    {
+      "epoch": 4.15872,
+      "grad_norm": 0.7066096663475037,
+      "learning_rate": 4.3519407763105245e-05,
+      "loss": 0.5779,
+      "step": 3249
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.6899321675300598,
+      "learning_rate": 4.3517406962785116e-05,
+      "loss": 0.6098,
+      "step": 3250
+    },
+    {
+      "epoch": 4.16128,
+      "grad_norm": 0.7460212707519531,
+      "learning_rate": 4.351540616246499e-05,
+      "loss": 0.6969,
+      "step": 3251
+    },
+    {
+      "epoch": 4.16256,
+      "grad_norm": 0.6735222339630127,
+      "learning_rate": 4.351340536214486e-05,
+      "loss": 0.5961,
+      "step": 3252
+    },
+    {
+      "epoch": 4.16384,
+      "grad_norm": 0.668106198310852,
+      "learning_rate": 4.351140456182473e-05,
+      "loss": 0.588,
+      "step": 3253
+    },
+    {
+      "epoch": 4.16512,
+      "grad_norm": 0.7419319748878479,
+      "learning_rate": 4.3509403761504604e-05,
+      "loss": 0.6707,
+      "step": 3254
+    },
+    {
+      "epoch": 4.1664,
+      "grad_norm": 0.7336167693138123,
+      "learning_rate": 4.3507402961184476e-05,
+      "loss": 0.6756,
+      "step": 3255
+    },
+    {
+      "epoch": 4.16768,
+      "grad_norm": 0.7352126240730286,
+      "learning_rate": 4.350540216086435e-05,
+      "loss": 0.6582,
+      "step": 3256
+    },
+    {
+      "epoch": 4.16896,
+      "grad_norm": 0.6964594721794128,
+      "learning_rate": 4.350340136054422e-05,
+      "loss": 0.6481,
+      "step": 3257
+    },
+    {
+      "epoch": 4.17024,
+      "grad_norm": 0.6809775233268738,
+      "learning_rate": 4.350140056022409e-05,
+      "loss": 0.6216,
+      "step": 3258
+    },
+    {
+      "epoch": 4.17152,
+      "grad_norm": 0.7008639574050903,
+      "learning_rate": 4.349939975990396e-05,
+      "loss": 0.6148,
+      "step": 3259
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.7400866150856018,
+      "learning_rate": 4.3497398959583835e-05,
+      "loss": 0.6512,
+      "step": 3260
+    },
+    {
+      "epoch": 4.17408,
+      "grad_norm": 0.6811804175376892,
+      "learning_rate": 4.349539815926371e-05,
+      "loss": 0.6134,
+      "step": 3261
+    },
+    {
+      "epoch": 4.17536,
+      "grad_norm": 0.6632633805274963,
+      "learning_rate": 4.349339735894358e-05,
+      "loss": 0.6315,
+      "step": 3262
+    },
+    {
+      "epoch": 4.17664,
+      "grad_norm": 0.6872732043266296,
+      "learning_rate": 4.349139655862345e-05,
+      "loss": 0.6406,
+      "step": 3263
+    },
+    {
+      "epoch": 4.17792,
+      "grad_norm": 0.7408565282821655,
+      "learning_rate": 4.348939575830332e-05,
+      "loss": 0.6544,
+      "step": 3264
+    },
+    {
+      "epoch": 4.1792,
+      "grad_norm": 0.7121074199676514,
+      "learning_rate": 4.3487394957983194e-05,
+      "loss": 0.6336,
+      "step": 3265
+    },
+    {
+      "epoch": 4.18048,
+      "grad_norm": 0.7045162916183472,
+      "learning_rate": 4.3485394157663066e-05,
+      "loss": 0.7007,
+      "step": 3266
+    },
+    {
+      "epoch": 4.18176,
+      "grad_norm": 0.7155641317367554,
+      "learning_rate": 4.348339335734294e-05,
+      "loss": 0.7133,
+      "step": 3267
+    },
+    {
+      "epoch": 4.18304,
+      "grad_norm": 0.7041541934013367,
+      "learning_rate": 4.348139255702281e-05,
+      "loss": 0.6453,
+      "step": 3268
+    },
+    {
+      "epoch": 4.18432,
+      "grad_norm": 0.7294322848320007,
+      "learning_rate": 4.347939175670268e-05,
+      "loss": 0.6818,
+      "step": 3269
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.6842989325523376,
+      "learning_rate": 4.347739095638256e-05,
+      "loss": 0.654,
+      "step": 3270
+    },
+    {
+      "epoch": 4.18688,
+      "grad_norm": 0.7636930346488953,
+      "learning_rate": 4.3475390156062425e-05,
+      "loss": 0.6433,
+      "step": 3271
+    },
+    {
+      "epoch": 4.18816,
+      "grad_norm": 0.7250775694847107,
+      "learning_rate": 4.34733893557423e-05,
+      "loss": 0.6537,
+      "step": 3272
+    },
+    {
+      "epoch": 4.18944,
+      "grad_norm": 0.7232015132904053,
+      "learning_rate": 4.347138855542217e-05,
+      "loss": 0.6162,
+      "step": 3273
+    },
+    {
+      "epoch": 4.19072,
+      "grad_norm": 0.7661274671554565,
+      "learning_rate": 4.346938775510204e-05,
+      "loss": 0.6299,
+      "step": 3274
+    },
+    {
+      "epoch": 4.192,
+      "grad_norm": 0.6594750285148621,
+      "learning_rate": 4.346738695478191e-05,
+      "loss": 0.6141,
+      "step": 3275
+    },
+    {
+      "epoch": 4.19328,
+      "grad_norm": 0.7808811664581299,
+      "learning_rate": 4.3465386154461785e-05,
+      "loss": 0.7105,
+      "step": 3276
+    },
+    {
+      "epoch": 4.19456,
+      "grad_norm": 0.6678184866905212,
+      "learning_rate": 4.346338535414166e-05,
+      "loss": 0.618,
+      "step": 3277
+    },
+    {
+      "epoch": 4.19584,
+      "grad_norm": 0.7320860624313354,
+      "learning_rate": 4.3461384553821535e-05,
+      "loss": 0.6352,
+      "step": 3278
+    },
+    {
+      "epoch": 4.19712,
+      "grad_norm": 0.7203635573387146,
+      "learning_rate": 4.34593837535014e-05,
+      "loss": 0.6398,
+      "step": 3279
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.714809238910675,
+      "learning_rate": 4.345738295318127e-05,
+      "loss": 0.6567,
+      "step": 3280
+    },
+    {
+      "epoch": 4.19968,
+      "grad_norm": 0.6897768378257751,
+      "learning_rate": 4.3455382152861144e-05,
+      "loss": 0.6413,
+      "step": 3281
+    },
+    {
+      "epoch": 4.20096,
+      "grad_norm": 0.7241793274879456,
+      "learning_rate": 4.3453381352541016e-05,
+      "loss": 0.6331,
+      "step": 3282
+    },
+    {
+      "epoch": 4.20224,
+      "grad_norm": 0.6980943083763123,
+      "learning_rate": 4.345138055222089e-05,
+      "loss": 0.6244,
+      "step": 3283
+    },
+    {
+      "epoch": 4.20352,
+      "grad_norm": 0.7093860507011414,
+      "learning_rate": 4.3449379751900766e-05,
+      "loss": 0.6415,
+      "step": 3284
+    },
+    {
+      "epoch": 4.2048,
+      "grad_norm": 0.7445345520973206,
+      "learning_rate": 4.344737895158064e-05,
+      "loss": 0.6459,
+      "step": 3285
+    },
+    {
+      "epoch": 4.20608,
+      "grad_norm": 0.6987217664718628,
+      "learning_rate": 4.344537815126051e-05,
+      "loss": 0.6554,
+      "step": 3286
+    },
+    {
+      "epoch": 4.2073599999999995,
+      "grad_norm": 0.7271133065223694,
+      "learning_rate": 4.3443377350940375e-05,
+      "loss": 0.6412,
+      "step": 3287
+    },
+    {
+      "epoch": 4.20864,
+      "grad_norm": 0.7157135605812073,
+      "learning_rate": 4.344137655062025e-05,
+      "loss": 0.6508,
+      "step": 3288
+    },
+    {
+      "epoch": 4.20992,
+      "grad_norm": 0.666031539440155,
+      "learning_rate": 4.343937575030012e-05,
+      "loss": 0.6324,
+      "step": 3289
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.7204000353813171,
+      "learning_rate": 4.343737494997999e-05,
+      "loss": 0.6359,
+      "step": 3290
+    },
+    {
+      "epoch": 4.21248,
+      "grad_norm": 0.6386025547981262,
+      "learning_rate": 4.343537414965987e-05,
+      "loss": 0.5998,
+      "step": 3291
+    },
+    {
+      "epoch": 4.21376,
+      "grad_norm": 0.7287147641181946,
+      "learning_rate": 4.343337334933974e-05,
+      "loss": 0.6929,
+      "step": 3292
+    },
+    {
+      "epoch": 4.21504,
+      "grad_norm": 0.7094889283180237,
+      "learning_rate": 4.343137254901961e-05,
+      "loss": 0.6417,
+      "step": 3293
+    },
+    {
+      "epoch": 4.21632,
+      "grad_norm": 0.6859930753707886,
+      "learning_rate": 4.3429371748699485e-05,
+      "loss": 0.6024,
+      "step": 3294
+    },
+    {
+      "epoch": 4.2176,
+      "grad_norm": 0.6752464771270752,
+      "learning_rate": 4.342737094837935e-05,
+      "loss": 0.6713,
+      "step": 3295
+    },
+    {
+      "epoch": 4.21888,
+      "grad_norm": 0.6904641389846802,
+      "learning_rate": 4.342537014805922e-05,
+      "loss": 0.636,
+      "step": 3296
+    },
+    {
+      "epoch": 4.22016,
+      "grad_norm": 0.6931844353675842,
+      "learning_rate": 4.3423369347739094e-05,
+      "loss": 0.6189,
+      "step": 3297
+    },
+    {
+      "epoch": 4.22144,
+      "grad_norm": 0.6663493514060974,
+      "learning_rate": 4.342136854741897e-05,
+      "loss": 0.5673,
+      "step": 3298
+    },
+    {
+      "epoch": 4.22272,
+      "grad_norm": 0.7185217142105103,
+      "learning_rate": 4.3419367747098844e-05,
+      "loss": 0.6528,
+      "step": 3299
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.7160610556602478,
+      "learning_rate": 4.3417366946778716e-05,
+      "loss": 0.6577,
+      "step": 3300
+    },
+    {
+      "epoch": 4.22528,
+      "grad_norm": 0.6835800409317017,
+      "learning_rate": 4.341536614645859e-05,
+      "loss": 0.6558,
+      "step": 3301
+    },
+    {
+      "epoch": 4.22656,
+      "grad_norm": 0.7105692028999329,
+      "learning_rate": 4.341336534613846e-05,
+      "loss": 0.6875,
+      "step": 3302
+    },
+    {
+      "epoch": 4.22784,
+      "grad_norm": 0.7384511232376099,
+      "learning_rate": 4.3411364545818325e-05,
+      "loss": 0.6484,
+      "step": 3303
+    },
+    {
+      "epoch": 4.22912,
+      "grad_norm": 0.7124598622322083,
+      "learning_rate": 4.34093637454982e-05,
+      "loss": 0.6571,
+      "step": 3304
+    },
+    {
+      "epoch": 4.2304,
+      "grad_norm": 0.718474268913269,
+      "learning_rate": 4.340736294517807e-05,
+      "loss": 0.6457,
+      "step": 3305
+    },
+    {
+      "epoch": 4.23168,
+      "grad_norm": 0.6663171648979187,
+      "learning_rate": 4.340536214485795e-05,
+      "loss": 0.592,
+      "step": 3306
+    },
+    {
+      "epoch": 4.23296,
+      "grad_norm": 0.6899667382240295,
+      "learning_rate": 4.340336134453782e-05,
+      "loss": 0.6845,
+      "step": 3307
+    },
+    {
+      "epoch": 4.23424,
+      "grad_norm": 0.7246834635734558,
+      "learning_rate": 4.340136054421769e-05,
+      "loss": 0.6739,
+      "step": 3308
+    },
+    {
+      "epoch": 4.23552,
+      "grad_norm": 0.7182645797729492,
+      "learning_rate": 4.339935974389756e-05,
+      "loss": 0.6249,
+      "step": 3309
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.706695020198822,
+      "learning_rate": 4.3397358943577435e-05,
+      "loss": 0.6775,
+      "step": 3310
+    },
+    {
+      "epoch": 4.23808,
+      "grad_norm": 0.7705980539321899,
+      "learning_rate": 4.33953581432573e-05,
+      "loss": 0.7309,
+      "step": 3311
+    },
+    {
+      "epoch": 4.23936,
+      "grad_norm": 0.7092674970626831,
+      "learning_rate": 4.339335734293717e-05,
+      "loss": 0.6438,
+      "step": 3312
+    },
+    {
+      "epoch": 4.24064,
+      "grad_norm": 0.7281712889671326,
+      "learning_rate": 4.339135654261705e-05,
+      "loss": 0.6273,
+      "step": 3313
+    },
+    {
+      "epoch": 4.24192,
+      "grad_norm": 0.7132616639137268,
+      "learning_rate": 4.338935574229692e-05,
+      "loss": 0.6189,
+      "step": 3314
+    },
+    {
+      "epoch": 4.2432,
+      "grad_norm": 0.707158625125885,
+      "learning_rate": 4.3387354941976794e-05,
+      "loss": 0.6604,
+      "step": 3315
+    },
+    {
+      "epoch": 4.24448,
+      "grad_norm": 0.7403045892715454,
+      "learning_rate": 4.3385354141656666e-05,
+      "loss": 0.6983,
+      "step": 3316
+    },
+    {
+      "epoch": 4.24576,
+      "grad_norm": 0.7587504982948303,
+      "learning_rate": 4.338335334133654e-05,
+      "loss": 0.6572,
+      "step": 3317
+    },
+    {
+      "epoch": 4.24704,
+      "grad_norm": 0.682956337928772,
+      "learning_rate": 4.338135254101641e-05,
+      "loss": 0.6519,
+      "step": 3318
+    },
+    {
+      "epoch": 4.24832,
+      "grad_norm": 0.6858289241790771,
+      "learning_rate": 4.3379351740696275e-05,
+      "loss": 0.6171,
+      "step": 3319
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.7262416481971741,
+      "learning_rate": 4.337735094037615e-05,
+      "loss": 0.6585,
+      "step": 3320
+    },
+    {
+      "epoch": 4.25088,
+      "grad_norm": 0.6875229477882385,
+      "learning_rate": 4.3375350140056025e-05,
+      "loss": 0.6579,
+      "step": 3321
+    },
+    {
+      "epoch": 4.25216,
+      "grad_norm": 0.7535980343818665,
+      "learning_rate": 4.33733493397359e-05,
+      "loss": 0.7405,
+      "step": 3322
+    },
+    {
+      "epoch": 4.25344,
+      "grad_norm": 0.6897569298744202,
+      "learning_rate": 4.337134853941577e-05,
+      "loss": 0.6639,
+      "step": 3323
+    },
+    {
+      "epoch": 4.25472,
+      "grad_norm": 0.7273359298706055,
+      "learning_rate": 4.336934773909564e-05,
+      "loss": 0.6645,
+      "step": 3324
+    },
+    {
+      "epoch": 4.256,
+      "grad_norm": 0.7021262645721436,
+      "learning_rate": 4.336734693877551e-05,
+      "loss": 0.5955,
+      "step": 3325
+    },
+    {
+      "epoch": 4.25728,
+      "grad_norm": 0.6609035134315491,
+      "learning_rate": 4.3365346138455384e-05,
+      "loss": 0.6416,
+      "step": 3326
+    },
+    {
+      "epoch": 4.25856,
+      "grad_norm": 0.8022797703742981,
+      "learning_rate": 4.3363345338135256e-05,
+      "loss": 0.6363,
+      "step": 3327
+    },
+    {
+      "epoch": 4.25984,
+      "grad_norm": 0.7287960648536682,
+      "learning_rate": 4.336134453781513e-05,
+      "loss": 0.6815,
+      "step": 3328
+    },
+    {
+      "epoch": 4.26112,
+      "grad_norm": 0.6661964654922485,
+      "learning_rate": 4.3359343737495e-05,
+      "loss": 0.6532,
+      "step": 3329
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.7338867783546448,
+      "learning_rate": 4.335734293717487e-05,
+      "loss": 0.7156,
+      "step": 3330
+    },
+    {
+      "epoch": 4.26368,
+      "grad_norm": 0.7184217572212219,
+      "learning_rate": 4.3355342136854744e-05,
+      "loss": 0.6892,
+      "step": 3331
+    },
+    {
+      "epoch": 4.26496,
+      "grad_norm": 0.715719997882843,
+      "learning_rate": 4.3353341336534615e-05,
+      "loss": 0.6508,
+      "step": 3332
+    },
+    {
+      "epoch": 4.26624,
+      "grad_norm": 0.6484602689743042,
+      "learning_rate": 4.335134053621449e-05,
+      "loss": 0.6111,
+      "step": 3333
+    },
+    {
+      "epoch": 4.26752,
+      "grad_norm": 0.6997168064117432,
+      "learning_rate": 4.334933973589436e-05,
+      "loss": 0.6312,
+      "step": 3334
+    },
+    {
+      "epoch": 4.2688,
+      "grad_norm": 0.7127143740653992,
+      "learning_rate": 4.334733893557423e-05,
+      "loss": 0.6445,
+      "step": 3335
+    },
+    {
+      "epoch": 4.27008,
+      "grad_norm": 0.736754298210144,
+      "learning_rate": 4.33453381352541e-05,
+      "loss": 0.6678,
+      "step": 3336
+    },
+    {
+      "epoch": 4.27136,
+      "grad_norm": 0.7001957893371582,
+      "learning_rate": 4.3343337334933975e-05,
+      "loss": 0.6383,
+      "step": 3337
+    },
+    {
+      "epoch": 4.27264,
+      "grad_norm": 0.6919710040092468,
+      "learning_rate": 4.334133653461385e-05,
+      "loss": 0.6978,
+      "step": 3338
+    },
+    {
+      "epoch": 4.27392,
+      "grad_norm": 0.7082419991493225,
+      "learning_rate": 4.333933573429372e-05,
+      "loss": 0.6756,
+      "step": 3339
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.7254000306129456,
+      "learning_rate": 4.333733493397359e-05,
+      "loss": 0.6301,
+      "step": 3340
+    },
+    {
+      "epoch": 4.27648,
+      "grad_norm": 0.7085341811180115,
+      "learning_rate": 4.333533413365347e-05,
+      "loss": 0.6495,
+      "step": 3341
+    },
+    {
+      "epoch": 4.27776,
+      "grad_norm": 0.699360728263855,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.6426,
+      "step": 3342
+    },
+    {
+      "epoch": 4.27904,
+      "grad_norm": 0.7209745049476624,
+      "learning_rate": 4.3331332533013206e-05,
+      "loss": 0.6308,
+      "step": 3343
+    },
+    {
+      "epoch": 4.28032,
+      "grad_norm": 0.7370860576629639,
+      "learning_rate": 4.332933173269308e-05,
+      "loss": 0.623,
+      "step": 3344
+    },
+    {
+      "epoch": 4.2816,
+      "grad_norm": 0.7051844596862793,
+      "learning_rate": 4.332733093237295e-05,
+      "loss": 0.6661,
+      "step": 3345
+    },
+    {
+      "epoch": 4.2828800000000005,
+      "grad_norm": 0.6612274050712585,
+      "learning_rate": 4.332533013205282e-05,
+      "loss": 0.5542,
+      "step": 3346
+    },
+    {
+      "epoch": 4.28416,
+      "grad_norm": 0.7362515926361084,
+      "learning_rate": 4.332332933173269e-05,
+      "loss": 0.7012,
+      "step": 3347
+    },
+    {
+      "epoch": 4.28544,
+      "grad_norm": 0.7105290293693542,
+      "learning_rate": 4.332132853141257e-05,
+      "loss": 0.6536,
+      "step": 3348
+    },
+    {
+      "epoch": 4.28672,
+      "grad_norm": 0.7252454161643982,
+      "learning_rate": 4.3319327731092444e-05,
+      "loss": 0.6018,
+      "step": 3349
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.706749677658081,
+      "learning_rate": 4.331732693077231e-05,
+      "loss": 0.6721,
+      "step": 3350
+    },
+    {
+      "epoch": 4.28928,
+      "grad_norm": 0.6935198307037354,
+      "learning_rate": 4.331532613045218e-05,
+      "loss": 0.6268,
+      "step": 3351
+    },
+    {
+      "epoch": 4.29056,
+      "grad_norm": 0.708602786064148,
+      "learning_rate": 4.331332533013205e-05,
+      "loss": 0.6309,
+      "step": 3352
+    },
+    {
+      "epoch": 4.29184,
+      "grad_norm": 0.6960015892982483,
+      "learning_rate": 4.3311324529811924e-05,
+      "loss": 0.6418,
+      "step": 3353
+    },
+    {
+      "epoch": 4.29312,
+      "grad_norm": 0.7291163206100464,
+      "learning_rate": 4.3309323729491796e-05,
+      "loss": 0.6651,
+      "step": 3354
+    },
+    {
+      "epoch": 4.2943999999999996,
+      "grad_norm": 0.7136040925979614,
+      "learning_rate": 4.3307322929171675e-05,
+      "loss": 0.6568,
+      "step": 3355
+    },
+    {
+      "epoch": 4.29568,
+      "grad_norm": 0.7023734450340271,
+      "learning_rate": 4.330532212885155e-05,
+      "loss": 0.6535,
+      "step": 3356
+    },
+    {
+      "epoch": 4.29696,
+      "grad_norm": 0.706221878528595,
+      "learning_rate": 4.330332132853142e-05,
+      "loss": 0.6287,
+      "step": 3357
+    },
+    {
+      "epoch": 4.29824,
+      "grad_norm": 0.6975945830345154,
+      "learning_rate": 4.3301320528211284e-05,
+      "loss": 0.635,
+      "step": 3358
+    },
+    {
+      "epoch": 4.29952,
+      "grad_norm": 0.7090380787849426,
+      "learning_rate": 4.3299319727891156e-05,
+      "loss": 0.6462,
+      "step": 3359
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.7626847624778748,
+      "learning_rate": 4.329731892757103e-05,
+      "loss": 0.6438,
+      "step": 3360
+    },
+    {
+      "epoch": 4.30208,
+      "grad_norm": 0.6872966289520264,
+      "learning_rate": 4.32953181272509e-05,
+      "loss": 0.6682,
+      "step": 3361
+    },
+    {
+      "epoch": 4.30336,
+      "grad_norm": 0.6952762603759766,
+      "learning_rate": 4.329331732693078e-05,
+      "loss": 0.6368,
+      "step": 3362
+    },
+    {
+      "epoch": 4.30464,
+      "grad_norm": 0.7490139007568359,
+      "learning_rate": 4.329131652661065e-05,
+      "loss": 0.6352,
+      "step": 3363
+    },
+    {
+      "epoch": 4.30592,
+      "grad_norm": 0.6900007128715515,
+      "learning_rate": 4.328931572629052e-05,
+      "loss": 0.652,
+      "step": 3364
+    },
+    {
+      "epoch": 4.3072,
+      "grad_norm": 0.7102410197257996,
+      "learning_rate": 4.3287314925970394e-05,
+      "loss": 0.6628,
+      "step": 3365
+    },
+    {
+      "epoch": 4.30848,
+      "grad_norm": 0.7227296233177185,
+      "learning_rate": 4.328531412565026e-05,
+      "loss": 0.6377,
+      "step": 3366
+    },
+    {
+      "epoch": 4.30976,
+      "grad_norm": 0.7517803311347961,
+      "learning_rate": 4.328331332533013e-05,
+      "loss": 0.7112,
+      "step": 3367
+    },
+    {
+      "epoch": 4.31104,
+      "grad_norm": 0.6790129542350769,
+      "learning_rate": 4.328131252501e-05,
+      "loss": 0.5814,
+      "step": 3368
+    },
+    {
+      "epoch": 4.31232,
+      "grad_norm": 0.6734822988510132,
+      "learning_rate": 4.327931172468988e-05,
+      "loss": 0.6569,
+      "step": 3369
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.6671125888824463,
+      "learning_rate": 4.327731092436975e-05,
+      "loss": 0.6558,
+      "step": 3370
+    },
+    {
+      "epoch": 4.31488,
+      "grad_norm": 0.6953800916671753,
+      "learning_rate": 4.3275310124049625e-05,
+      "loss": 0.6398,
+      "step": 3371
+    },
+    {
+      "epoch": 4.31616,
+      "grad_norm": 0.669097900390625,
+      "learning_rate": 4.3273309323729497e-05,
+      "loss": 0.6175,
+      "step": 3372
+    },
+    {
+      "epoch": 4.31744,
+      "grad_norm": 0.6786286234855652,
+      "learning_rate": 4.327130852340937e-05,
+      "loss": 0.6225,
+      "step": 3373
+    },
+    {
+      "epoch": 4.31872,
+      "grad_norm": 0.7077885866165161,
+      "learning_rate": 4.3269307723089233e-05,
+      "loss": 0.6297,
+      "step": 3374
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.7308838367462158,
+      "learning_rate": 4.3267306922769105e-05,
+      "loss": 0.6341,
+      "step": 3375
+    },
+    {
+      "epoch": 4.32128,
+      "grad_norm": 0.7163229584693909,
+      "learning_rate": 4.3265306122448984e-05,
+      "loss": 0.6429,
+      "step": 3376
+    },
+    {
+      "epoch": 4.32256,
+      "grad_norm": 0.7708077430725098,
+      "learning_rate": 4.3263305322128856e-05,
+      "loss": 0.7227,
+      "step": 3377
+    },
+    {
+      "epoch": 4.32384,
+      "grad_norm": 0.694496214389801,
+      "learning_rate": 4.326130452180873e-05,
+      "loss": 0.6162,
+      "step": 3378
+    },
+    {
+      "epoch": 4.32512,
+      "grad_norm": 0.666400671005249,
+      "learning_rate": 4.32593037214886e-05,
+      "loss": 0.5697,
+      "step": 3379
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.7009501457214355,
+      "learning_rate": 4.325730292116847e-05,
+      "loss": 0.664,
+      "step": 3380
+    },
+    {
+      "epoch": 4.32768,
+      "grad_norm": 0.7188718914985657,
+      "learning_rate": 4.325530212084834e-05,
+      "loss": 0.6408,
+      "step": 3381
+    },
+    {
+      "epoch": 4.32896,
+      "grad_norm": 0.6969630122184753,
+      "learning_rate": 4.325330132052821e-05,
+      "loss": 0.6023,
+      "step": 3382
+    },
+    {
+      "epoch": 4.33024,
+      "grad_norm": 0.6989743113517761,
+      "learning_rate": 4.325130052020809e-05,
+      "loss": 0.652,
+      "step": 3383
+    },
+    {
+      "epoch": 4.33152,
+      "grad_norm": 0.7306556701660156,
+      "learning_rate": 4.324929971988796e-05,
+      "loss": 0.7016,
+      "step": 3384
+    },
+    {
+      "epoch": 4.3328,
+      "grad_norm": 0.6765083074569702,
+      "learning_rate": 4.324729891956783e-05,
+      "loss": 0.6264,
+      "step": 3385
+    },
+    {
+      "epoch": 4.33408,
+      "grad_norm": 0.6959035396575928,
+      "learning_rate": 4.32452981192477e-05,
+      "loss": 0.5973,
+      "step": 3386
+    },
+    {
+      "epoch": 4.33536,
+      "grad_norm": 0.6781784892082214,
+      "learning_rate": 4.3243297318927574e-05,
+      "loss": 0.6569,
+      "step": 3387
+    },
+    {
+      "epoch": 4.33664,
+      "grad_norm": 0.7198589444160461,
+      "learning_rate": 4.3241296518607446e-05,
+      "loss": 0.6785,
+      "step": 3388
+    },
+    {
+      "epoch": 4.33792,
+      "grad_norm": 0.74173504114151,
+      "learning_rate": 4.323929571828732e-05,
+      "loss": 0.6489,
+      "step": 3389
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.7088286876678467,
+      "learning_rate": 4.323729491796719e-05,
+      "loss": 0.6311,
+      "step": 3390
+    },
+    {
+      "epoch": 4.34048,
+      "grad_norm": 0.7243875861167908,
+      "learning_rate": 4.323529411764706e-05,
+      "loss": 0.64,
+      "step": 3391
+    },
+    {
+      "epoch": 4.34176,
+      "grad_norm": 0.7260725498199463,
+      "learning_rate": 4.3233293317326934e-05,
+      "loss": 0.6259,
+      "step": 3392
+    },
+    {
+      "epoch": 4.34304,
+      "grad_norm": 0.6951209902763367,
+      "learning_rate": 4.3231292517006806e-05,
+      "loss": 0.6162,
+      "step": 3393
+    },
+    {
+      "epoch": 4.34432,
+      "grad_norm": 0.6746131181716919,
+      "learning_rate": 4.322929171668668e-05,
+      "loss": 0.6122,
+      "step": 3394
+    },
+    {
+      "epoch": 4.3456,
+      "grad_norm": 0.7057126760482788,
+      "learning_rate": 4.322729091636655e-05,
+      "loss": 0.6491,
+      "step": 3395
+    },
+    {
+      "epoch": 4.34688,
+      "grad_norm": 0.7594478726387024,
+      "learning_rate": 4.322529011604642e-05,
+      "loss": 0.6743,
+      "step": 3396
+    },
+    {
+      "epoch": 4.34816,
+      "grad_norm": 0.6961939930915833,
+      "learning_rate": 4.322328931572629e-05,
+      "loss": 0.6814,
+      "step": 3397
+    },
+    {
+      "epoch": 4.3494399999999995,
+      "grad_norm": 0.7641561627388,
+      "learning_rate": 4.3221288515406165e-05,
+      "loss": 0.6787,
+      "step": 3398
+    },
+    {
+      "epoch": 4.35072,
+      "grad_norm": 0.7088523507118225,
+      "learning_rate": 4.321928771508604e-05,
+      "loss": 0.6936,
+      "step": 3399
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.7125248908996582,
+      "learning_rate": 4.321728691476591e-05,
+      "loss": 0.6512,
+      "step": 3400
+    },
+    {
+      "epoch": 4.35328,
+      "grad_norm": 0.7159749269485474,
+      "learning_rate": 4.321528611444578e-05,
+      "loss": 0.6013,
+      "step": 3401
+    },
+    {
+      "epoch": 4.35456,
+      "grad_norm": 0.7057827115058899,
+      "learning_rate": 4.321328531412565e-05,
+      "loss": 0.6754,
+      "step": 3402
+    },
+    {
+      "epoch": 4.35584,
+      "grad_norm": 0.7021535038948059,
+      "learning_rate": 4.3211284513805524e-05,
+      "loss": 0.6203,
+      "step": 3403
+    },
+    {
+      "epoch": 4.35712,
+      "grad_norm": 0.6961078643798828,
+      "learning_rate": 4.3209283713485396e-05,
+      "loss": 0.6115,
+      "step": 3404
+    },
+    {
+      "epoch": 4.3584,
+      "grad_norm": 0.6545549035072327,
+      "learning_rate": 4.320728291316527e-05,
+      "loss": 0.5919,
+      "step": 3405
+    },
+    {
+      "epoch": 4.35968,
+      "grad_norm": 0.7263978123664856,
+      "learning_rate": 4.320528211284514e-05,
+      "loss": 0.6691,
+      "step": 3406
+    },
+    {
+      "epoch": 4.36096,
+      "grad_norm": 0.7528616189956665,
+      "learning_rate": 4.320328131252501e-05,
+      "loss": 0.7051,
+      "step": 3407
+    },
+    {
+      "epoch": 4.36224,
+      "grad_norm": 0.6853641867637634,
+      "learning_rate": 4.320128051220488e-05,
+      "loss": 0.6242,
+      "step": 3408
+    },
+    {
+      "epoch": 4.36352,
+      "grad_norm": 0.7724465727806091,
+      "learning_rate": 4.3199279711884755e-05,
+      "loss": 0.5945,
+      "step": 3409
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.6866872310638428,
+      "learning_rate": 4.319727891156463e-05,
+      "loss": 0.6668,
+      "step": 3410
+    },
+    {
+      "epoch": 4.36608,
+      "grad_norm": 0.7362887263298035,
+      "learning_rate": 4.3195278111244506e-05,
+      "loss": 0.6773,
+      "step": 3411
+    },
+    {
+      "epoch": 4.36736,
+      "grad_norm": 0.7267001867294312,
+      "learning_rate": 4.319327731092437e-05,
+      "loss": 0.651,
+      "step": 3412
+    },
+    {
+      "epoch": 4.36864,
+      "grad_norm": 0.7169755697250366,
+      "learning_rate": 4.319127651060424e-05,
+      "loss": 0.7017,
+      "step": 3413
+    },
+    {
+      "epoch": 4.3699200000000005,
+      "grad_norm": 0.7027722597122192,
+      "learning_rate": 4.3189275710284115e-05,
+      "loss": 0.5944,
+      "step": 3414
+    },
+    {
+      "epoch": 4.3712,
+      "grad_norm": 0.7033854722976685,
+      "learning_rate": 4.3187274909963986e-05,
+      "loss": 0.6433,
+      "step": 3415
+    },
+    {
+      "epoch": 4.37248,
+      "grad_norm": 0.7184159755706787,
+      "learning_rate": 4.318527410964386e-05,
+      "loss": 0.6733,
+      "step": 3416
+    },
+    {
+      "epoch": 4.37376,
+      "grad_norm": 0.7299154996871948,
+      "learning_rate": 4.318327330932373e-05,
+      "loss": 0.7094,
+      "step": 3417
+    },
+    {
+      "epoch": 4.37504,
+      "grad_norm": 0.6535490155220032,
+      "learning_rate": 4.31812725090036e-05,
+      "loss": 0.6,
+      "step": 3418
+    },
+    {
+      "epoch": 4.37632,
+      "grad_norm": 0.6924900412559509,
+      "learning_rate": 4.317927170868348e-05,
+      "loss": 0.632,
+      "step": 3419
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.7481140494346619,
+      "learning_rate": 4.3177270908363346e-05,
+      "loss": 0.6669,
+      "step": 3420
+    },
+    {
+      "epoch": 4.37888,
+      "grad_norm": 0.6912330985069275,
+      "learning_rate": 4.317527010804322e-05,
+      "loss": 0.6514,
+      "step": 3421
+    },
+    {
+      "epoch": 4.38016,
+      "grad_norm": 0.7058601379394531,
+      "learning_rate": 4.317326930772309e-05,
+      "loss": 0.66,
+      "step": 3422
+    },
+    {
+      "epoch": 4.38144,
+      "grad_norm": 0.6704264879226685,
+      "learning_rate": 4.317126850740296e-05,
+      "loss": 0.6386,
+      "step": 3423
+    },
+    {
+      "epoch": 4.38272,
+      "grad_norm": 0.6729428172111511,
+      "learning_rate": 4.316926770708283e-05,
+      "loss": 0.6369,
+      "step": 3424
+    },
+    {
+      "epoch": 4.384,
+      "grad_norm": 0.6993991732597351,
+      "learning_rate": 4.3167266906762705e-05,
+      "loss": 0.6756,
+      "step": 3425
+    },
+    {
+      "epoch": 4.38528,
+      "grad_norm": 0.7253281474113464,
+      "learning_rate": 4.3165266106442584e-05,
+      "loss": 0.6467,
+      "step": 3426
+    },
+    {
+      "epoch": 4.38656,
+      "grad_norm": 0.7546789646148682,
+      "learning_rate": 4.3163265306122455e-05,
+      "loss": 0.6914,
+      "step": 3427
+    },
+    {
+      "epoch": 4.38784,
+      "grad_norm": 0.7461205720901489,
+      "learning_rate": 4.316126450580232e-05,
+      "loss": 0.6633,
+      "step": 3428
+    },
+    {
+      "epoch": 4.38912,
+      "grad_norm": 0.7659752368927002,
+      "learning_rate": 4.315926370548219e-05,
+      "loss": 0.6514,
+      "step": 3429
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.7351868152618408,
+      "learning_rate": 4.3157262905162064e-05,
+      "loss": 0.7025,
+      "step": 3430
+    },
+    {
+      "epoch": 4.39168,
+      "grad_norm": 0.7607386708259583,
+      "learning_rate": 4.3155262104841936e-05,
+      "loss": 0.6971,
+      "step": 3431
+    },
+    {
+      "epoch": 4.39296,
+      "grad_norm": 0.7106710076332092,
+      "learning_rate": 4.315326130452181e-05,
+      "loss": 0.6642,
+      "step": 3432
+    },
+    {
+      "epoch": 4.39424,
+      "grad_norm": 0.716480553150177,
+      "learning_rate": 4.3151260504201687e-05,
+      "loss": 0.6401,
+      "step": 3433
+    },
+    {
+      "epoch": 4.39552,
+      "grad_norm": 0.6813743710517883,
+      "learning_rate": 4.314925970388156e-05,
+      "loss": 0.6335,
+      "step": 3434
+    },
+    {
+      "epoch": 4.3968,
+      "grad_norm": 0.6941514611244202,
+      "learning_rate": 4.314725890356143e-05,
+      "loss": 0.6294,
+      "step": 3435
+    },
+    {
+      "epoch": 4.39808,
+      "grad_norm": 0.75897616147995,
+      "learning_rate": 4.3145258103241295e-05,
+      "loss": 0.674,
+      "step": 3436
+    },
+    {
+      "epoch": 4.39936,
+      "grad_norm": 0.6969784498214722,
+      "learning_rate": 4.314325730292117e-05,
+      "loss": 0.6424,
+      "step": 3437
+    },
+    {
+      "epoch": 4.40064,
+      "grad_norm": 0.6732988953590393,
+      "learning_rate": 4.314125650260104e-05,
+      "loss": 0.6206,
+      "step": 3438
+    },
+    {
+      "epoch": 4.40192,
+      "grad_norm": 0.6822874546051025,
+      "learning_rate": 4.313925570228091e-05,
+      "loss": 0.5994,
+      "step": 3439
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.6838374733924866,
+      "learning_rate": 4.313725490196079e-05,
+      "loss": 0.6224,
+      "step": 3440
+    },
+    {
+      "epoch": 4.40448,
+      "grad_norm": 0.7581614851951599,
+      "learning_rate": 4.313525410164066e-05,
+      "loss": 0.6899,
+      "step": 3441
+    },
+    {
+      "epoch": 4.40576,
+      "grad_norm": 0.7927963733673096,
+      "learning_rate": 4.313325330132053e-05,
+      "loss": 0.6671,
+      "step": 3442
+    },
+    {
+      "epoch": 4.40704,
+      "grad_norm": 0.7558853030204773,
+      "learning_rate": 4.3131252501000405e-05,
+      "loss": 0.6714,
+      "step": 3443
+    },
+    {
+      "epoch": 4.40832,
+      "grad_norm": 0.6957927346229553,
+      "learning_rate": 4.312925170068027e-05,
+      "loss": 0.6465,
+      "step": 3444
+    },
+    {
+      "epoch": 4.4096,
+      "grad_norm": 0.7612014412879944,
+      "learning_rate": 4.312725090036014e-05,
+      "loss": 0.6855,
+      "step": 3445
+    },
+    {
+      "epoch": 4.41088,
+      "grad_norm": 0.735772430896759,
+      "learning_rate": 4.3125250100040014e-05,
+      "loss": 0.6411,
+      "step": 3446
+    },
+    {
+      "epoch": 4.41216,
+      "grad_norm": 0.6750103235244751,
+      "learning_rate": 4.312324929971989e-05,
+      "loss": 0.5845,
+      "step": 3447
+    },
+    {
+      "epoch": 4.41344,
+      "grad_norm": 0.7323096990585327,
+      "learning_rate": 4.3121248499399764e-05,
+      "loss": 0.6724,
+      "step": 3448
+    },
+    {
+      "epoch": 4.41472,
+      "grad_norm": 0.706743597984314,
+      "learning_rate": 4.3119247699079636e-05,
+      "loss": 0.6313,
+      "step": 3449
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.6990499496459961,
+      "learning_rate": 4.311724689875951e-05,
+      "loss": 0.6688,
+      "step": 3450
+    },
+    {
+      "epoch": 4.41728,
+      "grad_norm": 0.7238355875015259,
+      "learning_rate": 4.311524609843938e-05,
+      "loss": 0.6343,
+      "step": 3451
+    },
+    {
+      "epoch": 4.41856,
+      "grad_norm": 0.6958813667297363,
+      "learning_rate": 4.3113245298119245e-05,
+      "loss": 0.6341,
+      "step": 3452
+    },
+    {
+      "epoch": 4.41984,
+      "grad_norm": 0.7154257893562317,
+      "learning_rate": 4.311124449779912e-05,
+      "loss": 0.6938,
+      "step": 3453
+    },
+    {
+      "epoch": 4.42112,
+      "grad_norm": 0.6367705464363098,
+      "learning_rate": 4.3109243697478996e-05,
+      "loss": 0.5721,
+      "step": 3454
+    },
+    {
+      "epoch": 4.4224,
+      "grad_norm": 0.6861926913261414,
+      "learning_rate": 4.310724289715887e-05,
+      "loss": 0.6292,
+      "step": 3455
+    },
+    {
+      "epoch": 4.42368,
+      "grad_norm": 0.7064390182495117,
+      "learning_rate": 4.310524209683874e-05,
+      "loss": 0.6467,
+      "step": 3456
+    },
+    {
+      "epoch": 4.4249600000000004,
+      "grad_norm": 0.6897706985473633,
+      "learning_rate": 4.310324129651861e-05,
+      "loss": 0.6367,
+      "step": 3457
+    },
+    {
+      "epoch": 4.42624,
+      "grad_norm": 0.7397942543029785,
+      "learning_rate": 4.310124049619848e-05,
+      "loss": 0.7199,
+      "step": 3458
+    },
+    {
+      "epoch": 4.42752,
+      "grad_norm": 0.7196327447891235,
+      "learning_rate": 4.3099239695878355e-05,
+      "loss": 0.6154,
+      "step": 3459
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.7079979777336121,
+      "learning_rate": 4.309723889555822e-05,
+      "loss": 0.6571,
+      "step": 3460
+    },
+    {
+      "epoch": 4.43008,
+      "grad_norm": 0.7943414449691772,
+      "learning_rate": 4.30952380952381e-05,
+      "loss": 0.7078,
+      "step": 3461
+    },
+    {
+      "epoch": 4.43136,
+      "grad_norm": 0.6748960614204407,
+      "learning_rate": 4.309323729491797e-05,
+      "loss": 0.6341,
+      "step": 3462
+    },
+    {
+      "epoch": 4.43264,
+      "grad_norm": 0.7459009885787964,
+      "learning_rate": 4.309123649459784e-05,
+      "loss": 0.72,
+      "step": 3463
+    },
+    {
+      "epoch": 4.43392,
+      "grad_norm": 0.7245545983314514,
+      "learning_rate": 4.3089235694277714e-05,
+      "loss": 0.6007,
+      "step": 3464
+    },
+    {
+      "epoch": 4.4352,
+      "grad_norm": 0.6897006630897522,
+      "learning_rate": 4.3087234893957586e-05,
+      "loss": 0.6911,
+      "step": 3465
+    },
+    {
+      "epoch": 4.4364799999999995,
+      "grad_norm": 0.684449315071106,
+      "learning_rate": 4.308523409363746e-05,
+      "loss": 0.627,
+      "step": 3466
+    },
+    {
+      "epoch": 4.43776,
+      "grad_norm": 0.7077491283416748,
+      "learning_rate": 4.308323329331733e-05,
+      "loss": 0.6295,
+      "step": 3467
+    },
+    {
+      "epoch": 4.43904,
+      "grad_norm": 0.716648280620575,
+      "learning_rate": 4.30812324929972e-05,
+      "loss": 0.601,
+      "step": 3468
+    },
+    {
+      "epoch": 4.44032,
+      "grad_norm": 0.7259345650672913,
+      "learning_rate": 4.3079231692677073e-05,
+      "loss": 0.6443,
+      "step": 3469
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.7068425416946411,
+      "learning_rate": 4.3077230892356945e-05,
+      "loss": 0.6623,
+      "step": 3470
+    },
+    {
+      "epoch": 4.44288,
+      "grad_norm": 0.713869571685791,
+      "learning_rate": 4.307523009203682e-05,
+      "loss": 0.6596,
+      "step": 3471
+    },
+    {
+      "epoch": 4.44416,
+      "grad_norm": 0.6864603757858276,
+      "learning_rate": 4.307322929171669e-05,
+      "loss": 0.6143,
+      "step": 3472
+    },
+    {
+      "epoch": 4.44544,
+      "grad_norm": 0.7068229913711548,
+      "learning_rate": 4.307122849139656e-05,
+      "loss": 0.6255,
+      "step": 3473
+    },
+    {
+      "epoch": 4.44672,
+      "grad_norm": 0.7183898091316223,
+      "learning_rate": 4.306922769107643e-05,
+      "loss": 0.705,
+      "step": 3474
+    },
+    {
+      "epoch": 4.448,
+      "grad_norm": 0.6923708915710449,
+      "learning_rate": 4.3067226890756305e-05,
+      "loss": 0.6048,
+      "step": 3475
+    },
+    {
+      "epoch": 4.44928,
+      "grad_norm": 0.7035982608795166,
+      "learning_rate": 4.3065226090436176e-05,
+      "loss": 0.6131,
+      "step": 3476
+    },
+    {
+      "epoch": 4.45056,
+      "grad_norm": 0.7768038511276245,
+      "learning_rate": 4.306322529011605e-05,
+      "loss": 0.7338,
+      "step": 3477
+    },
+    {
+      "epoch": 4.45184,
+      "grad_norm": 0.7020729780197144,
+      "learning_rate": 4.306122448979592e-05,
+      "loss": 0.6507,
+      "step": 3478
+    },
+    {
+      "epoch": 4.45312,
+      "grad_norm": 0.6677333116531372,
+      "learning_rate": 4.305922368947579e-05,
+      "loss": 0.6101,
+      "step": 3479
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.6806728839874268,
+      "learning_rate": 4.3057222889155664e-05,
+      "loss": 0.6016,
+      "step": 3480
+    },
+    {
+      "epoch": 4.45568,
+      "grad_norm": 0.7218868732452393,
+      "learning_rate": 4.3055222088835536e-05,
+      "loss": 0.6874,
+      "step": 3481
+    },
+    {
+      "epoch": 4.45696,
+      "grad_norm": 0.757403552532196,
+      "learning_rate": 4.305322128851541e-05,
+      "loss": 0.7041,
+      "step": 3482
+    },
+    {
+      "epoch": 4.45824,
+      "grad_norm": 0.725426435470581,
+      "learning_rate": 4.305122048819528e-05,
+      "loss": 0.6705,
+      "step": 3483
+    },
+    {
+      "epoch": 4.45952,
+      "grad_norm": 0.7142929434776306,
+      "learning_rate": 4.304921968787515e-05,
+      "loss": 0.6241,
+      "step": 3484
+    },
+    {
+      "epoch": 4.4608,
+      "grad_norm": 0.7178435325622559,
+      "learning_rate": 4.304721888755502e-05,
+      "loss": 0.6537,
+      "step": 3485
+    },
+    {
+      "epoch": 4.46208,
+      "grad_norm": 0.7174624800682068,
+      "learning_rate": 4.3045218087234895e-05,
+      "loss": 0.6261,
+      "step": 3486
+    },
+    {
+      "epoch": 4.46336,
+      "grad_norm": 0.7432010173797607,
+      "learning_rate": 4.304321728691477e-05,
+      "loss": 0.6189,
+      "step": 3487
+    },
+    {
+      "epoch": 4.46464,
+      "grad_norm": 0.7320737838745117,
+      "learning_rate": 4.304121648659464e-05,
+      "loss": 0.6203,
+      "step": 3488
+    },
+    {
+      "epoch": 4.46592,
+      "grad_norm": 0.7159371972084045,
+      "learning_rate": 4.303921568627452e-05,
+      "loss": 0.6565,
+      "step": 3489
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.6766809225082397,
+      "learning_rate": 4.303721488595438e-05,
+      "loss": 0.6285,
+      "step": 3490
+    },
+    {
+      "epoch": 4.46848,
+      "grad_norm": 0.7048515677452087,
+      "learning_rate": 4.3035214085634254e-05,
+      "loss": 0.6259,
+      "step": 3491
+    },
+    {
+      "epoch": 4.46976,
+      "grad_norm": 0.7184696197509766,
+      "learning_rate": 4.3033213285314126e-05,
+      "loss": 0.6429,
+      "step": 3492
+    },
+    {
+      "epoch": 4.47104,
+      "grad_norm": 0.7331928014755249,
+      "learning_rate": 4.3031212484994e-05,
+      "loss": 0.6189,
+      "step": 3493
+    },
+    {
+      "epoch": 4.47232,
+      "grad_norm": 0.6746116876602173,
+      "learning_rate": 4.302921168467387e-05,
+      "loss": 0.6499,
+      "step": 3494
+    },
+    {
+      "epoch": 4.4736,
+      "grad_norm": 0.6947497725486755,
+      "learning_rate": 4.302721088435374e-05,
+      "loss": 0.6577,
+      "step": 3495
+    },
+    {
+      "epoch": 4.47488,
+      "grad_norm": 0.7183555960655212,
+      "learning_rate": 4.302521008403362e-05,
+      "loss": 0.6201,
+      "step": 3496
+    },
+    {
+      "epoch": 4.47616,
+      "grad_norm": 0.7647238373756409,
+      "learning_rate": 4.302320928371349e-05,
+      "loss": 0.6564,
+      "step": 3497
+    },
+    {
+      "epoch": 4.47744,
+      "grad_norm": 0.6781284809112549,
+      "learning_rate": 4.302120848339336e-05,
+      "loss": 0.6367,
+      "step": 3498
+    },
+    {
+      "epoch": 4.47872,
+      "grad_norm": 0.7044297456741333,
+      "learning_rate": 4.301920768307323e-05,
+      "loss": 0.6508,
+      "step": 3499
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.7338101863861084,
+      "learning_rate": 4.30172068827531e-05,
+      "loss": 0.6807,
+      "step": 3500
+    },
+    {
+      "epoch": 4.48128,
+      "grad_norm": 0.7334948182106018,
+      "learning_rate": 4.301520608243297e-05,
+      "loss": 0.6962,
+      "step": 3501
+    },
+    {
+      "epoch": 4.48256,
+      "grad_norm": 0.7243069410324097,
+      "learning_rate": 4.3013205282112845e-05,
+      "loss": 0.6448,
+      "step": 3502
+    },
+    {
+      "epoch": 4.48384,
+      "grad_norm": 0.7103859782218933,
+      "learning_rate": 4.301120448179272e-05,
+      "loss": 0.6347,
+      "step": 3503
+    },
+    {
+      "epoch": 4.48512,
+      "grad_norm": 0.7307097911834717,
+      "learning_rate": 4.3009203681472595e-05,
+      "loss": 0.6947,
+      "step": 3504
+    },
+    {
+      "epoch": 4.4864,
+      "grad_norm": 0.6719437837600708,
+      "learning_rate": 4.300720288115247e-05,
+      "loss": 0.6101,
+      "step": 3505
+    },
+    {
+      "epoch": 4.48768,
+      "grad_norm": 0.6755344867706299,
+      "learning_rate": 4.300520208083233e-05,
+      "loss": 0.6134,
+      "step": 3506
+    },
+    {
+      "epoch": 4.48896,
+      "grad_norm": 0.7292615175247192,
+      "learning_rate": 4.3003201280512204e-05,
+      "loss": 0.7104,
+      "step": 3507
+    },
+    {
+      "epoch": 4.49024,
+      "grad_norm": 0.7445333003997803,
+      "learning_rate": 4.3001200480192076e-05,
+      "loss": 0.702,
+      "step": 3508
+    },
+    {
+      "epoch": 4.49152,
+      "grad_norm": 0.7050941586494446,
+      "learning_rate": 4.299919967987195e-05,
+      "loss": 0.6471,
+      "step": 3509
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.7550082206726074,
+      "learning_rate": 4.2997198879551826e-05,
+      "loss": 0.6283,
+      "step": 3510
+    },
+    {
+      "epoch": 4.49408,
+      "grad_norm": 0.6943626999855042,
+      "learning_rate": 4.29951980792317e-05,
+      "loss": 0.6155,
+      "step": 3511
+    },
+    {
+      "epoch": 4.49536,
+      "grad_norm": 0.7097110152244568,
+      "learning_rate": 4.299319727891157e-05,
+      "loss": 0.6762,
+      "step": 3512
+    },
+    {
+      "epoch": 4.49664,
+      "grad_norm": 0.7257136106491089,
+      "learning_rate": 4.299119647859144e-05,
+      "loss": 0.6559,
+      "step": 3513
+    },
+    {
+      "epoch": 4.49792,
+      "grad_norm": 0.7039051651954651,
+      "learning_rate": 4.298919567827131e-05,
+      "loss": 0.6605,
+      "step": 3514
+    },
+    {
+      "epoch": 4.4992,
+      "grad_norm": 0.7470154762268066,
+      "learning_rate": 4.298719487795118e-05,
+      "loss": 0.6569,
+      "step": 3515
+    },
+    {
+      "epoch": 4.50048,
+      "grad_norm": 0.7026588916778564,
+      "learning_rate": 4.298519407763105e-05,
+      "loss": 0.6678,
+      "step": 3516
+    },
+    {
+      "epoch": 4.50176,
+      "grad_norm": 0.7094910144805908,
+      "learning_rate": 4.298319327731093e-05,
+      "loss": 0.6832,
+      "step": 3517
+    },
+    {
+      "epoch": 4.50304,
+      "grad_norm": 0.701600968837738,
+      "learning_rate": 4.29811924769908e-05,
+      "loss": 0.6583,
+      "step": 3518
+    },
+    {
+      "epoch": 4.50432,
+      "grad_norm": 0.6822985410690308,
+      "learning_rate": 4.297919167667067e-05,
+      "loss": 0.6335,
+      "step": 3519
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.7292996048927307,
+      "learning_rate": 4.2977190876350545e-05,
+      "loss": 0.6355,
+      "step": 3520
+    },
+    {
+      "epoch": 4.50688,
+      "grad_norm": 0.6827433705329895,
+      "learning_rate": 4.297519007603042e-05,
+      "loss": 0.6288,
+      "step": 3521
+    },
+    {
+      "epoch": 4.50816,
+      "grad_norm": 0.6858897805213928,
+      "learning_rate": 4.297318927571028e-05,
+      "loss": 0.6333,
+      "step": 3522
+    },
+    {
+      "epoch": 4.50944,
+      "grad_norm": 0.7058318853378296,
+      "learning_rate": 4.2971188475390154e-05,
+      "loss": 0.6742,
+      "step": 3523
+    },
+    {
+      "epoch": 4.51072,
+      "grad_norm": 0.7074992060661316,
+      "learning_rate": 4.296918767507003e-05,
+      "loss": 0.6459,
+      "step": 3524
+    },
+    {
+      "epoch": 4.5120000000000005,
+      "grad_norm": 0.7430311441421509,
+      "learning_rate": 4.2967186874749904e-05,
+      "loss": 0.6278,
+      "step": 3525
+    },
+    {
+      "epoch": 4.51328,
+      "grad_norm": 0.6989936828613281,
+      "learning_rate": 4.2965186074429776e-05,
+      "loss": 0.6302,
+      "step": 3526
+    },
+    {
+      "epoch": 4.51456,
+      "grad_norm": 0.7324742674827576,
+      "learning_rate": 4.296318527410965e-05,
+      "loss": 0.6347,
+      "step": 3527
+    },
+    {
+      "epoch": 4.51584,
+      "grad_norm": 0.8047772645950317,
+      "learning_rate": 4.296118447378952e-05,
+      "loss": 0.6524,
+      "step": 3528
+    },
+    {
+      "epoch": 4.51712,
+      "grad_norm": 0.7454036474227905,
+      "learning_rate": 4.295918367346939e-05,
+      "loss": 0.6636,
+      "step": 3529
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.6896642446517944,
+      "learning_rate": 4.295718287314926e-05,
+      "loss": 0.6145,
+      "step": 3530
+    },
+    {
+      "epoch": 4.51968,
+      "grad_norm": 0.7129753828048706,
+      "learning_rate": 4.295518207282913e-05,
+      "loss": 0.6929,
+      "step": 3531
+    },
+    {
+      "epoch": 4.52096,
+      "grad_norm": 0.6888583898544312,
+      "learning_rate": 4.295318127250901e-05,
+      "loss": 0.6411,
+      "step": 3532
+    },
+    {
+      "epoch": 4.52224,
+      "grad_norm": 0.7414783239364624,
+      "learning_rate": 4.295118047218888e-05,
+      "loss": 0.7311,
+      "step": 3533
+    },
+    {
+      "epoch": 4.5235199999999995,
+      "grad_norm": 0.7330256104469299,
+      "learning_rate": 4.294917967186875e-05,
+      "loss": 0.6781,
+      "step": 3534
+    },
+    {
+      "epoch": 4.5248,
+      "grad_norm": 0.6906652450561523,
+      "learning_rate": 4.294717887154862e-05,
+      "loss": 0.6435,
+      "step": 3535
+    },
+    {
+      "epoch": 4.52608,
+      "grad_norm": 0.677453875541687,
+      "learning_rate": 4.2945178071228495e-05,
+      "loss": 0.6618,
+      "step": 3536
+    },
+    {
+      "epoch": 4.52736,
+      "grad_norm": 0.7467184066772461,
+      "learning_rate": 4.2943177270908366e-05,
+      "loss": 0.6991,
+      "step": 3537
+    },
+    {
+      "epoch": 4.52864,
+      "grad_norm": 0.7100308537483215,
+      "learning_rate": 4.294117647058823e-05,
+      "loss": 0.6298,
+      "step": 3538
+    },
+    {
+      "epoch": 4.52992,
+      "grad_norm": 0.6982895135879517,
+      "learning_rate": 4.293917567026811e-05,
+      "loss": 0.6826,
+      "step": 3539
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.6765713095664978,
+      "learning_rate": 4.293717486994798e-05,
+      "loss": 0.6307,
+      "step": 3540
+    },
+    {
+      "epoch": 4.53248,
+      "grad_norm": 0.6806417107582092,
+      "learning_rate": 4.2935174069627854e-05,
+      "loss": 0.6495,
+      "step": 3541
+    },
+    {
+      "epoch": 4.53376,
+      "grad_norm": 0.7208904027938843,
+      "learning_rate": 4.2933173269307726e-05,
+      "loss": 0.6566,
+      "step": 3542
+    },
+    {
+      "epoch": 4.53504,
+      "grad_norm": 0.7589842677116394,
+      "learning_rate": 4.29311724689876e-05,
+      "loss": 0.7152,
+      "step": 3543
+    },
+    {
+      "epoch": 4.53632,
+      "grad_norm": 0.7217769622802734,
+      "learning_rate": 4.292917166866747e-05,
+      "loss": 0.6694,
+      "step": 3544
+    },
+    {
+      "epoch": 4.5376,
+      "grad_norm": 0.681645393371582,
+      "learning_rate": 4.292717086834734e-05,
+      "loss": 0.6182,
+      "step": 3545
+    },
+    {
+      "epoch": 4.53888,
+      "grad_norm": 0.76906818151474,
+      "learning_rate": 4.292517006802721e-05,
+      "loss": 0.6863,
+      "step": 3546
+    },
+    {
+      "epoch": 4.54016,
+      "grad_norm": 0.6864603161811829,
+      "learning_rate": 4.2923169267707085e-05,
+      "loss": 0.5935,
+      "step": 3547
+    },
+    {
+      "epoch": 4.54144,
+      "grad_norm": 0.7130399942398071,
+      "learning_rate": 4.292116846738696e-05,
+      "loss": 0.6145,
+      "step": 3548
+    },
+    {
+      "epoch": 4.54272,
+      "grad_norm": 0.7169683575630188,
+      "learning_rate": 4.291916766706683e-05,
+      "loss": 0.6562,
+      "step": 3549
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.7166752815246582,
+      "learning_rate": 4.29171668667467e-05,
+      "loss": 0.6539,
+      "step": 3550
+    },
+    {
+      "epoch": 4.54528,
+      "grad_norm": 0.681341290473938,
+      "learning_rate": 4.291516606642657e-05,
+      "loss": 0.6518,
+      "step": 3551
+    },
+    {
+      "epoch": 4.54656,
+      "grad_norm": 0.6968530416488647,
+      "learning_rate": 4.2913165266106444e-05,
+      "loss": 0.6345,
+      "step": 3552
+    },
+    {
+      "epoch": 4.54784,
+      "grad_norm": 0.7314958572387695,
+      "learning_rate": 4.2911164465786316e-05,
+      "loss": 0.7076,
+      "step": 3553
+    },
+    {
+      "epoch": 4.54912,
+      "grad_norm": 0.7102290987968445,
+      "learning_rate": 4.290916366546619e-05,
+      "loss": 0.6552,
+      "step": 3554
+    },
+    {
+      "epoch": 4.5504,
+      "grad_norm": 0.7079906463623047,
+      "learning_rate": 4.290716286514606e-05,
+      "loss": 0.6323,
+      "step": 3555
+    },
+    {
+      "epoch": 4.55168,
+      "grad_norm": 0.6791037917137146,
+      "learning_rate": 4.290516206482593e-05,
+      "loss": 0.6335,
+      "step": 3556
+    },
+    {
+      "epoch": 4.55296,
+      "grad_norm": 0.6969685554504395,
+      "learning_rate": 4.2903161264505804e-05,
+      "loss": 0.6266,
+      "step": 3557
+    },
+    {
+      "epoch": 4.55424,
+      "grad_norm": 0.7246193885803223,
+      "learning_rate": 4.2901160464185675e-05,
+      "loss": 0.6518,
+      "step": 3558
+    },
+    {
+      "epoch": 4.55552,
+      "grad_norm": 0.679791271686554,
+      "learning_rate": 4.289915966386555e-05,
+      "loss": 0.6229,
+      "step": 3559
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.7322858572006226,
+      "learning_rate": 4.289715886354542e-05,
+      "loss": 0.6321,
+      "step": 3560
+    },
+    {
+      "epoch": 4.55808,
+      "grad_norm": 0.6904955506324768,
+      "learning_rate": 4.289515806322529e-05,
+      "loss": 0.6772,
+      "step": 3561
+    },
+    {
+      "epoch": 4.55936,
+      "grad_norm": 0.746525764465332,
+      "learning_rate": 4.289315726290516e-05,
+      "loss": 0.6534,
+      "step": 3562
+    },
+    {
+      "epoch": 4.56064,
+      "grad_norm": 0.7122935056686401,
+      "learning_rate": 4.2891156462585035e-05,
+      "loss": 0.6796,
+      "step": 3563
+    },
+    {
+      "epoch": 4.56192,
+      "grad_norm": 0.7172386050224304,
+      "learning_rate": 4.2889155662264907e-05,
+      "loss": 0.6768,
+      "step": 3564
+    },
+    {
+      "epoch": 4.5632,
+      "grad_norm": 0.7053015232086182,
+      "learning_rate": 4.288715486194478e-05,
+      "loss": 0.5911,
+      "step": 3565
+    },
+    {
+      "epoch": 4.56448,
+      "grad_norm": 0.6957263946533203,
+      "learning_rate": 4.288515406162465e-05,
+      "loss": 0.6538,
+      "step": 3566
+    },
+    {
+      "epoch": 4.56576,
+      "grad_norm": 0.6875421404838562,
+      "learning_rate": 4.288315326130453e-05,
+      "loss": 0.6839,
+      "step": 3567
+    },
+    {
+      "epoch": 4.56704,
+      "grad_norm": 0.7036089301109314,
+      "learning_rate": 4.2881152460984394e-05,
+      "loss": 0.6077,
+      "step": 3568
+    },
+    {
+      "epoch": 4.56832,
+      "grad_norm": 0.6996341943740845,
+      "learning_rate": 4.2879151660664266e-05,
+      "loss": 0.6304,
+      "step": 3569
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.7085795402526855,
+      "learning_rate": 4.287715086034414e-05,
+      "loss": 0.6497,
+      "step": 3570
+    },
+    {
+      "epoch": 4.57088,
+      "grad_norm": 0.6893650889396667,
+      "learning_rate": 4.287515006002401e-05,
+      "loss": 0.6503,
+      "step": 3571
+    },
+    {
+      "epoch": 4.57216,
+      "grad_norm": 0.6926917433738708,
+      "learning_rate": 4.287314925970388e-05,
+      "loss": 0.6245,
+      "step": 3572
+    },
+    {
+      "epoch": 4.57344,
+      "grad_norm": 0.7077941298484802,
+      "learning_rate": 4.287114845938375e-05,
+      "loss": 0.6522,
+      "step": 3573
+    },
+    {
+      "epoch": 4.57472,
+      "grad_norm": 0.7533298134803772,
+      "learning_rate": 4.286914765906363e-05,
+      "loss": 0.6637,
+      "step": 3574
+    },
+    {
+      "epoch": 4.576,
+      "grad_norm": 0.7265738844871521,
+      "learning_rate": 4.2867146858743504e-05,
+      "loss": 0.6432,
+      "step": 3575
+    },
+    {
+      "epoch": 4.57728,
+      "grad_norm": 0.7495373487472534,
+      "learning_rate": 4.286514605842337e-05,
+      "loss": 0.6487,
+      "step": 3576
+    },
+    {
+      "epoch": 4.5785599999999995,
+      "grad_norm": 0.7241270542144775,
+      "learning_rate": 4.286314525810324e-05,
+      "loss": 0.6962,
+      "step": 3577
+    },
+    {
+      "epoch": 4.57984,
+      "grad_norm": 0.6941715478897095,
+      "learning_rate": 4.286114445778311e-05,
+      "loss": 0.6421,
+      "step": 3578
+    },
+    {
+      "epoch": 4.58112,
+      "grad_norm": 0.7148897647857666,
+      "learning_rate": 4.2859143657462984e-05,
+      "loss": 0.6584,
+      "step": 3579
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.7130563855171204,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.6817,
+      "step": 3580
+    },
+    {
+      "epoch": 4.58368,
+      "grad_norm": 0.7111728191375732,
+      "learning_rate": 4.2855142056822735e-05,
+      "loss": 0.6443,
+      "step": 3581
+    },
+    {
+      "epoch": 4.58496,
+      "grad_norm": 0.7034164667129517,
+      "learning_rate": 4.285314125650261e-05,
+      "loss": 0.6662,
+      "step": 3582
+    },
+    {
+      "epoch": 4.58624,
+      "grad_norm": 0.6821185946464539,
+      "learning_rate": 4.285114045618248e-05,
+      "loss": 0.6381,
+      "step": 3583
+    },
+    {
+      "epoch": 4.58752,
+      "grad_norm": 0.6892850995063782,
+      "learning_rate": 4.2849139655862344e-05,
+      "loss": 0.6515,
+      "step": 3584
+    },
+    {
+      "epoch": 4.5888,
+      "grad_norm": 0.7385504245758057,
+      "learning_rate": 4.2847138855542216e-05,
+      "loss": 0.6632,
+      "step": 3585
+    },
+    {
+      "epoch": 4.59008,
+      "grad_norm": 0.6548694372177124,
+      "learning_rate": 4.284513805522209e-05,
+      "loss": 0.609,
+      "step": 3586
+    },
+    {
+      "epoch": 4.59136,
+      "grad_norm": 0.7125710248947144,
+      "learning_rate": 4.284313725490196e-05,
+      "loss": 0.6506,
+      "step": 3587
+    },
+    {
+      "epoch": 4.59264,
+      "grad_norm": 0.7181362509727478,
+      "learning_rate": 4.284113645458184e-05,
+      "loss": 0.5887,
+      "step": 3588
+    },
+    {
+      "epoch": 4.59392,
+      "grad_norm": 0.6801038980484009,
+      "learning_rate": 4.283913565426171e-05,
+      "loss": 0.6322,
+      "step": 3589
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.7283415198326111,
+      "learning_rate": 4.283713485394158e-05,
+      "loss": 0.6793,
+      "step": 3590
+    },
+    {
+      "epoch": 4.59648,
+      "grad_norm": 0.7576428055763245,
+      "learning_rate": 4.2835134053621453e-05,
+      "loss": 0.6391,
+      "step": 3591
+    },
+    {
+      "epoch": 4.59776,
+      "grad_norm": 0.7399201393127441,
+      "learning_rate": 4.283313325330132e-05,
+      "loss": 0.6719,
+      "step": 3592
+    },
+    {
+      "epoch": 4.5990400000000005,
+      "grad_norm": 0.7038607001304626,
+      "learning_rate": 4.283113245298119e-05,
+      "loss": 0.6885,
+      "step": 3593
+    },
+    {
+      "epoch": 4.60032,
+      "grad_norm": 0.7106660604476929,
+      "learning_rate": 4.282913165266106e-05,
+      "loss": 0.6382,
+      "step": 3594
+    },
+    {
+      "epoch": 4.6016,
+      "grad_norm": 0.7221371531486511,
+      "learning_rate": 4.282713085234094e-05,
+      "loss": 0.6697,
+      "step": 3595
+    },
+    {
+      "epoch": 4.60288,
+      "grad_norm": 0.702472984790802,
+      "learning_rate": 4.282513005202081e-05,
+      "loss": 0.6555,
+      "step": 3596
+    },
+    {
+      "epoch": 4.60416,
+      "grad_norm": 0.6985082626342773,
+      "learning_rate": 4.2823129251700685e-05,
+      "loss": 0.6591,
+      "step": 3597
+    },
+    {
+      "epoch": 4.60544,
+      "grad_norm": 0.6914473176002502,
+      "learning_rate": 4.2821128451380556e-05,
+      "loss": 0.6284,
+      "step": 3598
+    },
+    {
+      "epoch": 4.60672,
+      "grad_norm": 0.7164996266365051,
+      "learning_rate": 4.281912765106043e-05,
+      "loss": 0.6409,
+      "step": 3599
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.7067193388938904,
+      "learning_rate": 4.2817126850740293e-05,
+      "loss": 0.6891,
+      "step": 3600
+    },
+    {
+      "epoch": 4.60928,
+      "grad_norm": 0.672056257724762,
+      "learning_rate": 4.2815126050420165e-05,
+      "loss": 0.6533,
+      "step": 3601
+    },
+    {
+      "epoch": 4.6105599999999995,
+      "grad_norm": 0.6592856049537659,
+      "learning_rate": 4.2813125250100044e-05,
+      "loss": 0.5726,
+      "step": 3602
+    },
+    {
+      "epoch": 4.61184,
+      "grad_norm": 0.7327611446380615,
+      "learning_rate": 4.2811124449779916e-05,
+      "loss": 0.6566,
+      "step": 3603
+    },
+    {
+      "epoch": 4.61312,
+      "grad_norm": 0.7556269764900208,
+      "learning_rate": 4.280912364945979e-05,
+      "loss": 0.7096,
+      "step": 3604
+    },
+    {
+      "epoch": 4.6144,
+      "grad_norm": 0.7422052025794983,
+      "learning_rate": 4.280712284913966e-05,
+      "loss": 0.7039,
+      "step": 3605
+    },
+    {
+      "epoch": 4.61568,
+      "grad_norm": 0.7220710515975952,
+      "learning_rate": 4.280512204881953e-05,
+      "loss": 0.6709,
+      "step": 3606
+    },
+    {
+      "epoch": 4.61696,
+      "grad_norm": 0.7694050669670105,
+      "learning_rate": 4.28031212484994e-05,
+      "loss": 0.6938,
+      "step": 3607
+    },
+    {
+      "epoch": 4.61824,
+      "grad_norm": 0.7003862261772156,
+      "learning_rate": 4.280112044817927e-05,
+      "loss": 0.6526,
+      "step": 3608
+    },
+    {
+      "epoch": 4.61952,
+      "grad_norm": 0.7154176831245422,
+      "learning_rate": 4.279911964785915e-05,
+      "loss": 0.6106,
+      "step": 3609
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.6721168160438538,
+      "learning_rate": 4.279711884753902e-05,
+      "loss": 0.6439,
+      "step": 3610
+    },
+    {
+      "epoch": 4.62208,
+      "grad_norm": 0.7174552083015442,
+      "learning_rate": 4.279511804721889e-05,
+      "loss": 0.6373,
+      "step": 3611
+    },
+    {
+      "epoch": 4.62336,
+      "grad_norm": 0.7415494918823242,
+      "learning_rate": 4.279311724689876e-05,
+      "loss": 0.6797,
+      "step": 3612
+    },
+    {
+      "epoch": 4.62464,
+      "grad_norm": 0.6769813299179077,
+      "learning_rate": 4.2791116446578634e-05,
+      "loss": 0.5868,
+      "step": 3613
+    },
+    {
+      "epoch": 4.62592,
+      "grad_norm": 0.6615371108055115,
+      "learning_rate": 4.2789115646258506e-05,
+      "loss": 0.5537,
+      "step": 3614
+    },
+    {
+      "epoch": 4.6272,
+      "grad_norm": 0.7274612188339233,
+      "learning_rate": 4.278711484593838e-05,
+      "loss": 0.6443,
+      "step": 3615
+    },
+    {
+      "epoch": 4.62848,
+      "grad_norm": 0.7486844658851624,
+      "learning_rate": 4.278511404561825e-05,
+      "loss": 0.6367,
+      "step": 3616
+    },
+    {
+      "epoch": 4.62976,
+      "grad_norm": 0.7376373410224915,
+      "learning_rate": 4.278311324529812e-05,
+      "loss": 0.6621,
+      "step": 3617
+    },
+    {
+      "epoch": 4.6310400000000005,
+      "grad_norm": 0.6872167587280273,
+      "learning_rate": 4.2781112444977994e-05,
+      "loss": 0.6174,
+      "step": 3618
+    },
+    {
+      "epoch": 4.63232,
+      "grad_norm": 0.7116283774375916,
+      "learning_rate": 4.2779111644657865e-05,
+      "loss": 0.6833,
+      "step": 3619
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.7109349966049194,
+      "learning_rate": 4.277711084433774e-05,
+      "loss": 0.6226,
+      "step": 3620
+    },
+    {
+      "epoch": 4.63488,
+      "grad_norm": 0.6518250703811646,
+      "learning_rate": 4.277511004401761e-05,
+      "loss": 0.6072,
+      "step": 3621
+    },
+    {
+      "epoch": 4.63616,
+      "grad_norm": 0.7209012508392334,
+      "learning_rate": 4.277310924369748e-05,
+      "loss": 0.6737,
+      "step": 3622
+    },
+    {
+      "epoch": 4.63744,
+      "grad_norm": 0.7458158731460571,
+      "learning_rate": 4.277110844337735e-05,
+      "loss": 0.702,
+      "step": 3623
+    },
+    {
+      "epoch": 4.63872,
+      "grad_norm": 0.7005342245101929,
+      "learning_rate": 4.2769107643057225e-05,
+      "loss": 0.6082,
+      "step": 3624
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.7359268665313721,
+      "learning_rate": 4.27671068427371e-05,
+      "loss": 0.6727,
+      "step": 3625
+    },
+    {
+      "epoch": 4.64128,
+      "grad_norm": 0.7191507816314697,
+      "learning_rate": 4.276510604241697e-05,
+      "loss": 0.6743,
+      "step": 3626
+    },
+    {
+      "epoch": 4.64256,
+      "grad_norm": 0.7470155954360962,
+      "learning_rate": 4.276310524209684e-05,
+      "loss": 0.6833,
+      "step": 3627
+    },
+    {
+      "epoch": 4.64384,
+      "grad_norm": 0.7260749936103821,
+      "learning_rate": 4.276110444177671e-05,
+      "loss": 0.6241,
+      "step": 3628
+    },
+    {
+      "epoch": 4.64512,
+      "grad_norm": 0.7276033759117126,
+      "learning_rate": 4.2759103641456584e-05,
+      "loss": 0.6384,
+      "step": 3629
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.7192178964614868,
+      "learning_rate": 4.275710284113646e-05,
+      "loss": 0.6837,
+      "step": 3630
+    },
+    {
+      "epoch": 4.64768,
+      "grad_norm": 0.7400460839271545,
+      "learning_rate": 4.275510204081633e-05,
+      "loss": 0.6588,
+      "step": 3631
+    },
+    {
+      "epoch": 4.64896,
+      "grad_norm": 0.7040890455245972,
+      "learning_rate": 4.27531012404962e-05,
+      "loss": 0.6355,
+      "step": 3632
+    },
+    {
+      "epoch": 4.65024,
+      "grad_norm": 0.7764822840690613,
+      "learning_rate": 4.275110044017607e-05,
+      "loss": 0.6686,
+      "step": 3633
+    },
+    {
+      "epoch": 4.65152,
+      "grad_norm": 0.72344571352005,
+      "learning_rate": 4.274909963985594e-05,
+      "loss": 0.6094,
+      "step": 3634
+    },
+    {
+      "epoch": 4.6528,
+      "grad_norm": 0.7050177454948425,
+      "learning_rate": 4.2747098839535815e-05,
+      "loss": 0.6337,
+      "step": 3635
+    },
+    {
+      "epoch": 4.65408,
+      "grad_norm": 0.7142139673233032,
+      "learning_rate": 4.274509803921569e-05,
+      "loss": 0.6422,
+      "step": 3636
+    },
+    {
+      "epoch": 4.65536,
+      "grad_norm": 0.7412136793136597,
+      "learning_rate": 4.274309723889556e-05,
+      "loss": 0.6324,
+      "step": 3637
+    },
+    {
+      "epoch": 4.65664,
+      "grad_norm": 0.7400050163269043,
+      "learning_rate": 4.274109643857544e-05,
+      "loss": 0.6371,
+      "step": 3638
+    },
+    {
+      "epoch": 4.65792,
+      "grad_norm": 0.6862472295761108,
+      "learning_rate": 4.27390956382553e-05,
+      "loss": 0.6114,
+      "step": 3639
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.6816151142120361,
+      "learning_rate": 4.2737094837935174e-05,
+      "loss": 0.6169,
+      "step": 3640
+    },
+    {
+      "epoch": 4.66048,
+      "grad_norm": 0.7262451648712158,
+      "learning_rate": 4.2735094037615046e-05,
+      "loss": 0.6691,
+      "step": 3641
+    },
+    {
+      "epoch": 4.66176,
+      "grad_norm": 0.6490700244903564,
+      "learning_rate": 4.273309323729492e-05,
+      "loss": 0.5776,
+      "step": 3642
+    },
+    {
+      "epoch": 4.66304,
+      "grad_norm": 0.701505184173584,
+      "learning_rate": 4.273109243697479e-05,
+      "loss": 0.6501,
+      "step": 3643
+    },
+    {
+      "epoch": 4.66432,
+      "grad_norm": 0.731262743473053,
+      "learning_rate": 4.272909163665466e-05,
+      "loss": 0.6893,
+      "step": 3644
+    },
+    {
+      "epoch": 4.6655999999999995,
+      "grad_norm": 0.6694958209991455,
+      "learning_rate": 4.272709083633454e-05,
+      "loss": 0.6047,
+      "step": 3645
+    },
+    {
+      "epoch": 4.66688,
+      "grad_norm": 0.7336912155151367,
+      "learning_rate": 4.272509003601441e-05,
+      "loss": 0.6718,
+      "step": 3646
+    },
+    {
+      "epoch": 4.66816,
+      "grad_norm": 0.7041752338409424,
+      "learning_rate": 4.272308923569428e-05,
+      "loss": 0.6913,
+      "step": 3647
+    },
+    {
+      "epoch": 4.66944,
+      "grad_norm": 0.7393541932106018,
+      "learning_rate": 4.272108843537415e-05,
+      "loss": 0.6946,
+      "step": 3648
+    },
+    {
+      "epoch": 4.67072,
+      "grad_norm": 0.6891679167747498,
+      "learning_rate": 4.271908763505402e-05,
+      "loss": 0.5869,
+      "step": 3649
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.6804254055023193,
+      "learning_rate": 4.271708683473389e-05,
+      "loss": 0.6375,
+      "step": 3650
+    },
+    {
+      "epoch": 4.67328,
+      "grad_norm": 0.7413250803947449,
+      "learning_rate": 4.2715086034413765e-05,
+      "loss": 0.6526,
+      "step": 3651
+    },
+    {
+      "epoch": 4.67456,
+      "grad_norm": 0.7123025059700012,
+      "learning_rate": 4.2713085234093644e-05,
+      "loss": 0.6193,
+      "step": 3652
+    },
+    {
+      "epoch": 4.67584,
+      "grad_norm": 0.6694692373275757,
+      "learning_rate": 4.2711084433773515e-05,
+      "loss": 0.6437,
+      "step": 3653
+    },
+    {
+      "epoch": 4.67712,
+      "grad_norm": 0.744706392288208,
+      "learning_rate": 4.270908363345339e-05,
+      "loss": 0.648,
+      "step": 3654
+    },
+    {
+      "epoch": 4.6784,
+      "grad_norm": 0.7176006436347961,
+      "learning_rate": 4.270708283313325e-05,
+      "loss": 0.6077,
+      "step": 3655
+    },
+    {
+      "epoch": 4.67968,
+      "grad_norm": 0.7272926568984985,
+      "learning_rate": 4.2705082032813124e-05,
+      "loss": 0.6571,
+      "step": 3656
+    },
+    {
+      "epoch": 4.68096,
+      "grad_norm": 0.6795682907104492,
+      "learning_rate": 4.2703081232492996e-05,
+      "loss": 0.6645,
+      "step": 3657
+    },
+    {
+      "epoch": 4.68224,
+      "grad_norm": 0.7655096650123596,
+      "learning_rate": 4.270108043217287e-05,
+      "loss": 0.6812,
+      "step": 3658
+    },
+    {
+      "epoch": 4.68352,
+      "grad_norm": 0.7039996981620789,
+      "learning_rate": 4.2699079631852747e-05,
+      "loss": 0.6863,
+      "step": 3659
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.7228350639343262,
+      "learning_rate": 4.269707883153262e-05,
+      "loss": 0.6519,
+      "step": 3660
+    },
+    {
+      "epoch": 4.6860800000000005,
+      "grad_norm": 0.658035397529602,
+      "learning_rate": 4.269507803121249e-05,
+      "loss": 0.6366,
+      "step": 3661
+    },
+    {
+      "epoch": 4.68736,
+      "grad_norm": 0.7360743880271912,
+      "learning_rate": 4.269307723089236e-05,
+      "loss": 0.6543,
+      "step": 3662
+    },
+    {
+      "epoch": 4.68864,
+      "grad_norm": 0.6831920146942139,
+      "learning_rate": 4.269107643057223e-05,
+      "loss": 0.6334,
+      "step": 3663
+    },
+    {
+      "epoch": 4.68992,
+      "grad_norm": 0.7199382185935974,
+      "learning_rate": 4.26890756302521e-05,
+      "loss": 0.6388,
+      "step": 3664
+    },
+    {
+      "epoch": 4.6912,
+      "grad_norm": 0.6735736131668091,
+      "learning_rate": 4.268707482993197e-05,
+      "loss": 0.6249,
+      "step": 3665
+    },
+    {
+      "epoch": 4.69248,
+      "grad_norm": 0.7085421681404114,
+      "learning_rate": 4.268507402961185e-05,
+      "loss": 0.6752,
+      "step": 3666
+    },
+    {
+      "epoch": 4.69376,
+      "grad_norm": 0.7214930057525635,
+      "learning_rate": 4.268307322929172e-05,
+      "loss": 0.6651,
+      "step": 3667
+    },
+    {
+      "epoch": 4.69504,
+      "grad_norm": 0.7339044213294983,
+      "learning_rate": 4.268107242897159e-05,
+      "loss": 0.7262,
+      "step": 3668
+    },
+    {
+      "epoch": 4.69632,
+      "grad_norm": 0.707331120967865,
+      "learning_rate": 4.2679071628651465e-05,
+      "loss": 0.5875,
+      "step": 3669
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.7132313847541809,
+      "learning_rate": 4.267707082833134e-05,
+      "loss": 0.6412,
+      "step": 3670
+    },
+    {
+      "epoch": 4.69888,
+      "grad_norm": 0.6542635560035706,
+      "learning_rate": 4.26750700280112e-05,
+      "loss": 0.6243,
+      "step": 3671
+    },
+    {
+      "epoch": 4.70016,
+      "grad_norm": 0.6929904818534851,
+      "learning_rate": 4.2673069227691074e-05,
+      "loss": 0.6621,
+      "step": 3672
+    },
+    {
+      "epoch": 4.70144,
+      "grad_norm": 0.6636141538619995,
+      "learning_rate": 4.267106842737095e-05,
+      "loss": 0.6205,
+      "step": 3673
+    },
+    {
+      "epoch": 4.70272,
+      "grad_norm": 0.7498050332069397,
+      "learning_rate": 4.2669067627050824e-05,
+      "loss": 0.6606,
+      "step": 3674
+    },
+    {
+      "epoch": 4.704,
+      "grad_norm": 0.6959406137466431,
+      "learning_rate": 4.2667066826730696e-05,
+      "loss": 0.627,
+      "step": 3675
+    },
+    {
+      "epoch": 4.70528,
+      "grad_norm": 0.7800446152687073,
+      "learning_rate": 4.266506602641057e-05,
+      "loss": 0.6928,
+      "step": 3676
+    },
+    {
+      "epoch": 4.70656,
+      "grad_norm": 0.7218673825263977,
+      "learning_rate": 4.266306522609044e-05,
+      "loss": 0.6479,
+      "step": 3677
+    },
+    {
+      "epoch": 4.70784,
+      "grad_norm": 0.6821399331092834,
+      "learning_rate": 4.266106442577031e-05,
+      "loss": 0.6622,
+      "step": 3678
+    },
+    {
+      "epoch": 4.70912,
+      "grad_norm": 0.6904776692390442,
+      "learning_rate": 4.265906362545018e-05,
+      "loss": 0.6627,
+      "step": 3679
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.6957124471664429,
+      "learning_rate": 4.2657062825130056e-05,
+      "loss": 0.6226,
+      "step": 3680
+    },
+    {
+      "epoch": 4.71168,
+      "grad_norm": 0.7129904627799988,
+      "learning_rate": 4.265506202480993e-05,
+      "loss": 0.6552,
+      "step": 3681
+    },
+    {
+      "epoch": 4.71296,
+      "grad_norm": 0.7037960886955261,
+      "learning_rate": 4.26530612244898e-05,
+      "loss": 0.6661,
+      "step": 3682
+    },
+    {
+      "epoch": 4.71424,
+      "grad_norm": 0.7117835283279419,
+      "learning_rate": 4.265106042416967e-05,
+      "loss": 0.6382,
+      "step": 3683
+    },
+    {
+      "epoch": 4.71552,
+      "grad_norm": 0.7245675325393677,
+      "learning_rate": 4.264905962384954e-05,
+      "loss": 0.625,
+      "step": 3684
+    },
+    {
+      "epoch": 4.7168,
+      "grad_norm": 0.7148410677909851,
+      "learning_rate": 4.2647058823529415e-05,
+      "loss": 0.6358,
+      "step": 3685
+    },
+    {
+      "epoch": 4.7180800000000005,
+      "grad_norm": 0.7175400257110596,
+      "learning_rate": 4.264505802320929e-05,
+      "loss": 0.6145,
+      "step": 3686
+    },
+    {
+      "epoch": 4.71936,
+      "grad_norm": 0.7203708291053772,
+      "learning_rate": 4.264305722288916e-05,
+      "loss": 0.6923,
+      "step": 3687
+    },
+    {
+      "epoch": 4.7206399999999995,
+      "grad_norm": 0.7105974555015564,
+      "learning_rate": 4.264105642256903e-05,
+      "loss": 0.6218,
+      "step": 3688
+    },
+    {
+      "epoch": 4.72192,
+      "grad_norm": 0.7175532579421997,
+      "learning_rate": 4.26390556222489e-05,
+      "loss": 0.6648,
+      "step": 3689
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.7026408910751343,
+      "learning_rate": 4.2637054821928774e-05,
+      "loss": 0.6619,
+      "step": 3690
+    },
+    {
+      "epoch": 4.72448,
+      "grad_norm": 0.7409774661064148,
+      "learning_rate": 4.2635054021608646e-05,
+      "loss": 0.7161,
+      "step": 3691
+    },
+    {
+      "epoch": 4.72576,
+      "grad_norm": 0.6791418194770813,
+      "learning_rate": 4.263305322128852e-05,
+      "loss": 0.6204,
+      "step": 3692
+    },
+    {
+      "epoch": 4.72704,
+      "grad_norm": 0.6994455456733704,
+      "learning_rate": 4.263105242096839e-05,
+      "loss": 0.6185,
+      "step": 3693
+    },
+    {
+      "epoch": 4.72832,
+      "grad_norm": 0.7309840321540833,
+      "learning_rate": 4.262905162064826e-05,
+      "loss": 0.7147,
+      "step": 3694
+    },
+    {
+      "epoch": 4.7296,
+      "grad_norm": 0.6696488261222839,
+      "learning_rate": 4.262705082032813e-05,
+      "loss": 0.5993,
+      "step": 3695
+    },
+    {
+      "epoch": 4.73088,
+      "grad_norm": 0.7723877429962158,
+      "learning_rate": 4.2625050020008005e-05,
+      "loss": 0.6397,
+      "step": 3696
+    },
+    {
+      "epoch": 4.73216,
+      "grad_norm": 0.6929299831390381,
+      "learning_rate": 4.262304921968788e-05,
+      "loss": 0.6304,
+      "step": 3697
+    },
+    {
+      "epoch": 4.73344,
+      "grad_norm": 0.6986497640609741,
+      "learning_rate": 4.262104841936775e-05,
+      "loss": 0.6173,
+      "step": 3698
+    },
+    {
+      "epoch": 4.73472,
+      "grad_norm": 0.7609420418739319,
+      "learning_rate": 4.261904761904762e-05,
+      "loss": 0.6912,
+      "step": 3699
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.7289432287216187,
+      "learning_rate": 4.261704681872749e-05,
+      "loss": 0.6354,
+      "step": 3700
+    },
+    {
+      "epoch": 4.73728,
+      "grad_norm": 0.7391143441200256,
+      "learning_rate": 4.2615046018407365e-05,
+      "loss": 0.6755,
+      "step": 3701
+    },
+    {
+      "epoch": 4.73856,
+      "grad_norm": 0.7171617746353149,
+      "learning_rate": 4.2613045218087236e-05,
+      "loss": 0.6639,
+      "step": 3702
+    },
+    {
+      "epoch": 4.73984,
+      "grad_norm": 0.7035807371139526,
+      "learning_rate": 4.261104441776711e-05,
+      "loss": 0.6142,
+      "step": 3703
+    },
+    {
+      "epoch": 4.7411200000000004,
+      "grad_norm": 0.6329298615455627,
+      "learning_rate": 4.260904361744698e-05,
+      "loss": 0.6249,
+      "step": 3704
+    },
+    {
+      "epoch": 4.7424,
+      "grad_norm": 0.6702702045440674,
+      "learning_rate": 4.260704281712685e-05,
+      "loss": 0.5998,
+      "step": 3705
+    },
+    {
+      "epoch": 4.74368,
+      "grad_norm": 0.7044887542724609,
+      "learning_rate": 4.2605042016806724e-05,
+      "loss": 0.6308,
+      "step": 3706
+    },
+    {
+      "epoch": 4.74496,
+      "grad_norm": 0.6639620065689087,
+      "learning_rate": 4.2603041216486596e-05,
+      "loss": 0.6565,
+      "step": 3707
+    },
+    {
+      "epoch": 4.74624,
+      "grad_norm": 0.72665935754776,
+      "learning_rate": 4.2601040416166474e-05,
+      "loss": 0.6819,
+      "step": 3708
+    },
+    {
+      "epoch": 4.74752,
+      "grad_norm": 0.7186412811279297,
+      "learning_rate": 4.259903961584634e-05,
+      "loss": 0.6577,
+      "step": 3709
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.6942617893218994,
+      "learning_rate": 4.259703881552621e-05,
+      "loss": 0.6603,
+      "step": 3710
+    },
+    {
+      "epoch": 4.75008,
+      "grad_norm": 0.6832451224327087,
+      "learning_rate": 4.259503801520608e-05,
+      "loss": 0.6013,
+      "step": 3711
+    },
+    {
+      "epoch": 4.75136,
+      "grad_norm": 0.765796959400177,
+      "learning_rate": 4.2593037214885955e-05,
+      "loss": 0.6965,
+      "step": 3712
+    },
+    {
+      "epoch": 4.7526399999999995,
+      "grad_norm": 0.7410770654678345,
+      "learning_rate": 4.259103641456583e-05,
+      "loss": 0.6428,
+      "step": 3713
+    },
+    {
+      "epoch": 4.75392,
+      "grad_norm": 0.7163513898849487,
+      "learning_rate": 4.25890356142457e-05,
+      "loss": 0.6708,
+      "step": 3714
+    },
+    {
+      "epoch": 4.7552,
+      "grad_norm": 0.7363733053207397,
+      "learning_rate": 4.258703481392558e-05,
+      "loss": 0.6724,
+      "step": 3715
+    },
+    {
+      "epoch": 4.75648,
+      "grad_norm": 0.6866661310195923,
+      "learning_rate": 4.258503401360545e-05,
+      "loss": 0.6056,
+      "step": 3716
+    },
+    {
+      "epoch": 4.75776,
+      "grad_norm": 0.723304271697998,
+      "learning_rate": 4.2583033213285314e-05,
+      "loss": 0.6742,
+      "step": 3717
+    },
+    {
+      "epoch": 4.75904,
+      "grad_norm": 0.730629026889801,
+      "learning_rate": 4.2581032412965186e-05,
+      "loss": 0.6486,
+      "step": 3718
+    },
+    {
+      "epoch": 4.76032,
+      "grad_norm": 0.6811598539352417,
+      "learning_rate": 4.257903161264506e-05,
+      "loss": 0.6246,
+      "step": 3719
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.7193059921264648,
+      "learning_rate": 4.257703081232493e-05,
+      "loss": 0.6695,
+      "step": 3720
+    },
+    {
+      "epoch": 4.76288,
+      "grad_norm": 0.7182266116142273,
+      "learning_rate": 4.25750300120048e-05,
+      "loss": 0.6379,
+      "step": 3721
+    },
+    {
+      "epoch": 4.76416,
+      "grad_norm": 0.6965410113334656,
+      "learning_rate": 4.257302921168468e-05,
+      "loss": 0.6782,
+      "step": 3722
+    },
+    {
+      "epoch": 4.76544,
+      "grad_norm": 0.7410017251968384,
+      "learning_rate": 4.257102841136455e-05,
+      "loss": 0.6226,
+      "step": 3723
+    },
+    {
+      "epoch": 4.76672,
+      "grad_norm": 0.6835784316062927,
+      "learning_rate": 4.2569027611044424e-05,
+      "loss": 0.6244,
+      "step": 3724
+    },
+    {
+      "epoch": 4.768,
+      "grad_norm": 0.7124270796775818,
+      "learning_rate": 4.256702681072429e-05,
+      "loss": 0.6385,
+      "step": 3725
+    },
+    {
+      "epoch": 4.76928,
+      "grad_norm": 0.6720713376998901,
+      "learning_rate": 4.256502601040416e-05,
+      "loss": 0.6582,
+      "step": 3726
+    },
+    {
+      "epoch": 4.77056,
+      "grad_norm": 0.690466046333313,
+      "learning_rate": 4.256302521008403e-05,
+      "loss": 0.6574,
+      "step": 3727
+    },
+    {
+      "epoch": 4.77184,
+      "grad_norm": 0.7075904011726379,
+      "learning_rate": 4.2561024409763905e-05,
+      "loss": 0.6623,
+      "step": 3728
+    },
+    {
+      "epoch": 4.7731200000000005,
+      "grad_norm": 0.7090795636177063,
+      "learning_rate": 4.255902360944378e-05,
+      "loss": 0.6237,
+      "step": 3729
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.6967689394950867,
+      "learning_rate": 4.2557022809123655e-05,
+      "loss": 0.6401,
+      "step": 3730
+    },
+    {
+      "epoch": 4.77568,
+      "grad_norm": 0.7227605581283569,
+      "learning_rate": 4.255502200880353e-05,
+      "loss": 0.6841,
+      "step": 3731
+    },
+    {
+      "epoch": 4.77696,
+      "grad_norm": 0.7367991209030151,
+      "learning_rate": 4.25530212084834e-05,
+      "loss": 0.6267,
+      "step": 3732
+    },
+    {
+      "epoch": 4.77824,
+      "grad_norm": 0.655534565448761,
+      "learning_rate": 4.2551020408163264e-05,
+      "loss": 0.5815,
+      "step": 3733
+    },
+    {
+      "epoch": 4.77952,
+      "grad_norm": 0.7497329115867615,
+      "learning_rate": 4.2549019607843136e-05,
+      "loss": 0.6308,
+      "step": 3734
+    },
+    {
+      "epoch": 4.7808,
+      "grad_norm": 0.7356445789337158,
+      "learning_rate": 4.254701880752301e-05,
+      "loss": 0.6944,
+      "step": 3735
+    },
+    {
+      "epoch": 4.78208,
+      "grad_norm": 0.7261192202568054,
+      "learning_rate": 4.2545018007202886e-05,
+      "loss": 0.6922,
+      "step": 3736
+    },
+    {
+      "epoch": 4.78336,
+      "grad_norm": 0.727255642414093,
+      "learning_rate": 4.254301720688276e-05,
+      "loss": 0.6582,
+      "step": 3737
+    },
+    {
+      "epoch": 4.78464,
+      "grad_norm": 0.7310218811035156,
+      "learning_rate": 4.254101640656263e-05,
+      "loss": 0.6671,
+      "step": 3738
+    },
+    {
+      "epoch": 4.78592,
+      "grad_norm": 0.7008660435676575,
+      "learning_rate": 4.25390156062425e-05,
+      "loss": 0.6518,
+      "step": 3739
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.6646680235862732,
+      "learning_rate": 4.2537014805922374e-05,
+      "loss": 0.6003,
+      "step": 3740
+    },
+    {
+      "epoch": 4.78848,
+      "grad_norm": 0.6931464076042175,
+      "learning_rate": 4.253501400560224e-05,
+      "loss": 0.624,
+      "step": 3741
+    },
+    {
+      "epoch": 4.78976,
+      "grad_norm": 0.7339210510253906,
+      "learning_rate": 4.253301320528211e-05,
+      "loss": 0.6385,
+      "step": 3742
+    },
+    {
+      "epoch": 4.79104,
+      "grad_norm": 0.7324719429016113,
+      "learning_rate": 4.253101240496199e-05,
+      "loss": 0.6915,
+      "step": 3743
+    },
+    {
+      "epoch": 4.79232,
+      "grad_norm": 0.7005221843719482,
+      "learning_rate": 4.252901160464186e-05,
+      "loss": 0.6452,
+      "step": 3744
+    },
+    {
+      "epoch": 4.7936,
+      "grad_norm": 0.7293890118598938,
+      "learning_rate": 4.252701080432173e-05,
+      "loss": 0.6334,
+      "step": 3745
+    },
+    {
+      "epoch": 4.79488,
+      "grad_norm": 0.709510862827301,
+      "learning_rate": 4.2525010004001605e-05,
+      "loss": 0.6718,
+      "step": 3746
+    },
+    {
+      "epoch": 4.79616,
+      "grad_norm": 0.7241347432136536,
+      "learning_rate": 4.252300920368148e-05,
+      "loss": 0.6326,
+      "step": 3747
+    },
+    {
+      "epoch": 4.79744,
+      "grad_norm": 0.6769399642944336,
+      "learning_rate": 4.252100840336135e-05,
+      "loss": 0.6442,
+      "step": 3748
+    },
+    {
+      "epoch": 4.79872,
+      "grad_norm": 0.7203426361083984,
+      "learning_rate": 4.2519007603041214e-05,
+      "loss": 0.644,
+      "step": 3749
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.734929084777832,
+      "learning_rate": 4.2517006802721085e-05,
+      "loss": 0.6644,
+      "step": 3750
+    },
+    {
+      "epoch": 4.80128,
+      "grad_norm": 0.7082551717758179,
+      "learning_rate": 4.2515006002400964e-05,
+      "loss": 0.6084,
+      "step": 3751
+    },
+    {
+      "epoch": 4.80256,
+      "grad_norm": 0.6955980062484741,
+      "learning_rate": 4.2513005202080836e-05,
+      "loss": 0.6303,
+      "step": 3752
+    },
+    {
+      "epoch": 4.80384,
+      "grad_norm": 0.6883904933929443,
+      "learning_rate": 4.251100440176071e-05,
+      "loss": 0.6464,
+      "step": 3753
+    },
+    {
+      "epoch": 4.80512,
+      "grad_norm": 0.6787061095237732,
+      "learning_rate": 4.250900360144058e-05,
+      "loss": 0.6351,
+      "step": 3754
+    },
+    {
+      "epoch": 4.8064,
+      "grad_norm": 0.6913610696792603,
+      "learning_rate": 4.250700280112045e-05,
+      "loss": 0.5745,
+      "step": 3755
+    },
+    {
+      "epoch": 4.8076799999999995,
+      "grad_norm": 0.7253073453903198,
+      "learning_rate": 4.2505002000800323e-05,
+      "loss": 0.6662,
+      "step": 3756
+    },
+    {
+      "epoch": 4.80896,
+      "grad_norm": 0.7172198295593262,
+      "learning_rate": 4.250300120048019e-05,
+      "loss": 0.6726,
+      "step": 3757
+    },
+    {
+      "epoch": 4.81024,
+      "grad_norm": 0.7208426594734192,
+      "learning_rate": 4.250100040016007e-05,
+      "loss": 0.6478,
+      "step": 3758
+    },
+    {
+      "epoch": 4.81152,
+      "grad_norm": 0.6597497463226318,
+      "learning_rate": 4.249899959983994e-05,
+      "loss": 0.6372,
+      "step": 3759
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.7280398011207581,
+      "learning_rate": 4.249699879951981e-05,
+      "loss": 0.6449,
+      "step": 3760
+    },
+    {
+      "epoch": 4.81408,
+      "grad_norm": 0.7056030035018921,
+      "learning_rate": 4.249499799919968e-05,
+      "loss": 0.6638,
+      "step": 3761
+    },
+    {
+      "epoch": 4.81536,
+      "grad_norm": 0.7291399836540222,
+      "learning_rate": 4.2492997198879555e-05,
+      "loss": 0.6717,
+      "step": 3762
+    },
+    {
+      "epoch": 4.81664,
+      "grad_norm": 0.7240449786186218,
+      "learning_rate": 4.2490996398559426e-05,
+      "loss": 0.6896,
+      "step": 3763
+    },
+    {
+      "epoch": 4.81792,
+      "grad_norm": 0.6927871704101562,
+      "learning_rate": 4.24889955982393e-05,
+      "loss": 0.6627,
+      "step": 3764
+    },
+    {
+      "epoch": 4.8192,
+      "grad_norm": 0.6421213150024414,
+      "learning_rate": 4.248699479791917e-05,
+      "loss": 0.6177,
+      "step": 3765
+    },
+    {
+      "epoch": 4.82048,
+      "grad_norm": 0.6733725666999817,
+      "learning_rate": 4.248499399759904e-05,
+      "loss": 0.5987,
+      "step": 3766
+    },
+    {
+      "epoch": 4.82176,
+      "grad_norm": 0.7016422152519226,
+      "learning_rate": 4.2482993197278914e-05,
+      "loss": 0.6563,
+      "step": 3767
+    },
+    {
+      "epoch": 4.82304,
+      "grad_norm": 0.662580132484436,
+      "learning_rate": 4.2480992396958786e-05,
+      "loss": 0.6167,
+      "step": 3768
+    },
+    {
+      "epoch": 4.82432,
+      "grad_norm": 0.6980850696563721,
+      "learning_rate": 4.247899159663866e-05,
+      "loss": 0.5772,
+      "step": 3769
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.6765894293785095,
+      "learning_rate": 4.247699079631853e-05,
+      "loss": 0.6503,
+      "step": 3770
+    },
+    {
+      "epoch": 4.82688,
+      "grad_norm": 0.739733099937439,
+      "learning_rate": 4.24749899959984e-05,
+      "loss": 0.7014,
+      "step": 3771
+    },
+    {
+      "epoch": 4.8281600000000005,
+      "grad_norm": 0.706731915473938,
+      "learning_rate": 4.247298919567827e-05,
+      "loss": 0.6687,
+      "step": 3772
+    },
+    {
+      "epoch": 4.82944,
+      "grad_norm": 0.7055926322937012,
+      "learning_rate": 4.2470988395358145e-05,
+      "loss": 0.6354,
+      "step": 3773
+    },
+    {
+      "epoch": 4.83072,
+      "grad_norm": 0.7980726361274719,
+      "learning_rate": 4.246898759503802e-05,
+      "loss": 0.7241,
+      "step": 3774
+    },
+    {
+      "epoch": 4.832,
+      "grad_norm": 0.6939478516578674,
+      "learning_rate": 4.246698679471789e-05,
+      "loss": 0.621,
+      "step": 3775
+    },
+    {
+      "epoch": 4.83328,
+      "grad_norm": 0.6984978318214417,
+      "learning_rate": 4.246498599439776e-05,
+      "loss": 0.591,
+      "step": 3776
+    },
+    {
+      "epoch": 4.83456,
+      "grad_norm": 0.7327542901039124,
+      "learning_rate": 4.246298519407763e-05,
+      "loss": 0.6889,
+      "step": 3777
+    },
+    {
+      "epoch": 4.83584,
+      "grad_norm": 0.7143282890319824,
+      "learning_rate": 4.2460984393757504e-05,
+      "loss": 0.6558,
+      "step": 3778
+    },
+    {
+      "epoch": 4.83712,
+      "grad_norm": 0.7296421527862549,
+      "learning_rate": 4.2458983593437376e-05,
+      "loss": 0.6599,
+      "step": 3779
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.7196743488311768,
+      "learning_rate": 4.245698279311725e-05,
+      "loss": 0.6421,
+      "step": 3780
+    },
+    {
+      "epoch": 4.8396799999999995,
+      "grad_norm": 0.6790671944618225,
+      "learning_rate": 4.245498199279712e-05,
+      "loss": 0.6075,
+      "step": 3781
+    },
+    {
+      "epoch": 4.84096,
+      "grad_norm": 0.7062405347824097,
+      "learning_rate": 4.245298119247699e-05,
+      "loss": 0.6633,
+      "step": 3782
+    },
+    {
+      "epoch": 4.84224,
+      "grad_norm": 0.766210675239563,
+      "learning_rate": 4.2450980392156864e-05,
+      "loss": 0.6746,
+      "step": 3783
+    },
+    {
+      "epoch": 4.84352,
+      "grad_norm": 0.6822754740715027,
+      "learning_rate": 4.2448979591836735e-05,
+      "loss": 0.6352,
+      "step": 3784
+    },
+    {
+      "epoch": 4.8448,
+      "grad_norm": 0.6726470589637756,
+      "learning_rate": 4.244697879151661e-05,
+      "loss": 0.6153,
+      "step": 3785
+    },
+    {
+      "epoch": 4.84608,
+      "grad_norm": 0.7688694596290588,
+      "learning_rate": 4.2444977991196486e-05,
+      "loss": 0.7037,
+      "step": 3786
+    },
+    {
+      "epoch": 4.84736,
+      "grad_norm": 0.6525343060493469,
+      "learning_rate": 4.244297719087635e-05,
+      "loss": 0.6385,
+      "step": 3787
+    },
+    {
+      "epoch": 4.84864,
+      "grad_norm": 0.7031245827674866,
+      "learning_rate": 4.244097639055622e-05,
+      "loss": 0.6051,
+      "step": 3788
+    },
+    {
+      "epoch": 4.84992,
+      "grad_norm": 0.7337315082550049,
+      "learning_rate": 4.2438975590236095e-05,
+      "loss": 0.6638,
+      "step": 3789
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.7109333276748657,
+      "learning_rate": 4.2436974789915967e-05,
+      "loss": 0.6656,
+      "step": 3790
+    },
+    {
+      "epoch": 4.85248,
+      "grad_norm": 0.7319778800010681,
+      "learning_rate": 4.243497398959584e-05,
+      "loss": 0.6261,
+      "step": 3791
+    },
+    {
+      "epoch": 4.85376,
+      "grad_norm": 0.7297192811965942,
+      "learning_rate": 4.243297318927571e-05,
+      "loss": 0.703,
+      "step": 3792
+    },
+    {
+      "epoch": 4.85504,
+      "grad_norm": 0.7348429560661316,
+      "learning_rate": 4.243097238895559e-05,
+      "loss": 0.6583,
+      "step": 3793
+    },
+    {
+      "epoch": 4.85632,
+      "grad_norm": 0.7114925384521484,
+      "learning_rate": 4.242897158863546e-05,
+      "loss": 0.6621,
+      "step": 3794
+    },
+    {
+      "epoch": 4.8576,
+      "grad_norm": 0.679014265537262,
+      "learning_rate": 4.2426970788315326e-05,
+      "loss": 0.615,
+      "step": 3795
+    },
+    {
+      "epoch": 4.85888,
+      "grad_norm": 0.7080205678939819,
+      "learning_rate": 4.24249699879952e-05,
+      "loss": 0.653,
+      "step": 3796
+    },
+    {
+      "epoch": 4.8601600000000005,
+      "grad_norm": 0.7360637187957764,
+      "learning_rate": 4.242296918767507e-05,
+      "loss": 0.6624,
+      "step": 3797
+    },
+    {
+      "epoch": 4.86144,
+      "grad_norm": 0.7157953977584839,
+      "learning_rate": 4.242096838735494e-05,
+      "loss": 0.6692,
+      "step": 3798
+    },
+    {
+      "epoch": 4.86272,
+      "grad_norm": 0.7022566199302673,
+      "learning_rate": 4.241896758703481e-05,
+      "loss": 0.6586,
+      "step": 3799
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.7275310158729553,
+      "learning_rate": 4.241696678671469e-05,
+      "loss": 0.6602,
+      "step": 3800
+    },
+    {
+      "epoch": 4.86528,
+      "grad_norm": 0.7159553170204163,
+      "learning_rate": 4.2414965986394564e-05,
+      "loss": 0.6299,
+      "step": 3801
+    },
+    {
+      "epoch": 4.86656,
+      "grad_norm": 0.7265822887420654,
+      "learning_rate": 4.2412965186074436e-05,
+      "loss": 0.5922,
+      "step": 3802
+    },
+    {
+      "epoch": 4.86784,
+      "grad_norm": 0.6696851253509521,
+      "learning_rate": 4.24109643857543e-05,
+      "loss": 0.6128,
+      "step": 3803
+    },
+    {
+      "epoch": 4.86912,
+      "grad_norm": 0.7184013724327087,
+      "learning_rate": 4.240896358543417e-05,
+      "loss": 0.6601,
+      "step": 3804
+    },
+    {
+      "epoch": 4.8704,
+      "grad_norm": 0.7109546661376953,
+      "learning_rate": 4.2406962785114044e-05,
+      "loss": 0.6546,
+      "step": 3805
+    },
+    {
+      "epoch": 4.87168,
+      "grad_norm": 0.7122465372085571,
+      "learning_rate": 4.2404961984793916e-05,
+      "loss": 0.6324,
+      "step": 3806
+    },
+    {
+      "epoch": 4.87296,
+      "grad_norm": 0.7610680460929871,
+      "learning_rate": 4.2402961184473795e-05,
+      "loss": 0.6355,
+      "step": 3807
+    },
+    {
+      "epoch": 4.87424,
+      "grad_norm": 0.7229766845703125,
+      "learning_rate": 4.240096038415367e-05,
+      "loss": 0.6396,
+      "step": 3808
+    },
+    {
+      "epoch": 4.87552,
+      "grad_norm": 0.7144048810005188,
+      "learning_rate": 4.239895958383354e-05,
+      "loss": 0.6485,
+      "step": 3809
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.7222286462783813,
+      "learning_rate": 4.239695878351341e-05,
+      "loss": 0.6919,
+      "step": 3810
+    },
+    {
+      "epoch": 4.87808,
+      "grad_norm": 0.700534462928772,
+      "learning_rate": 4.2394957983193276e-05,
+      "loss": 0.6468,
+      "step": 3811
+    },
+    {
+      "epoch": 4.87936,
+      "grad_norm": 0.696043074131012,
+      "learning_rate": 4.239295718287315e-05,
+      "loss": 0.589,
+      "step": 3812
+    },
+    {
+      "epoch": 4.88064,
+      "grad_norm": 0.6939215064048767,
+      "learning_rate": 4.239095638255302e-05,
+      "loss": 0.6358,
+      "step": 3813
+    },
+    {
+      "epoch": 4.88192,
+      "grad_norm": 0.6854797601699829,
+      "learning_rate": 4.23889555822329e-05,
+      "loss": 0.6574,
+      "step": 3814
+    },
+    {
+      "epoch": 4.8832,
+      "grad_norm": 0.7024257183074951,
+      "learning_rate": 4.238695478191277e-05,
+      "loss": 0.6688,
+      "step": 3815
+    },
+    {
+      "epoch": 4.88448,
+      "grad_norm": 0.7299116849899292,
+      "learning_rate": 4.238495398159264e-05,
+      "loss": 0.6921,
+      "step": 3816
+    },
+    {
+      "epoch": 4.88576,
+      "grad_norm": 0.7275465130805969,
+      "learning_rate": 4.2382953181272513e-05,
+      "loss": 0.6454,
+      "step": 3817
+    },
+    {
+      "epoch": 4.88704,
+      "grad_norm": 0.7085564136505127,
+      "learning_rate": 4.2380952380952385e-05,
+      "loss": 0.6523,
+      "step": 3818
+    },
+    {
+      "epoch": 4.88832,
+      "grad_norm": 0.7188389897346497,
+      "learning_rate": 4.237895158063225e-05,
+      "loss": 0.6946,
+      "step": 3819
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.72255939245224,
+      "learning_rate": 4.237695078031212e-05,
+      "loss": 0.6166,
+      "step": 3820
+    },
+    {
+      "epoch": 4.89088,
+      "grad_norm": 0.6671111583709717,
+      "learning_rate": 4.2374949979992e-05,
+      "loss": 0.6173,
+      "step": 3821
+    },
+    {
+      "epoch": 4.89216,
+      "grad_norm": 0.7809231877326965,
+      "learning_rate": 4.237294917967187e-05,
+      "loss": 0.7262,
+      "step": 3822
+    },
+    {
+      "epoch": 4.89344,
+      "grad_norm": 0.6766269207000732,
+      "learning_rate": 4.2370948379351745e-05,
+      "loss": 0.6216,
+      "step": 3823
+    },
+    {
+      "epoch": 4.8947199999999995,
+      "grad_norm": 0.7346859574317932,
+      "learning_rate": 4.2368947579031616e-05,
+      "loss": 0.6529,
+      "step": 3824
+    },
+    {
+      "epoch": 4.896,
+      "grad_norm": 0.7254279851913452,
+      "learning_rate": 4.236694677871149e-05,
+      "loss": 0.6319,
+      "step": 3825
+    },
+    {
+      "epoch": 4.89728,
+      "grad_norm": 0.7144622802734375,
+      "learning_rate": 4.236494597839136e-05,
+      "loss": 0.6586,
+      "step": 3826
+    },
+    {
+      "epoch": 4.89856,
+      "grad_norm": 0.7005541324615479,
+      "learning_rate": 4.2362945178071225e-05,
+      "loss": 0.67,
+      "step": 3827
+    },
+    {
+      "epoch": 4.89984,
+      "grad_norm": 0.7724364995956421,
+      "learning_rate": 4.2360944377751104e-05,
+      "loss": 0.663,
+      "step": 3828
+    },
+    {
+      "epoch": 4.90112,
+      "grad_norm": 0.7261497974395752,
+      "learning_rate": 4.2358943577430976e-05,
+      "loss": 0.6781,
+      "step": 3829
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.673323929309845,
+      "learning_rate": 4.235694277711085e-05,
+      "loss": 0.6236,
+      "step": 3830
+    },
+    {
+      "epoch": 4.90368,
+      "grad_norm": 0.6874783635139465,
+      "learning_rate": 4.235494197679072e-05,
+      "loss": 0.644,
+      "step": 3831
+    },
+    {
+      "epoch": 4.90496,
+      "grad_norm": 0.7447608709335327,
+      "learning_rate": 4.235294117647059e-05,
+      "loss": 0.7078,
+      "step": 3832
+    },
+    {
+      "epoch": 4.90624,
+      "grad_norm": 0.7306890487670898,
+      "learning_rate": 4.235094037615046e-05,
+      "loss": 0.649,
+      "step": 3833
+    },
+    {
+      "epoch": 4.90752,
+      "grad_norm": 0.7219198942184448,
+      "learning_rate": 4.2348939575830335e-05,
+      "loss": 0.6817,
+      "step": 3834
+    },
+    {
+      "epoch": 4.9088,
+      "grad_norm": 0.7355913519859314,
+      "learning_rate": 4.234693877551021e-05,
+      "loss": 0.6291,
+      "step": 3835
+    },
+    {
+      "epoch": 4.91008,
+      "grad_norm": 0.7445414662361145,
+      "learning_rate": 4.234493797519008e-05,
+      "loss": 0.6903,
+      "step": 3836
+    },
+    {
+      "epoch": 4.91136,
+      "grad_norm": 0.6713249087333679,
+      "learning_rate": 4.234293717486995e-05,
+      "loss": 0.6459,
+      "step": 3837
+    },
+    {
+      "epoch": 4.91264,
+      "grad_norm": 0.7056488990783691,
+      "learning_rate": 4.234093637454982e-05,
+      "loss": 0.6648,
+      "step": 3838
+    },
+    {
+      "epoch": 4.91392,
+      "grad_norm": 0.6989180445671082,
+      "learning_rate": 4.2338935574229694e-05,
+      "loss": 0.705,
+      "step": 3839
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.753159761428833,
+      "learning_rate": 4.2336934773909566e-05,
+      "loss": 0.6622,
+      "step": 3840
+    },
+    {
+      "epoch": 4.91648,
+      "grad_norm": 0.7392045855522156,
+      "learning_rate": 4.233493397358944e-05,
+      "loss": 0.6205,
+      "step": 3841
+    },
+    {
+      "epoch": 4.91776,
+      "grad_norm": 0.7536989450454712,
+      "learning_rate": 4.233293317326931e-05,
+      "loss": 0.6748,
+      "step": 3842
+    },
+    {
+      "epoch": 4.91904,
+      "grad_norm": 0.7107492089271545,
+      "learning_rate": 4.233093237294918e-05,
+      "loss": 0.6372,
+      "step": 3843
+    },
+    {
+      "epoch": 4.92032,
+      "grad_norm": 0.6968305110931396,
+      "learning_rate": 4.2328931572629054e-05,
+      "loss": 0.6463,
+      "step": 3844
+    },
+    {
+      "epoch": 4.9216,
+      "grad_norm": 0.7517611980438232,
+      "learning_rate": 4.2326930772308925e-05,
+      "loss": 0.6505,
+      "step": 3845
+    },
+    {
+      "epoch": 4.92288,
+      "grad_norm": 0.7284044623374939,
+      "learning_rate": 4.23249299719888e-05,
+      "loss": 0.6767,
+      "step": 3846
+    },
+    {
+      "epoch": 4.92416,
+      "grad_norm": 0.676810085773468,
+      "learning_rate": 4.232292917166867e-05,
+      "loss": 0.6505,
+      "step": 3847
+    },
+    {
+      "epoch": 4.92544,
+      "grad_norm": 0.7427054643630981,
+      "learning_rate": 4.232092837134854e-05,
+      "loss": 0.684,
+      "step": 3848
+    },
+    {
+      "epoch": 4.9267199999999995,
+      "grad_norm": 0.7146655917167664,
+      "learning_rate": 4.231892757102841e-05,
+      "loss": 0.6527,
+      "step": 3849
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.690930187702179,
+      "learning_rate": 4.2316926770708285e-05,
+      "loss": 0.6584,
+      "step": 3850
+    },
+    {
+      "epoch": 4.92928,
+      "grad_norm": 0.742597222328186,
+      "learning_rate": 4.2314925970388157e-05,
+      "loss": 0.6644,
+      "step": 3851
+    },
+    {
+      "epoch": 4.93056,
+      "grad_norm": 0.7021815776824951,
+      "learning_rate": 4.231292517006803e-05,
+      "loss": 0.6765,
+      "step": 3852
+    },
+    {
+      "epoch": 4.93184,
+      "grad_norm": 0.7302278876304626,
+      "learning_rate": 4.23109243697479e-05,
+      "loss": 0.6875,
+      "step": 3853
+    },
+    {
+      "epoch": 4.93312,
+      "grad_norm": 0.7111605405807495,
+      "learning_rate": 4.230892356942777e-05,
+      "loss": 0.6808,
+      "step": 3854
+    },
+    {
+      "epoch": 4.9344,
+      "grad_norm": 0.7146978378295898,
+      "learning_rate": 4.2306922769107644e-05,
+      "loss": 0.598,
+      "step": 3855
+    },
+    {
+      "epoch": 4.93568,
+      "grad_norm": 0.6885932683944702,
+      "learning_rate": 4.230492196878752e-05,
+      "loss": 0.6331,
+      "step": 3856
+    },
+    {
+      "epoch": 4.93696,
+      "grad_norm": 0.7933337092399597,
+      "learning_rate": 4.230292116846739e-05,
+      "loss": 0.7657,
+      "step": 3857
+    },
+    {
+      "epoch": 4.93824,
+      "grad_norm": 0.7346404790878296,
+      "learning_rate": 4.230092036814726e-05,
+      "loss": 0.7045,
+      "step": 3858
+    },
+    {
+      "epoch": 4.93952,
+      "grad_norm": 0.7179169058799744,
+      "learning_rate": 4.229891956782713e-05,
+      "loss": 0.6378,
+      "step": 3859
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.7097733020782471,
+      "learning_rate": 4.2296918767507e-05,
+      "loss": 0.6371,
+      "step": 3860
+    },
+    {
+      "epoch": 4.94208,
+      "grad_norm": 0.7483527064323425,
+      "learning_rate": 4.2294917967186875e-05,
+      "loss": 0.6334,
+      "step": 3861
+    },
+    {
+      "epoch": 4.94336,
+      "grad_norm": 0.7159331440925598,
+      "learning_rate": 4.229291716686675e-05,
+      "loss": 0.6409,
+      "step": 3862
+    },
+    {
+      "epoch": 4.94464,
+      "grad_norm": 0.7627670764923096,
+      "learning_rate": 4.229091636654662e-05,
+      "loss": 0.6671,
+      "step": 3863
+    },
+    {
+      "epoch": 4.94592,
+      "grad_norm": 0.7190044522285461,
+      "learning_rate": 4.22889155662265e-05,
+      "loss": 0.6906,
+      "step": 3864
+    },
+    {
+      "epoch": 4.9472000000000005,
+      "grad_norm": 0.7264943718910217,
+      "learning_rate": 4.228691476590636e-05,
+      "loss": 0.6485,
+      "step": 3865
+    },
+    {
+      "epoch": 4.94848,
+      "grad_norm": 0.7001417875289917,
+      "learning_rate": 4.2284913965586234e-05,
+      "loss": 0.6528,
+      "step": 3866
+    },
+    {
+      "epoch": 4.94976,
+      "grad_norm": 0.7295548319816589,
+      "learning_rate": 4.2282913165266106e-05,
+      "loss": 0.6047,
+      "step": 3867
+    },
+    {
+      "epoch": 4.95104,
+      "grad_norm": 0.7287285327911377,
+      "learning_rate": 4.228091236494598e-05,
+      "loss": 0.6203,
+      "step": 3868
+    },
+    {
+      "epoch": 4.95232,
+      "grad_norm": 0.6717093586921692,
+      "learning_rate": 4.227891156462585e-05,
+      "loss": 0.642,
+      "step": 3869
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.7274335622787476,
+      "learning_rate": 4.227691076430572e-05,
+      "loss": 0.7194,
+      "step": 3870
+    },
+    {
+      "epoch": 4.95488,
+      "grad_norm": 0.7339951395988464,
+      "learning_rate": 4.22749099639856e-05,
+      "loss": 0.6511,
+      "step": 3871
+    },
+    {
+      "epoch": 4.95616,
+      "grad_norm": 0.7088412046432495,
+      "learning_rate": 4.227290916366547e-05,
+      "loss": 0.6424,
+      "step": 3872
+    },
+    {
+      "epoch": 4.95744,
+      "grad_norm": 0.6983089447021484,
+      "learning_rate": 4.227090836334534e-05,
+      "loss": 0.6186,
+      "step": 3873
+    },
+    {
+      "epoch": 4.95872,
+      "grad_norm": 0.727528989315033,
+      "learning_rate": 4.226890756302521e-05,
+      "loss": 0.616,
+      "step": 3874
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.732072651386261,
+      "learning_rate": 4.226690676270508e-05,
+      "loss": 0.6259,
+      "step": 3875
+    },
+    {
+      "epoch": 4.96128,
+      "grad_norm": 0.6999890208244324,
+      "learning_rate": 4.226490596238495e-05,
+      "loss": 0.677,
+      "step": 3876
+    },
+    {
+      "epoch": 4.96256,
+      "grad_norm": 0.6825425028800964,
+      "learning_rate": 4.2262905162064825e-05,
+      "loss": 0.6053,
+      "step": 3877
+    },
+    {
+      "epoch": 4.96384,
+      "grad_norm": 0.7285329699516296,
+      "learning_rate": 4.2260904361744703e-05,
+      "loss": 0.696,
+      "step": 3878
+    },
+    {
+      "epoch": 4.96512,
+      "grad_norm": 0.6932041645050049,
+      "learning_rate": 4.2258903561424575e-05,
+      "loss": 0.6412,
+      "step": 3879
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.7156229019165039,
+      "learning_rate": 4.225690276110445e-05,
+      "loss": 0.648,
+      "step": 3880
+    },
+    {
+      "epoch": 4.96768,
+      "grad_norm": 0.7821218371391296,
+      "learning_rate": 4.225490196078431e-05,
+      "loss": 0.72,
+      "step": 3881
+    },
+    {
+      "epoch": 4.96896,
+      "grad_norm": 0.6983861327171326,
+      "learning_rate": 4.2252901160464184e-05,
+      "loss": 0.6516,
+      "step": 3882
+    },
+    {
+      "epoch": 4.97024,
+      "grad_norm": 0.7392165660858154,
+      "learning_rate": 4.2250900360144056e-05,
+      "loss": 0.6855,
+      "step": 3883
+    },
+    {
+      "epoch": 4.97152,
+      "grad_norm": 0.7489770650863647,
+      "learning_rate": 4.224889955982393e-05,
+      "loss": 0.658,
+      "step": 3884
+    },
+    {
+      "epoch": 4.9728,
+      "grad_norm": 0.6763161420822144,
+      "learning_rate": 4.2246898759503806e-05,
+      "loss": 0.6016,
+      "step": 3885
+    },
+    {
+      "epoch": 4.97408,
+      "grad_norm": 0.7260777354240417,
+      "learning_rate": 4.224489795918368e-05,
+      "loss": 0.6908,
+      "step": 3886
+    },
+    {
+      "epoch": 4.97536,
+      "grad_norm": 0.7529964447021484,
+      "learning_rate": 4.224289715886355e-05,
+      "loss": 0.6736,
+      "step": 3887
+    },
+    {
+      "epoch": 4.97664,
+      "grad_norm": 0.6947912573814392,
+      "learning_rate": 4.224089635854342e-05,
+      "loss": 0.6518,
+      "step": 3888
+    },
+    {
+      "epoch": 4.97792,
+      "grad_norm": 0.6871297955513,
+      "learning_rate": 4.223889555822329e-05,
+      "loss": 0.6536,
+      "step": 3889
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.7125989198684692,
+      "learning_rate": 4.223689475790316e-05,
+      "loss": 0.7032,
+      "step": 3890
+    },
+    {
+      "epoch": 4.98048,
+      "grad_norm": 0.7379283308982849,
+      "learning_rate": 4.223489395758303e-05,
+      "loss": 0.6589,
+      "step": 3891
+    },
+    {
+      "epoch": 4.9817599999999995,
+      "grad_norm": 0.7186259031295776,
+      "learning_rate": 4.223289315726291e-05,
+      "loss": 0.6572,
+      "step": 3892
+    },
+    {
+      "epoch": 4.98304,
+      "grad_norm": 0.7020161151885986,
+      "learning_rate": 4.223089235694278e-05,
+      "loss": 0.6087,
+      "step": 3893
+    },
+    {
+      "epoch": 4.98432,
+      "grad_norm": 0.7505249977111816,
+      "learning_rate": 4.222889155662265e-05,
+      "loss": 0.7485,
+      "step": 3894
+    },
+    {
+      "epoch": 4.9856,
+      "grad_norm": 0.7452868223190308,
+      "learning_rate": 4.2226890756302525e-05,
+      "loss": 0.6414,
+      "step": 3895
+    },
+    {
+      "epoch": 4.98688,
+      "grad_norm": 0.7297542691230774,
+      "learning_rate": 4.22248899559824e-05,
+      "loss": 0.6662,
+      "step": 3896
+    },
+    {
+      "epoch": 4.98816,
+      "grad_norm": 0.7233591079711914,
+      "learning_rate": 4.222288915566226e-05,
+      "loss": 0.6443,
+      "step": 3897
+    },
+    {
+      "epoch": 4.98944,
+      "grad_norm": 0.6811563968658447,
+      "learning_rate": 4.2220888355342134e-05,
+      "loss": 0.6349,
+      "step": 3898
+    },
+    {
+      "epoch": 4.99072,
+      "grad_norm": 0.7110039591789246,
+      "learning_rate": 4.221888755502201e-05,
+      "loss": 0.5868,
+      "step": 3899
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.685947060585022,
+      "learning_rate": 4.2216886754701884e-05,
+      "loss": 0.6747,
+      "step": 3900
+    },
+    {
+      "epoch": 4.99328,
+      "grad_norm": 0.6793825626373291,
+      "learning_rate": 4.2214885954381756e-05,
+      "loss": 0.6335,
+      "step": 3901
+    },
+    {
+      "epoch": 4.99456,
+      "grad_norm": 0.7450732588768005,
+      "learning_rate": 4.221288515406163e-05,
+      "loss": 0.6845,
+      "step": 3902
+    },
+    {
+      "epoch": 4.99584,
+      "grad_norm": 0.7151440978050232,
+      "learning_rate": 4.22108843537415e-05,
+      "loss": 0.6479,
+      "step": 3903
+    },
+    {
+      "epoch": 4.99712,
+      "grad_norm": 0.7348343133926392,
+      "learning_rate": 4.220888355342137e-05,
+      "loss": 0.6799,
+      "step": 3904
+    },
+    {
+      "epoch": 4.9984,
+      "grad_norm": 0.6973733305931091,
+      "learning_rate": 4.220688275310124e-05,
+      "loss": 0.6555,
+      "step": 3905
+    },
+    {
+      "epoch": 4.99968,
+      "grad_norm": 0.7058578133583069,
+      "learning_rate": 4.2204881952781115e-05,
+      "loss": 0.6924,
+      "step": 3906
+    },
+    {
+      "epoch": 5.00096,
+      "grad_norm": 1.5164827108383179,
+      "learning_rate": 4.220288115246099e-05,
+      "loss": 1.0917,
+      "step": 3907
+    },
+    {
+      "epoch": 5.00224,
+      "grad_norm": 0.7053406834602356,
+      "learning_rate": 4.220088035214086e-05,
+      "loss": 0.6068,
+      "step": 3908
+    },
+    {
+      "epoch": 5.00352,
+      "grad_norm": 0.6764478087425232,
+      "learning_rate": 4.219887955182073e-05,
+      "loss": 0.6047,
+      "step": 3909
+    },
+    {
+      "epoch": 5.0048,
+      "grad_norm": 0.6720277070999146,
+      "learning_rate": 4.21968787515006e-05,
+      "loss": 0.6328,
+      "step": 3910
+    },
+    {
+      "epoch": 5.00608,
+      "grad_norm": 0.715262234210968,
+      "learning_rate": 4.2194877951180475e-05,
+      "loss": 0.6962,
+      "step": 3911
+    },
+    {
+      "epoch": 5.00736,
+      "grad_norm": 0.6942645907402039,
+      "learning_rate": 4.219287715086035e-05,
+      "loss": 0.6243,
+      "step": 3912
+    },
+    {
+      "epoch": 5.00864,
+      "grad_norm": 0.7017389535903931,
+      "learning_rate": 4.219087635054022e-05,
+      "loss": 0.6112,
+      "step": 3913
+    },
+    {
+      "epoch": 5.00992,
+      "grad_norm": 0.7179716229438782,
+      "learning_rate": 4.218887555022009e-05,
+      "loss": 0.6935,
+      "step": 3914
+    },
+    {
+      "epoch": 5.0112,
+      "grad_norm": 0.7195164561271667,
+      "learning_rate": 4.218687474989996e-05,
+      "loss": 0.6682,
+      "step": 3915
+    },
+    {
+      "epoch": 5.01248,
+      "grad_norm": 0.7197250723838806,
+      "learning_rate": 4.2184873949579834e-05,
+      "loss": 0.6436,
+      "step": 3916
+    },
+    {
+      "epoch": 5.01376,
+      "grad_norm": 0.7025637626647949,
+      "learning_rate": 4.2182873149259706e-05,
+      "loss": 0.6336,
+      "step": 3917
+    },
+    {
+      "epoch": 5.01504,
+      "grad_norm": 0.7276149392127991,
+      "learning_rate": 4.218087234893958e-05,
+      "loss": 0.6207,
+      "step": 3918
+    },
+    {
+      "epoch": 5.01632,
+      "grad_norm": 0.7081112265586853,
+      "learning_rate": 4.217887154861945e-05,
+      "loss": 0.6396,
+      "step": 3919
+    },
+    {
+      "epoch": 5.0176,
+      "grad_norm": 0.7482070922851562,
+      "learning_rate": 4.217687074829932e-05,
+      "loss": 0.632,
+      "step": 3920
+    },
+    {
+      "epoch": 5.01888,
+      "grad_norm": 0.7229052186012268,
+      "learning_rate": 4.217486994797919e-05,
+      "loss": 0.5806,
+      "step": 3921
+    },
+    {
+      "epoch": 5.02016,
+      "grad_norm": 0.678459644317627,
+      "learning_rate": 4.2172869147659065e-05,
+      "loss": 0.5644,
+      "step": 3922
+    },
+    {
+      "epoch": 5.02144,
+      "grad_norm": 0.7244837880134583,
+      "learning_rate": 4.217086834733894e-05,
+      "loss": 0.6354,
+      "step": 3923
+    },
+    {
+      "epoch": 5.02272,
+      "grad_norm": 0.752399206161499,
+      "learning_rate": 4.216886754701881e-05,
+      "loss": 0.6326,
+      "step": 3924
+    },
+    {
+      "epoch": 5.024,
+      "grad_norm": 0.7514381408691406,
+      "learning_rate": 4.216686674669868e-05,
+      "loss": 0.6534,
+      "step": 3925
+    },
+    {
+      "epoch": 5.02528,
+      "grad_norm": 0.720856249332428,
+      "learning_rate": 4.216486594637855e-05,
+      "loss": 0.6045,
+      "step": 3926
+    },
+    {
+      "epoch": 5.02656,
+      "grad_norm": 0.7397941946983337,
+      "learning_rate": 4.2162865146058424e-05,
+      "loss": 0.6208,
+      "step": 3927
+    },
+    {
+      "epoch": 5.02784,
+      "grad_norm": 0.7684844732284546,
+      "learning_rate": 4.2160864345738296e-05,
+      "loss": 0.6398,
+      "step": 3928
+    },
+    {
+      "epoch": 5.02912,
+      "grad_norm": 0.7364361882209778,
+      "learning_rate": 4.215886354541817e-05,
+      "loss": 0.6167,
+      "step": 3929
+    },
+    {
+      "epoch": 5.0304,
+      "grad_norm": 0.6819130182266235,
+      "learning_rate": 4.215686274509804e-05,
+      "loss": 0.5566,
+      "step": 3930
+    },
+    {
+      "epoch": 5.03168,
+      "grad_norm": 0.7268734574317932,
+      "learning_rate": 4.215486194477791e-05,
+      "loss": 0.6476,
+      "step": 3931
+    },
+    {
+      "epoch": 5.03296,
+      "grad_norm": 0.7218887805938721,
+      "learning_rate": 4.2152861144457784e-05,
+      "loss": 0.6149,
+      "step": 3932
+    },
+    {
+      "epoch": 5.03424,
+      "grad_norm": 0.7900970578193665,
+      "learning_rate": 4.2150860344137656e-05,
+      "loss": 0.6828,
+      "step": 3933
+    },
+    {
+      "epoch": 5.03552,
+      "grad_norm": 0.6771689653396606,
+      "learning_rate": 4.2148859543817534e-05,
+      "loss": 0.5821,
+      "step": 3934
+    },
+    {
+      "epoch": 5.0368,
+      "grad_norm": 0.7154511213302612,
+      "learning_rate": 4.21468587434974e-05,
+      "loss": 0.5964,
+      "step": 3935
+    },
+    {
+      "epoch": 5.03808,
+      "grad_norm": 0.7221072912216187,
+      "learning_rate": 4.214485794317727e-05,
+      "loss": 0.6088,
+      "step": 3936
+    },
+    {
+      "epoch": 5.03936,
+      "grad_norm": 0.7826116681098938,
+      "learning_rate": 4.214285714285714e-05,
+      "loss": 0.6514,
+      "step": 3937
+    },
+    {
+      "epoch": 5.04064,
+      "grad_norm": 0.7515047192573547,
+      "learning_rate": 4.2140856342537015e-05,
+      "loss": 0.6776,
+      "step": 3938
+    },
+    {
+      "epoch": 5.04192,
+      "grad_norm": 0.7475928664207458,
+      "learning_rate": 4.213885554221689e-05,
+      "loss": 0.6478,
+      "step": 3939
+    },
+    {
+      "epoch": 5.0432,
+      "grad_norm": 0.7153226733207703,
+      "learning_rate": 4.213685474189676e-05,
+      "loss": 0.6474,
+      "step": 3940
+    },
+    {
+      "epoch": 5.04448,
+      "grad_norm": 0.6863680481910706,
+      "learning_rate": 4.213485394157664e-05,
+      "loss": 0.6281,
+      "step": 3941
+    },
+    {
+      "epoch": 5.04576,
+      "grad_norm": 0.6896254420280457,
+      "learning_rate": 4.213285314125651e-05,
+      "loss": 0.6058,
+      "step": 3942
+    },
+    {
+      "epoch": 5.04704,
+      "grad_norm": 0.7509266138076782,
+      "learning_rate": 4.2130852340936374e-05,
+      "loss": 0.6445,
+      "step": 3943
+    },
+    {
+      "epoch": 5.04832,
+      "grad_norm": 0.7177383303642273,
+      "learning_rate": 4.2128851540616246e-05,
+      "loss": 0.6522,
+      "step": 3944
+    },
+    {
+      "epoch": 5.0496,
+      "grad_norm": 0.7128506302833557,
+      "learning_rate": 4.212685074029612e-05,
+      "loss": 0.6011,
+      "step": 3945
+    },
+    {
+      "epoch": 5.05088,
+      "grad_norm": 0.6913168430328369,
+      "learning_rate": 4.212484993997599e-05,
+      "loss": 0.6598,
+      "step": 3946
+    },
+    {
+      "epoch": 5.05216,
+      "grad_norm": 0.740969181060791,
+      "learning_rate": 4.212284913965586e-05,
+      "loss": 0.6409,
+      "step": 3947
+    },
+    {
+      "epoch": 5.05344,
+      "grad_norm": 0.7747688293457031,
+      "learning_rate": 4.212084833933574e-05,
+      "loss": 0.6727,
+      "step": 3948
+    },
+    {
+      "epoch": 5.05472,
+      "grad_norm": 0.7710026502609253,
+      "learning_rate": 4.211884753901561e-05,
+      "loss": 0.6585,
+      "step": 3949
+    },
+    {
+      "epoch": 5.056,
+      "grad_norm": 0.713817834854126,
+      "learning_rate": 4.2116846738695484e-05,
+      "loss": 0.6174,
+      "step": 3950
+    },
+    {
+      "epoch": 5.05728,
+      "grad_norm": 0.7268139123916626,
+      "learning_rate": 4.211484593837535e-05,
+      "loss": 0.6679,
+      "step": 3951
+    },
+    {
+      "epoch": 5.05856,
+      "grad_norm": 0.7151246666908264,
+      "learning_rate": 4.211284513805522e-05,
+      "loss": 0.5802,
+      "step": 3952
+    },
+    {
+      "epoch": 5.05984,
+      "grad_norm": 0.6851016283035278,
+      "learning_rate": 4.211084433773509e-05,
+      "loss": 0.6258,
+      "step": 3953
+    },
+    {
+      "epoch": 5.06112,
+      "grad_norm": 0.7177205681800842,
+      "learning_rate": 4.2108843537414965e-05,
+      "loss": 0.6276,
+      "step": 3954
+    },
+    {
+      "epoch": 5.0624,
+      "grad_norm": 0.7391200661659241,
+      "learning_rate": 4.210684273709484e-05,
+      "loss": 0.6175,
+      "step": 3955
+    },
+    {
+      "epoch": 5.06368,
+      "grad_norm": 0.7397950887680054,
+      "learning_rate": 4.2104841936774715e-05,
+      "loss": 0.6149,
+      "step": 3956
+    },
+    {
+      "epoch": 5.06496,
+      "grad_norm": 0.7378024458885193,
+      "learning_rate": 4.210284113645459e-05,
+      "loss": 0.6524,
+      "step": 3957
+    },
+    {
+      "epoch": 5.06624,
+      "grad_norm": 0.7197895050048828,
+      "learning_rate": 4.210084033613446e-05,
+      "loss": 0.6747,
+      "step": 3958
+    },
+    {
+      "epoch": 5.06752,
+      "grad_norm": 0.7385547161102295,
+      "learning_rate": 4.2098839535814324e-05,
+      "loss": 0.6381,
+      "step": 3959
+    },
+    {
+      "epoch": 5.0688,
+      "grad_norm": 0.7747622728347778,
+      "learning_rate": 4.2096838735494196e-05,
+      "loss": 0.6198,
+      "step": 3960
+    },
+    {
+      "epoch": 5.07008,
+      "grad_norm": 0.7808424830436707,
+      "learning_rate": 4.209483793517407e-05,
+      "loss": 0.6496,
+      "step": 3961
+    },
+    {
+      "epoch": 5.07136,
+      "grad_norm": 0.7174487709999084,
+      "learning_rate": 4.2092837134853946e-05,
+      "loss": 0.5677,
+      "step": 3962
+    },
+    {
+      "epoch": 5.07264,
+      "grad_norm": 0.7417564392089844,
+      "learning_rate": 4.209083633453382e-05,
+      "loss": 0.6681,
+      "step": 3963
+    },
+    {
+      "epoch": 5.07392,
+      "grad_norm": 0.654253363609314,
+      "learning_rate": 4.208883553421369e-05,
+      "loss": 0.5682,
+      "step": 3964
+    },
+    {
+      "epoch": 5.0752,
+      "grad_norm": 0.7092409133911133,
+      "learning_rate": 4.208683473389356e-05,
+      "loss": 0.6277,
+      "step": 3965
+    },
+    {
+      "epoch": 5.07648,
+      "grad_norm": 0.7257412672042847,
+      "learning_rate": 4.2084833933573434e-05,
+      "loss": 0.6708,
+      "step": 3966
+    },
+    {
+      "epoch": 5.07776,
+      "grad_norm": 0.7401608228683472,
+      "learning_rate": 4.20828331332533e-05,
+      "loss": 0.6563,
+      "step": 3967
+    },
+    {
+      "epoch": 5.07904,
+      "grad_norm": 0.7241030335426331,
+      "learning_rate": 4.208083233293317e-05,
+      "loss": 0.6504,
+      "step": 3968
+    },
+    {
+      "epoch": 5.08032,
+      "grad_norm": 0.7237942218780518,
+      "learning_rate": 4.207883153261305e-05,
+      "loss": 0.6548,
+      "step": 3969
+    },
+    {
+      "epoch": 5.0816,
+      "grad_norm": 0.7459864616394043,
+      "learning_rate": 4.207683073229292e-05,
+      "loss": 0.6404,
+      "step": 3970
+    },
+    {
+      "epoch": 5.08288,
+      "grad_norm": 0.7194232940673828,
+      "learning_rate": 4.207482993197279e-05,
+      "loss": 0.6405,
+      "step": 3971
+    },
+    {
+      "epoch": 5.08416,
+      "grad_norm": 0.7170876860618591,
+      "learning_rate": 4.2072829131652665e-05,
+      "loss": 0.6667,
+      "step": 3972
+    },
+    {
+      "epoch": 5.08544,
+      "grad_norm": 0.7066750526428223,
+      "learning_rate": 4.207082833133254e-05,
+      "loss": 0.6285,
+      "step": 3973
+    },
+    {
+      "epoch": 5.08672,
+      "grad_norm": 0.7601994276046753,
+      "learning_rate": 4.206882753101241e-05,
+      "loss": 0.7017,
+      "step": 3974
+    },
+    {
+      "epoch": 5.088,
+      "grad_norm": 0.7173680663108826,
+      "learning_rate": 4.2066826730692274e-05,
+      "loss": 0.6869,
+      "step": 3975
+    },
+    {
+      "epoch": 5.08928,
+      "grad_norm": 0.7139931321144104,
+      "learning_rate": 4.2064825930372145e-05,
+      "loss": 0.6175,
+      "step": 3976
+    },
+    {
+      "epoch": 5.09056,
+      "grad_norm": 0.7550209760665894,
+      "learning_rate": 4.2062825130052024e-05,
+      "loss": 0.6521,
+      "step": 3977
+    },
+    {
+      "epoch": 5.09184,
+      "grad_norm": 0.7335852384567261,
+      "learning_rate": 4.2060824329731896e-05,
+      "loss": 0.6432,
+      "step": 3978
+    },
+    {
+      "epoch": 5.09312,
+      "grad_norm": 0.7130416035652161,
+      "learning_rate": 4.205882352941177e-05,
+      "loss": 0.616,
+      "step": 3979
+    },
+    {
+      "epoch": 5.0944,
+      "grad_norm": 0.7009579539299011,
+      "learning_rate": 4.205682272909164e-05,
+      "loss": 0.6022,
+      "step": 3980
+    },
+    {
+      "epoch": 5.09568,
+      "grad_norm": 0.7348793148994446,
+      "learning_rate": 4.205482192877151e-05,
+      "loss": 0.6118,
+      "step": 3981
+    },
+    {
+      "epoch": 5.09696,
+      "grad_norm": 0.7403352856636047,
+      "learning_rate": 4.205282112845138e-05,
+      "loss": 0.6397,
+      "step": 3982
+    },
+    {
+      "epoch": 5.09824,
+      "grad_norm": 0.7160905003547668,
+      "learning_rate": 4.205082032813125e-05,
+      "loss": 0.6391,
+      "step": 3983
+    },
+    {
+      "epoch": 5.09952,
+      "grad_norm": 0.7039486765861511,
+      "learning_rate": 4.204881952781113e-05,
+      "loss": 0.6062,
+      "step": 3984
+    },
+    {
+      "epoch": 5.1008,
+      "grad_norm": 0.752713680267334,
+      "learning_rate": 4.2046818727491e-05,
+      "loss": 0.6827,
+      "step": 3985
+    },
+    {
+      "epoch": 5.10208,
+      "grad_norm": 0.7083015441894531,
+      "learning_rate": 4.204481792717087e-05,
+      "loss": 0.6033,
+      "step": 3986
+    },
+    {
+      "epoch": 5.10336,
+      "grad_norm": 0.7464533448219299,
+      "learning_rate": 4.204281712685074e-05,
+      "loss": 0.6603,
+      "step": 3987
+    },
+    {
+      "epoch": 5.10464,
+      "grad_norm": 0.713887095451355,
+      "learning_rate": 4.2040816326530615e-05,
+      "loss": 0.6253,
+      "step": 3988
+    },
+    {
+      "epoch": 5.10592,
+      "grad_norm": 0.7454844117164612,
+      "learning_rate": 4.2038815526210486e-05,
+      "loss": 0.6518,
+      "step": 3989
+    },
+    {
+      "epoch": 5.1072,
+      "grad_norm": 0.770743191242218,
+      "learning_rate": 4.203681472589036e-05,
+      "loss": 0.6907,
+      "step": 3990
+    },
+    {
+      "epoch": 5.10848,
+      "grad_norm": 0.7465303540229797,
+      "learning_rate": 4.203481392557023e-05,
+      "loss": 0.7005,
+      "step": 3991
+    },
+    {
+      "epoch": 5.10976,
+      "grad_norm": 0.745933473110199,
+      "learning_rate": 4.20328131252501e-05,
+      "loss": 0.6115,
+      "step": 3992
+    },
+    {
+      "epoch": 5.11104,
+      "grad_norm": 0.6949446797370911,
+      "learning_rate": 4.2030812324929974e-05,
+      "loss": 0.5874,
+      "step": 3993
+    },
+    {
+      "epoch": 5.11232,
+      "grad_norm": 0.7319580912590027,
+      "learning_rate": 4.2028811524609846e-05,
+      "loss": 0.6294,
+      "step": 3994
+    },
+    {
+      "epoch": 5.1136,
+      "grad_norm": 0.7834770679473877,
+      "learning_rate": 4.202681072428972e-05,
+      "loss": 0.7396,
+      "step": 3995
+    },
+    {
+      "epoch": 5.11488,
+      "grad_norm": 0.7234769463539124,
+      "learning_rate": 4.202480992396959e-05,
+      "loss": 0.6397,
+      "step": 3996
+    },
+    {
+      "epoch": 5.11616,
+      "grad_norm": 0.7542564868927002,
+      "learning_rate": 4.202280912364946e-05,
+      "loss": 0.6028,
+      "step": 3997
+    },
+    {
+      "epoch": 5.11744,
+      "grad_norm": 0.7732244729995728,
+      "learning_rate": 4.202080832332933e-05,
+      "loss": 0.6938,
+      "step": 3998
+    },
+    {
+      "epoch": 5.11872,
+      "grad_norm": 0.6997548937797546,
+      "learning_rate": 4.2018807523009205e-05,
+      "loss": 0.6206,
+      "step": 3999
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.7281977534294128,
+      "learning_rate": 4.201680672268908e-05,
+      "loss": 0.6279,
+      "step": 4000
+    },
+    {
+      "epoch": 5.12128,
+      "grad_norm": 0.7627147436141968,
+      "learning_rate": 4.201480592236895e-05,
+      "loss": 0.6297,
+      "step": 4001
+    },
+    {
+      "epoch": 5.12256,
+      "grad_norm": 0.7351235151290894,
+      "learning_rate": 4.201280512204882e-05,
+      "loss": 0.6132,
+      "step": 4002
+    },
+    {
+      "epoch": 5.12384,
+      "grad_norm": 0.7031277418136597,
+      "learning_rate": 4.201080432172869e-05,
+      "loss": 0.6561,
+      "step": 4003
+    },
+    {
+      "epoch": 5.12512,
+      "grad_norm": 0.7446048259735107,
+      "learning_rate": 4.2008803521408564e-05,
+      "loss": 0.6526,
+      "step": 4004
+    },
+    {
+      "epoch": 5.1264,
+      "grad_norm": 0.7449467778205872,
+      "learning_rate": 4.200680272108844e-05,
+      "loss": 0.6027,
+      "step": 4005
+    },
+    {
+      "epoch": 5.12768,
+      "grad_norm": 0.8109471201896667,
+      "learning_rate": 4.200480192076831e-05,
+      "loss": 0.6602,
+      "step": 4006
+    },
+    {
+      "epoch": 5.12896,
+      "grad_norm": 0.7701667547225952,
+      "learning_rate": 4.200280112044818e-05,
+      "loss": 0.6192,
+      "step": 4007
+    },
+    {
+      "epoch": 5.13024,
+      "grad_norm": 0.7170472741127014,
+      "learning_rate": 4.200080032012805e-05,
+      "loss": 0.5924,
+      "step": 4008
+    },
+    {
+      "epoch": 5.13152,
+      "grad_norm": 0.7401517033576965,
+      "learning_rate": 4.1998799519807924e-05,
+      "loss": 0.6098,
+      "step": 4009
+    },
+    {
+      "epoch": 5.1328,
+      "grad_norm": 0.7323094606399536,
+      "learning_rate": 4.1996798719487795e-05,
+      "loss": 0.6426,
+      "step": 4010
+    },
+    {
+      "epoch": 5.13408,
+      "grad_norm": 0.73775315284729,
+      "learning_rate": 4.199479791916767e-05,
+      "loss": 0.6091,
+      "step": 4011
+    },
+    {
+      "epoch": 5.13536,
+      "grad_norm": 0.7261705994606018,
+      "learning_rate": 4.1992797118847546e-05,
+      "loss": 0.6217,
+      "step": 4012
+    },
+    {
+      "epoch": 5.13664,
+      "grad_norm": 0.7077757716178894,
+      "learning_rate": 4.199079631852742e-05,
+      "loss": 0.6288,
+      "step": 4013
+    },
+    {
+      "epoch": 5.13792,
+      "grad_norm": 0.7402573823928833,
+      "learning_rate": 4.198879551820728e-05,
+      "loss": 0.6528,
+      "step": 4014
+    },
+    {
+      "epoch": 5.1392,
+      "grad_norm": 0.7423076629638672,
+      "learning_rate": 4.1986794717887155e-05,
+      "loss": 0.6715,
+      "step": 4015
+    },
+    {
+      "epoch": 5.14048,
+      "grad_norm": 0.76578289270401,
+      "learning_rate": 4.1984793917567026e-05,
+      "loss": 0.6479,
+      "step": 4016
+    },
+    {
+      "epoch": 5.14176,
+      "grad_norm": 0.723196268081665,
+      "learning_rate": 4.19827931172469e-05,
+      "loss": 0.6415,
+      "step": 4017
+    },
+    {
+      "epoch": 5.14304,
+      "grad_norm": 0.7332058548927307,
+      "learning_rate": 4.198079231692677e-05,
+      "loss": 0.6373,
+      "step": 4018
+    },
+    {
+      "epoch": 5.1443200000000004,
+      "grad_norm": 0.7015236616134644,
+      "learning_rate": 4.197879151660665e-05,
+      "loss": 0.5996,
+      "step": 4019
+    },
+    {
+      "epoch": 5.1456,
+      "grad_norm": 0.743124783039093,
+      "learning_rate": 4.197679071628652e-05,
+      "loss": 0.5971,
+      "step": 4020
+    },
+    {
+      "epoch": 5.14688,
+      "grad_norm": 0.7382779717445374,
+      "learning_rate": 4.197478991596639e-05,
+      "loss": 0.6285,
+      "step": 4021
+    },
+    {
+      "epoch": 5.14816,
+      "grad_norm": 0.7928086519241333,
+      "learning_rate": 4.197278911564626e-05,
+      "loss": 0.6687,
+      "step": 4022
+    },
+    {
+      "epoch": 5.14944,
+      "grad_norm": 0.7693018913269043,
+      "learning_rate": 4.197078831532613e-05,
+      "loss": 0.6553,
+      "step": 4023
+    },
+    {
+      "epoch": 5.15072,
+      "grad_norm": 0.7531712055206299,
+      "learning_rate": 4.1968787515006e-05,
+      "loss": 0.642,
+      "step": 4024
+    },
+    {
+      "epoch": 5.152,
+      "grad_norm": 0.7181296944618225,
+      "learning_rate": 4.196678671468587e-05,
+      "loss": 0.5862,
+      "step": 4025
+    },
+    {
+      "epoch": 5.15328,
+      "grad_norm": 0.724727988243103,
+      "learning_rate": 4.196478591436575e-05,
+      "loss": 0.6467,
+      "step": 4026
+    },
+    {
+      "epoch": 5.15456,
+      "grad_norm": 0.7461944222450256,
+      "learning_rate": 4.1962785114045624e-05,
+      "loss": 0.6054,
+      "step": 4027
+    },
+    {
+      "epoch": 5.15584,
+      "grad_norm": 0.7383855581283569,
+      "learning_rate": 4.1960784313725496e-05,
+      "loss": 0.621,
+      "step": 4028
+    },
+    {
+      "epoch": 5.15712,
+      "grad_norm": 0.6796069145202637,
+      "learning_rate": 4.195878351340537e-05,
+      "loss": 0.5679,
+      "step": 4029
+    },
+    {
+      "epoch": 5.1584,
+      "grad_norm": 0.7139230370521545,
+      "learning_rate": 4.195678271308523e-05,
+      "loss": 0.5988,
+      "step": 4030
+    },
+    {
+      "epoch": 5.15968,
+      "grad_norm": 0.7442138195037842,
+      "learning_rate": 4.1954781912765104e-05,
+      "loss": 0.5973,
+      "step": 4031
+    },
+    {
+      "epoch": 5.16096,
+      "grad_norm": 0.7515295147895813,
+      "learning_rate": 4.1952781112444976e-05,
+      "loss": 0.5769,
+      "step": 4032
+    },
+    {
+      "epoch": 5.16224,
+      "grad_norm": 0.7667818665504456,
+      "learning_rate": 4.1950780312124855e-05,
+      "loss": 0.6609,
+      "step": 4033
+    },
+    {
+      "epoch": 5.16352,
+      "grad_norm": 0.7336476445198059,
+      "learning_rate": 4.194877951180473e-05,
+      "loss": 0.6163,
+      "step": 4034
+    },
+    {
+      "epoch": 5.1648,
+      "grad_norm": 0.6899265646934509,
+      "learning_rate": 4.19467787114846e-05,
+      "loss": 0.5985,
+      "step": 4035
+    },
+    {
+      "epoch": 5.16608,
+      "grad_norm": 0.6779667735099792,
+      "learning_rate": 4.194477791116447e-05,
+      "loss": 0.57,
+      "step": 4036
+    },
+    {
+      "epoch": 5.16736,
+      "grad_norm": 0.6793948411941528,
+      "learning_rate": 4.194277711084434e-05,
+      "loss": 0.5694,
+      "step": 4037
+    },
+    {
+      "epoch": 5.16864,
+      "grad_norm": 0.763278067111969,
+      "learning_rate": 4.194077631052421e-05,
+      "loss": 0.6917,
+      "step": 4038
+    },
+    {
+      "epoch": 5.16992,
+      "grad_norm": 0.699497640132904,
+      "learning_rate": 4.193877551020408e-05,
+      "loss": 0.5704,
+      "step": 4039
+    },
+    {
+      "epoch": 5.1712,
+      "grad_norm": 0.7154186964035034,
+      "learning_rate": 4.193677470988396e-05,
+      "loss": 0.6079,
+      "step": 4040
+    },
+    {
+      "epoch": 5.17248,
+      "grad_norm": 0.7278900146484375,
+      "learning_rate": 4.193477390956383e-05,
+      "loss": 0.6887,
+      "step": 4041
+    },
+    {
+      "epoch": 5.17376,
+      "grad_norm": 0.7596189379692078,
+      "learning_rate": 4.19327731092437e-05,
+      "loss": 0.7018,
+      "step": 4042
+    },
+    {
+      "epoch": 5.17504,
+      "grad_norm": 0.7543389201164246,
+      "learning_rate": 4.1930772308923573e-05,
+      "loss": 0.6161,
+      "step": 4043
+    },
+    {
+      "epoch": 5.17632,
+      "grad_norm": 0.7368336915969849,
+      "learning_rate": 4.1928771508603445e-05,
+      "loss": 0.642,
+      "step": 4044
+    },
+    {
+      "epoch": 5.1776,
+      "grad_norm": 0.7439759969711304,
+      "learning_rate": 4.192677070828332e-05,
+      "loss": 0.6999,
+      "step": 4045
+    },
+    {
+      "epoch": 5.17888,
+      "grad_norm": 0.695544958114624,
+      "learning_rate": 4.192476990796318e-05,
+      "loss": 0.6256,
+      "step": 4046
+    },
+    {
+      "epoch": 5.18016,
+      "grad_norm": 0.7475451231002808,
+      "learning_rate": 4.192276910764306e-05,
+      "loss": 0.648,
+      "step": 4047
+    },
+    {
+      "epoch": 5.18144,
+      "grad_norm": 0.740695059299469,
+      "learning_rate": 4.192076830732293e-05,
+      "loss": 0.598,
+      "step": 4048
+    },
+    {
+      "epoch": 5.18272,
+      "grad_norm": 0.7205529808998108,
+      "learning_rate": 4.1918767507002805e-05,
+      "loss": 0.6135,
+      "step": 4049
+    },
+    {
+      "epoch": 5.184,
+      "grad_norm": 0.6971973776817322,
+      "learning_rate": 4.1916766706682676e-05,
+      "loss": 0.6237,
+      "step": 4050
+    },
+    {
+      "epoch": 5.18528,
+      "grad_norm": 0.7059743404388428,
+      "learning_rate": 4.191476590636255e-05,
+      "loss": 0.6197,
+      "step": 4051
+    },
+    {
+      "epoch": 5.18656,
+      "grad_norm": 0.7417049407958984,
+      "learning_rate": 4.191276510604242e-05,
+      "loss": 0.6321,
+      "step": 4052
+    },
+    {
+      "epoch": 5.18784,
+      "grad_norm": 0.726333498954773,
+      "learning_rate": 4.191076430572229e-05,
+      "loss": 0.5983,
+      "step": 4053
+    },
+    {
+      "epoch": 5.18912,
+      "grad_norm": 0.7621744275093079,
+      "learning_rate": 4.1908763505402164e-05,
+      "loss": 0.675,
+      "step": 4054
+    },
+    {
+      "epoch": 5.1904,
+      "grad_norm": 0.7966973185539246,
+      "learning_rate": 4.1906762705082036e-05,
+      "loss": 0.6198,
+      "step": 4055
+    },
+    {
+      "epoch": 5.19168,
+      "grad_norm": 0.7543626427650452,
+      "learning_rate": 4.190476190476191e-05,
+      "loss": 0.6282,
+      "step": 4056
+    },
+    {
+      "epoch": 5.19296,
+      "grad_norm": 0.7392343282699585,
+      "learning_rate": 4.190276110444178e-05,
+      "loss": 0.6496,
+      "step": 4057
+    },
+    {
+      "epoch": 5.19424,
+      "grad_norm": 0.7553439140319824,
+      "learning_rate": 4.190076030412165e-05,
+      "loss": 0.6604,
+      "step": 4058
+    },
+    {
+      "epoch": 5.19552,
+      "grad_norm": 0.7107070088386536,
+      "learning_rate": 4.189875950380152e-05,
+      "loss": 0.5846,
+      "step": 4059
+    },
+    {
+      "epoch": 5.1968,
+      "grad_norm": 0.7690718173980713,
+      "learning_rate": 4.1896758703481395e-05,
+      "loss": 0.6315,
+      "step": 4060
+    },
+    {
+      "epoch": 5.19808,
+      "grad_norm": 0.7583000063896179,
+      "learning_rate": 4.189475790316127e-05,
+      "loss": 0.6565,
+      "step": 4061
+    },
+    {
+      "epoch": 5.19936,
+      "grad_norm": 0.735431432723999,
+      "learning_rate": 4.189275710284114e-05,
+      "loss": 0.6382,
+      "step": 4062
+    },
+    {
+      "epoch": 5.20064,
+      "grad_norm": 0.749560534954071,
+      "learning_rate": 4.189075630252101e-05,
+      "loss": 0.6199,
+      "step": 4063
+    },
+    {
+      "epoch": 5.20192,
+      "grad_norm": 0.7637155055999756,
+      "learning_rate": 4.188875550220088e-05,
+      "loss": 0.6392,
+      "step": 4064
+    },
+    {
+      "epoch": 5.2032,
+      "grad_norm": 0.7183476686477661,
+      "learning_rate": 4.1886754701880754e-05,
+      "loss": 0.6105,
+      "step": 4065
+    },
+    {
+      "epoch": 5.20448,
+      "grad_norm": 0.7134882807731628,
+      "learning_rate": 4.1884753901560626e-05,
+      "loss": 0.6176,
+      "step": 4066
+    },
+    {
+      "epoch": 5.20576,
+      "grad_norm": 0.6925643682479858,
+      "learning_rate": 4.18827531012405e-05,
+      "loss": 0.6148,
+      "step": 4067
+    },
+    {
+      "epoch": 5.20704,
+      "grad_norm": 0.6939037442207336,
+      "learning_rate": 4.188075230092037e-05,
+      "loss": 0.5917,
+      "step": 4068
+    },
+    {
+      "epoch": 5.20832,
+      "grad_norm": 0.7910222411155701,
+      "learning_rate": 4.187875150060024e-05,
+      "loss": 0.7064,
+      "step": 4069
+    },
+    {
+      "epoch": 5.2096,
+      "grad_norm": 0.7466183304786682,
+      "learning_rate": 4.1876750700280114e-05,
+      "loss": 0.6069,
+      "step": 4070
+    },
+    {
+      "epoch": 5.21088,
+      "grad_norm": 0.7494035363197327,
+      "learning_rate": 4.1874749899959985e-05,
+      "loss": 0.6614,
+      "step": 4071
+    },
+    {
+      "epoch": 5.21216,
+      "grad_norm": 0.7295969724655151,
+      "learning_rate": 4.187274909963986e-05,
+      "loss": 0.6364,
+      "step": 4072
+    },
+    {
+      "epoch": 5.21344,
+      "grad_norm": 0.7598521709442139,
+      "learning_rate": 4.187074829931973e-05,
+      "loss": 0.6466,
+      "step": 4073
+    },
+    {
+      "epoch": 5.21472,
+      "grad_norm": 0.7255356907844543,
+      "learning_rate": 4.18687474989996e-05,
+      "loss": 0.6205,
+      "step": 4074
+    },
+    {
+      "epoch": 5.216,
+      "grad_norm": 0.7490739822387695,
+      "learning_rate": 4.186674669867948e-05,
+      "loss": 0.6438,
+      "step": 4075
+    },
+    {
+      "epoch": 5.21728,
+      "grad_norm": 0.7887585759162903,
+      "learning_rate": 4.1864745898359345e-05,
+      "loss": 0.5891,
+      "step": 4076
+    },
+    {
+      "epoch": 5.21856,
+      "grad_norm": 0.7423108220100403,
+      "learning_rate": 4.1862745098039217e-05,
+      "loss": 0.6278,
+      "step": 4077
+    },
+    {
+      "epoch": 5.21984,
+      "grad_norm": 0.747302234172821,
+      "learning_rate": 4.186074429771909e-05,
+      "loss": 0.6105,
+      "step": 4078
+    },
+    {
+      "epoch": 5.22112,
+      "grad_norm": 0.7417611479759216,
+      "learning_rate": 4.185874349739896e-05,
+      "loss": 0.6077,
+      "step": 4079
+    },
+    {
+      "epoch": 5.2224,
+      "grad_norm": 0.7022831439971924,
+      "learning_rate": 4.185674269707883e-05,
+      "loss": 0.5837,
+      "step": 4080
+    },
+    {
+      "epoch": 5.22368,
+      "grad_norm": 0.7441087365150452,
+      "learning_rate": 4.1854741896758704e-05,
+      "loss": 0.6113,
+      "step": 4081
+    },
+    {
+      "epoch": 5.22496,
+      "grad_norm": 0.7969503998756409,
+      "learning_rate": 4.185274109643858e-05,
+      "loss": 0.6077,
+      "step": 4082
+    },
+    {
+      "epoch": 5.22624,
+      "grad_norm": 0.7303327918052673,
+      "learning_rate": 4.1850740296118454e-05,
+      "loss": 0.5847,
+      "step": 4083
+    },
+    {
+      "epoch": 5.22752,
+      "grad_norm": 0.7135294675827026,
+      "learning_rate": 4.184873949579832e-05,
+      "loss": 0.5898,
+      "step": 4084
+    },
+    {
+      "epoch": 5.2288,
+      "grad_norm": 0.6808853149414062,
+      "learning_rate": 4.184673869547819e-05,
+      "loss": 0.6139,
+      "step": 4085
+    },
+    {
+      "epoch": 5.23008,
+      "grad_norm": 0.7298513054847717,
+      "learning_rate": 4.184473789515806e-05,
+      "loss": 0.6338,
+      "step": 4086
+    },
+    {
+      "epoch": 5.2313600000000005,
+      "grad_norm": 0.7220539450645447,
+      "learning_rate": 4.1842737094837935e-05,
+      "loss": 0.5868,
+      "step": 4087
+    },
+    {
+      "epoch": 5.23264,
+      "grad_norm": 0.7567940354347229,
+      "learning_rate": 4.184073629451781e-05,
+      "loss": 0.626,
+      "step": 4088
+    },
+    {
+      "epoch": 5.23392,
+      "grad_norm": 0.7327399253845215,
+      "learning_rate": 4.183873549419768e-05,
+      "loss": 0.6274,
+      "step": 4089
+    },
+    {
+      "epoch": 5.2352,
+      "grad_norm": 0.7534685134887695,
+      "learning_rate": 4.183673469387756e-05,
+      "loss": 0.6031,
+      "step": 4090
+    },
+    {
+      "epoch": 5.23648,
+      "grad_norm": 0.7302762269973755,
+      "learning_rate": 4.183473389355743e-05,
+      "loss": 0.5695,
+      "step": 4091
+    },
+    {
+      "epoch": 5.23776,
+      "grad_norm": 0.737391471862793,
+      "learning_rate": 4.1832733093237294e-05,
+      "loss": 0.6894,
+      "step": 4092
+    },
+    {
+      "epoch": 5.23904,
+      "grad_norm": 0.7553057670593262,
+      "learning_rate": 4.1830732292917166e-05,
+      "loss": 0.6557,
+      "step": 4093
+    },
+    {
+      "epoch": 5.24032,
+      "grad_norm": 0.7426884770393372,
+      "learning_rate": 4.182873149259704e-05,
+      "loss": 0.6648,
+      "step": 4094
+    },
+    {
+      "epoch": 5.2416,
+      "grad_norm": 0.7494625449180603,
+      "learning_rate": 4.182673069227691e-05,
+      "loss": 0.6473,
+      "step": 4095
+    },
+    {
+      "epoch": 5.24288,
+      "grad_norm": 0.6882435083389282,
+      "learning_rate": 4.182472989195678e-05,
+      "loss": 0.5835,
+      "step": 4096
+    },
+    {
+      "epoch": 5.24416,
+      "grad_norm": 0.7787857055664062,
+      "learning_rate": 4.182272909163666e-05,
+      "loss": 0.6566,
+      "step": 4097
+    },
+    {
+      "epoch": 5.24544,
+      "grad_norm": 0.8027628660202026,
+      "learning_rate": 4.182072829131653e-05,
+      "loss": 0.6464,
+      "step": 4098
+    },
+    {
+      "epoch": 5.24672,
+      "grad_norm": 0.7405392527580261,
+      "learning_rate": 4.1818727490996404e-05,
+      "loss": 0.6006,
+      "step": 4099
+    },
+    {
+      "epoch": 5.248,
+      "grad_norm": 0.7443488836288452,
+      "learning_rate": 4.181672669067627e-05,
+      "loss": 0.646,
+      "step": 4100
+    },
+    {
+      "epoch": 5.24928,
+      "grad_norm": 0.8020368218421936,
+      "learning_rate": 4.181472589035614e-05,
+      "loss": 0.6962,
+      "step": 4101
+    },
+    {
+      "epoch": 5.25056,
+      "grad_norm": 0.7345948219299316,
+      "learning_rate": 4.181272509003601e-05,
+      "loss": 0.6171,
+      "step": 4102
+    },
+    {
+      "epoch": 5.25184,
+      "grad_norm": 0.7310644388198853,
+      "learning_rate": 4.1810724289715885e-05,
+      "loss": 0.6516,
+      "step": 4103
+    },
+    {
+      "epoch": 5.25312,
+      "grad_norm": 0.6996403336524963,
+      "learning_rate": 4.1808723489395763e-05,
+      "loss": 0.5719,
+      "step": 4104
+    },
+    {
+      "epoch": 5.2544,
+      "grad_norm": 0.683502197265625,
+      "learning_rate": 4.1806722689075635e-05,
+      "loss": 0.5978,
+      "step": 4105
+    },
+    {
+      "epoch": 5.25568,
+      "grad_norm": 0.7105028629302979,
+      "learning_rate": 4.180472188875551e-05,
+      "loss": 0.6489,
+      "step": 4106
+    },
+    {
+      "epoch": 5.25696,
+      "grad_norm": 0.7716310620307922,
+      "learning_rate": 4.180272108843538e-05,
+      "loss": 0.6468,
+      "step": 4107
+    },
+    {
+      "epoch": 5.25824,
+      "grad_norm": 0.7370600700378418,
+      "learning_rate": 4.1800720288115244e-05,
+      "loss": 0.5911,
+      "step": 4108
+    },
+    {
+      "epoch": 5.25952,
+      "grad_norm": 0.7595857381820679,
+      "learning_rate": 4.1798719487795116e-05,
+      "loss": 0.6627,
+      "step": 4109
+    },
+    {
+      "epoch": 5.2608,
+      "grad_norm": 0.7154218554496765,
+      "learning_rate": 4.179671868747499e-05,
+      "loss": 0.6439,
+      "step": 4110
+    },
+    {
+      "epoch": 5.26208,
+      "grad_norm": 0.7370497584342957,
+      "learning_rate": 4.1794717887154866e-05,
+      "loss": 0.6511,
+      "step": 4111
+    },
+    {
+      "epoch": 5.26336,
+      "grad_norm": 0.7522472143173218,
+      "learning_rate": 4.179271708683474e-05,
+      "loss": 0.6171,
+      "step": 4112
+    },
+    {
+      "epoch": 5.26464,
+      "grad_norm": 0.7183180451393127,
+      "learning_rate": 4.179071628651461e-05,
+      "loss": 0.6296,
+      "step": 4113
+    },
+    {
+      "epoch": 5.26592,
+      "grad_norm": 0.7467809319496155,
+      "learning_rate": 4.178871548619448e-05,
+      "loss": 0.6705,
+      "step": 4114
+    },
+    {
+      "epoch": 5.2672,
+      "grad_norm": 0.7425380349159241,
+      "learning_rate": 4.1786714685874354e-05,
+      "loss": 0.6152,
+      "step": 4115
+    },
+    {
+      "epoch": 5.26848,
+      "grad_norm": 0.729131281375885,
+      "learning_rate": 4.178471388555422e-05,
+      "loss": 0.6485,
+      "step": 4116
+    },
+    {
+      "epoch": 5.26976,
+      "grad_norm": 0.7531429529190063,
+      "learning_rate": 4.178271308523409e-05,
+      "loss": 0.702,
+      "step": 4117
+    },
+    {
+      "epoch": 5.27104,
+      "grad_norm": 0.6844903826713562,
+      "learning_rate": 4.178071228491397e-05,
+      "loss": 0.5557,
+      "step": 4118
+    },
+    {
+      "epoch": 5.27232,
+      "grad_norm": 0.6600164175033569,
+      "learning_rate": 4.177871148459384e-05,
+      "loss": 0.5779,
+      "step": 4119
+    },
+    {
+      "epoch": 5.2736,
+      "grad_norm": 0.7316297888755798,
+      "learning_rate": 4.177671068427371e-05,
+      "loss": 0.6198,
+      "step": 4120
+    },
+    {
+      "epoch": 5.27488,
+      "grad_norm": 0.7616431713104248,
+      "learning_rate": 4.1774709883953585e-05,
+      "loss": 0.6581,
+      "step": 4121
+    },
+    {
+      "epoch": 5.27616,
+      "grad_norm": 0.7071192264556885,
+      "learning_rate": 4.177270908363346e-05,
+      "loss": 0.616,
+      "step": 4122
+    },
+    {
+      "epoch": 5.27744,
+      "grad_norm": 0.717705249786377,
+      "learning_rate": 4.177070828331333e-05,
+      "loss": 0.6081,
+      "step": 4123
+    },
+    {
+      "epoch": 5.27872,
+      "grad_norm": 0.7297554612159729,
+      "learning_rate": 4.1768707482993194e-05,
+      "loss": 0.6139,
+      "step": 4124
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.7143238186836243,
+      "learning_rate": 4.176670668267307e-05,
+      "loss": 0.6375,
+      "step": 4125
+    },
+    {
+      "epoch": 5.28128,
+      "grad_norm": 0.701789915561676,
+      "learning_rate": 4.1764705882352944e-05,
+      "loss": 0.5887,
+      "step": 4126
+    },
+    {
+      "epoch": 5.28256,
+      "grad_norm": 0.7739312052726746,
+      "learning_rate": 4.1762705082032816e-05,
+      "loss": 0.6815,
+      "step": 4127
+    },
+    {
+      "epoch": 5.28384,
+      "grad_norm": 0.714019238948822,
+      "learning_rate": 4.176070428171269e-05,
+      "loss": 0.6148,
+      "step": 4128
+    },
+    {
+      "epoch": 5.28512,
+      "grad_norm": 0.7301841378211975,
+      "learning_rate": 4.175870348139256e-05,
+      "loss": 0.6268,
+      "step": 4129
+    },
+    {
+      "epoch": 5.2864,
+      "grad_norm": 0.7254041433334351,
+      "learning_rate": 4.175670268107243e-05,
+      "loss": 0.6933,
+      "step": 4130
+    },
+    {
+      "epoch": 5.28768,
+      "grad_norm": 0.7699323892593384,
+      "learning_rate": 4.1754701880752304e-05,
+      "loss": 0.7007,
+      "step": 4131
+    },
+    {
+      "epoch": 5.28896,
+      "grad_norm": 0.7059085965156555,
+      "learning_rate": 4.1752701080432175e-05,
+      "loss": 0.6194,
+      "step": 4132
+    },
+    {
+      "epoch": 5.29024,
+      "grad_norm": 0.7336937785148621,
+      "learning_rate": 4.175070028011205e-05,
+      "loss": 0.6656,
+      "step": 4133
+    },
+    {
+      "epoch": 5.29152,
+      "grad_norm": 0.7493408918380737,
+      "learning_rate": 4.174869947979192e-05,
+      "loss": 0.6766,
+      "step": 4134
+    },
+    {
+      "epoch": 5.2928,
+      "grad_norm": 0.7297886610031128,
+      "learning_rate": 4.174669867947179e-05,
+      "loss": 0.6352,
+      "step": 4135
+    },
+    {
+      "epoch": 5.29408,
+      "grad_norm": 0.7195264101028442,
+      "learning_rate": 4.174469787915166e-05,
+      "loss": 0.6102,
+      "step": 4136
+    },
+    {
+      "epoch": 5.29536,
+      "grad_norm": 0.7068727612495422,
+      "learning_rate": 4.1742697078831535e-05,
+      "loss": 0.556,
+      "step": 4137
+    },
+    {
+      "epoch": 5.29664,
+      "grad_norm": 0.7307595014572144,
+      "learning_rate": 4.1740696278511407e-05,
+      "loss": 0.6653,
+      "step": 4138
+    },
+    {
+      "epoch": 5.29792,
+      "grad_norm": 0.7298409938812256,
+      "learning_rate": 4.173869547819128e-05,
+      "loss": 0.6587,
+      "step": 4139
+    },
+    {
+      "epoch": 5.2992,
+      "grad_norm": 0.7297065258026123,
+      "learning_rate": 4.173669467787115e-05,
+      "loss": 0.63,
+      "step": 4140
+    },
+    {
+      "epoch": 5.30048,
+      "grad_norm": 0.7633817791938782,
+      "learning_rate": 4.173469387755102e-05,
+      "loss": 0.6555,
+      "step": 4141
+    },
+    {
+      "epoch": 5.30176,
+      "grad_norm": 0.7700405716896057,
+      "learning_rate": 4.1732693077230894e-05,
+      "loss": 0.6576,
+      "step": 4142
+    },
+    {
+      "epoch": 5.30304,
+      "grad_norm": 0.7559324502944946,
+      "learning_rate": 4.1730692276910766e-05,
+      "loss": 0.6633,
+      "step": 4143
+    },
+    {
+      "epoch": 5.30432,
+      "grad_norm": 0.782576322555542,
+      "learning_rate": 4.172869147659064e-05,
+      "loss": 0.639,
+      "step": 4144
+    },
+    {
+      "epoch": 5.3056,
+      "grad_norm": 0.7763882875442505,
+      "learning_rate": 4.172669067627051e-05,
+      "loss": 0.6591,
+      "step": 4145
+    },
+    {
+      "epoch": 5.30688,
+      "grad_norm": 0.7644296288490295,
+      "learning_rate": 4.172468987595038e-05,
+      "loss": 0.5824,
+      "step": 4146
+    },
+    {
+      "epoch": 5.30816,
+      "grad_norm": 0.7676565647125244,
+      "learning_rate": 4.172268907563025e-05,
+      "loss": 0.6401,
+      "step": 4147
+    },
+    {
+      "epoch": 5.30944,
+      "grad_norm": 0.7719744443893433,
+      "learning_rate": 4.1720688275310125e-05,
+      "loss": 0.701,
+      "step": 4148
+    },
+    {
+      "epoch": 5.31072,
+      "grad_norm": 0.7409470677375793,
+      "learning_rate": 4.171868747499e-05,
+      "loss": 0.596,
+      "step": 4149
+    },
+    {
+      "epoch": 5.312,
+      "grad_norm": 0.7712857723236084,
+      "learning_rate": 4.171668667466987e-05,
+      "loss": 0.6428,
+      "step": 4150
+    },
+    {
+      "epoch": 5.31328,
+      "grad_norm": 0.8165924549102783,
+      "learning_rate": 4.171468587434974e-05,
+      "loss": 0.6534,
+      "step": 4151
+    },
+    {
+      "epoch": 5.31456,
+      "grad_norm": 0.7560482025146484,
+      "learning_rate": 4.171268507402961e-05,
+      "loss": 0.6532,
+      "step": 4152
+    },
+    {
+      "epoch": 5.31584,
+      "grad_norm": 0.740051805973053,
+      "learning_rate": 4.171068427370949e-05,
+      "loss": 0.5908,
+      "step": 4153
+    },
+    {
+      "epoch": 5.31712,
+      "grad_norm": 0.7237354516983032,
+      "learning_rate": 4.1708683473389356e-05,
+      "loss": 0.6267,
+      "step": 4154
+    },
+    {
+      "epoch": 5.3184000000000005,
+      "grad_norm": 0.7441720962524414,
+      "learning_rate": 4.170668267306923e-05,
+      "loss": 0.5965,
+      "step": 4155
+    },
+    {
+      "epoch": 5.31968,
+      "grad_norm": 0.770770788192749,
+      "learning_rate": 4.17046818727491e-05,
+      "loss": 0.6421,
+      "step": 4156
+    },
+    {
+      "epoch": 5.32096,
+      "grad_norm": 0.7334791421890259,
+      "learning_rate": 4.170268107242897e-05,
+      "loss": 0.6324,
+      "step": 4157
+    },
+    {
+      "epoch": 5.32224,
+      "grad_norm": 0.7998023629188538,
+      "learning_rate": 4.1700680272108844e-05,
+      "loss": 0.6521,
+      "step": 4158
+    },
+    {
+      "epoch": 5.32352,
+      "grad_norm": 0.7167679071426392,
+      "learning_rate": 4.1698679471788716e-05,
+      "loss": 0.6568,
+      "step": 4159
+    },
+    {
+      "epoch": 5.3248,
+      "grad_norm": 0.6926052570343018,
+      "learning_rate": 4.1696678671468594e-05,
+      "loss": 0.5946,
+      "step": 4160
+    },
+    {
+      "epoch": 5.32608,
+      "grad_norm": 0.704468846321106,
+      "learning_rate": 4.1694677871148466e-05,
+      "loss": 0.5969,
+      "step": 4161
+    },
+    {
+      "epoch": 5.32736,
+      "grad_norm": 0.7154397368431091,
+      "learning_rate": 4.169267707082833e-05,
+      "loss": 0.6305,
+      "step": 4162
+    },
+    {
+      "epoch": 5.32864,
+      "grad_norm": 0.7278268337249756,
+      "learning_rate": 4.16906762705082e-05,
+      "loss": 0.6246,
+      "step": 4163
+    },
+    {
+      "epoch": 5.3299199999999995,
+      "grad_norm": 0.7837611436843872,
+      "learning_rate": 4.1688675470188075e-05,
+      "loss": 0.6799,
+      "step": 4164
+    },
+    {
+      "epoch": 5.3312,
+      "grad_norm": 0.7822878360748291,
+      "learning_rate": 4.168667466986795e-05,
+      "loss": 0.6387,
+      "step": 4165
+    },
+    {
+      "epoch": 5.33248,
+      "grad_norm": 0.7381898760795593,
+      "learning_rate": 4.168467386954782e-05,
+      "loss": 0.6071,
+      "step": 4166
+    },
+    {
+      "epoch": 5.33376,
+      "grad_norm": 0.7611213326454163,
+      "learning_rate": 4.16826730692277e-05,
+      "loss": 0.5542,
+      "step": 4167
+    },
+    {
+      "epoch": 5.33504,
+      "grad_norm": 0.7357604503631592,
+      "learning_rate": 4.168067226890757e-05,
+      "loss": 0.6474,
+      "step": 4168
+    },
+    {
+      "epoch": 5.33632,
+      "grad_norm": 0.7721554636955261,
+      "learning_rate": 4.167867146858744e-05,
+      "loss": 0.61,
+      "step": 4169
+    },
+    {
+      "epoch": 5.3376,
+      "grad_norm": 0.7680994868278503,
+      "learning_rate": 4.1676670668267306e-05,
+      "loss": 0.672,
+      "step": 4170
+    },
+    {
+      "epoch": 5.33888,
+      "grad_norm": 0.7215784192085266,
+      "learning_rate": 4.167466986794718e-05,
+      "loss": 0.6394,
+      "step": 4171
+    },
+    {
+      "epoch": 5.34016,
+      "grad_norm": 0.7135915160179138,
+      "learning_rate": 4.167266906762705e-05,
+      "loss": 0.6101,
+      "step": 4172
+    },
+    {
+      "epoch": 5.34144,
+      "grad_norm": 0.7465267181396484,
+      "learning_rate": 4.167066826730692e-05,
+      "loss": 0.6649,
+      "step": 4173
+    },
+    {
+      "epoch": 5.34272,
+      "grad_norm": 0.7213309407234192,
+      "learning_rate": 4.16686674669868e-05,
+      "loss": 0.595,
+      "step": 4174
+    },
+    {
+      "epoch": 5.344,
+      "grad_norm": 0.7122322916984558,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.643,
+      "step": 4175
+    },
+    {
+      "epoch": 5.34528,
+      "grad_norm": 0.7451832890510559,
+      "learning_rate": 4.1664665866346544e-05,
+      "loss": 0.6457,
+      "step": 4176
+    },
+    {
+      "epoch": 5.34656,
+      "grad_norm": 0.7480894923210144,
+      "learning_rate": 4.1662665066026416e-05,
+      "loss": 0.6788,
+      "step": 4177
+    },
+    {
+      "epoch": 5.34784,
+      "grad_norm": 0.742680013179779,
+      "learning_rate": 4.166066426570628e-05,
+      "loss": 0.6303,
+      "step": 4178
+    },
+    {
+      "epoch": 5.34912,
+      "grad_norm": 0.7494868636131287,
+      "learning_rate": 4.165866346538615e-05,
+      "loss": 0.6294,
+      "step": 4179
+    },
+    {
+      "epoch": 5.3504,
+      "grad_norm": 0.7331695556640625,
+      "learning_rate": 4.1656662665066025e-05,
+      "loss": 0.6088,
+      "step": 4180
+    },
+    {
+      "epoch": 5.35168,
+      "grad_norm": 0.7193648219108582,
+      "learning_rate": 4.16546618647459e-05,
+      "loss": 0.624,
+      "step": 4181
+    },
+    {
+      "epoch": 5.35296,
+      "grad_norm": 0.7384651303291321,
+      "learning_rate": 4.1652661064425775e-05,
+      "loss": 0.6541,
+      "step": 4182
+    },
+    {
+      "epoch": 5.35424,
+      "grad_norm": 0.7119219899177551,
+      "learning_rate": 4.165066026410565e-05,
+      "loss": 0.5827,
+      "step": 4183
+    },
+    {
+      "epoch": 5.35552,
+      "grad_norm": 0.7343322038650513,
+      "learning_rate": 4.164865946378552e-05,
+      "loss": 0.5935,
+      "step": 4184
+    },
+    {
+      "epoch": 5.3568,
+      "grad_norm": 0.7368735671043396,
+      "learning_rate": 4.164665866346539e-05,
+      "loss": 0.6512,
+      "step": 4185
+    },
+    {
+      "epoch": 5.35808,
+      "grad_norm": 0.7362673282623291,
+      "learning_rate": 4.1644657863145256e-05,
+      "loss": 0.5961,
+      "step": 4186
+    },
+    {
+      "epoch": 5.35936,
+      "grad_norm": 0.7422880530357361,
+      "learning_rate": 4.164265706282513e-05,
+      "loss": 0.5847,
+      "step": 4187
+    },
+    {
+      "epoch": 5.36064,
+      "grad_norm": 0.7207329869270325,
+      "learning_rate": 4.1640656262505006e-05,
+      "loss": 0.645,
+      "step": 4188
+    },
+    {
+      "epoch": 5.36192,
+      "grad_norm": 0.7351892590522766,
+      "learning_rate": 4.163865546218488e-05,
+      "loss": 0.6344,
+      "step": 4189
+    },
+    {
+      "epoch": 5.3632,
+      "grad_norm": 0.7149160504341125,
+      "learning_rate": 4.163665466186475e-05,
+      "loss": 0.6381,
+      "step": 4190
+    },
+    {
+      "epoch": 5.36448,
+      "grad_norm": 0.7507062554359436,
+      "learning_rate": 4.163465386154462e-05,
+      "loss": 0.6487,
+      "step": 4191
+    },
+    {
+      "epoch": 5.36576,
+      "grad_norm": 0.718169629573822,
+      "learning_rate": 4.1632653061224494e-05,
+      "loss": 0.5903,
+      "step": 4192
+    },
+    {
+      "epoch": 5.36704,
+      "grad_norm": 0.6719139218330383,
+      "learning_rate": 4.1630652260904365e-05,
+      "loss": 0.5836,
+      "step": 4193
+    },
+    {
+      "epoch": 5.36832,
+      "grad_norm": 0.704992413520813,
+      "learning_rate": 4.162865146058423e-05,
+      "loss": 0.6264,
+      "step": 4194
+    },
+    {
+      "epoch": 5.3696,
+      "grad_norm": 0.7397636771202087,
+      "learning_rate": 4.16266506602641e-05,
+      "loss": 0.6152,
+      "step": 4195
+    },
+    {
+      "epoch": 5.37088,
+      "grad_norm": 0.7924968600273132,
+      "learning_rate": 4.162464985994398e-05,
+      "loss": 0.6927,
+      "step": 4196
+    },
+    {
+      "epoch": 5.37216,
+      "grad_norm": 0.7544575929641724,
+      "learning_rate": 4.162264905962385e-05,
+      "loss": 0.6515,
+      "step": 4197
+    },
+    {
+      "epoch": 5.37344,
+      "grad_norm": 0.743186891078949,
+      "learning_rate": 4.1620648259303725e-05,
+      "loss": 0.5953,
+      "step": 4198
+    },
+    {
+      "epoch": 5.37472,
+      "grad_norm": 0.708406925201416,
+      "learning_rate": 4.16186474589836e-05,
+      "loss": 0.6146,
+      "step": 4199
+    },
+    {
+      "epoch": 5.376,
+      "grad_norm": 0.7445898652076721,
+      "learning_rate": 4.161664665866347e-05,
+      "loss": 0.6632,
+      "step": 4200
+    },
+    {
+      "epoch": 5.37728,
+      "grad_norm": 0.7561652660369873,
+      "learning_rate": 4.161464585834334e-05,
+      "loss": 0.678,
+      "step": 4201
+    },
+    {
+      "epoch": 5.37856,
+      "grad_norm": 0.7312771081924438,
+      "learning_rate": 4.1612645058023205e-05,
+      "loss": 0.5748,
+      "step": 4202
+    },
+    {
+      "epoch": 5.37984,
+      "grad_norm": 0.7716453671455383,
+      "learning_rate": 4.1610644257703084e-05,
+      "loss": 0.6249,
+      "step": 4203
+    },
+    {
+      "epoch": 5.38112,
+      "grad_norm": 0.7312607765197754,
+      "learning_rate": 4.1608643457382956e-05,
+      "loss": 0.66,
+      "step": 4204
+    },
+    {
+      "epoch": 5.3824,
+      "grad_norm": 0.7225040197372437,
+      "learning_rate": 4.160664265706283e-05,
+      "loss": 0.6272,
+      "step": 4205
+    },
+    {
+      "epoch": 5.38368,
+      "grad_norm": 0.7225370407104492,
+      "learning_rate": 4.16046418567427e-05,
+      "loss": 0.6611,
+      "step": 4206
+    },
+    {
+      "epoch": 5.38496,
+      "grad_norm": 0.7278191447257996,
+      "learning_rate": 4.160264105642257e-05,
+      "loss": 0.562,
+      "step": 4207
+    },
+    {
+      "epoch": 5.38624,
+      "grad_norm": 0.7607426047325134,
+      "learning_rate": 4.160064025610244e-05,
+      "loss": 0.6992,
+      "step": 4208
+    },
+    {
+      "epoch": 5.38752,
+      "grad_norm": 0.7173566818237305,
+      "learning_rate": 4.1598639455782315e-05,
+      "loss": 0.6655,
+      "step": 4209
+    },
+    {
+      "epoch": 5.3888,
+      "grad_norm": 0.6946642398834229,
+      "learning_rate": 4.159663865546219e-05,
+      "loss": 0.5802,
+      "step": 4210
+    },
+    {
+      "epoch": 5.39008,
+      "grad_norm": 0.7513164281845093,
+      "learning_rate": 4.159463785514206e-05,
+      "loss": 0.6538,
+      "step": 4211
+    },
+    {
+      "epoch": 5.39136,
+      "grad_norm": 0.7534035444259644,
+      "learning_rate": 4.159263705482193e-05,
+      "loss": 0.6571,
+      "step": 4212
+    },
+    {
+      "epoch": 5.39264,
+      "grad_norm": 0.7139204740524292,
+      "learning_rate": 4.15906362545018e-05,
+      "loss": 0.5496,
+      "step": 4213
+    },
+    {
+      "epoch": 5.39392,
+      "grad_norm": 0.7388843894004822,
+      "learning_rate": 4.1588635454181674e-05,
+      "loss": 0.6338,
+      "step": 4214
+    },
+    {
+      "epoch": 5.3952,
+      "grad_norm": 0.7447206974029541,
+      "learning_rate": 4.1586634653861546e-05,
+      "loss": 0.5863,
+      "step": 4215
+    },
+    {
+      "epoch": 5.39648,
+      "grad_norm": 0.7396455407142639,
+      "learning_rate": 4.158463385354142e-05,
+      "loss": 0.6237,
+      "step": 4216
+    },
+    {
+      "epoch": 5.39776,
+      "grad_norm": 0.7606258392333984,
+      "learning_rate": 4.158263305322129e-05,
+      "loss": 0.66,
+      "step": 4217
+    },
+    {
+      "epoch": 5.39904,
+      "grad_norm": 0.8394240140914917,
+      "learning_rate": 4.158063225290116e-05,
+      "loss": 0.6537,
+      "step": 4218
+    },
+    {
+      "epoch": 5.40032,
+      "grad_norm": 0.7464375495910645,
+      "learning_rate": 4.1578631452581034e-05,
+      "loss": 0.6092,
+      "step": 4219
+    },
+    {
+      "epoch": 5.4016,
+      "grad_norm": 0.7182469367980957,
+      "learning_rate": 4.1576630652260906e-05,
+      "loss": 0.6298,
+      "step": 4220
+    },
+    {
+      "epoch": 5.40288,
+      "grad_norm": 0.772256076335907,
+      "learning_rate": 4.157462985194078e-05,
+      "loss": 0.6736,
+      "step": 4221
+    },
+    {
+      "epoch": 5.40416,
+      "grad_norm": 0.7597767114639282,
+      "learning_rate": 4.157262905162065e-05,
+      "loss": 0.6706,
+      "step": 4222
+    },
+    {
+      "epoch": 5.4054400000000005,
+      "grad_norm": 0.7585070729255676,
+      "learning_rate": 4.157062825130052e-05,
+      "loss": 0.6311,
+      "step": 4223
+    },
+    {
+      "epoch": 5.40672,
+      "grad_norm": 0.7339140772819519,
+      "learning_rate": 4.156862745098039e-05,
+      "loss": 0.6276,
+      "step": 4224
+    },
+    {
+      "epoch": 5.408,
+      "grad_norm": 0.7652080059051514,
+      "learning_rate": 4.1566626650660265e-05,
+      "loss": 0.6573,
+      "step": 4225
+    },
+    {
+      "epoch": 5.40928,
+      "grad_norm": 0.7553825974464417,
+      "learning_rate": 4.156462585034014e-05,
+      "loss": 0.6619,
+      "step": 4226
+    },
+    {
+      "epoch": 5.41056,
+      "grad_norm": 0.753115177154541,
+      "learning_rate": 4.156262505002001e-05,
+      "loss": 0.6916,
+      "step": 4227
+    },
+    {
+      "epoch": 5.41184,
+      "grad_norm": 0.7845631241798401,
+      "learning_rate": 4.156062424969988e-05,
+      "loss": 0.6693,
+      "step": 4228
+    },
+    {
+      "epoch": 5.41312,
+      "grad_norm": 0.7271414995193481,
+      "learning_rate": 4.155862344937975e-05,
+      "loss": 0.5909,
+      "step": 4229
+    },
+    {
+      "epoch": 5.4144,
+      "grad_norm": 0.7683008313179016,
+      "learning_rate": 4.1556622649059624e-05,
+      "loss": 0.617,
+      "step": 4230
+    },
+    {
+      "epoch": 5.41568,
+      "grad_norm": 0.7922500967979431,
+      "learning_rate": 4.15546218487395e-05,
+      "loss": 0.6502,
+      "step": 4231
+    },
+    {
+      "epoch": 5.4169599999999996,
+      "grad_norm": 0.7644626498222351,
+      "learning_rate": 4.155262104841937e-05,
+      "loss": 0.6364,
+      "step": 4232
+    },
+    {
+      "epoch": 5.41824,
+      "grad_norm": 0.7368267774581909,
+      "learning_rate": 4.155062024809924e-05,
+      "loss": 0.6217,
+      "step": 4233
+    },
+    {
+      "epoch": 5.41952,
+      "grad_norm": 0.7373183369636536,
+      "learning_rate": 4.154861944777911e-05,
+      "loss": 0.6186,
+      "step": 4234
+    },
+    {
+      "epoch": 5.4208,
+      "grad_norm": 0.7321687936782837,
+      "learning_rate": 4.1546618647458983e-05,
+      "loss": 0.6367,
+      "step": 4235
+    },
+    {
+      "epoch": 5.42208,
+      "grad_norm": 0.7199650406837463,
+      "learning_rate": 4.1544617847138855e-05,
+      "loss": 0.5642,
+      "step": 4236
+    },
+    {
+      "epoch": 5.42336,
+      "grad_norm": 0.7745632529258728,
+      "learning_rate": 4.154261704681873e-05,
+      "loss": 0.6663,
+      "step": 4237
+    },
+    {
+      "epoch": 5.42464,
+      "grad_norm": 0.7231164574623108,
+      "learning_rate": 4.1540616246498606e-05,
+      "loss": 0.5956,
+      "step": 4238
+    },
+    {
+      "epoch": 5.42592,
+      "grad_norm": 0.7279819250106812,
+      "learning_rate": 4.153861544617848e-05,
+      "loss": 0.6507,
+      "step": 4239
+    },
+    {
+      "epoch": 5.4272,
+      "grad_norm": 0.7030388116836548,
+      "learning_rate": 4.153661464585834e-05,
+      "loss": 0.616,
+      "step": 4240
+    },
+    {
+      "epoch": 5.42848,
+      "grad_norm": 0.7477138042449951,
+      "learning_rate": 4.1534613845538215e-05,
+      "loss": 0.6875,
+      "step": 4241
+    },
+    {
+      "epoch": 5.42976,
+      "grad_norm": 0.6878876686096191,
+      "learning_rate": 4.1532613045218086e-05,
+      "loss": 0.5856,
+      "step": 4242
+    },
+    {
+      "epoch": 5.43104,
+      "grad_norm": 0.7027300596237183,
+      "learning_rate": 4.153061224489796e-05,
+      "loss": 0.6155,
+      "step": 4243
+    },
+    {
+      "epoch": 5.43232,
+      "grad_norm": 0.7319774627685547,
+      "learning_rate": 4.152861144457783e-05,
+      "loss": 0.6515,
+      "step": 4244
+    },
+    {
+      "epoch": 5.4336,
+      "grad_norm": 0.7408891916275024,
+      "learning_rate": 4.152661064425771e-05,
+      "loss": 0.6834,
+      "step": 4245
+    },
+    {
+      "epoch": 5.43488,
+      "grad_norm": 0.7682157754898071,
+      "learning_rate": 4.152460984393758e-05,
+      "loss": 0.6244,
+      "step": 4246
+    },
+    {
+      "epoch": 5.43616,
+      "grad_norm": 0.7219270467758179,
+      "learning_rate": 4.152260904361745e-05,
+      "loss": 0.631,
+      "step": 4247
+    },
+    {
+      "epoch": 5.43744,
+      "grad_norm": 0.7537044882774353,
+      "learning_rate": 4.152060824329732e-05,
+      "loss": 0.6456,
+      "step": 4248
+    },
+    {
+      "epoch": 5.43872,
+      "grad_norm": 0.7409535050392151,
+      "learning_rate": 4.151860744297719e-05,
+      "loss": 0.6145,
+      "step": 4249
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.7161388993263245,
+      "learning_rate": 4.151660664265706e-05,
+      "loss": 0.608,
+      "step": 4250
+    },
+    {
+      "epoch": 5.44128,
+      "grad_norm": 0.7184132933616638,
+      "learning_rate": 4.151460584233693e-05,
+      "loss": 0.6225,
+      "step": 4251
+    },
+    {
+      "epoch": 5.44256,
+      "grad_norm": 0.7418175935745239,
+      "learning_rate": 4.151260504201681e-05,
+      "loss": 0.6434,
+      "step": 4252
+    },
+    {
+      "epoch": 5.44384,
+      "grad_norm": 0.791469395160675,
+      "learning_rate": 4.1510604241696684e-05,
+      "loss": 0.7016,
+      "step": 4253
+    },
+    {
+      "epoch": 5.44512,
+      "grad_norm": 0.7191916108131409,
+      "learning_rate": 4.1508603441376556e-05,
+      "loss": 0.5732,
+      "step": 4254
+    },
+    {
+      "epoch": 5.4464,
+      "grad_norm": 0.723224937915802,
+      "learning_rate": 4.150660264105643e-05,
+      "loss": 0.6206,
+      "step": 4255
+    },
+    {
+      "epoch": 5.44768,
+      "grad_norm": 0.7462792992591858,
+      "learning_rate": 4.150460184073629e-05,
+      "loss": 0.6278,
+      "step": 4256
+    },
+    {
+      "epoch": 5.44896,
+      "grad_norm": 0.7277225255966187,
+      "learning_rate": 4.1502601040416164e-05,
+      "loss": 0.6239,
+      "step": 4257
+    },
+    {
+      "epoch": 5.45024,
+      "grad_norm": 0.7746177911758423,
+      "learning_rate": 4.1500600240096036e-05,
+      "loss": 0.6549,
+      "step": 4258
+    },
+    {
+      "epoch": 5.45152,
+      "grad_norm": 0.7281146049499512,
+      "learning_rate": 4.1498599439775915e-05,
+      "loss": 0.6261,
+      "step": 4259
+    },
+    {
+      "epoch": 5.4528,
+      "grad_norm": 0.751801609992981,
+      "learning_rate": 4.149659863945579e-05,
+      "loss": 0.6514,
+      "step": 4260
+    },
+    {
+      "epoch": 5.45408,
+      "grad_norm": 0.7374839782714844,
+      "learning_rate": 4.149459783913566e-05,
+      "loss": 0.6498,
+      "step": 4261
+    },
+    {
+      "epoch": 5.45536,
+      "grad_norm": 0.7305448055267334,
+      "learning_rate": 4.149259703881553e-05,
+      "loss": 0.5913,
+      "step": 4262
+    },
+    {
+      "epoch": 5.45664,
+      "grad_norm": 0.7178943753242493,
+      "learning_rate": 4.14905962384954e-05,
+      "loss": 0.6412,
+      "step": 4263
+    },
+    {
+      "epoch": 5.45792,
+      "grad_norm": 0.7705517411231995,
+      "learning_rate": 4.148859543817527e-05,
+      "loss": 0.6451,
+      "step": 4264
+    },
+    {
+      "epoch": 5.4592,
+      "grad_norm": 0.7343666553497314,
+      "learning_rate": 4.148659463785514e-05,
+      "loss": 0.632,
+      "step": 4265
+    },
+    {
+      "epoch": 5.4604800000000004,
+      "grad_norm": 0.7479756474494934,
+      "learning_rate": 4.148459383753502e-05,
+      "loss": 0.6502,
+      "step": 4266
+    },
+    {
+      "epoch": 5.46176,
+      "grad_norm": 0.7750630378723145,
+      "learning_rate": 4.148259303721489e-05,
+      "loss": 0.6519,
+      "step": 4267
+    },
+    {
+      "epoch": 5.46304,
+      "grad_norm": 0.7741780281066895,
+      "learning_rate": 4.148059223689476e-05,
+      "loss": 0.6251,
+      "step": 4268
+    },
+    {
+      "epoch": 5.46432,
+      "grad_norm": 0.7677576541900635,
+      "learning_rate": 4.147859143657463e-05,
+      "loss": 0.6463,
+      "step": 4269
+    },
+    {
+      "epoch": 5.4656,
+      "grad_norm": 0.7549638748168945,
+      "learning_rate": 4.1476590636254505e-05,
+      "loss": 0.6795,
+      "step": 4270
+    },
+    {
+      "epoch": 5.46688,
+      "grad_norm": 0.7305397391319275,
+      "learning_rate": 4.147458983593438e-05,
+      "loss": 0.6243,
+      "step": 4271
+    },
+    {
+      "epoch": 5.46816,
+      "grad_norm": 0.7237127423286438,
+      "learning_rate": 4.147258903561424e-05,
+      "loss": 0.6224,
+      "step": 4272
+    },
+    {
+      "epoch": 5.46944,
+      "grad_norm": 0.6945486068725586,
+      "learning_rate": 4.147058823529412e-05,
+      "loss": 0.6099,
+      "step": 4273
+    },
+    {
+      "epoch": 5.47072,
+      "grad_norm": 0.7361446619033813,
+      "learning_rate": 4.146858743497399e-05,
+      "loss": 0.5984,
+      "step": 4274
+    },
+    {
+      "epoch": 5.4719999999999995,
+      "grad_norm": 0.6917027235031128,
+      "learning_rate": 4.1466586634653865e-05,
+      "loss": 0.5837,
+      "step": 4275
+    },
+    {
+      "epoch": 5.47328,
+      "grad_norm": 0.7172991633415222,
+      "learning_rate": 4.1464585834333736e-05,
+      "loss": 0.6568,
+      "step": 4276
+    },
+    {
+      "epoch": 5.47456,
+      "grad_norm": 0.6892058253288269,
+      "learning_rate": 4.146258503401361e-05,
+      "loss": 0.6219,
+      "step": 4277
+    },
+    {
+      "epoch": 5.47584,
+      "grad_norm": 0.7407313585281372,
+      "learning_rate": 4.146058423369348e-05,
+      "loss": 0.6509,
+      "step": 4278
+    },
+    {
+      "epoch": 5.47712,
+      "grad_norm": 0.7310953736305237,
+      "learning_rate": 4.145858343337335e-05,
+      "loss": 0.6227,
+      "step": 4279
+    },
+    {
+      "epoch": 5.4784,
+      "grad_norm": 0.7653133273124695,
+      "learning_rate": 4.1456582633053224e-05,
+      "loss": 0.6519,
+      "step": 4280
+    },
+    {
+      "epoch": 5.47968,
+      "grad_norm": 0.7355585098266602,
+      "learning_rate": 4.1454581832733096e-05,
+      "loss": 0.6163,
+      "step": 4281
+    },
+    {
+      "epoch": 5.48096,
+      "grad_norm": 0.7199105620384216,
+      "learning_rate": 4.145258103241297e-05,
+      "loss": 0.6104,
+      "step": 4282
+    },
+    {
+      "epoch": 5.48224,
+      "grad_norm": 0.7551198601722717,
+      "learning_rate": 4.145058023209284e-05,
+      "loss": 0.6599,
+      "step": 4283
+    },
+    {
+      "epoch": 5.48352,
+      "grad_norm": 0.7141267657279968,
+      "learning_rate": 4.144857943177271e-05,
+      "loss": 0.5802,
+      "step": 4284
+    },
+    {
+      "epoch": 5.4848,
+      "grad_norm": 0.7375865578651428,
+      "learning_rate": 4.144657863145258e-05,
+      "loss": 0.6395,
+      "step": 4285
+    },
+    {
+      "epoch": 5.48608,
+      "grad_norm": 0.7291966676712036,
+      "learning_rate": 4.1444577831132455e-05,
+      "loss": 0.6074,
+      "step": 4286
+    },
+    {
+      "epoch": 5.48736,
+      "grad_norm": 0.7387133836746216,
+      "learning_rate": 4.144257703081233e-05,
+      "loss": 0.6659,
+      "step": 4287
+    },
+    {
+      "epoch": 5.48864,
+      "grad_norm": 0.7290977835655212,
+      "learning_rate": 4.14405762304922e-05,
+      "loss": 0.5999,
+      "step": 4288
+    },
+    {
+      "epoch": 5.48992,
+      "grad_norm": 0.7163506746292114,
+      "learning_rate": 4.143857543017207e-05,
+      "loss": 0.5901,
+      "step": 4289
+    },
+    {
+      "epoch": 5.4912,
+      "grad_norm": 0.7519278526306152,
+      "learning_rate": 4.143657462985194e-05,
+      "loss": 0.6746,
+      "step": 4290
+    },
+    {
+      "epoch": 5.49248,
+      "grad_norm": 0.7682675123214722,
+      "learning_rate": 4.1434573829531814e-05,
+      "loss": 0.6694,
+      "step": 4291
+    },
+    {
+      "epoch": 5.49376,
+      "grad_norm": 0.7485939264297485,
+      "learning_rate": 4.1432573029211686e-05,
+      "loss": 0.6073,
+      "step": 4292
+    },
+    {
+      "epoch": 5.49504,
+      "grad_norm": 0.7766785621643066,
+      "learning_rate": 4.143057222889156e-05,
+      "loss": 0.5799,
+      "step": 4293
+    },
+    {
+      "epoch": 5.49632,
+      "grad_norm": 0.7536609768867493,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 0.6169,
+      "step": 4294
+    },
+    {
+      "epoch": 5.4976,
+      "grad_norm": 0.725071907043457,
+      "learning_rate": 4.14265706282513e-05,
+      "loss": 0.6393,
+      "step": 4295
+    },
+    {
+      "epoch": 5.49888,
+      "grad_norm": 0.7046527862548828,
+      "learning_rate": 4.1424569827931173e-05,
+      "loss": 0.5988,
+      "step": 4296
+    },
+    {
+      "epoch": 5.50016,
+      "grad_norm": 0.7888898849487305,
+      "learning_rate": 4.1422569027611045e-05,
+      "loss": 0.6408,
+      "step": 4297
+    },
+    {
+      "epoch": 5.50144,
+      "grad_norm": 0.7429561614990234,
+      "learning_rate": 4.142056822729092e-05,
+      "loss": 0.6463,
+      "step": 4298
+    },
+    {
+      "epoch": 5.50272,
+      "grad_norm": 0.7440699934959412,
+      "learning_rate": 4.141856742697079e-05,
+      "loss": 0.652,
+      "step": 4299
+    },
+    {
+      "epoch": 5.504,
+      "grad_norm": 0.7715908885002136,
+      "learning_rate": 4.141656662665066e-05,
+      "loss": 0.6796,
+      "step": 4300
+    },
+    {
+      "epoch": 5.50528,
+      "grad_norm": 0.714020311832428,
+      "learning_rate": 4.141456582633054e-05,
+      "loss": 0.5743,
+      "step": 4301
+    },
+    {
+      "epoch": 5.50656,
+      "grad_norm": 0.7244009375572205,
+      "learning_rate": 4.141256502601041e-05,
+      "loss": 0.6498,
+      "step": 4302
+    },
+    {
+      "epoch": 5.50784,
+      "grad_norm": 0.7662729620933533,
+      "learning_rate": 4.1410564225690276e-05,
+      "loss": 0.6456,
+      "step": 4303
+    },
+    {
+      "epoch": 5.50912,
+      "grad_norm": 0.6930511593818665,
+      "learning_rate": 4.140856342537015e-05,
+      "loss": 0.5915,
+      "step": 4304
+    },
+    {
+      "epoch": 5.5104,
+      "grad_norm": 0.7424132227897644,
+      "learning_rate": 4.140656262505002e-05,
+      "loss": 0.6082,
+      "step": 4305
+    },
+    {
+      "epoch": 5.51168,
+      "grad_norm": 0.7727630734443665,
+      "learning_rate": 4.140456182472989e-05,
+      "loss": 0.6735,
+      "step": 4306
+    },
+    {
+      "epoch": 5.51296,
+      "grad_norm": 0.6813573837280273,
+      "learning_rate": 4.1402561024409764e-05,
+      "loss": 0.5995,
+      "step": 4307
+    },
+    {
+      "epoch": 5.51424,
+      "grad_norm": 0.7176674008369446,
+      "learning_rate": 4.1400560224089636e-05,
+      "loss": 0.6404,
+      "step": 4308
+    },
+    {
+      "epoch": 5.51552,
+      "grad_norm": 0.74003666639328,
+      "learning_rate": 4.1398559423769514e-05,
+      "loss": 0.6219,
+      "step": 4309
+    },
+    {
+      "epoch": 5.5168,
+      "grad_norm": 0.705366313457489,
+      "learning_rate": 4.1396558623449386e-05,
+      "loss": 0.6144,
+      "step": 4310
+    },
+    {
+      "epoch": 5.51808,
+      "grad_norm": 0.70579594373703,
+      "learning_rate": 4.139455782312925e-05,
+      "loss": 0.6345,
+      "step": 4311
+    },
+    {
+      "epoch": 5.51936,
+      "grad_norm": 0.776077926158905,
+      "learning_rate": 4.139255702280912e-05,
+      "loss": 0.6859,
+      "step": 4312
+    },
+    {
+      "epoch": 5.52064,
+      "grad_norm": 0.7283969521522522,
+      "learning_rate": 4.1390556222488995e-05,
+      "loss": 0.5956,
+      "step": 4313
+    },
+    {
+      "epoch": 5.52192,
+      "grad_norm": 0.7156528234481812,
+      "learning_rate": 4.138855542216887e-05,
+      "loss": 0.6284,
+      "step": 4314
+    },
+    {
+      "epoch": 5.5232,
+      "grad_norm": 0.7174420356750488,
+      "learning_rate": 4.138655462184874e-05,
+      "loss": 0.5826,
+      "step": 4315
+    },
+    {
+      "epoch": 5.52448,
+      "grad_norm": 0.7152808904647827,
+      "learning_rate": 4.138455382152862e-05,
+      "loss": 0.5983,
+      "step": 4316
+    },
+    {
+      "epoch": 5.52576,
+      "grad_norm": 0.7197295427322388,
+      "learning_rate": 4.138255302120849e-05,
+      "loss": 0.6156,
+      "step": 4317
+    },
+    {
+      "epoch": 5.5270399999999995,
+      "grad_norm": 0.7645232677459717,
+      "learning_rate": 4.138055222088836e-05,
+      "loss": 0.6525,
+      "step": 4318
+    },
+    {
+      "epoch": 5.52832,
+      "grad_norm": 0.6872177720069885,
+      "learning_rate": 4.1378551420568226e-05,
+      "loss": 0.6044,
+      "step": 4319
+    },
+    {
+      "epoch": 5.5296,
+      "grad_norm": 0.7118226289749146,
+      "learning_rate": 4.13765506202481e-05,
+      "loss": 0.6176,
+      "step": 4320
+    },
+    {
+      "epoch": 5.53088,
+      "grad_norm": 0.7357558608055115,
+      "learning_rate": 4.137454981992797e-05,
+      "loss": 0.6369,
+      "step": 4321
+    },
+    {
+      "epoch": 5.53216,
+      "grad_norm": 0.7684040069580078,
+      "learning_rate": 4.137254901960784e-05,
+      "loss": 0.639,
+      "step": 4322
+    },
+    {
+      "epoch": 5.53344,
+      "grad_norm": 0.7693250179290771,
+      "learning_rate": 4.137054821928772e-05,
+      "loss": 0.6801,
+      "step": 4323
+    },
+    {
+      "epoch": 5.53472,
+      "grad_norm": 0.6828309893608093,
+      "learning_rate": 4.136854741896759e-05,
+      "loss": 0.6232,
+      "step": 4324
+    },
+    {
+      "epoch": 5.536,
+      "grad_norm": 0.7307888865470886,
+      "learning_rate": 4.1366546618647464e-05,
+      "loss": 0.5854,
+      "step": 4325
+    },
+    {
+      "epoch": 5.53728,
+      "grad_norm": 0.7778929471969604,
+      "learning_rate": 4.1364545818327336e-05,
+      "loss": 0.6335,
+      "step": 4326
+    },
+    {
+      "epoch": 5.53856,
+      "grad_norm": 0.7381669878959656,
+      "learning_rate": 4.13625450180072e-05,
+      "loss": 0.6113,
+      "step": 4327
+    },
+    {
+      "epoch": 5.53984,
+      "grad_norm": 0.7228230834007263,
+      "learning_rate": 4.136054421768707e-05,
+      "loss": 0.6167,
+      "step": 4328
+    },
+    {
+      "epoch": 5.54112,
+      "grad_norm": 0.7991678714752197,
+      "learning_rate": 4.1358543417366945e-05,
+      "loss": 0.6445,
+      "step": 4329
+    },
+    {
+      "epoch": 5.5424,
+      "grad_norm": 0.7551771402359009,
+      "learning_rate": 4.1356542617046823e-05,
+      "loss": 0.6068,
+      "step": 4330
+    },
+    {
+      "epoch": 5.54368,
+      "grad_norm": 0.7122798562049866,
+      "learning_rate": 4.1354541816726695e-05,
+      "loss": 0.6084,
+      "step": 4331
+    },
+    {
+      "epoch": 5.54496,
+      "grad_norm": 0.7480770349502563,
+      "learning_rate": 4.135254101640657e-05,
+      "loss": 0.6652,
+      "step": 4332
+    },
+    {
+      "epoch": 5.54624,
+      "grad_norm": 0.7483392953872681,
+      "learning_rate": 4.135054021608644e-05,
+      "loss": 0.6547,
+      "step": 4333
+    },
+    {
+      "epoch": 5.5475200000000005,
+      "grad_norm": 0.7390307784080505,
+      "learning_rate": 4.134853941576631e-05,
+      "loss": 0.6251,
+      "step": 4334
+    },
+    {
+      "epoch": 5.5488,
+      "grad_norm": 0.7169946432113647,
+      "learning_rate": 4.1346538615446176e-05,
+      "loss": 0.5833,
+      "step": 4335
+    },
+    {
+      "epoch": 5.55008,
+      "grad_norm": 0.799982488155365,
+      "learning_rate": 4.134453781512605e-05,
+      "loss": 0.6722,
+      "step": 4336
+    },
+    {
+      "epoch": 5.55136,
+      "grad_norm": 0.7108681201934814,
+      "learning_rate": 4.1342537014805926e-05,
+      "loss": 0.5692,
+      "step": 4337
+    },
+    {
+      "epoch": 5.55264,
+      "grad_norm": 0.7432281374931335,
+      "learning_rate": 4.13405362144858e-05,
+      "loss": 0.6642,
+      "step": 4338
+    },
+    {
+      "epoch": 5.55392,
+      "grad_norm": 0.7395031452178955,
+      "learning_rate": 4.133853541416567e-05,
+      "loss": 0.6601,
+      "step": 4339
+    },
+    {
+      "epoch": 5.5552,
+      "grad_norm": 0.7047814130783081,
+      "learning_rate": 4.133653461384554e-05,
+      "loss": 0.6566,
+      "step": 4340
+    },
+    {
+      "epoch": 5.55648,
+      "grad_norm": 0.7198272347450256,
+      "learning_rate": 4.1334533813525414e-05,
+      "loss": 0.6494,
+      "step": 4341
+    },
+    {
+      "epoch": 5.55776,
+      "grad_norm": 0.690862238407135,
+      "learning_rate": 4.1332533013205286e-05,
+      "loss": 0.5978,
+      "step": 4342
+    },
+    {
+      "epoch": 5.5590399999999995,
+      "grad_norm": 0.7230415940284729,
+      "learning_rate": 4.133053221288515e-05,
+      "loss": 0.6449,
+      "step": 4343
+    },
+    {
+      "epoch": 5.56032,
+      "grad_norm": 0.7639696002006531,
+      "learning_rate": 4.132853141256503e-05,
+      "loss": 0.659,
+      "step": 4344
+    },
+    {
+      "epoch": 5.5616,
+      "grad_norm": 0.7363530397415161,
+      "learning_rate": 4.13265306122449e-05,
+      "loss": 0.5586,
+      "step": 4345
+    },
+    {
+      "epoch": 5.56288,
+      "grad_norm": 0.7204784750938416,
+      "learning_rate": 4.132452981192477e-05,
+      "loss": 0.6629,
+      "step": 4346
+    },
+    {
+      "epoch": 5.56416,
+      "grad_norm": 0.787613570690155,
+      "learning_rate": 4.1322529011604645e-05,
+      "loss": 0.6596,
+      "step": 4347
+    },
+    {
+      "epoch": 5.56544,
+      "grad_norm": 0.7643619179725647,
+      "learning_rate": 4.132052821128452e-05,
+      "loss": 0.6955,
+      "step": 4348
+    },
+    {
+      "epoch": 5.56672,
+      "grad_norm": 0.7102790474891663,
+      "learning_rate": 4.131852741096439e-05,
+      "loss": 0.6573,
+      "step": 4349
+    },
+    {
+      "epoch": 5.568,
+      "grad_norm": 0.7346861362457275,
+      "learning_rate": 4.131652661064426e-05,
+      "loss": 0.638,
+      "step": 4350
+    },
+    {
+      "epoch": 5.56928,
+      "grad_norm": 0.7016168832778931,
+      "learning_rate": 4.131452581032413e-05,
+      "loss": 0.641,
+      "step": 4351
+    },
+    {
+      "epoch": 5.57056,
+      "grad_norm": 0.7079355120658875,
+      "learning_rate": 4.1312525010004004e-05,
+      "loss": 0.6284,
+      "step": 4352
+    },
+    {
+      "epoch": 5.57184,
+      "grad_norm": 0.7500919699668884,
+      "learning_rate": 4.1310524209683876e-05,
+      "loss": 0.6272,
+      "step": 4353
+    },
+    {
+      "epoch": 5.57312,
+      "grad_norm": 0.708759605884552,
+      "learning_rate": 4.130852340936375e-05,
+      "loss": 0.5711,
+      "step": 4354
+    },
+    {
+      "epoch": 5.5744,
+      "grad_norm": 0.7522294521331787,
+      "learning_rate": 4.130652260904362e-05,
+      "loss": 0.6849,
+      "step": 4355
+    },
+    {
+      "epoch": 5.57568,
+      "grad_norm": 0.7236372828483582,
+      "learning_rate": 4.130452180872349e-05,
+      "loss": 0.612,
+      "step": 4356
+    },
+    {
+      "epoch": 5.57696,
+      "grad_norm": 0.7170467972755432,
+      "learning_rate": 4.1302521008403364e-05,
+      "loss": 0.6623,
+      "step": 4357
+    },
+    {
+      "epoch": 5.57824,
+      "grad_norm": 0.7366024255752563,
+      "learning_rate": 4.1300520208083235e-05,
+      "loss": 0.6025,
+      "step": 4358
+    },
+    {
+      "epoch": 5.5795200000000005,
+      "grad_norm": 0.7273461222648621,
+      "learning_rate": 4.129851940776311e-05,
+      "loss": 0.6107,
+      "step": 4359
+    },
+    {
+      "epoch": 5.5808,
+      "grad_norm": 0.7220586538314819,
+      "learning_rate": 4.129651860744298e-05,
+      "loss": 0.5956,
+      "step": 4360
+    },
+    {
+      "epoch": 5.58208,
+      "grad_norm": 0.6963488459587097,
+      "learning_rate": 4.129451780712285e-05,
+      "loss": 0.6292,
+      "step": 4361
+    },
+    {
+      "epoch": 5.58336,
+      "grad_norm": 0.7539929747581482,
+      "learning_rate": 4.129251700680272e-05,
+      "loss": 0.6307,
+      "step": 4362
+    },
+    {
+      "epoch": 5.58464,
+      "grad_norm": 0.7906867265701294,
+      "learning_rate": 4.1290516206482595e-05,
+      "loss": 0.6993,
+      "step": 4363
+    },
+    {
+      "epoch": 5.58592,
+      "grad_norm": 0.7354369163513184,
+      "learning_rate": 4.1288515406162467e-05,
+      "loss": 0.6817,
+      "step": 4364
+    },
+    {
+      "epoch": 5.5872,
+      "grad_norm": 0.7213559150695801,
+      "learning_rate": 4.128651460584234e-05,
+      "loss": 0.601,
+      "step": 4365
+    },
+    {
+      "epoch": 5.58848,
+      "grad_norm": 0.782337486743927,
+      "learning_rate": 4.128451380552221e-05,
+      "loss": 0.6586,
+      "step": 4366
+    },
+    {
+      "epoch": 5.58976,
+      "grad_norm": 0.720512330532074,
+      "learning_rate": 4.128251300520208e-05,
+      "loss": 0.5993,
+      "step": 4367
+    },
+    {
+      "epoch": 5.59104,
+      "grad_norm": 0.7558559775352478,
+      "learning_rate": 4.1280512204881954e-05,
+      "loss": 0.6544,
+      "step": 4368
+    },
+    {
+      "epoch": 5.59232,
+      "grad_norm": 0.7189001441001892,
+      "learning_rate": 4.1278511404561826e-05,
+      "loss": 0.5955,
+      "step": 4369
+    },
+    {
+      "epoch": 5.5936,
+      "grad_norm": 0.7266587018966675,
+      "learning_rate": 4.12765106042417e-05,
+      "loss": 0.6626,
+      "step": 4370
+    },
+    {
+      "epoch": 5.59488,
+      "grad_norm": 0.7241358160972595,
+      "learning_rate": 4.127450980392157e-05,
+      "loss": 0.5832,
+      "step": 4371
+    },
+    {
+      "epoch": 5.59616,
+      "grad_norm": 0.732850968837738,
+      "learning_rate": 4.127250900360145e-05,
+      "loss": 0.6341,
+      "step": 4372
+    },
+    {
+      "epoch": 5.59744,
+      "grad_norm": 0.7497940063476562,
+      "learning_rate": 4.127050820328131e-05,
+      "loss": 0.6735,
+      "step": 4373
+    },
+    {
+      "epoch": 5.59872,
+      "grad_norm": 0.7972435355186462,
+      "learning_rate": 4.1268507402961185e-05,
+      "loss": 0.6529,
+      "step": 4374
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.7178497314453125,
+      "learning_rate": 4.126650660264106e-05,
+      "loss": 0.5903,
+      "step": 4375
+    },
+    {
+      "epoch": 5.60128,
+      "grad_norm": 0.7551731467247009,
+      "learning_rate": 4.126450580232093e-05,
+      "loss": 0.6286,
+      "step": 4376
+    },
+    {
+      "epoch": 5.60256,
+      "grad_norm": 0.7458837032318115,
+      "learning_rate": 4.12625050020008e-05,
+      "loss": 0.6696,
+      "step": 4377
+    },
+    {
+      "epoch": 5.60384,
+      "grad_norm": 0.7230226993560791,
+      "learning_rate": 4.126050420168067e-05,
+      "loss": 0.6121,
+      "step": 4378
+    },
+    {
+      "epoch": 5.60512,
+      "grad_norm": 0.714648425579071,
+      "learning_rate": 4.125850340136055e-05,
+      "loss": 0.6175,
+      "step": 4379
+    },
+    {
+      "epoch": 5.6064,
+      "grad_norm": 0.713045060634613,
+      "learning_rate": 4.125650260104042e-05,
+      "loss": 0.6181,
+      "step": 4380
+    },
+    {
+      "epoch": 5.60768,
+      "grad_norm": 0.7239891886711121,
+      "learning_rate": 4.125450180072029e-05,
+      "loss": 0.6454,
+      "step": 4381
+    },
+    {
+      "epoch": 5.60896,
+      "grad_norm": 0.7390968799591064,
+      "learning_rate": 4.125250100040016e-05,
+      "loss": 0.598,
+      "step": 4382
+    },
+    {
+      "epoch": 5.61024,
+      "grad_norm": 0.7643166780471802,
+      "learning_rate": 4.125050020008003e-05,
+      "loss": 0.6299,
+      "step": 4383
+    },
+    {
+      "epoch": 5.61152,
+      "grad_norm": 0.7632148861885071,
+      "learning_rate": 4.1248499399759904e-05,
+      "loss": 0.6601,
+      "step": 4384
+    },
+    {
+      "epoch": 5.6128,
+      "grad_norm": 0.74639493227005,
+      "learning_rate": 4.1246498599439776e-05,
+      "loss": 0.7099,
+      "step": 4385
+    },
+    {
+      "epoch": 5.6140799999999995,
+      "grad_norm": 0.7428092956542969,
+      "learning_rate": 4.1244497799119654e-05,
+      "loss": 0.6596,
+      "step": 4386
+    },
+    {
+      "epoch": 5.61536,
+      "grad_norm": 0.7230287790298462,
+      "learning_rate": 4.1242496998799526e-05,
+      "loss": 0.603,
+      "step": 4387
+    },
+    {
+      "epoch": 5.61664,
+      "grad_norm": 0.7466754913330078,
+      "learning_rate": 4.12404961984794e-05,
+      "loss": 0.702,
+      "step": 4388
+    },
+    {
+      "epoch": 5.61792,
+      "grad_norm": 0.7360090017318726,
+      "learning_rate": 4.123849539815926e-05,
+      "loss": 0.6036,
+      "step": 4389
+    },
+    {
+      "epoch": 5.6192,
+      "grad_norm": 0.7444376945495605,
+      "learning_rate": 4.1236494597839135e-05,
+      "loss": 0.6554,
+      "step": 4390
+    },
+    {
+      "epoch": 5.62048,
+      "grad_norm": 0.7407775521278381,
+      "learning_rate": 4.123449379751901e-05,
+      "loss": 0.6132,
+      "step": 4391
+    },
+    {
+      "epoch": 5.62176,
+      "grad_norm": 0.6796741485595703,
+      "learning_rate": 4.123249299719888e-05,
+      "loss": 0.5356,
+      "step": 4392
+    },
+    {
+      "epoch": 5.62304,
+      "grad_norm": 0.7613147497177124,
+      "learning_rate": 4.123049219687876e-05,
+      "loss": 0.6522,
+      "step": 4393
+    },
+    {
+      "epoch": 5.62432,
+      "grad_norm": 0.7880995273590088,
+      "learning_rate": 4.122849139655863e-05,
+      "loss": 0.6211,
+      "step": 4394
+    },
+    {
+      "epoch": 5.6256,
+      "grad_norm": 0.7633781433105469,
+      "learning_rate": 4.12264905962385e-05,
+      "loss": 0.6276,
+      "step": 4395
+    },
+    {
+      "epoch": 5.62688,
+      "grad_norm": 0.7847978472709656,
+      "learning_rate": 4.122448979591837e-05,
+      "loss": 0.6798,
+      "step": 4396
+    },
+    {
+      "epoch": 5.62816,
+      "grad_norm": 0.7588475346565247,
+      "learning_rate": 4.122248899559824e-05,
+      "loss": 0.642,
+      "step": 4397
+    },
+    {
+      "epoch": 5.62944,
+      "grad_norm": 0.7453870177268982,
+      "learning_rate": 4.122048819527811e-05,
+      "loss": 0.623,
+      "step": 4398
+    },
+    {
+      "epoch": 5.63072,
+      "grad_norm": 0.7075188755989075,
+      "learning_rate": 4.121848739495798e-05,
+      "loss": 0.6164,
+      "step": 4399
+    },
+    {
+      "epoch": 5.632,
+      "grad_norm": 0.7136649489402771,
+      "learning_rate": 4.121648659463786e-05,
+      "loss": 0.6288,
+      "step": 4400
+    },
+    {
+      "epoch": 5.63328,
+      "grad_norm": 0.7545991539955139,
+      "learning_rate": 4.121448579431773e-05,
+      "loss": 0.6107,
+      "step": 4401
+    },
+    {
+      "epoch": 5.6345600000000005,
+      "grad_norm": 0.7157647013664246,
+      "learning_rate": 4.1212484993997604e-05,
+      "loss": 0.6143,
+      "step": 4402
+    },
+    {
+      "epoch": 5.63584,
+      "grad_norm": 0.6989755630493164,
+      "learning_rate": 4.1210484193677476e-05,
+      "loss": 0.5937,
+      "step": 4403
+    },
+    {
+      "epoch": 5.63712,
+      "grad_norm": 0.7509260177612305,
+      "learning_rate": 4.120848339335735e-05,
+      "loss": 0.6746,
+      "step": 4404
+    },
+    {
+      "epoch": 5.6384,
+      "grad_norm": 0.7333345413208008,
+      "learning_rate": 4.120648259303721e-05,
+      "loss": 0.6164,
+      "step": 4405
+    },
+    {
+      "epoch": 5.63968,
+      "grad_norm": 0.739132285118103,
+      "learning_rate": 4.1204481792717085e-05,
+      "loss": 0.6531,
+      "step": 4406
+    },
+    {
+      "epoch": 5.64096,
+      "grad_norm": 0.7224278450012207,
+      "learning_rate": 4.120248099239696e-05,
+      "loss": 0.5819,
+      "step": 4407
+    },
+    {
+      "epoch": 5.64224,
+      "grad_norm": 0.7341471910476685,
+      "learning_rate": 4.1200480192076835e-05,
+      "loss": 0.607,
+      "step": 4408
+    },
+    {
+      "epoch": 5.64352,
+      "grad_norm": 0.7280572056770325,
+      "learning_rate": 4.119847939175671e-05,
+      "loss": 0.6057,
+      "step": 4409
+    },
+    {
+      "epoch": 5.6448,
+      "grad_norm": 0.759056568145752,
+      "learning_rate": 4.119647859143658e-05,
+      "loss": 0.629,
+      "step": 4410
+    },
+    {
+      "epoch": 5.6460799999999995,
+      "grad_norm": 0.7521056532859802,
+      "learning_rate": 4.119447779111645e-05,
+      "loss": 0.6231,
+      "step": 4411
+    },
+    {
+      "epoch": 5.64736,
+      "grad_norm": 0.7467377185821533,
+      "learning_rate": 4.119247699079632e-05,
+      "loss": 0.6442,
+      "step": 4412
+    },
+    {
+      "epoch": 5.64864,
+      "grad_norm": 0.7342444658279419,
+      "learning_rate": 4.119047619047619e-05,
+      "loss": 0.6292,
+      "step": 4413
+    },
+    {
+      "epoch": 5.64992,
+      "grad_norm": 0.7450172901153564,
+      "learning_rate": 4.1188475390156066e-05,
+      "loss": 0.5998,
+      "step": 4414
+    },
+    {
+      "epoch": 5.6512,
+      "grad_norm": 0.7140272855758667,
+      "learning_rate": 4.118647458983594e-05,
+      "loss": 0.6113,
+      "step": 4415
+    },
+    {
+      "epoch": 5.65248,
+      "grad_norm": 0.6929570436477661,
+      "learning_rate": 4.118447378951581e-05,
+      "loss": 0.6088,
+      "step": 4416
+    },
+    {
+      "epoch": 5.65376,
+      "grad_norm": 0.7283996939659119,
+      "learning_rate": 4.118247298919568e-05,
+      "loss": 0.6608,
+      "step": 4417
+    },
+    {
+      "epoch": 5.65504,
+      "grad_norm": 0.7235706448554993,
+      "learning_rate": 4.1180472188875554e-05,
+      "loss": 0.653,
+      "step": 4418
+    },
+    {
+      "epoch": 5.65632,
+      "grad_norm": 0.7216994166374207,
+      "learning_rate": 4.1178471388555425e-05,
+      "loss": 0.6264,
+      "step": 4419
+    },
+    {
+      "epoch": 5.6576,
+      "grad_norm": 0.7781568169593811,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 0.6419,
+      "step": 4420
+    },
+    {
+      "epoch": 5.65888,
+      "grad_norm": 0.8012334704399109,
+      "learning_rate": 4.117446978791516e-05,
+      "loss": 0.6591,
+      "step": 4421
+    },
+    {
+      "epoch": 5.66016,
+      "grad_norm": 0.7411988973617554,
+      "learning_rate": 4.117246898759504e-05,
+      "loss": 0.6353,
+      "step": 4422
+    },
+    {
+      "epoch": 5.66144,
+      "grad_norm": 0.7188833355903625,
+      "learning_rate": 4.117046818727491e-05,
+      "loss": 0.5684,
+      "step": 4423
+    },
+    {
+      "epoch": 5.66272,
+      "grad_norm": 0.7834116816520691,
+      "learning_rate": 4.1168467386954785e-05,
+      "loss": 0.6605,
+      "step": 4424
+    },
+    {
+      "epoch": 5.664,
+      "grad_norm": 0.7423588037490845,
+      "learning_rate": 4.1166466586634657e-05,
+      "loss": 0.6259,
+      "step": 4425
+    },
+    {
+      "epoch": 5.66528,
+      "grad_norm": 0.732661783695221,
+      "learning_rate": 4.116446578631453e-05,
+      "loss": 0.5846,
+      "step": 4426
+    },
+    {
+      "epoch": 5.6665600000000005,
+      "grad_norm": 0.6886220574378967,
+      "learning_rate": 4.11624649859944e-05,
+      "loss": 0.5771,
+      "step": 4427
+    },
+    {
+      "epoch": 5.66784,
+      "grad_norm": 0.7514188885688782,
+      "learning_rate": 4.116046418567427e-05,
+      "loss": 0.6466,
+      "step": 4428
+    },
+    {
+      "epoch": 5.66912,
+      "grad_norm": 0.7303938269615173,
+      "learning_rate": 4.1158463385354144e-05,
+      "loss": 0.6405,
+      "step": 4429
+    },
+    {
+      "epoch": 5.6704,
+      "grad_norm": 0.7757360935211182,
+      "learning_rate": 4.1156462585034016e-05,
+      "loss": 0.6433,
+      "step": 4430
+    },
+    {
+      "epoch": 5.67168,
+      "grad_norm": 0.7613855004310608,
+      "learning_rate": 4.115446178471389e-05,
+      "loss": 0.6125,
+      "step": 4431
+    },
+    {
+      "epoch": 5.67296,
+      "grad_norm": 0.739494800567627,
+      "learning_rate": 4.115246098439376e-05,
+      "loss": 0.5834,
+      "step": 4432
+    },
+    {
+      "epoch": 5.67424,
+      "grad_norm": 0.7127769589424133,
+      "learning_rate": 4.115046018407363e-05,
+      "loss": 0.5963,
+      "step": 4433
+    },
+    {
+      "epoch": 5.67552,
+      "grad_norm": 0.7135427594184875,
+      "learning_rate": 4.11484593837535e-05,
+      "loss": 0.6307,
+      "step": 4434
+    },
+    {
+      "epoch": 5.6768,
+      "grad_norm": 0.7762594819068909,
+      "learning_rate": 4.1146458583433375e-05,
+      "loss": 0.666,
+      "step": 4435
+    },
+    {
+      "epoch": 5.67808,
+      "grad_norm": 0.7698591351509094,
+      "learning_rate": 4.114445778311325e-05,
+      "loss": 0.5811,
+      "step": 4436
+    },
+    {
+      "epoch": 5.67936,
+      "grad_norm": 0.7814648151397705,
+      "learning_rate": 4.114245698279312e-05,
+      "loss": 0.6203,
+      "step": 4437
+    },
+    {
+      "epoch": 5.68064,
+      "grad_norm": 0.7443010210990906,
+      "learning_rate": 4.114045618247299e-05,
+      "loss": 0.6657,
+      "step": 4438
+    },
+    {
+      "epoch": 5.68192,
+      "grad_norm": 0.691595196723938,
+      "learning_rate": 4.113845538215286e-05,
+      "loss": 0.6411,
+      "step": 4439
+    },
+    {
+      "epoch": 5.6832,
+      "grad_norm": 0.7466495037078857,
+      "learning_rate": 4.1136454581832734e-05,
+      "loss": 0.6843,
+      "step": 4440
+    },
+    {
+      "epoch": 5.68448,
+      "grad_norm": 0.7032425999641418,
+      "learning_rate": 4.1134453781512606e-05,
+      "loss": 0.5646,
+      "step": 4441
+    },
+    {
+      "epoch": 5.68576,
+      "grad_norm": 0.7595518827438354,
+      "learning_rate": 4.113245298119248e-05,
+      "loss": 0.6562,
+      "step": 4442
+    },
+    {
+      "epoch": 5.68704,
+      "grad_norm": 0.7430517077445984,
+      "learning_rate": 4.113045218087235e-05,
+      "loss": 0.6283,
+      "step": 4443
+    },
+    {
+      "epoch": 5.68832,
+      "grad_norm": 0.7672387957572937,
+      "learning_rate": 4.112845138055222e-05,
+      "loss": 0.6236,
+      "step": 4444
+    },
+    {
+      "epoch": 5.6896,
+      "grad_norm": 0.7632207274436951,
+      "learning_rate": 4.1126450580232094e-05,
+      "loss": 0.6524,
+      "step": 4445
+    },
+    {
+      "epoch": 5.69088,
+      "grad_norm": 0.7333277463912964,
+      "learning_rate": 4.1124449779911966e-05,
+      "loss": 0.6461,
+      "step": 4446
+    },
+    {
+      "epoch": 5.69216,
+      "grad_norm": 0.7459308505058289,
+      "learning_rate": 4.112244897959184e-05,
+      "loss": 0.6702,
+      "step": 4447
+    },
+    {
+      "epoch": 5.69344,
+      "grad_norm": 0.7271379232406616,
+      "learning_rate": 4.112044817927171e-05,
+      "loss": 0.6392,
+      "step": 4448
+    },
+    {
+      "epoch": 5.69472,
+      "grad_norm": 0.7518323659896851,
+      "learning_rate": 4.111844737895158e-05,
+      "loss": 0.6591,
+      "step": 4449
+    },
+    {
+      "epoch": 5.696,
+      "grad_norm": 0.7861558794975281,
+      "learning_rate": 4.111644657863146e-05,
+      "loss": 0.6533,
+      "step": 4450
+    },
+    {
+      "epoch": 5.69728,
+      "grad_norm": 0.7315780520439148,
+      "learning_rate": 4.1114445778311325e-05,
+      "loss": 0.6908,
+      "step": 4451
+    },
+    {
+      "epoch": 5.69856,
+      "grad_norm": 0.7502129673957825,
+      "learning_rate": 4.11124449779912e-05,
+      "loss": 0.6766,
+      "step": 4452
+    },
+    {
+      "epoch": 5.69984,
+      "grad_norm": 0.7338371872901917,
+      "learning_rate": 4.111044417767107e-05,
+      "loss": 0.6115,
+      "step": 4453
+    },
+    {
+      "epoch": 5.7011199999999995,
+      "grad_norm": 0.7228209972381592,
+      "learning_rate": 4.110844337735094e-05,
+      "loss": 0.6202,
+      "step": 4454
+    },
+    {
+      "epoch": 5.7024,
+      "grad_norm": 0.7461691498756409,
+      "learning_rate": 4.110644257703081e-05,
+      "loss": 0.6349,
+      "step": 4455
+    },
+    {
+      "epoch": 5.70368,
+      "grad_norm": 0.7166462540626526,
+      "learning_rate": 4.1104441776710684e-05,
+      "loss": 0.6712,
+      "step": 4456
+    },
+    {
+      "epoch": 5.70496,
+      "grad_norm": 0.7871848940849304,
+      "learning_rate": 4.110244097639056e-05,
+      "loss": 0.6437,
+      "step": 4457
+    },
+    {
+      "epoch": 5.70624,
+      "grad_norm": 0.7596803307533264,
+      "learning_rate": 4.1100440176070435e-05,
+      "loss": 0.6566,
+      "step": 4458
+    },
+    {
+      "epoch": 5.70752,
+      "grad_norm": 0.7356720566749573,
+      "learning_rate": 4.10984393757503e-05,
+      "loss": 0.6138,
+      "step": 4459
+    },
+    {
+      "epoch": 5.7088,
+      "grad_norm": 0.7193202376365662,
+      "learning_rate": 4.109643857543017e-05,
+      "loss": 0.6187,
+      "step": 4460
+    },
+    {
+      "epoch": 5.71008,
+      "grad_norm": 0.7889663577079773,
+      "learning_rate": 4.1094437775110043e-05,
+      "loss": 0.6713,
+      "step": 4461
+    },
+    {
+      "epoch": 5.71136,
+      "grad_norm": 0.7016068696975708,
+      "learning_rate": 4.1092436974789915e-05,
+      "loss": 0.5817,
+      "step": 4462
+    },
+    {
+      "epoch": 5.71264,
+      "grad_norm": 0.7515783905982971,
+      "learning_rate": 4.109043617446979e-05,
+      "loss": 0.6497,
+      "step": 4463
+    },
+    {
+      "epoch": 5.71392,
+      "grad_norm": 0.7173634171485901,
+      "learning_rate": 4.1088435374149666e-05,
+      "loss": 0.5583,
+      "step": 4464
+    },
+    {
+      "epoch": 5.7152,
+      "grad_norm": 0.7629589438438416,
+      "learning_rate": 4.108643457382954e-05,
+      "loss": 0.6278,
+      "step": 4465
+    },
+    {
+      "epoch": 5.71648,
+      "grad_norm": 0.7493870854377747,
+      "learning_rate": 4.108443377350941e-05,
+      "loss": 0.6373,
+      "step": 4466
+    },
+    {
+      "epoch": 5.71776,
+      "grad_norm": 0.7296896576881409,
+      "learning_rate": 4.1082432973189275e-05,
+      "loss": 0.6011,
+      "step": 4467
+    },
+    {
+      "epoch": 5.71904,
+      "grad_norm": 0.7143272757530212,
+      "learning_rate": 4.1080432172869146e-05,
+      "loss": 0.6313,
+      "step": 4468
+    },
+    {
+      "epoch": 5.72032,
+      "grad_norm": 0.7367825508117676,
+      "learning_rate": 4.107843137254902e-05,
+      "loss": 0.6217,
+      "step": 4469
+    },
+    {
+      "epoch": 5.7216000000000005,
+      "grad_norm": 0.7655727863311768,
+      "learning_rate": 4.107643057222889e-05,
+      "loss": 0.6714,
+      "step": 4470
+    },
+    {
+      "epoch": 5.72288,
+      "grad_norm": 0.7177988290786743,
+      "learning_rate": 4.107442977190877e-05,
+      "loss": 0.5762,
+      "step": 4471
+    },
+    {
+      "epoch": 5.72416,
+      "grad_norm": 0.7632842063903809,
+      "learning_rate": 4.107242897158864e-05,
+      "loss": 0.6587,
+      "step": 4472
+    },
+    {
+      "epoch": 5.72544,
+      "grad_norm": 0.7221323251724243,
+      "learning_rate": 4.107042817126851e-05,
+      "loss": 0.6062,
+      "step": 4473
+    },
+    {
+      "epoch": 5.72672,
+      "grad_norm": 0.7250041365623474,
+      "learning_rate": 4.1068427370948384e-05,
+      "loss": 0.609,
+      "step": 4474
+    },
+    {
+      "epoch": 5.728,
+      "grad_norm": 0.7287482619285583,
+      "learning_rate": 4.106642657062825e-05,
+      "loss": 0.6225,
+      "step": 4475
+    },
+    {
+      "epoch": 5.72928,
+      "grad_norm": 0.7629007697105408,
+      "learning_rate": 4.106442577030812e-05,
+      "loss": 0.6546,
+      "step": 4476
+    },
+    {
+      "epoch": 5.73056,
+      "grad_norm": 0.7124512791633606,
+      "learning_rate": 4.106242496998799e-05,
+      "loss": 0.6365,
+      "step": 4477
+    },
+    {
+      "epoch": 5.73184,
+      "grad_norm": 0.7133152484893799,
+      "learning_rate": 4.106042416966787e-05,
+      "loss": 0.6063,
+      "step": 4478
+    },
+    {
+      "epoch": 5.7331199999999995,
+      "grad_norm": 0.7545233964920044,
+      "learning_rate": 4.1058423369347744e-05,
+      "loss": 0.6563,
+      "step": 4479
+    },
+    {
+      "epoch": 5.7344,
+      "grad_norm": 0.755072832107544,
+      "learning_rate": 4.1056422569027615e-05,
+      "loss": 0.6586,
+      "step": 4480
+    },
+    {
+      "epoch": 5.73568,
+      "grad_norm": 0.7463574409484863,
+      "learning_rate": 4.105442176870749e-05,
+      "loss": 0.6427,
+      "step": 4481
+    },
+    {
+      "epoch": 5.73696,
+      "grad_norm": 0.7250651121139526,
+      "learning_rate": 4.105242096838736e-05,
+      "loss": 0.6494,
+      "step": 4482
+    },
+    {
+      "epoch": 5.73824,
+      "grad_norm": 0.7188123464584351,
+      "learning_rate": 4.1050420168067224e-05,
+      "loss": 0.6417,
+      "step": 4483
+    },
+    {
+      "epoch": 5.73952,
+      "grad_norm": 0.758231520652771,
+      "learning_rate": 4.1048419367747096e-05,
+      "loss": 0.6194,
+      "step": 4484
+    },
+    {
+      "epoch": 5.7408,
+      "grad_norm": 0.7206960320472717,
+      "learning_rate": 4.1046418567426975e-05,
+      "loss": 0.6222,
+      "step": 4485
+    },
+    {
+      "epoch": 5.74208,
+      "grad_norm": 0.7939612865447998,
+      "learning_rate": 4.1044417767106847e-05,
+      "loss": 0.6451,
+      "step": 4486
+    },
+    {
+      "epoch": 5.74336,
+      "grad_norm": 0.7785453796386719,
+      "learning_rate": 4.104241696678672e-05,
+      "loss": 0.6497,
+      "step": 4487
+    },
+    {
+      "epoch": 5.74464,
+      "grad_norm": 0.7556130290031433,
+      "learning_rate": 4.104041616646659e-05,
+      "loss": 0.6284,
+      "step": 4488
+    },
+    {
+      "epoch": 5.74592,
+      "grad_norm": 0.6738141775131226,
+      "learning_rate": 4.103841536614646e-05,
+      "loss": 0.6161,
+      "step": 4489
+    },
+    {
+      "epoch": 5.7472,
+      "grad_norm": 0.7317538857460022,
+      "learning_rate": 4.1036414565826334e-05,
+      "loss": 0.6297,
+      "step": 4490
+    },
+    {
+      "epoch": 5.74848,
+      "grad_norm": 0.7428886890411377,
+      "learning_rate": 4.10344137655062e-05,
+      "loss": 0.5963,
+      "step": 4491
+    },
+    {
+      "epoch": 5.74976,
+      "grad_norm": 0.7947216629981995,
+      "learning_rate": 4.103241296518608e-05,
+      "loss": 0.695,
+      "step": 4492
+    },
+    {
+      "epoch": 5.75104,
+      "grad_norm": 0.8010162115097046,
+      "learning_rate": 4.103041216486595e-05,
+      "loss": 0.7321,
+      "step": 4493
+    },
+    {
+      "epoch": 5.75232,
+      "grad_norm": 0.7321368455886841,
+      "learning_rate": 4.102841136454582e-05,
+      "loss": 0.6789,
+      "step": 4494
+    },
+    {
+      "epoch": 5.7536000000000005,
+      "grad_norm": 0.7535264492034912,
+      "learning_rate": 4.102641056422569e-05,
+      "loss": 0.6474,
+      "step": 4495
+    },
+    {
+      "epoch": 5.75488,
+      "grad_norm": 0.7041172385215759,
+      "learning_rate": 4.1024409763905565e-05,
+      "loss": 0.5921,
+      "step": 4496
+    },
+    {
+      "epoch": 5.75616,
+      "grad_norm": 0.7587586641311646,
+      "learning_rate": 4.102240896358544e-05,
+      "loss": 0.6006,
+      "step": 4497
+    },
+    {
+      "epoch": 5.75744,
+      "grad_norm": 0.7653579115867615,
+      "learning_rate": 4.102040816326531e-05,
+      "loss": 0.6526,
+      "step": 4498
+    },
+    {
+      "epoch": 5.75872,
+      "grad_norm": 0.7497640252113342,
+      "learning_rate": 4.101840736294518e-05,
+      "loss": 0.6286,
+      "step": 4499
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.7529040575027466,
+      "learning_rate": 4.101640656262505e-05,
+      "loss": 0.6229,
+      "step": 4500
+    },
+    {
+      "epoch": 5.76128,
+      "grad_norm": 0.7931720018386841,
+      "learning_rate": 4.1014405762304924e-05,
+      "loss": 0.6423,
+      "step": 4501
+    },
+    {
+      "epoch": 5.76256,
+      "grad_norm": 0.7556605935096741,
+      "learning_rate": 4.1012404961984796e-05,
+      "loss": 0.5925,
+      "step": 4502
+    },
+    {
+      "epoch": 5.76384,
+      "grad_norm": 0.7954467535018921,
+      "learning_rate": 4.101040416166467e-05,
+      "loss": 0.6246,
+      "step": 4503
+    },
+    {
+      "epoch": 5.76512,
+      "grad_norm": 0.7651516199111938,
+      "learning_rate": 4.100840336134454e-05,
+      "loss": 0.6147,
+      "step": 4504
+    },
+    {
+      "epoch": 5.7664,
+      "grad_norm": 0.758766770362854,
+      "learning_rate": 4.100640256102441e-05,
+      "loss": 0.6635,
+      "step": 4505
+    },
+    {
+      "epoch": 5.76768,
+      "grad_norm": 0.7414392828941345,
+      "learning_rate": 4.1004401760704284e-05,
+      "loss": 0.6027,
+      "step": 4506
+    },
+    {
+      "epoch": 5.76896,
+      "grad_norm": 0.7388911247253418,
+      "learning_rate": 4.1002400960384156e-05,
+      "loss": 0.6582,
+      "step": 4507
+    },
+    {
+      "epoch": 5.77024,
+      "grad_norm": 0.7597686052322388,
+      "learning_rate": 4.100040016006403e-05,
+      "loss": 0.6223,
+      "step": 4508
+    },
+    {
+      "epoch": 5.77152,
+      "grad_norm": 0.7913456559181213,
+      "learning_rate": 4.09983993597439e-05,
+      "loss": 0.6457,
+      "step": 4509
+    },
+    {
+      "epoch": 5.7728,
+      "grad_norm": 0.7759582996368408,
+      "learning_rate": 4.099639855942377e-05,
+      "loss": 0.6386,
+      "step": 4510
+    },
+    {
+      "epoch": 5.77408,
+      "grad_norm": 0.7403820753097534,
+      "learning_rate": 4.099439775910364e-05,
+      "loss": 0.6217,
+      "step": 4511
+    },
+    {
+      "epoch": 5.77536,
+      "grad_norm": 0.7305609583854675,
+      "learning_rate": 4.0992396958783515e-05,
+      "loss": 0.6371,
+      "step": 4512
+    },
+    {
+      "epoch": 5.77664,
+      "grad_norm": 0.7057348489761353,
+      "learning_rate": 4.099039615846339e-05,
+      "loss": 0.62,
+      "step": 4513
+    },
+    {
+      "epoch": 5.77792,
+      "grad_norm": 0.7573639750480652,
+      "learning_rate": 4.098839535814326e-05,
+      "loss": 0.639,
+      "step": 4514
+    },
+    {
+      "epoch": 5.7792,
+      "grad_norm": 0.7069675326347351,
+      "learning_rate": 4.098639455782313e-05,
+      "loss": 0.5887,
+      "step": 4515
+    },
+    {
+      "epoch": 5.78048,
+      "grad_norm": 0.7522568702697754,
+      "learning_rate": 4.0984393757503e-05,
+      "loss": 0.6604,
+      "step": 4516
+    },
+    {
+      "epoch": 5.78176,
+      "grad_norm": 0.7098150253295898,
+      "learning_rate": 4.0982392957182874e-05,
+      "loss": 0.6172,
+      "step": 4517
+    },
+    {
+      "epoch": 5.78304,
+      "grad_norm": 0.7658361792564392,
+      "learning_rate": 4.0980392156862746e-05,
+      "loss": 0.7318,
+      "step": 4518
+    },
+    {
+      "epoch": 5.78432,
+      "grad_norm": 0.7216033935546875,
+      "learning_rate": 4.097839135654262e-05,
+      "loss": 0.6195,
+      "step": 4519
+    },
+    {
+      "epoch": 5.7856,
+      "grad_norm": 0.7166069149971008,
+      "learning_rate": 4.0976390556222497e-05,
+      "loss": 0.5746,
+      "step": 4520
+    },
+    {
+      "epoch": 5.78688,
+      "grad_norm": 0.7510432004928589,
+      "learning_rate": 4.097438975590236e-05,
+      "loss": 0.6373,
+      "step": 4521
+    },
+    {
+      "epoch": 5.7881599999999995,
+      "grad_norm": 0.700448751449585,
+      "learning_rate": 4.0972388955582233e-05,
+      "loss": 0.5698,
+      "step": 4522
+    },
+    {
+      "epoch": 5.78944,
+      "grad_norm": 0.7070586681365967,
+      "learning_rate": 4.0970388155262105e-05,
+      "loss": 0.6247,
+      "step": 4523
+    },
+    {
+      "epoch": 5.79072,
+      "grad_norm": 0.6846281886100769,
+      "learning_rate": 4.096838735494198e-05,
+      "loss": 0.625,
+      "step": 4524
+    },
+    {
+      "epoch": 5.792,
+      "grad_norm": 0.745806872844696,
+      "learning_rate": 4.096638655462185e-05,
+      "loss": 0.6629,
+      "step": 4525
+    },
+    {
+      "epoch": 5.79328,
+      "grad_norm": 0.7132704854011536,
+      "learning_rate": 4.096438575430172e-05,
+      "loss": 0.6222,
+      "step": 4526
+    },
+    {
+      "epoch": 5.79456,
+      "grad_norm": 0.751122236251831,
+      "learning_rate": 4.09623849539816e-05,
+      "loss": 0.6406,
+      "step": 4527
+    },
+    {
+      "epoch": 5.79584,
+      "grad_norm": 0.7574059367179871,
+      "learning_rate": 4.096038415366147e-05,
+      "loss": 0.6449,
+      "step": 4528
+    },
+    {
+      "epoch": 5.79712,
+      "grad_norm": 0.6792533993721008,
+      "learning_rate": 4.0958383353341336e-05,
+      "loss": 0.5987,
+      "step": 4529
+    },
+    {
+      "epoch": 5.7984,
+      "grad_norm": 0.6878370046615601,
+      "learning_rate": 4.095638255302121e-05,
+      "loss": 0.5901,
+      "step": 4530
+    },
+    {
+      "epoch": 5.79968,
+      "grad_norm": 0.7586196660995483,
+      "learning_rate": 4.095438175270108e-05,
+      "loss": 0.6344,
+      "step": 4531
+    },
+    {
+      "epoch": 5.80096,
+      "grad_norm": 0.7338054180145264,
+      "learning_rate": 4.095238095238095e-05,
+      "loss": 0.6223,
+      "step": 4532
+    },
+    {
+      "epoch": 5.80224,
+      "grad_norm": 0.7303552627563477,
+      "learning_rate": 4.0950380152060824e-05,
+      "loss": 0.5915,
+      "step": 4533
+    },
+    {
+      "epoch": 5.80352,
+      "grad_norm": 0.7648298740386963,
+      "learning_rate": 4.0948379351740696e-05,
+      "loss": 0.615,
+      "step": 4534
+    },
+    {
+      "epoch": 5.8048,
+      "grad_norm": 0.8030759692192078,
+      "learning_rate": 4.0946378551420574e-05,
+      "loss": 0.6604,
+      "step": 4535
+    },
+    {
+      "epoch": 5.80608,
+      "grad_norm": 0.8049156069755554,
+      "learning_rate": 4.0944377751100446e-05,
+      "loss": 0.7182,
+      "step": 4536
+    },
+    {
+      "epoch": 5.80736,
+      "grad_norm": 0.7284945845603943,
+      "learning_rate": 4.094237695078031e-05,
+      "loss": 0.6061,
+      "step": 4537
+    },
+    {
+      "epoch": 5.8086400000000005,
+      "grad_norm": 0.6970748901367188,
+      "learning_rate": 4.094037615046018e-05,
+      "loss": 0.6016,
+      "step": 4538
+    },
+    {
+      "epoch": 5.80992,
+      "grad_norm": 0.7719080448150635,
+      "learning_rate": 4.0938375350140055e-05,
+      "loss": 0.7058,
+      "step": 4539
+    },
+    {
+      "epoch": 5.8112,
+      "grad_norm": 0.7578266859054565,
+      "learning_rate": 4.093637454981993e-05,
+      "loss": 0.6276,
+      "step": 4540
+    },
+    {
+      "epoch": 5.81248,
+      "grad_norm": 0.7241922616958618,
+      "learning_rate": 4.09343737494998e-05,
+      "loss": 0.6167,
+      "step": 4541
+    },
+    {
+      "epoch": 5.81376,
+      "grad_norm": 0.8004947304725647,
+      "learning_rate": 4.093237294917968e-05,
+      "loss": 0.7253,
+      "step": 4542
+    },
+    {
+      "epoch": 5.81504,
+      "grad_norm": 0.7189851999282837,
+      "learning_rate": 4.093037214885955e-05,
+      "loss": 0.6462,
+      "step": 4543
+    },
+    {
+      "epoch": 5.81632,
+      "grad_norm": 0.7319324016571045,
+      "learning_rate": 4.092837134853942e-05,
+      "loss": 0.6575,
+      "step": 4544
+    },
+    {
+      "epoch": 5.8176,
+      "grad_norm": 0.7121530175209045,
+      "learning_rate": 4.0926370548219286e-05,
+      "loss": 0.6186,
+      "step": 4545
+    },
+    {
+      "epoch": 5.81888,
+      "grad_norm": 0.7266112565994263,
+      "learning_rate": 4.092436974789916e-05,
+      "loss": 0.6121,
+      "step": 4546
+    },
+    {
+      "epoch": 5.82016,
+      "grad_norm": 0.7524951696395874,
+      "learning_rate": 4.092236894757903e-05,
+      "loss": 0.6035,
+      "step": 4547
+    },
+    {
+      "epoch": 5.82144,
+      "grad_norm": 0.7526006698608398,
+      "learning_rate": 4.09203681472589e-05,
+      "loss": 0.6494,
+      "step": 4548
+    },
+    {
+      "epoch": 5.82272,
+      "grad_norm": 0.7238689064979553,
+      "learning_rate": 4.091836734693878e-05,
+      "loss": 0.6186,
+      "step": 4549
+    },
+    {
+      "epoch": 5.824,
+      "grad_norm": 0.7418443560600281,
+      "learning_rate": 4.091636654661865e-05,
+      "loss": 0.6611,
+      "step": 4550
+    },
+    {
+      "epoch": 5.82528,
+      "grad_norm": 0.7511699795722961,
+      "learning_rate": 4.0914365746298524e-05,
+      "loss": 0.6564,
+      "step": 4551
+    },
+    {
+      "epoch": 5.82656,
+      "grad_norm": 0.6964154243469238,
+      "learning_rate": 4.0912364945978396e-05,
+      "loss": 0.6026,
+      "step": 4552
+    },
+    {
+      "epoch": 5.82784,
+      "grad_norm": 0.7538017630577087,
+      "learning_rate": 4.091036414565826e-05,
+      "loss": 0.7241,
+      "step": 4553
+    },
+    {
+      "epoch": 5.82912,
+      "grad_norm": 0.7111532092094421,
+      "learning_rate": 4.090836334533813e-05,
+      "loss": 0.6199,
+      "step": 4554
+    },
+    {
+      "epoch": 5.8304,
+      "grad_norm": 0.7004513740539551,
+      "learning_rate": 4.0906362545018005e-05,
+      "loss": 0.6281,
+      "step": 4555
+    },
+    {
+      "epoch": 5.83168,
+      "grad_norm": 0.743412971496582,
+      "learning_rate": 4.090436174469788e-05,
+      "loss": 0.6546,
+      "step": 4556
+    },
+    {
+      "epoch": 5.83296,
+      "grad_norm": 0.7199406623840332,
+      "learning_rate": 4.0902360944377755e-05,
+      "loss": 0.6265,
+      "step": 4557
+    },
+    {
+      "epoch": 5.83424,
+      "grad_norm": 0.7323152422904968,
+      "learning_rate": 4.090036014405763e-05,
+      "loss": 0.6174,
+      "step": 4558
+    },
+    {
+      "epoch": 5.83552,
+      "grad_norm": 0.7763957381248474,
+      "learning_rate": 4.08983593437375e-05,
+      "loss": 0.6362,
+      "step": 4559
+    },
+    {
+      "epoch": 5.8368,
+      "grad_norm": 0.7239270806312561,
+      "learning_rate": 4.089635854341737e-05,
+      "loss": 0.594,
+      "step": 4560
+    },
+    {
+      "epoch": 5.83808,
+      "grad_norm": 0.7253373265266418,
+      "learning_rate": 4.0894357743097236e-05,
+      "loss": 0.6342,
+      "step": 4561
+    },
+    {
+      "epoch": 5.83936,
+      "grad_norm": 0.7696877717971802,
+      "learning_rate": 4.089235694277711e-05,
+      "loss": 0.6545,
+      "step": 4562
+    },
+    {
+      "epoch": 5.8406400000000005,
+      "grad_norm": 0.7727090716362,
+      "learning_rate": 4.0890356142456986e-05,
+      "loss": 0.6611,
+      "step": 4563
+    },
+    {
+      "epoch": 5.84192,
+      "grad_norm": 0.6735355854034424,
+      "learning_rate": 4.088835534213686e-05,
+      "loss": 0.5948,
+      "step": 4564
+    },
+    {
+      "epoch": 5.8431999999999995,
+      "grad_norm": 0.7272473573684692,
+      "learning_rate": 4.088635454181673e-05,
+      "loss": 0.6286,
+      "step": 4565
+    },
+    {
+      "epoch": 5.84448,
+      "grad_norm": 0.760036289691925,
+      "learning_rate": 4.08843537414966e-05,
+      "loss": 0.6463,
+      "step": 4566
+    },
+    {
+      "epoch": 5.84576,
+      "grad_norm": 0.7171865701675415,
+      "learning_rate": 4.0882352941176474e-05,
+      "loss": 0.6485,
+      "step": 4567
+    },
+    {
+      "epoch": 5.84704,
+      "grad_norm": 0.7382311820983887,
+      "learning_rate": 4.0880352140856346e-05,
+      "loss": 0.6391,
+      "step": 4568
+    },
+    {
+      "epoch": 5.84832,
+      "grad_norm": 0.7612113356590271,
+      "learning_rate": 4.087835134053621e-05,
+      "loss": 0.6734,
+      "step": 4569
+    },
+    {
+      "epoch": 5.8496,
+      "grad_norm": 0.7614792585372925,
+      "learning_rate": 4.087635054021609e-05,
+      "loss": 0.635,
+      "step": 4570
+    },
+    {
+      "epoch": 5.85088,
+      "grad_norm": 0.8073767423629761,
+      "learning_rate": 4.087434973989596e-05,
+      "loss": 0.6494,
+      "step": 4571
+    },
+    {
+      "epoch": 5.85216,
+      "grad_norm": 0.7585006952285767,
+      "learning_rate": 4.087234893957583e-05,
+      "loss": 0.6522,
+      "step": 4572
+    },
+    {
+      "epoch": 5.85344,
+      "grad_norm": 0.7650215029716492,
+      "learning_rate": 4.0870348139255705e-05,
+      "loss": 0.6874,
+      "step": 4573
+    },
+    {
+      "epoch": 5.85472,
+      "grad_norm": 0.7307962775230408,
+      "learning_rate": 4.086834733893558e-05,
+      "loss": 0.6017,
+      "step": 4574
+    },
+    {
+      "epoch": 5.856,
+      "grad_norm": 0.7022853493690491,
+      "learning_rate": 4.086634653861545e-05,
+      "loss": 0.6255,
+      "step": 4575
+    },
+    {
+      "epoch": 5.85728,
+      "grad_norm": 0.7452864050865173,
+      "learning_rate": 4.086434573829532e-05,
+      "loss": 0.6482,
+      "step": 4576
+    },
+    {
+      "epoch": 5.85856,
+      "grad_norm": 0.686421811580658,
+      "learning_rate": 4.086234493797519e-05,
+      "loss": 0.5918,
+      "step": 4577
+    },
+    {
+      "epoch": 5.85984,
+      "grad_norm": 0.7287219166755676,
+      "learning_rate": 4.0860344137655064e-05,
+      "loss": 0.6691,
+      "step": 4578
+    },
+    {
+      "epoch": 5.86112,
+      "grad_norm": 0.7395564913749695,
+      "learning_rate": 4.0858343337334936e-05,
+      "loss": 0.6626,
+      "step": 4579
+    },
+    {
+      "epoch": 5.8624,
+      "grad_norm": 0.727815568447113,
+      "learning_rate": 4.085634253701481e-05,
+      "loss": 0.6524,
+      "step": 4580
+    },
+    {
+      "epoch": 5.8636800000000004,
+      "grad_norm": 0.7862719893455505,
+      "learning_rate": 4.085434173669468e-05,
+      "loss": 0.6451,
+      "step": 4581
+    },
+    {
+      "epoch": 5.86496,
+      "grad_norm": 0.7924978137016296,
+      "learning_rate": 4.085234093637455e-05,
+      "loss": 0.6689,
+      "step": 4582
+    },
+    {
+      "epoch": 5.86624,
+      "grad_norm": 0.7216591835021973,
+      "learning_rate": 4.0850340136054423e-05,
+      "loss": 0.6297,
+      "step": 4583
+    },
+    {
+      "epoch": 5.86752,
+      "grad_norm": 0.7293557524681091,
+      "learning_rate": 4.0848339335734295e-05,
+      "loss": 0.65,
+      "step": 4584
+    },
+    {
+      "epoch": 5.8688,
+      "grad_norm": 0.7389594912528992,
+      "learning_rate": 4.084633853541417e-05,
+      "loss": 0.6201,
+      "step": 4585
+    },
+    {
+      "epoch": 5.87008,
+      "grad_norm": 0.7208465337753296,
+      "learning_rate": 4.084433773509404e-05,
+      "loss": 0.5968,
+      "step": 4586
+    },
+    {
+      "epoch": 5.87136,
+      "grad_norm": 0.7426589131355286,
+      "learning_rate": 4.084233693477391e-05,
+      "loss": 0.6313,
+      "step": 4587
+    },
+    {
+      "epoch": 5.87264,
+      "grad_norm": 0.7646098732948303,
+      "learning_rate": 4.084033613445378e-05,
+      "loss": 0.7042,
+      "step": 4588
+    },
+    {
+      "epoch": 5.87392,
+      "grad_norm": 0.7568240165710449,
+      "learning_rate": 4.0838335334133655e-05,
+      "loss": 0.6097,
+      "step": 4589
+    },
+    {
+      "epoch": 5.8751999999999995,
+      "grad_norm": 0.7447047829627991,
+      "learning_rate": 4.0836334533813526e-05,
+      "loss": 0.6462,
+      "step": 4590
+    },
+    {
+      "epoch": 5.87648,
+      "grad_norm": 0.7372913360595703,
+      "learning_rate": 4.0834333733493405e-05,
+      "loss": 0.6434,
+      "step": 4591
+    },
+    {
+      "epoch": 5.87776,
+      "grad_norm": 0.7013605833053589,
+      "learning_rate": 4.083233293317327e-05,
+      "loss": 0.6086,
+      "step": 4592
+    },
+    {
+      "epoch": 5.87904,
+      "grad_norm": 0.7706699967384338,
+      "learning_rate": 4.083033213285314e-05,
+      "loss": 0.6739,
+      "step": 4593
+    },
+    {
+      "epoch": 5.88032,
+      "grad_norm": 0.7354162931442261,
+      "learning_rate": 4.0828331332533014e-05,
+      "loss": 0.6037,
+      "step": 4594
+    },
+    {
+      "epoch": 5.8816,
+      "grad_norm": 0.7437216639518738,
+      "learning_rate": 4.0826330532212886e-05,
+      "loss": 0.6416,
+      "step": 4595
+    },
+    {
+      "epoch": 5.88288,
+      "grad_norm": 0.7383934259414673,
+      "learning_rate": 4.082432973189276e-05,
+      "loss": 0.6556,
+      "step": 4596
+    },
+    {
+      "epoch": 5.88416,
+      "grad_norm": 0.7448809146881104,
+      "learning_rate": 4.082232893157263e-05,
+      "loss": 0.6923,
+      "step": 4597
+    },
+    {
+      "epoch": 5.88544,
+      "grad_norm": 0.7986266613006592,
+      "learning_rate": 4.082032813125251e-05,
+      "loss": 0.6663,
+      "step": 4598
+    },
+    {
+      "epoch": 5.88672,
+      "grad_norm": 0.7175683975219727,
+      "learning_rate": 4.081832733093238e-05,
+      "loss": 0.6453,
+      "step": 4599
+    },
+    {
+      "epoch": 5.888,
+      "grad_norm": 0.7115585803985596,
+      "learning_rate": 4.0816326530612245e-05,
+      "loss": 0.5905,
+      "step": 4600
+    },
+    {
+      "epoch": 5.88928,
+      "grad_norm": 0.6863852739334106,
+      "learning_rate": 4.081432573029212e-05,
+      "loss": 0.5642,
+      "step": 4601
+    },
+    {
+      "epoch": 5.89056,
+      "grad_norm": 0.6771440505981445,
+      "learning_rate": 4.081232492997199e-05,
+      "loss": 0.565,
+      "step": 4602
+    },
+    {
+      "epoch": 5.89184,
+      "grad_norm": 0.6936319470405579,
+      "learning_rate": 4.081032412965186e-05,
+      "loss": 0.5993,
+      "step": 4603
+    },
+    {
+      "epoch": 5.89312,
+      "grad_norm": 0.7660918831825256,
+      "learning_rate": 4.080832332933173e-05,
+      "loss": 0.6337,
+      "step": 4604
+    },
+    {
+      "epoch": 5.8944,
+      "grad_norm": 0.7964367270469666,
+      "learning_rate": 4.080632252901161e-05,
+      "loss": 0.6357,
+      "step": 4605
+    },
+    {
+      "epoch": 5.8956800000000005,
+      "grad_norm": 0.7397758364677429,
+      "learning_rate": 4.080432172869148e-05,
+      "loss": 0.5938,
+      "step": 4606
+    },
+    {
+      "epoch": 5.89696,
+      "grad_norm": 0.7152984738349915,
+      "learning_rate": 4.0802320928371355e-05,
+      "loss": 0.5801,
+      "step": 4607
+    },
+    {
+      "epoch": 5.89824,
+      "grad_norm": 0.7661724090576172,
+      "learning_rate": 4.080032012805122e-05,
+      "loss": 0.6988,
+      "step": 4608
+    },
+    {
+      "epoch": 5.89952,
+      "grad_norm": 0.7536041140556335,
+      "learning_rate": 4.079831932773109e-05,
+      "loss": 0.6331,
+      "step": 4609
+    },
+    {
+      "epoch": 5.9008,
+      "grad_norm": 0.7481626868247986,
+      "learning_rate": 4.0796318527410964e-05,
+      "loss": 0.6484,
+      "step": 4610
+    },
+    {
+      "epoch": 5.90208,
+      "grad_norm": 0.7638240456581116,
+      "learning_rate": 4.0794317727090835e-05,
+      "loss": 0.6959,
+      "step": 4611
+    },
+    {
+      "epoch": 5.90336,
+      "grad_norm": 0.7237250804901123,
+      "learning_rate": 4.0792316926770714e-05,
+      "loss": 0.605,
+      "step": 4612
+    },
+    {
+      "epoch": 5.90464,
+      "grad_norm": 0.7751877903938293,
+      "learning_rate": 4.0790316126450586e-05,
+      "loss": 0.6425,
+      "step": 4613
+    },
+    {
+      "epoch": 5.90592,
+      "grad_norm": 0.7242938280105591,
+      "learning_rate": 4.078831532613046e-05,
+      "loss": 0.616,
+      "step": 4614
+    },
+    {
+      "epoch": 5.9072,
+      "grad_norm": 0.7469364404678345,
+      "learning_rate": 4.078631452581033e-05,
+      "loss": 0.6072,
+      "step": 4615
+    },
+    {
+      "epoch": 5.90848,
+      "grad_norm": 0.728795051574707,
+      "learning_rate": 4.0784313725490195e-05,
+      "loss": 0.6552,
+      "step": 4616
+    },
+    {
+      "epoch": 5.90976,
+      "grad_norm": 0.7057138085365295,
+      "learning_rate": 4.078231292517007e-05,
+      "loss": 0.6434,
+      "step": 4617
+    },
+    {
+      "epoch": 5.91104,
+      "grad_norm": 0.7345725893974304,
+      "learning_rate": 4.078031212484994e-05,
+      "loss": 0.6371,
+      "step": 4618
+    },
+    {
+      "epoch": 5.91232,
+      "grad_norm": 0.7135785818099976,
+      "learning_rate": 4.077831132452982e-05,
+      "loss": 0.6094,
+      "step": 4619
+    },
+    {
+      "epoch": 5.9136,
+      "grad_norm": 0.7003117799758911,
+      "learning_rate": 4.077631052420969e-05,
+      "loss": 0.6538,
+      "step": 4620
+    },
+    {
+      "epoch": 5.91488,
+      "grad_norm": 0.7230803966522217,
+      "learning_rate": 4.077430972388956e-05,
+      "loss": 0.6399,
+      "step": 4621
+    },
+    {
+      "epoch": 5.91616,
+      "grad_norm": 0.7477660179138184,
+      "learning_rate": 4.077230892356943e-05,
+      "loss": 0.6334,
+      "step": 4622
+    },
+    {
+      "epoch": 5.91744,
+      "grad_norm": 0.7172017693519592,
+      "learning_rate": 4.0770308123249305e-05,
+      "loss": 0.6714,
+      "step": 4623
+    },
+    {
+      "epoch": 5.91872,
+      "grad_norm": 0.7592531442642212,
+      "learning_rate": 4.076830732292917e-05,
+      "loss": 0.6453,
+      "step": 4624
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.6898649334907532,
+      "learning_rate": 4.076630652260904e-05,
+      "loss": 0.6417,
+      "step": 4625
+    },
+    {
+      "epoch": 5.92128,
+      "grad_norm": 0.7369380593299866,
+      "learning_rate": 4.076430572228892e-05,
+      "loss": 0.6391,
+      "step": 4626
+    },
+    {
+      "epoch": 5.92256,
+      "grad_norm": 0.7180305123329163,
+      "learning_rate": 4.076230492196879e-05,
+      "loss": 0.5914,
+      "step": 4627
+    },
+    {
+      "epoch": 5.92384,
+      "grad_norm": 0.7108614444732666,
+      "learning_rate": 4.0760304121648664e-05,
+      "loss": 0.629,
+      "step": 4628
+    },
+    {
+      "epoch": 5.92512,
+      "grad_norm": 0.753671407699585,
+      "learning_rate": 4.0758303321328536e-05,
+      "loss": 0.6227,
+      "step": 4629
+    },
+    {
+      "epoch": 5.9264,
+      "grad_norm": 0.7389430403709412,
+      "learning_rate": 4.075630252100841e-05,
+      "loss": 0.6287,
+      "step": 4630
+    },
+    {
+      "epoch": 5.92768,
+      "grad_norm": 0.7096794843673706,
+      "learning_rate": 4.075430172068828e-05,
+      "loss": 0.6025,
+      "step": 4631
+    },
+    {
+      "epoch": 5.92896,
+      "grad_norm": 0.7167031764984131,
+      "learning_rate": 4.0752300920368144e-05,
+      "loss": 0.5903,
+      "step": 4632
+    },
+    {
+      "epoch": 5.9302399999999995,
+      "grad_norm": 0.7445734143257141,
+      "learning_rate": 4.075030012004802e-05,
+      "loss": 0.6246,
+      "step": 4633
+    },
+    {
+      "epoch": 5.93152,
+      "grad_norm": 0.7259752750396729,
+      "learning_rate": 4.0748299319727895e-05,
+      "loss": 0.6677,
+      "step": 4634
+    },
+    {
+      "epoch": 5.9328,
+      "grad_norm": 0.7165055871009827,
+      "learning_rate": 4.074629851940777e-05,
+      "loss": 0.6135,
+      "step": 4635
+    },
+    {
+      "epoch": 5.93408,
+      "grad_norm": 0.7381249666213989,
+      "learning_rate": 4.074429771908764e-05,
+      "loss": 0.616,
+      "step": 4636
+    },
+    {
+      "epoch": 5.93536,
+      "grad_norm": 0.7970700263977051,
+      "learning_rate": 4.074229691876751e-05,
+      "loss": 0.7101,
+      "step": 4637
+    },
+    {
+      "epoch": 5.93664,
+      "grad_norm": 0.7447617053985596,
+      "learning_rate": 4.074029611844738e-05,
+      "loss": 0.6006,
+      "step": 4638
+    },
+    {
+      "epoch": 5.93792,
+      "grad_norm": 0.7397119402885437,
+      "learning_rate": 4.0738295318127254e-05,
+      "loss": 0.6462,
+      "step": 4639
+    },
+    {
+      "epoch": 5.9392,
+      "grad_norm": 0.6990648508071899,
+      "learning_rate": 4.073629451780712e-05,
+      "loss": 0.6027,
+      "step": 4640
+    },
+    {
+      "epoch": 5.94048,
+      "grad_norm": 0.739801824092865,
+      "learning_rate": 4.0734293717487e-05,
+      "loss": 0.6539,
+      "step": 4641
+    },
+    {
+      "epoch": 5.94176,
+      "grad_norm": 0.7326043248176575,
+      "learning_rate": 4.073229291716687e-05,
+      "loss": 0.6548,
+      "step": 4642
+    },
+    {
+      "epoch": 5.94304,
+      "grad_norm": 0.733765184879303,
+      "learning_rate": 4.073029211684674e-05,
+      "loss": 0.6622,
+      "step": 4643
+    },
+    {
+      "epoch": 5.94432,
+      "grad_norm": 0.7281367778778076,
+      "learning_rate": 4.0728291316526614e-05,
+      "loss": 0.6045,
+      "step": 4644
+    },
+    {
+      "epoch": 5.9456,
+      "grad_norm": 0.7234017848968506,
+      "learning_rate": 4.0726290516206485e-05,
+      "loss": 0.5911,
+      "step": 4645
+    },
+    {
+      "epoch": 5.94688,
+      "grad_norm": 0.7438223958015442,
+      "learning_rate": 4.072428971588636e-05,
+      "loss": 0.6348,
+      "step": 4646
+    },
+    {
+      "epoch": 5.94816,
+      "grad_norm": 0.7301828861236572,
+      "learning_rate": 4.072228891556623e-05,
+      "loss": 0.6204,
+      "step": 4647
+    },
+    {
+      "epoch": 5.94944,
+      "grad_norm": 0.7404515743255615,
+      "learning_rate": 4.07202881152461e-05,
+      "loss": 0.6279,
+      "step": 4648
+    },
+    {
+      "epoch": 5.9507200000000005,
+      "grad_norm": 0.6974015831947327,
+      "learning_rate": 4.071828731492597e-05,
+      "loss": 0.6196,
+      "step": 4649
+    },
+    {
+      "epoch": 5.952,
+      "grad_norm": 0.7563602924346924,
+      "learning_rate": 4.0716286514605845e-05,
+      "loss": 0.6058,
+      "step": 4650
+    },
+    {
+      "epoch": 5.95328,
+      "grad_norm": 0.7107272744178772,
+      "learning_rate": 4.0714285714285717e-05,
+      "loss": 0.6388,
+      "step": 4651
+    },
+    {
+      "epoch": 5.95456,
+      "grad_norm": 0.7473395466804504,
+      "learning_rate": 4.071228491396559e-05,
+      "loss": 0.586,
+      "step": 4652
+    },
+    {
+      "epoch": 5.95584,
+      "grad_norm": 0.6987488865852356,
+      "learning_rate": 4.071028411364546e-05,
+      "loss": 0.5951,
+      "step": 4653
+    },
+    {
+      "epoch": 5.95712,
+      "grad_norm": 0.7390433549880981,
+      "learning_rate": 4.070828331332533e-05,
+      "loss": 0.6402,
+      "step": 4654
+    },
+    {
+      "epoch": 5.9584,
+      "grad_norm": 0.7470793128013611,
+      "learning_rate": 4.0706282513005204e-05,
+      "loss": 0.6361,
+      "step": 4655
+    },
+    {
+      "epoch": 5.95968,
+      "grad_norm": 0.7746595740318298,
+      "learning_rate": 4.0704281712685076e-05,
+      "loss": 0.667,
+      "step": 4656
+    },
+    {
+      "epoch": 5.96096,
+      "grad_norm": 0.7226885557174683,
+      "learning_rate": 4.070228091236495e-05,
+      "loss": 0.6213,
+      "step": 4657
+    },
+    {
+      "epoch": 5.9622399999999995,
+      "grad_norm": 0.7263903021812439,
+      "learning_rate": 4.070028011204482e-05,
+      "loss": 0.6263,
+      "step": 4658
+    },
+    {
+      "epoch": 5.96352,
+      "grad_norm": 0.7590324282646179,
+      "learning_rate": 4.069827931172469e-05,
+      "loss": 0.6146,
+      "step": 4659
+    },
+    {
+      "epoch": 5.9648,
+      "grad_norm": 0.7753758430480957,
+      "learning_rate": 4.069627851140456e-05,
+      "loss": 0.6689,
+      "step": 4660
+    },
+    {
+      "epoch": 5.96608,
+      "grad_norm": 0.7328972220420837,
+      "learning_rate": 4.0694277711084435e-05,
+      "loss": 0.6338,
+      "step": 4661
+    },
+    {
+      "epoch": 5.96736,
+      "grad_norm": 0.7797688841819763,
+      "learning_rate": 4.069227691076431e-05,
+      "loss": 0.6802,
+      "step": 4662
+    },
+    {
+      "epoch": 5.96864,
+      "grad_norm": 0.741576075553894,
+      "learning_rate": 4.069027611044418e-05,
+      "loss": 0.6109,
+      "step": 4663
+    },
+    {
+      "epoch": 5.96992,
+      "grad_norm": 0.7838340997695923,
+      "learning_rate": 4.068827531012405e-05,
+      "loss": 0.6586,
+      "step": 4664
+    },
+    {
+      "epoch": 5.9712,
+      "grad_norm": 0.7517712712287903,
+      "learning_rate": 4.068627450980392e-05,
+      "loss": 0.6261,
+      "step": 4665
+    },
+    {
+      "epoch": 5.97248,
+      "grad_norm": 0.803371787071228,
+      "learning_rate": 4.0684273709483794e-05,
+      "loss": 0.6969,
+      "step": 4666
+    },
+    {
+      "epoch": 5.97376,
+      "grad_norm": 0.7221397161483765,
+      "learning_rate": 4.0682272909163666e-05,
+      "loss": 0.6129,
+      "step": 4667
+    },
+    {
+      "epoch": 5.97504,
+      "grad_norm": 0.7079233527183533,
+      "learning_rate": 4.068027210884354e-05,
+      "loss": 0.6356,
+      "step": 4668
+    },
+    {
+      "epoch": 5.97632,
+      "grad_norm": 0.6837298274040222,
+      "learning_rate": 4.067827130852342e-05,
+      "loss": 0.6188,
+      "step": 4669
+    },
+    {
+      "epoch": 5.9776,
+      "grad_norm": 0.7204573154449463,
+      "learning_rate": 4.067627050820328e-05,
+      "loss": 0.6068,
+      "step": 4670
+    },
+    {
+      "epoch": 5.97888,
+      "grad_norm": 0.7004908919334412,
+      "learning_rate": 4.0674269707883154e-05,
+      "loss": 0.5853,
+      "step": 4671
+    },
+    {
+      "epoch": 5.98016,
+      "grad_norm": 0.6996192932128906,
+      "learning_rate": 4.0672268907563026e-05,
+      "loss": 0.5958,
+      "step": 4672
+    },
+    {
+      "epoch": 5.98144,
+      "grad_norm": 0.7411110401153564,
+      "learning_rate": 4.06702681072429e-05,
+      "loss": 0.5999,
+      "step": 4673
+    },
+    {
+      "epoch": 5.9827200000000005,
+      "grad_norm": 0.7705451250076294,
+      "learning_rate": 4.066826730692277e-05,
+      "loss": 0.7017,
+      "step": 4674
+    },
+    {
+      "epoch": 5.984,
+      "grad_norm": 0.7225644588470459,
+      "learning_rate": 4.066626650660264e-05,
+      "loss": 0.5662,
+      "step": 4675
+    },
+    {
+      "epoch": 5.98528,
+      "grad_norm": 0.7521904706954956,
+      "learning_rate": 4.066426570628252e-05,
+      "loss": 0.6083,
+      "step": 4676
+    },
+    {
+      "epoch": 5.98656,
+      "grad_norm": 0.756615161895752,
+      "learning_rate": 4.066226490596239e-05,
+      "loss": 0.6185,
+      "step": 4677
+    },
+    {
+      "epoch": 5.98784,
+      "grad_norm": 0.733359158039093,
+      "learning_rate": 4.066026410564226e-05,
+      "loss": 0.6312,
+      "step": 4678
+    },
+    {
+      "epoch": 5.98912,
+      "grad_norm": 0.774350643157959,
+      "learning_rate": 4.065826330532213e-05,
+      "loss": 0.6328,
+      "step": 4679
+    },
+    {
+      "epoch": 5.9904,
+      "grad_norm": 0.7676503658294678,
+      "learning_rate": 4.0656262505002e-05,
+      "loss": 0.6359,
+      "step": 4680
+    },
+    {
+      "epoch": 5.99168,
+      "grad_norm": 0.7357990741729736,
+      "learning_rate": 4.065426170468187e-05,
+      "loss": 0.6147,
+      "step": 4681
+    },
+    {
+      "epoch": 5.99296,
+      "grad_norm": 0.7393483519554138,
+      "learning_rate": 4.0652260904361744e-05,
+      "loss": 0.6331,
+      "step": 4682
+    },
+    {
+      "epoch": 5.99424,
+      "grad_norm": 0.7861780524253845,
+      "learning_rate": 4.065026010404162e-05,
+      "loss": 0.6724,
+      "step": 4683
+    },
+    {
+      "epoch": 5.99552,
+      "grad_norm": 0.739687979221344,
+      "learning_rate": 4.0648259303721495e-05,
+      "loss": 0.6299,
+      "step": 4684
+    },
+    {
+      "epoch": 5.9968,
+      "grad_norm": 0.7430374622344971,
+      "learning_rate": 4.0646258503401366e-05,
+      "loss": 0.645,
+      "step": 4685
+    },
+    {
+      "epoch": 5.99808,
+      "grad_norm": 0.7406571507453918,
+      "learning_rate": 4.064425770308123e-05,
+      "loss": 0.6604,
+      "step": 4686
+    },
+    {
+      "epoch": 5.99936,
+      "grad_norm": 0.729308009147644,
+      "learning_rate": 4.06422569027611e-05,
+      "loss": 0.6399,
+      "step": 4687
+    },
+    {
+      "epoch": 6.00064,
+      "grad_norm": 1.4821006059646606,
+      "learning_rate": 4.0640256102440975e-05,
+      "loss": 1.1371,
+      "step": 4688
+    },
+    {
+      "epoch": 6.00192,
+      "grad_norm": 0.7121221423149109,
+      "learning_rate": 4.063825530212085e-05,
+      "loss": 0.6295,
+      "step": 4689
+    },
+    {
+      "epoch": 6.0032,
+      "grad_norm": 0.7151453495025635,
+      "learning_rate": 4.0636254501800726e-05,
+      "loss": 0.6071,
+      "step": 4690
+    },
+    {
+      "epoch": 6.00448,
+      "grad_norm": 0.7339667677879333,
+      "learning_rate": 4.06342537014806e-05,
+      "loss": 0.6585,
+      "step": 4691
+    },
+    {
+      "epoch": 6.00576,
+      "grad_norm": 0.7404779195785522,
+      "learning_rate": 4.063225290116047e-05,
+      "loss": 0.6038,
+      "step": 4692
+    },
+    {
+      "epoch": 6.00704,
+      "grad_norm": 0.7343531250953674,
+      "learning_rate": 4.063025210084034e-05,
+      "loss": 0.6758,
+      "step": 4693
+    },
+    {
+      "epoch": 6.00832,
+      "grad_norm": 0.7415891885757446,
+      "learning_rate": 4.0628251300520206e-05,
+      "loss": 0.6148,
+      "step": 4694
+    },
+    {
+      "epoch": 6.0096,
+      "grad_norm": 0.7926338911056519,
+      "learning_rate": 4.062625050020008e-05,
+      "loss": 0.6918,
+      "step": 4695
+    },
+    {
+      "epoch": 6.01088,
+      "grad_norm": 0.7479919791221619,
+      "learning_rate": 4.062424969987995e-05,
+      "loss": 0.6289,
+      "step": 4696
+    },
+    {
+      "epoch": 6.01216,
+      "grad_norm": 0.7507800459861755,
+      "learning_rate": 4.062224889955983e-05,
+      "loss": 0.6093,
+      "step": 4697
+    },
+    {
+      "epoch": 6.01344,
+      "grad_norm": 0.7076855301856995,
+      "learning_rate": 4.06202480992397e-05,
+      "loss": 0.5928,
+      "step": 4698
+    },
+    {
+      "epoch": 6.01472,
+      "grad_norm": 0.7334170341491699,
+      "learning_rate": 4.061824729891957e-05,
+      "loss": 0.6195,
+      "step": 4699
+    },
+    {
+      "epoch": 6.016,
+      "grad_norm": 0.7351385951042175,
+      "learning_rate": 4.0616246498599444e-05,
+      "loss": 0.6166,
+      "step": 4700
+    },
+    {
+      "epoch": 6.01728,
+      "grad_norm": 0.7694382071495056,
+      "learning_rate": 4.0614245698279316e-05,
+      "loss": 0.6092,
+      "step": 4701
+    },
+    {
+      "epoch": 6.01856,
+      "grad_norm": 0.7713647484779358,
+      "learning_rate": 4.061224489795918e-05,
+      "loss": 0.6331,
+      "step": 4702
+    },
+    {
+      "epoch": 6.01984,
+      "grad_norm": 0.7566975355148315,
+      "learning_rate": 4.061024409763905e-05,
+      "loss": 0.6276,
+      "step": 4703
+    },
+    {
+      "epoch": 6.02112,
+      "grad_norm": 0.6920115351676941,
+      "learning_rate": 4.060824329731893e-05,
+      "loss": 0.5527,
+      "step": 4704
+    },
+    {
+      "epoch": 6.0224,
+      "grad_norm": 0.7560464143753052,
+      "learning_rate": 4.0606242496998804e-05,
+      "loss": 0.6171,
+      "step": 4705
+    },
+    {
+      "epoch": 6.02368,
+      "grad_norm": 0.8139127492904663,
+      "learning_rate": 4.0604241696678675e-05,
+      "loss": 0.6635,
+      "step": 4706
+    },
+    {
+      "epoch": 6.02496,
+      "grad_norm": 0.7818084955215454,
+      "learning_rate": 4.060224089635855e-05,
+      "loss": 0.6318,
+      "step": 4707
+    },
+    {
+      "epoch": 6.02624,
+      "grad_norm": 0.7569645047187805,
+      "learning_rate": 4.060024009603842e-05,
+      "loss": 0.6523,
+      "step": 4708
+    },
+    {
+      "epoch": 6.02752,
+      "grad_norm": 0.7755677103996277,
+      "learning_rate": 4.059823929571829e-05,
+      "loss": 0.6469,
+      "step": 4709
+    },
+    {
+      "epoch": 6.0288,
+      "grad_norm": 0.762385904788971,
+      "learning_rate": 4.0596238495398156e-05,
+      "loss": 0.616,
+      "step": 4710
+    },
+    {
+      "epoch": 6.03008,
+      "grad_norm": 0.7784088253974915,
+      "learning_rate": 4.0594237695078035e-05,
+      "loss": 0.6624,
+      "step": 4711
+    },
+    {
+      "epoch": 6.03136,
+      "grad_norm": 0.7360923886299133,
+      "learning_rate": 4.0592236894757907e-05,
+      "loss": 0.5984,
+      "step": 4712
+    },
+    {
+      "epoch": 6.03264,
+      "grad_norm": 0.7154642343521118,
+      "learning_rate": 4.059023609443778e-05,
+      "loss": 0.6125,
+      "step": 4713
+    },
+    {
+      "epoch": 6.03392,
+      "grad_norm": 0.7827834486961365,
+      "learning_rate": 4.058823529411765e-05,
+      "loss": 0.5754,
+      "step": 4714
+    },
+    {
+      "epoch": 6.0352,
+      "grad_norm": 0.7146974205970764,
+      "learning_rate": 4.058623449379752e-05,
+      "loss": 0.584,
+      "step": 4715
+    },
+    {
+      "epoch": 6.03648,
+      "grad_norm": 0.7260125875473022,
+      "learning_rate": 4.0584233693477394e-05,
+      "loss": 0.6057,
+      "step": 4716
+    },
+    {
+      "epoch": 6.03776,
+      "grad_norm": 0.7503035068511963,
+      "learning_rate": 4.0582232893157266e-05,
+      "loss": 0.6242,
+      "step": 4717
+    },
+    {
+      "epoch": 6.03904,
+      "grad_norm": 0.7332813143730164,
+      "learning_rate": 4.058023209283714e-05,
+      "loss": 0.5681,
+      "step": 4718
+    },
+    {
+      "epoch": 6.04032,
+      "grad_norm": 0.7520899772644043,
+      "learning_rate": 4.057823129251701e-05,
+      "loss": 0.6185,
+      "step": 4719
+    },
+    {
+      "epoch": 6.0416,
+      "grad_norm": 0.7898109555244446,
+      "learning_rate": 4.057623049219688e-05,
+      "loss": 0.6065,
+      "step": 4720
+    },
+    {
+      "epoch": 6.04288,
+      "grad_norm": 0.7639850378036499,
+      "learning_rate": 4.057422969187675e-05,
+      "loss": 0.6498,
+      "step": 4721
+    },
+    {
+      "epoch": 6.04416,
+      "grad_norm": 0.7639596462249756,
+      "learning_rate": 4.0572228891556625e-05,
+      "loss": 0.6272,
+      "step": 4722
+    },
+    {
+      "epoch": 6.04544,
+      "grad_norm": 0.7639127969741821,
+      "learning_rate": 4.05702280912365e-05,
+      "loss": 0.6034,
+      "step": 4723
+    },
+    {
+      "epoch": 6.04672,
+      "grad_norm": 0.7917043566703796,
+      "learning_rate": 4.056822729091637e-05,
+      "loss": 0.6657,
+      "step": 4724
+    },
+    {
+      "epoch": 6.048,
+      "grad_norm": 0.7585735321044922,
+      "learning_rate": 4.056622649059624e-05,
+      "loss": 0.5913,
+      "step": 4725
+    },
+    {
+      "epoch": 6.04928,
+      "grad_norm": 0.7555089592933655,
+      "learning_rate": 4.056422569027611e-05,
+      "loss": 0.6295,
+      "step": 4726
+    },
+    {
+      "epoch": 6.05056,
+      "grad_norm": 0.7650047540664673,
+      "learning_rate": 4.0562224889955984e-05,
+      "loss": 0.5904,
+      "step": 4727
+    },
+    {
+      "epoch": 6.05184,
+      "grad_norm": 0.7548428177833557,
+      "learning_rate": 4.0560224089635856e-05,
+      "loss": 0.6117,
+      "step": 4728
+    },
+    {
+      "epoch": 6.05312,
+      "grad_norm": 0.7310490012168884,
+      "learning_rate": 4.055822328931573e-05,
+      "loss": 0.6228,
+      "step": 4729
+    },
+    {
+      "epoch": 6.0544,
+      "grad_norm": 0.7640831470489502,
+      "learning_rate": 4.05562224889956e-05,
+      "loss": 0.6002,
+      "step": 4730
+    },
+    {
+      "epoch": 6.05568,
+      "grad_norm": 0.7306190133094788,
+      "learning_rate": 4.055422168867547e-05,
+      "loss": 0.5617,
+      "step": 4731
+    },
+    {
+      "epoch": 6.05696,
+      "grad_norm": 0.767707109451294,
+      "learning_rate": 4.0552220888355344e-05,
+      "loss": 0.6073,
+      "step": 4732
+    },
+    {
+      "epoch": 6.05824,
+      "grad_norm": 0.7210080027580261,
+      "learning_rate": 4.0550220088035216e-05,
+      "loss": 0.5746,
+      "step": 4733
+    },
+    {
+      "epoch": 6.05952,
+      "grad_norm": 0.7745203971862793,
+      "learning_rate": 4.054821928771509e-05,
+      "loss": 0.6033,
+      "step": 4734
+    },
+    {
+      "epoch": 6.0608,
+      "grad_norm": 0.7430739998817444,
+      "learning_rate": 4.054621848739496e-05,
+      "loss": 0.6069,
+      "step": 4735
+    },
+    {
+      "epoch": 6.06208,
+      "grad_norm": 0.7837331891059875,
+      "learning_rate": 4.054421768707483e-05,
+      "loss": 0.6397,
+      "step": 4736
+    },
+    {
+      "epoch": 6.06336,
+      "grad_norm": 0.7566797137260437,
+      "learning_rate": 4.05422168867547e-05,
+      "loss": 0.6293,
+      "step": 4737
+    },
+    {
+      "epoch": 6.06464,
+      "grad_norm": 0.7480584979057312,
+      "learning_rate": 4.0540216086434575e-05,
+      "loss": 0.621,
+      "step": 4738
+    },
+    {
+      "epoch": 6.06592,
+      "grad_norm": 0.797978401184082,
+      "learning_rate": 4.0538215286114453e-05,
+      "loss": 0.6476,
+      "step": 4739
+    },
+    {
+      "epoch": 6.0672,
+      "grad_norm": 0.7844875454902649,
+      "learning_rate": 4.053621448579432e-05,
+      "loss": 0.6381,
+      "step": 4740
+    },
+    {
+      "epoch": 6.06848,
+      "grad_norm": 0.7457473278045654,
+      "learning_rate": 4.053421368547419e-05,
+      "loss": 0.6175,
+      "step": 4741
+    },
+    {
+      "epoch": 6.06976,
+      "grad_norm": 0.7767608761787415,
+      "learning_rate": 4.053221288515406e-05,
+      "loss": 0.6053,
+      "step": 4742
+    },
+    {
+      "epoch": 6.07104,
+      "grad_norm": 0.7705170512199402,
+      "learning_rate": 4.0530212084833934e-05,
+      "loss": 0.6364,
+      "step": 4743
+    },
+    {
+      "epoch": 6.07232,
+      "grad_norm": 0.7489408254623413,
+      "learning_rate": 4.0528211284513806e-05,
+      "loss": 0.6041,
+      "step": 4744
+    },
+    {
+      "epoch": 6.0736,
+      "grad_norm": 0.7729488611221313,
+      "learning_rate": 4.052621048419368e-05,
+      "loss": 0.6407,
+      "step": 4745
+    },
+    {
+      "epoch": 6.07488,
+      "grad_norm": 0.7681214213371277,
+      "learning_rate": 4.0524209683873556e-05,
+      "loss": 0.6271,
+      "step": 4746
+    },
+    {
+      "epoch": 6.07616,
+      "grad_norm": 0.7444661855697632,
+      "learning_rate": 4.052220888355343e-05,
+      "loss": 0.6305,
+      "step": 4747
+    },
+    {
+      "epoch": 6.07744,
+      "grad_norm": 0.7598939538002014,
+      "learning_rate": 4.0520208083233293e-05,
+      "loss": 0.6313,
+      "step": 4748
+    },
+    {
+      "epoch": 6.07872,
+      "grad_norm": 0.7359374761581421,
+      "learning_rate": 4.0518207282913165e-05,
+      "loss": 0.5767,
+      "step": 4749
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.711103081703186,
+      "learning_rate": 4.051620648259304e-05,
+      "loss": 0.5656,
+      "step": 4750
+    },
+    {
+      "epoch": 6.08128,
+      "grad_norm": 0.7639884352684021,
+      "learning_rate": 4.051420568227291e-05,
+      "loss": 0.6345,
+      "step": 4751
+    },
+    {
+      "epoch": 6.08256,
+      "grad_norm": 0.772283673286438,
+      "learning_rate": 4.051220488195278e-05,
+      "loss": 0.6508,
+      "step": 4752
+    },
+    {
+      "epoch": 6.08384,
+      "grad_norm": 0.7561479210853577,
+      "learning_rate": 4.051020408163265e-05,
+      "loss": 0.6016,
+      "step": 4753
+    },
+    {
+      "epoch": 6.08512,
+      "grad_norm": 0.7787504196166992,
+      "learning_rate": 4.050820328131253e-05,
+      "loss": 0.6232,
+      "step": 4754
+    },
+    {
+      "epoch": 6.0864,
+      "grad_norm": 0.7278993129730225,
+      "learning_rate": 4.05062024809924e-05,
+      "loss": 0.6724,
+      "step": 4755
+    },
+    {
+      "epoch": 6.08768,
+      "grad_norm": 0.7617911100387573,
+      "learning_rate": 4.050420168067227e-05,
+      "loss": 0.6296,
+      "step": 4756
+    },
+    {
+      "epoch": 6.08896,
+      "grad_norm": 0.7467544674873352,
+      "learning_rate": 4.050220088035214e-05,
+      "loss": 0.5833,
+      "step": 4757
+    },
+    {
+      "epoch": 6.09024,
+      "grad_norm": 0.8187177777290344,
+      "learning_rate": 4.050020008003201e-05,
+      "loss": 0.6642,
+      "step": 4758
+    },
+    {
+      "epoch": 6.09152,
+      "grad_norm": 0.7508170008659363,
+      "learning_rate": 4.0498199279711884e-05,
+      "loss": 0.5794,
+      "step": 4759
+    },
+    {
+      "epoch": 6.0928,
+      "grad_norm": 0.7760382294654846,
+      "learning_rate": 4.0496198479391756e-05,
+      "loss": 0.6437,
+      "step": 4760
+    },
+    {
+      "epoch": 6.09408,
+      "grad_norm": 0.762019693851471,
+      "learning_rate": 4.0494197679071634e-05,
+      "loss": 0.5662,
+      "step": 4761
+    },
+    {
+      "epoch": 6.09536,
+      "grad_norm": 0.7687118053436279,
+      "learning_rate": 4.0492196878751506e-05,
+      "loss": 0.6324,
+      "step": 4762
+    },
+    {
+      "epoch": 6.09664,
+      "grad_norm": 0.7786553502082825,
+      "learning_rate": 4.049019607843138e-05,
+      "loss": 0.5897,
+      "step": 4763
+    },
+    {
+      "epoch": 6.09792,
+      "grad_norm": 0.7534022927284241,
+      "learning_rate": 4.048819527811124e-05,
+      "loss": 0.5713,
+      "step": 4764
+    },
+    {
+      "epoch": 6.0992,
+      "grad_norm": 0.786078929901123,
+      "learning_rate": 4.0486194477791115e-05,
+      "loss": 0.5774,
+      "step": 4765
+    },
+    {
+      "epoch": 6.10048,
+      "grad_norm": 0.7434951663017273,
+      "learning_rate": 4.048419367747099e-05,
+      "loss": 0.6195,
+      "step": 4766
+    },
+    {
+      "epoch": 6.10176,
+      "grad_norm": 0.7410392761230469,
+      "learning_rate": 4.048219287715086e-05,
+      "loss": 0.6466,
+      "step": 4767
+    },
+    {
+      "epoch": 6.10304,
+      "grad_norm": 0.7867603302001953,
+      "learning_rate": 4.048019207683074e-05,
+      "loss": 0.6098,
+      "step": 4768
+    },
+    {
+      "epoch": 6.10432,
+      "grad_norm": 0.7912724614143372,
+      "learning_rate": 4.047819127651061e-05,
+      "loss": 0.66,
+      "step": 4769
+    },
+    {
+      "epoch": 6.1056,
+      "grad_norm": 0.6925088167190552,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.5844,
+      "step": 4770
+    },
+    {
+      "epoch": 6.10688,
+      "grad_norm": 0.7406015992164612,
+      "learning_rate": 4.047418967587035e-05,
+      "loss": 0.6517,
+      "step": 4771
+    },
+    {
+      "epoch": 6.10816,
+      "grad_norm": 0.7543079257011414,
+      "learning_rate": 4.047218887555022e-05,
+      "loss": 0.5988,
+      "step": 4772
+    },
+    {
+      "epoch": 6.10944,
+      "grad_norm": 0.7825755476951599,
+      "learning_rate": 4.047018807523009e-05,
+      "loss": 0.6316,
+      "step": 4773
+    },
+    {
+      "epoch": 6.11072,
+      "grad_norm": 0.7861376404762268,
+      "learning_rate": 4.046818727490996e-05,
+      "loss": 0.6427,
+      "step": 4774
+    },
+    {
+      "epoch": 6.112,
+      "grad_norm": 0.7529568672180176,
+      "learning_rate": 4.046618647458984e-05,
+      "loss": 0.6156,
+      "step": 4775
+    },
+    {
+      "epoch": 6.11328,
+      "grad_norm": 0.7438963055610657,
+      "learning_rate": 4.046418567426971e-05,
+      "loss": 0.6108,
+      "step": 4776
+    },
+    {
+      "epoch": 6.11456,
+      "grad_norm": 0.7456843852996826,
+      "learning_rate": 4.0462184873949584e-05,
+      "loss": 0.6296,
+      "step": 4777
+    },
+    {
+      "epoch": 6.11584,
+      "grad_norm": 0.7504041194915771,
+      "learning_rate": 4.0460184073629456e-05,
+      "loss": 0.5838,
+      "step": 4778
+    },
+    {
+      "epoch": 6.11712,
+      "grad_norm": 0.7819758653640747,
+      "learning_rate": 4.045818327330933e-05,
+      "loss": 0.5966,
+      "step": 4779
+    },
+    {
+      "epoch": 6.1184,
+      "grad_norm": 0.7668342590332031,
+      "learning_rate": 4.045618247298919e-05,
+      "loss": 0.6158,
+      "step": 4780
+    },
+    {
+      "epoch": 6.11968,
+      "grad_norm": 0.7753564715385437,
+      "learning_rate": 4.0454181672669065e-05,
+      "loss": 0.642,
+      "step": 4781
+    },
+    {
+      "epoch": 6.12096,
+      "grad_norm": 0.8022623062133789,
+      "learning_rate": 4.045218087234894e-05,
+      "loss": 0.6358,
+      "step": 4782
+    },
+    {
+      "epoch": 6.12224,
+      "grad_norm": 0.7484129667282104,
+      "learning_rate": 4.0450180072028815e-05,
+      "loss": 0.5627,
+      "step": 4783
+    },
+    {
+      "epoch": 6.12352,
+      "grad_norm": 0.7918726205825806,
+      "learning_rate": 4.044817927170869e-05,
+      "loss": 0.6377,
+      "step": 4784
+    },
+    {
+      "epoch": 6.1248,
+      "grad_norm": 0.7575948238372803,
+      "learning_rate": 4.044617847138856e-05,
+      "loss": 0.589,
+      "step": 4785
+    },
+    {
+      "epoch": 6.12608,
+      "grad_norm": 0.7408061623573303,
+      "learning_rate": 4.044417767106843e-05,
+      "loss": 0.6352,
+      "step": 4786
+    },
+    {
+      "epoch": 6.12736,
+      "grad_norm": 0.7478961944580078,
+      "learning_rate": 4.04421768707483e-05,
+      "loss": 0.5852,
+      "step": 4787
+    },
+    {
+      "epoch": 6.12864,
+      "grad_norm": 0.7341578006744385,
+      "learning_rate": 4.044017607042817e-05,
+      "loss": 0.6114,
+      "step": 4788
+    },
+    {
+      "epoch": 6.12992,
+      "grad_norm": 0.7750809192657471,
+      "learning_rate": 4.0438175270108046e-05,
+      "loss": 0.6121,
+      "step": 4789
+    },
+    {
+      "epoch": 6.1312,
+      "grad_norm": 0.7869114875793457,
+      "learning_rate": 4.043617446978792e-05,
+      "loss": 0.6637,
+      "step": 4790
+    },
+    {
+      "epoch": 6.13248,
+      "grad_norm": 0.7512006163597107,
+      "learning_rate": 4.043417366946779e-05,
+      "loss": 0.5679,
+      "step": 4791
+    },
+    {
+      "epoch": 6.13376,
+      "grad_norm": 0.7822096347808838,
+      "learning_rate": 4.043217286914766e-05,
+      "loss": 0.6191,
+      "step": 4792
+    },
+    {
+      "epoch": 6.13504,
+      "grad_norm": 0.7540210485458374,
+      "learning_rate": 4.0430172068827534e-05,
+      "loss": 0.6011,
+      "step": 4793
+    },
+    {
+      "epoch": 6.1363199999999996,
+      "grad_norm": 0.7164133191108704,
+      "learning_rate": 4.0428171268507406e-05,
+      "loss": 0.5114,
+      "step": 4794
+    },
+    {
+      "epoch": 6.1376,
+      "grad_norm": 0.8122145533561707,
+      "learning_rate": 4.042617046818728e-05,
+      "loss": 0.6451,
+      "step": 4795
+    },
+    {
+      "epoch": 6.13888,
+      "grad_norm": 0.8163143396377563,
+      "learning_rate": 4.042416966786715e-05,
+      "loss": 0.6086,
+      "step": 4796
+    },
+    {
+      "epoch": 6.14016,
+      "grad_norm": 0.7786663770675659,
+      "learning_rate": 4.042216886754702e-05,
+      "loss": 0.6447,
+      "step": 4797
+    },
+    {
+      "epoch": 6.14144,
+      "grad_norm": 0.7091884016990662,
+      "learning_rate": 4.042016806722689e-05,
+      "loss": 0.5588,
+      "step": 4798
+    },
+    {
+      "epoch": 6.14272,
+      "grad_norm": 0.7768360376358032,
+      "learning_rate": 4.0418167266906765e-05,
+      "loss": 0.6151,
+      "step": 4799
+    },
+    {
+      "epoch": 6.144,
+      "grad_norm": 0.7751429677009583,
+      "learning_rate": 4.041616646658664e-05,
+      "loss": 0.6076,
+      "step": 4800
+    },
+    {
+      "epoch": 6.14528,
+      "grad_norm": 0.7756529450416565,
+      "learning_rate": 4.041416566626651e-05,
+      "loss": 0.6237,
+      "step": 4801
+    },
+    {
+      "epoch": 6.14656,
+      "grad_norm": 0.7572968006134033,
+      "learning_rate": 4.041216486594638e-05,
+      "loss": 0.6407,
+      "step": 4802
+    },
+    {
+      "epoch": 6.14784,
+      "grad_norm": 0.7175725698471069,
+      "learning_rate": 4.041016406562625e-05,
+      "loss": 0.588,
+      "step": 4803
+    },
+    {
+      "epoch": 6.14912,
+      "grad_norm": 0.7753421068191528,
+      "learning_rate": 4.0408163265306124e-05,
+      "loss": 0.628,
+      "step": 4804
+    },
+    {
+      "epoch": 6.1504,
+      "grad_norm": 0.777458131313324,
+      "learning_rate": 4.0406162464985996e-05,
+      "loss": 0.5963,
+      "step": 4805
+    },
+    {
+      "epoch": 6.15168,
+      "grad_norm": 0.7486628890037537,
+      "learning_rate": 4.040416166466587e-05,
+      "loss": 0.5623,
+      "step": 4806
+    },
+    {
+      "epoch": 6.15296,
+      "grad_norm": 0.7569031715393066,
+      "learning_rate": 4.040216086434574e-05,
+      "loss": 0.6062,
+      "step": 4807
+    },
+    {
+      "epoch": 6.15424,
+      "grad_norm": 0.7664810419082642,
+      "learning_rate": 4.040016006402561e-05,
+      "loss": 0.6069,
+      "step": 4808
+    },
+    {
+      "epoch": 6.15552,
+      "grad_norm": 0.7901382446289062,
+      "learning_rate": 4.0398159263705483e-05,
+      "loss": 0.6281,
+      "step": 4809
+    },
+    {
+      "epoch": 6.1568,
+      "grad_norm": 0.7850843667984009,
+      "learning_rate": 4.0396158463385355e-05,
+      "loss": 0.6222,
+      "step": 4810
+    },
+    {
+      "epoch": 6.15808,
+      "grad_norm": 0.77830570936203,
+      "learning_rate": 4.039415766306523e-05,
+      "loss": 0.6369,
+      "step": 4811
+    },
+    {
+      "epoch": 6.15936,
+      "grad_norm": 0.7659288048744202,
+      "learning_rate": 4.03921568627451e-05,
+      "loss": 0.6099,
+      "step": 4812
+    },
+    {
+      "epoch": 6.16064,
+      "grad_norm": 0.7957006096839905,
+      "learning_rate": 4.039015606242497e-05,
+      "loss": 0.6153,
+      "step": 4813
+    },
+    {
+      "epoch": 6.16192,
+      "grad_norm": 0.7597688436508179,
+      "learning_rate": 4.038815526210484e-05,
+      "loss": 0.632,
+      "step": 4814
+    },
+    {
+      "epoch": 6.1632,
+      "grad_norm": 0.7524228692054749,
+      "learning_rate": 4.0386154461784715e-05,
+      "loss": 0.5995,
+      "step": 4815
+    },
+    {
+      "epoch": 6.16448,
+      "grad_norm": 0.7459036111831665,
+      "learning_rate": 4.0384153661464586e-05,
+      "loss": 0.6228,
+      "step": 4816
+    },
+    {
+      "epoch": 6.16576,
+      "grad_norm": 0.7470423579216003,
+      "learning_rate": 4.0382152861144465e-05,
+      "loss": 0.6378,
+      "step": 4817
+    },
+    {
+      "epoch": 6.16704,
+      "grad_norm": 0.7385801076889038,
+      "learning_rate": 4.038015206082433e-05,
+      "loss": 0.6017,
+      "step": 4818
+    },
+    {
+      "epoch": 6.16832,
+      "grad_norm": 0.7643359303474426,
+      "learning_rate": 4.03781512605042e-05,
+      "loss": 0.6293,
+      "step": 4819
+    },
+    {
+      "epoch": 6.1696,
+      "grad_norm": 0.7111765146255493,
+      "learning_rate": 4.0376150460184074e-05,
+      "loss": 0.582,
+      "step": 4820
+    },
+    {
+      "epoch": 6.17088,
+      "grad_norm": 0.780404806137085,
+      "learning_rate": 4.0374149659863946e-05,
+      "loss": 0.6365,
+      "step": 4821
+    },
+    {
+      "epoch": 6.17216,
+      "grad_norm": 0.7903060913085938,
+      "learning_rate": 4.037214885954382e-05,
+      "loss": 0.7153,
+      "step": 4822
+    },
+    {
+      "epoch": 6.17344,
+      "grad_norm": 0.7531589865684509,
+      "learning_rate": 4.037014805922369e-05,
+      "loss": 0.5813,
+      "step": 4823
+    },
+    {
+      "epoch": 6.17472,
+      "grad_norm": 0.7753871083259583,
+      "learning_rate": 4.036814725890357e-05,
+      "loss": 0.6235,
+      "step": 4824
+    },
+    {
+      "epoch": 6.176,
+      "grad_norm": 0.7611346244812012,
+      "learning_rate": 4.036614645858344e-05,
+      "loss": 0.6133,
+      "step": 4825
+    },
+    {
+      "epoch": 6.17728,
+      "grad_norm": 0.7633965611457825,
+      "learning_rate": 4.0364145658263305e-05,
+      "loss": 0.591,
+      "step": 4826
+    },
+    {
+      "epoch": 6.17856,
+      "grad_norm": 0.8028787970542908,
+      "learning_rate": 4.036214485794318e-05,
+      "loss": 0.6118,
+      "step": 4827
+    },
+    {
+      "epoch": 6.17984,
+      "grad_norm": 0.7600660920143127,
+      "learning_rate": 4.036014405762305e-05,
+      "loss": 0.6283,
+      "step": 4828
+    },
+    {
+      "epoch": 6.18112,
+      "grad_norm": 0.754428505897522,
+      "learning_rate": 4.035814325730292e-05,
+      "loss": 0.6278,
+      "step": 4829
+    },
+    {
+      "epoch": 6.1824,
+      "grad_norm": 0.7525807619094849,
+      "learning_rate": 4.035614245698279e-05,
+      "loss": 0.5935,
+      "step": 4830
+    },
+    {
+      "epoch": 6.18368,
+      "grad_norm": 0.7541505694389343,
+      "learning_rate": 4.035414165666267e-05,
+      "loss": 0.6048,
+      "step": 4831
+    },
+    {
+      "epoch": 6.18496,
+      "grad_norm": 0.7696326375007629,
+      "learning_rate": 4.035214085634254e-05,
+      "loss": 0.6422,
+      "step": 4832
+    },
+    {
+      "epoch": 6.18624,
+      "grad_norm": 0.7631934881210327,
+      "learning_rate": 4.0350140056022415e-05,
+      "loss": 0.6121,
+      "step": 4833
+    },
+    {
+      "epoch": 6.18752,
+      "grad_norm": 0.7651000022888184,
+      "learning_rate": 4.034813925570228e-05,
+      "loss": 0.5951,
+      "step": 4834
+    },
+    {
+      "epoch": 6.1888,
+      "grad_norm": 0.7359318137168884,
+      "learning_rate": 4.034613845538215e-05,
+      "loss": 0.5828,
+      "step": 4835
+    },
+    {
+      "epoch": 6.19008,
+      "grad_norm": 0.7644410729408264,
+      "learning_rate": 4.0344137655062024e-05,
+      "loss": 0.6218,
+      "step": 4836
+    },
+    {
+      "epoch": 6.19136,
+      "grad_norm": 0.7283311486244202,
+      "learning_rate": 4.0342136854741895e-05,
+      "loss": 0.5594,
+      "step": 4837
+    },
+    {
+      "epoch": 6.19264,
+      "grad_norm": 0.7427007555961609,
+      "learning_rate": 4.0340136054421774e-05,
+      "loss": 0.5902,
+      "step": 4838
+    },
+    {
+      "epoch": 6.19392,
+      "grad_norm": 0.7666735649108887,
+      "learning_rate": 4.0338135254101646e-05,
+      "loss": 0.6229,
+      "step": 4839
+    },
+    {
+      "epoch": 6.1952,
+      "grad_norm": 0.7470300197601318,
+      "learning_rate": 4.033613445378152e-05,
+      "loss": 0.6134,
+      "step": 4840
+    },
+    {
+      "epoch": 6.19648,
+      "grad_norm": 0.7347720265388489,
+      "learning_rate": 4.033413365346139e-05,
+      "loss": 0.5755,
+      "step": 4841
+    },
+    {
+      "epoch": 6.19776,
+      "grad_norm": 0.7179660797119141,
+      "learning_rate": 4.0332132853141255e-05,
+      "loss": 0.54,
+      "step": 4842
+    },
+    {
+      "epoch": 6.19904,
+      "grad_norm": 0.7514517903327942,
+      "learning_rate": 4.0330132052821127e-05,
+      "loss": 0.6363,
+      "step": 4843
+    },
+    {
+      "epoch": 6.20032,
+      "grad_norm": 0.7676318883895874,
+      "learning_rate": 4.0328131252501e-05,
+      "loss": 0.6533,
+      "step": 4844
+    },
+    {
+      "epoch": 6.2016,
+      "grad_norm": 0.830500602722168,
+      "learning_rate": 4.032613045218088e-05,
+      "loss": 0.7654,
+      "step": 4845
+    },
+    {
+      "epoch": 6.20288,
+      "grad_norm": 0.7603785395622253,
+      "learning_rate": 4.032412965186075e-05,
+      "loss": 0.5547,
+      "step": 4846
+    },
+    {
+      "epoch": 6.20416,
+      "grad_norm": 0.7514224648475647,
+      "learning_rate": 4.032212885154062e-05,
+      "loss": 0.5948,
+      "step": 4847
+    },
+    {
+      "epoch": 6.20544,
+      "grad_norm": 0.757417619228363,
+      "learning_rate": 4.032012805122049e-05,
+      "loss": 0.5691,
+      "step": 4848
+    },
+    {
+      "epoch": 6.20672,
+      "grad_norm": 0.719901978969574,
+      "learning_rate": 4.0318127250900364e-05,
+      "loss": 0.6315,
+      "step": 4849
+    },
+    {
+      "epoch": 6.208,
+      "grad_norm": 0.7711414098739624,
+      "learning_rate": 4.031612645058023e-05,
+      "loss": 0.5801,
+      "step": 4850
+    },
+    {
+      "epoch": 6.20928,
+      "grad_norm": 0.7761551737785339,
+      "learning_rate": 4.03141256502601e-05,
+      "loss": 0.5992,
+      "step": 4851
+    },
+    {
+      "epoch": 6.21056,
+      "grad_norm": 0.7462658286094666,
+      "learning_rate": 4.031212484993998e-05,
+      "loss": 0.6378,
+      "step": 4852
+    },
+    {
+      "epoch": 6.21184,
+      "grad_norm": 0.7466690540313721,
+      "learning_rate": 4.031012404961985e-05,
+      "loss": 0.6151,
+      "step": 4853
+    },
+    {
+      "epoch": 6.21312,
+      "grad_norm": 0.8062450885772705,
+      "learning_rate": 4.0308123249299724e-05,
+      "loss": 0.6893,
+      "step": 4854
+    },
+    {
+      "epoch": 6.2144,
+      "grad_norm": 0.8179064989089966,
+      "learning_rate": 4.0306122448979596e-05,
+      "loss": 0.6184,
+      "step": 4855
+    },
+    {
+      "epoch": 6.21568,
+      "grad_norm": 0.7624944448471069,
+      "learning_rate": 4.030412164865947e-05,
+      "loss": 0.6016,
+      "step": 4856
+    },
+    {
+      "epoch": 6.21696,
+      "grad_norm": 0.7305824756622314,
+      "learning_rate": 4.030212084833934e-05,
+      "loss": 0.6336,
+      "step": 4857
+    },
+    {
+      "epoch": 6.21824,
+      "grad_norm": 0.7892305850982666,
+      "learning_rate": 4.0300120048019204e-05,
+      "loss": 0.6435,
+      "step": 4858
+    },
+    {
+      "epoch": 6.21952,
+      "grad_norm": 0.7659081220626831,
+      "learning_rate": 4.029811924769908e-05,
+      "loss": 0.6336,
+      "step": 4859
+    },
+    {
+      "epoch": 6.2208,
+      "grad_norm": 0.729543149471283,
+      "learning_rate": 4.0296118447378955e-05,
+      "loss": 0.6063,
+      "step": 4860
+    },
+    {
+      "epoch": 6.22208,
+      "grad_norm": 0.7288227677345276,
+      "learning_rate": 4.029411764705883e-05,
+      "loss": 0.5951,
+      "step": 4861
+    },
+    {
+      "epoch": 6.22336,
+      "grad_norm": 0.7727043032646179,
+      "learning_rate": 4.02921168467387e-05,
+      "loss": 0.6412,
+      "step": 4862
+    },
+    {
+      "epoch": 6.22464,
+      "grad_norm": 0.7801682949066162,
+      "learning_rate": 4.029011604641857e-05,
+      "loss": 0.6209,
+      "step": 4863
+    },
+    {
+      "epoch": 6.22592,
+      "grad_norm": 0.7641760110855103,
+      "learning_rate": 4.028811524609844e-05,
+      "loss": 0.6342,
+      "step": 4864
+    },
+    {
+      "epoch": 6.2272,
+      "grad_norm": 0.7689356803894043,
+      "learning_rate": 4.0286114445778314e-05,
+      "loss": 0.6232,
+      "step": 4865
+    },
+    {
+      "epoch": 6.22848,
+      "grad_norm": 0.7414539456367493,
+      "learning_rate": 4.028411364545818e-05,
+      "loss": 0.6114,
+      "step": 4866
+    },
+    {
+      "epoch": 6.22976,
+      "grad_norm": 0.7376709580421448,
+      "learning_rate": 4.028211284513806e-05,
+      "loss": 0.5968,
+      "step": 4867
+    },
+    {
+      "epoch": 6.23104,
+      "grad_norm": 0.785620391368866,
+      "learning_rate": 4.028011204481793e-05,
+      "loss": 0.6389,
+      "step": 4868
+    },
+    {
+      "epoch": 6.23232,
+      "grad_norm": 0.768951416015625,
+      "learning_rate": 4.02781112444978e-05,
+      "loss": 0.5874,
+      "step": 4869
+    },
+    {
+      "epoch": 6.2336,
+      "grad_norm": 0.7968980669975281,
+      "learning_rate": 4.0276110444177673e-05,
+      "loss": 0.6834,
+      "step": 4870
+    },
+    {
+      "epoch": 6.23488,
+      "grad_norm": 0.769282341003418,
+      "learning_rate": 4.0274109643857545e-05,
+      "loss": 0.6257,
+      "step": 4871
+    },
+    {
+      "epoch": 6.23616,
+      "grad_norm": 0.7746549844741821,
+      "learning_rate": 4.027210884353742e-05,
+      "loss": 0.6179,
+      "step": 4872
+    },
+    {
+      "epoch": 6.23744,
+      "grad_norm": 0.7487301230430603,
+      "learning_rate": 4.027010804321729e-05,
+      "loss": 0.6463,
+      "step": 4873
+    },
+    {
+      "epoch": 6.23872,
+      "grad_norm": 0.7296451330184937,
+      "learning_rate": 4.026810724289716e-05,
+      "loss": 0.6646,
+      "step": 4874
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.7273462414741516,
+      "learning_rate": 4.026610644257703e-05,
+      "loss": 0.5974,
+      "step": 4875
+    },
+    {
+      "epoch": 6.24128,
+      "grad_norm": 0.7543179392814636,
+      "learning_rate": 4.0264105642256905e-05,
+      "loss": 0.6767,
+      "step": 4876
+    },
+    {
+      "epoch": 6.24256,
+      "grad_norm": 0.7834622859954834,
+      "learning_rate": 4.0262104841936776e-05,
+      "loss": 0.6277,
+      "step": 4877
+    },
+    {
+      "epoch": 6.24384,
+      "grad_norm": 0.7536554932594299,
+      "learning_rate": 4.026010404161665e-05,
+      "loss": 0.5714,
+      "step": 4878
+    },
+    {
+      "epoch": 6.24512,
+      "grad_norm": 0.7422727942466736,
+      "learning_rate": 4.025810324129652e-05,
+      "loss": 0.5911,
+      "step": 4879
+    },
+    {
+      "epoch": 6.2464,
+      "grad_norm": 0.7820923924446106,
+      "learning_rate": 4.025610244097639e-05,
+      "loss": 0.6481,
+      "step": 4880
+    },
+    {
+      "epoch": 6.24768,
+      "grad_norm": 0.6983485221862793,
+      "learning_rate": 4.0254101640656264e-05,
+      "loss": 0.5551,
+      "step": 4881
+    },
+    {
+      "epoch": 6.24896,
+      "grad_norm": 0.7593510746955872,
+      "learning_rate": 4.0252100840336136e-05,
+      "loss": 0.6049,
+      "step": 4882
+    },
+    {
+      "epoch": 6.25024,
+      "grad_norm": 0.803278386592865,
+      "learning_rate": 4.025010004001601e-05,
+      "loss": 0.6087,
+      "step": 4883
+    },
+    {
+      "epoch": 6.25152,
+      "grad_norm": 0.7537515163421631,
+      "learning_rate": 4.024809923969588e-05,
+      "loss": 0.5856,
+      "step": 4884
+    },
+    {
+      "epoch": 6.2528,
+      "grad_norm": 0.7161216735839844,
+      "learning_rate": 4.024609843937575e-05,
+      "loss": 0.5604,
+      "step": 4885
+    },
+    {
+      "epoch": 6.25408,
+      "grad_norm": 0.7833333015441895,
+      "learning_rate": 4.024409763905562e-05,
+      "loss": 0.6248,
+      "step": 4886
+    },
+    {
+      "epoch": 6.25536,
+      "grad_norm": 0.7813183069229126,
+      "learning_rate": 4.0242096838735495e-05,
+      "loss": 0.6205,
+      "step": 4887
+    },
+    {
+      "epoch": 6.25664,
+      "grad_norm": 0.8205659985542297,
+      "learning_rate": 4.024009603841537e-05,
+      "loss": 0.582,
+      "step": 4888
+    },
+    {
+      "epoch": 6.25792,
+      "grad_norm": 0.7965737581253052,
+      "learning_rate": 4.023809523809524e-05,
+      "loss": 0.6541,
+      "step": 4889
+    },
+    {
+      "epoch": 6.2592,
+      "grad_norm": 0.7237341403961182,
+      "learning_rate": 4.023609443777511e-05,
+      "loss": 0.6384,
+      "step": 4890
+    },
+    {
+      "epoch": 6.26048,
+      "grad_norm": 0.7347778081893921,
+      "learning_rate": 4.023409363745498e-05,
+      "loss": 0.6078,
+      "step": 4891
+    },
+    {
+      "epoch": 6.26176,
+      "grad_norm": 0.7442294359207153,
+      "learning_rate": 4.0232092837134854e-05,
+      "loss": 0.6534,
+      "step": 4892
+    },
+    {
+      "epoch": 6.26304,
+      "grad_norm": 0.7912835478782654,
+      "learning_rate": 4.0230092036814726e-05,
+      "loss": 0.5893,
+      "step": 4893
+    },
+    {
+      "epoch": 6.26432,
+      "grad_norm": 0.7610131502151489,
+      "learning_rate": 4.02280912364946e-05,
+      "loss": 0.6303,
+      "step": 4894
+    },
+    {
+      "epoch": 6.2656,
+      "grad_norm": 0.7433664798736572,
+      "learning_rate": 4.022609043617448e-05,
+      "loss": 0.6046,
+      "step": 4895
+    },
+    {
+      "epoch": 6.2668800000000005,
+      "grad_norm": 0.7865737080574036,
+      "learning_rate": 4.022408963585434e-05,
+      "loss": 0.6401,
+      "step": 4896
+    },
+    {
+      "epoch": 6.26816,
+      "grad_norm": 0.7448906898498535,
+      "learning_rate": 4.0222088835534214e-05,
+      "loss": 0.6217,
+      "step": 4897
+    },
+    {
+      "epoch": 6.26944,
+      "grad_norm": 0.7785893678665161,
+      "learning_rate": 4.0220088035214085e-05,
+      "loss": 0.6827,
+      "step": 4898
+    },
+    {
+      "epoch": 6.27072,
+      "grad_norm": 0.7872070670127869,
+      "learning_rate": 4.021808723489396e-05,
+      "loss": 0.6266,
+      "step": 4899
+    },
+    {
+      "epoch": 6.272,
+      "grad_norm": 0.7371118068695068,
+      "learning_rate": 4.021608643457383e-05,
+      "loss": 0.6029,
+      "step": 4900
+    },
+    {
+      "epoch": 6.27328,
+      "grad_norm": 0.7747698426246643,
+      "learning_rate": 4.02140856342537e-05,
+      "loss": 0.6047,
+      "step": 4901
+    },
+    {
+      "epoch": 6.27456,
+      "grad_norm": 0.7733140587806702,
+      "learning_rate": 4.021208483393358e-05,
+      "loss": 0.6443,
+      "step": 4902
+    },
+    {
+      "epoch": 6.27584,
+      "grad_norm": 0.7372586727142334,
+      "learning_rate": 4.021008403361345e-05,
+      "loss": 0.6,
+      "step": 4903
+    },
+    {
+      "epoch": 6.27712,
+      "grad_norm": 0.807420015335083,
+      "learning_rate": 4.020808323329332e-05,
+      "loss": 0.6279,
+      "step": 4904
+    },
+    {
+      "epoch": 6.2783999999999995,
+      "grad_norm": 0.7645445466041565,
+      "learning_rate": 4.020608243297319e-05,
+      "loss": 0.6336,
+      "step": 4905
+    },
+    {
+      "epoch": 6.27968,
+      "grad_norm": 0.7852516770362854,
+      "learning_rate": 4.020408163265306e-05,
+      "loss": 0.6219,
+      "step": 4906
+    },
+    {
+      "epoch": 6.28096,
+      "grad_norm": 0.7741815447807312,
+      "learning_rate": 4.020208083233293e-05,
+      "loss": 0.6526,
+      "step": 4907
+    },
+    {
+      "epoch": 6.28224,
+      "grad_norm": 0.7737951278686523,
+      "learning_rate": 4.0200080032012804e-05,
+      "loss": 0.6533,
+      "step": 4908
+    },
+    {
+      "epoch": 6.28352,
+      "grad_norm": 0.7810575366020203,
+      "learning_rate": 4.019807923169268e-05,
+      "loss": 0.6677,
+      "step": 4909
+    },
+    {
+      "epoch": 6.2848,
+      "grad_norm": 0.7242299914360046,
+      "learning_rate": 4.0196078431372555e-05,
+      "loss": 0.5908,
+      "step": 4910
+    },
+    {
+      "epoch": 6.28608,
+      "grad_norm": 0.7727022767066956,
+      "learning_rate": 4.0194077631052426e-05,
+      "loss": 0.592,
+      "step": 4911
+    },
+    {
+      "epoch": 6.28736,
+      "grad_norm": 0.7893999814987183,
+      "learning_rate": 4.019207683073229e-05,
+      "loss": 0.6288,
+      "step": 4912
+    },
+    {
+      "epoch": 6.28864,
+      "grad_norm": 0.7878124713897705,
+      "learning_rate": 4.019007603041216e-05,
+      "loss": 0.6547,
+      "step": 4913
+    },
+    {
+      "epoch": 6.28992,
+      "grad_norm": 0.7988905906677246,
+      "learning_rate": 4.0188075230092035e-05,
+      "loss": 0.6392,
+      "step": 4914
+    },
+    {
+      "epoch": 6.2912,
+      "grad_norm": 0.7619364261627197,
+      "learning_rate": 4.018607442977191e-05,
+      "loss": 0.6155,
+      "step": 4915
+    },
+    {
+      "epoch": 6.29248,
+      "grad_norm": 0.8226683735847473,
+      "learning_rate": 4.0184073629451786e-05,
+      "loss": 0.6236,
+      "step": 4916
+    },
+    {
+      "epoch": 6.29376,
+      "grad_norm": 0.77449631690979,
+      "learning_rate": 4.018207282913166e-05,
+      "loss": 0.5779,
+      "step": 4917
+    },
+    {
+      "epoch": 6.29504,
+      "grad_norm": 0.7637079358100891,
+      "learning_rate": 4.018007202881153e-05,
+      "loss": 0.5976,
+      "step": 4918
+    },
+    {
+      "epoch": 6.29632,
+      "grad_norm": 0.7683913707733154,
+      "learning_rate": 4.01780712284914e-05,
+      "loss": 0.6022,
+      "step": 4919
+    },
+    {
+      "epoch": 6.2976,
+      "grad_norm": 0.7907748818397522,
+      "learning_rate": 4.0176070428171266e-05,
+      "loss": 0.5968,
+      "step": 4920
+    },
+    {
+      "epoch": 6.29888,
+      "grad_norm": 0.7320314049720764,
+      "learning_rate": 4.017406962785114e-05,
+      "loss": 0.5732,
+      "step": 4921
+    },
+    {
+      "epoch": 6.30016,
+      "grad_norm": 0.8165443539619446,
+      "learning_rate": 4.017206882753101e-05,
+      "loss": 0.6384,
+      "step": 4922
+    },
+    {
+      "epoch": 6.30144,
+      "grad_norm": 0.8002688884735107,
+      "learning_rate": 4.017006802721089e-05,
+      "loss": 0.5705,
+      "step": 4923
+    },
+    {
+      "epoch": 6.30272,
+      "grad_norm": 0.7913237810134888,
+      "learning_rate": 4.016806722689076e-05,
+      "loss": 0.6072,
+      "step": 4924
+    },
+    {
+      "epoch": 6.304,
+      "grad_norm": 0.7377355098724365,
+      "learning_rate": 4.016606642657063e-05,
+      "loss": 0.6112,
+      "step": 4925
+    },
+    {
+      "epoch": 6.30528,
+      "grad_norm": 0.680272102355957,
+      "learning_rate": 4.0164065626250504e-05,
+      "loss": 0.5723,
+      "step": 4926
+    },
+    {
+      "epoch": 6.30656,
+      "grad_norm": 0.68758225440979,
+      "learning_rate": 4.0162064825930376e-05,
+      "loss": 0.5122,
+      "step": 4927
+    },
+    {
+      "epoch": 6.30784,
+      "grad_norm": 0.8090822100639343,
+      "learning_rate": 4.016006402561024e-05,
+      "loss": 0.6358,
+      "step": 4928
+    },
+    {
+      "epoch": 6.30912,
+      "grad_norm": 0.7640751600265503,
+      "learning_rate": 4.015806322529011e-05,
+      "loss": 0.6516,
+      "step": 4929
+    },
+    {
+      "epoch": 6.3104,
+      "grad_norm": 0.7689977288246155,
+      "learning_rate": 4.015606242496999e-05,
+      "loss": 0.602,
+      "step": 4930
+    },
+    {
+      "epoch": 6.31168,
+      "grad_norm": 0.7796737551689148,
+      "learning_rate": 4.0154061624649864e-05,
+      "loss": 0.5984,
+      "step": 4931
+    },
+    {
+      "epoch": 6.31296,
+      "grad_norm": 0.7571960687637329,
+      "learning_rate": 4.0152060824329735e-05,
+      "loss": 0.6333,
+      "step": 4932
+    },
+    {
+      "epoch": 6.31424,
+      "grad_norm": 0.773962676525116,
+      "learning_rate": 4.015006002400961e-05,
+      "loss": 0.5782,
+      "step": 4933
+    },
+    {
+      "epoch": 6.31552,
+      "grad_norm": 0.8556650280952454,
+      "learning_rate": 4.014805922368948e-05,
+      "loss": 0.692,
+      "step": 4934
+    },
+    {
+      "epoch": 6.3168,
+      "grad_norm": 0.7910232543945312,
+      "learning_rate": 4.014605842336935e-05,
+      "loss": 0.6353,
+      "step": 4935
+    },
+    {
+      "epoch": 6.31808,
+      "grad_norm": 0.7409533858299255,
+      "learning_rate": 4.0144057623049216e-05,
+      "loss": 0.5861,
+      "step": 4936
+    },
+    {
+      "epoch": 6.31936,
+      "grad_norm": 0.7902359366416931,
+      "learning_rate": 4.0142056822729095e-05,
+      "loss": 0.6325,
+      "step": 4937
+    },
+    {
+      "epoch": 6.32064,
+      "grad_norm": 0.7420294880867004,
+      "learning_rate": 4.0140056022408967e-05,
+      "loss": 0.5833,
+      "step": 4938
+    },
+    {
+      "epoch": 6.32192,
+      "grad_norm": 0.7837445139884949,
+      "learning_rate": 4.013805522208884e-05,
+      "loss": 0.6377,
+      "step": 4939
+    },
+    {
+      "epoch": 6.3232,
+      "grad_norm": 0.7776581645011902,
+      "learning_rate": 4.013605442176871e-05,
+      "loss": 0.6328,
+      "step": 4940
+    },
+    {
+      "epoch": 6.32448,
+      "grad_norm": 0.7739065885543823,
+      "learning_rate": 4.013405362144858e-05,
+      "loss": 0.5716,
+      "step": 4941
+    },
+    {
+      "epoch": 6.32576,
+      "grad_norm": 0.7609522342681885,
+      "learning_rate": 4.0132052821128454e-05,
+      "loss": 0.5454,
+      "step": 4942
+    },
+    {
+      "epoch": 6.32704,
+      "grad_norm": 0.7803019285202026,
+      "learning_rate": 4.0130052020808326e-05,
+      "loss": 0.6351,
+      "step": 4943
+    },
+    {
+      "epoch": 6.32832,
+      "grad_norm": 0.7924274206161499,
+      "learning_rate": 4.01280512204882e-05,
+      "loss": 0.6234,
+      "step": 4944
+    },
+    {
+      "epoch": 6.3296,
+      "grad_norm": 0.7934367060661316,
+      "learning_rate": 4.012605042016807e-05,
+      "loss": 0.632,
+      "step": 4945
+    },
+    {
+      "epoch": 6.33088,
+      "grad_norm": 0.7212319374084473,
+      "learning_rate": 4.012404961984794e-05,
+      "loss": 0.6216,
+      "step": 4946
+    },
+    {
+      "epoch": 6.33216,
+      "grad_norm": 0.7968099117279053,
+      "learning_rate": 4.012204881952781e-05,
+      "loss": 0.6441,
+      "step": 4947
+    },
+    {
+      "epoch": 6.33344,
+      "grad_norm": 0.7849327921867371,
+      "learning_rate": 4.0120048019207685e-05,
+      "loss": 0.6567,
+      "step": 4948
+    },
+    {
+      "epoch": 6.33472,
+      "grad_norm": 0.7716954350471497,
+      "learning_rate": 4.011804721888756e-05,
+      "loss": 0.6313,
+      "step": 4949
+    },
+    {
+      "epoch": 6.336,
+      "grad_norm": 0.7645626664161682,
+      "learning_rate": 4.011604641856743e-05,
+      "loss": 0.6403,
+      "step": 4950
+    },
+    {
+      "epoch": 6.33728,
+      "grad_norm": 0.7575722932815552,
+      "learning_rate": 4.01140456182473e-05,
+      "loss": 0.5777,
+      "step": 4951
+    },
+    {
+      "epoch": 6.33856,
+      "grad_norm": 0.7596458196640015,
+      "learning_rate": 4.011204481792717e-05,
+      "loss": 0.6296,
+      "step": 4952
+    },
+    {
+      "epoch": 6.33984,
+      "grad_norm": 0.7410492897033691,
+      "learning_rate": 4.0110044017607044e-05,
+      "loss": 0.5836,
+      "step": 4953
+    },
+    {
+      "epoch": 6.34112,
+      "grad_norm": 0.8087933659553528,
+      "learning_rate": 4.0108043217286916e-05,
+      "loss": 0.6035,
+      "step": 4954
+    },
+    {
+      "epoch": 6.3424,
+      "grad_norm": 0.7586066126823425,
+      "learning_rate": 4.010604241696679e-05,
+      "loss": 0.6121,
+      "step": 4955
+    },
+    {
+      "epoch": 6.34368,
+      "grad_norm": 0.7876436114311218,
+      "learning_rate": 4.010404161664666e-05,
+      "loss": 0.6206,
+      "step": 4956
+    },
+    {
+      "epoch": 6.34496,
+      "grad_norm": 0.7594673037528992,
+      "learning_rate": 4.010204081632653e-05,
+      "loss": 0.6085,
+      "step": 4957
+    },
+    {
+      "epoch": 6.34624,
+      "grad_norm": 0.7009496092796326,
+      "learning_rate": 4.010004001600641e-05,
+      "loss": 0.5725,
+      "step": 4958
+    },
+    {
+      "epoch": 6.34752,
+      "grad_norm": 0.7607222199440002,
+      "learning_rate": 4.0098039215686276e-05,
+      "loss": 0.6079,
+      "step": 4959
+    },
+    {
+      "epoch": 6.3488,
+      "grad_norm": 0.8146457076072693,
+      "learning_rate": 4.009603841536615e-05,
+      "loss": 0.6636,
+      "step": 4960
+    },
+    {
+      "epoch": 6.35008,
+      "grad_norm": 0.757946252822876,
+      "learning_rate": 4.009403761504602e-05,
+      "loss": 0.6275,
+      "step": 4961
+    },
+    {
+      "epoch": 6.35136,
+      "grad_norm": 0.7005823850631714,
+      "learning_rate": 4.009203681472589e-05,
+      "loss": 0.547,
+      "step": 4962
+    },
+    {
+      "epoch": 6.35264,
+      "grad_norm": 0.7344463467597961,
+      "learning_rate": 4.009003601440576e-05,
+      "loss": 0.6197,
+      "step": 4963
+    },
+    {
+      "epoch": 6.3539200000000005,
+      "grad_norm": 0.7441943883895874,
+      "learning_rate": 4.0088035214085635e-05,
+      "loss": 0.6031,
+      "step": 4964
+    },
+    {
+      "epoch": 6.3552,
+      "grad_norm": 0.7756435871124268,
+      "learning_rate": 4.0086034413765513e-05,
+      "loss": 0.6007,
+      "step": 4965
+    },
+    {
+      "epoch": 6.35648,
+      "grad_norm": 0.771628737449646,
+      "learning_rate": 4.0084033613445385e-05,
+      "loss": 0.6,
+      "step": 4966
+    },
+    {
+      "epoch": 6.35776,
+      "grad_norm": 0.7456242442131042,
+      "learning_rate": 4.008203281312525e-05,
+      "loss": 0.5856,
+      "step": 4967
+    },
+    {
+      "epoch": 6.35904,
+      "grad_norm": 0.7749770879745483,
+      "learning_rate": 4.008003201280512e-05,
+      "loss": 0.6364,
+      "step": 4968
+    },
+    {
+      "epoch": 6.36032,
+      "grad_norm": 0.7612026929855347,
+      "learning_rate": 4.0078031212484994e-05,
+      "loss": 0.5996,
+      "step": 4969
+    },
+    {
+      "epoch": 6.3616,
+      "grad_norm": 0.7761407494544983,
+      "learning_rate": 4.0076030412164866e-05,
+      "loss": 0.6143,
+      "step": 4970
+    },
+    {
+      "epoch": 6.36288,
+      "grad_norm": 0.7732192873954773,
+      "learning_rate": 4.007402961184474e-05,
+      "loss": 0.6296,
+      "step": 4971
+    },
+    {
+      "epoch": 6.36416,
+      "grad_norm": 0.7865023016929626,
+      "learning_rate": 4.0072028811524616e-05,
+      "loss": 0.6159,
+      "step": 4972
+    },
+    {
+      "epoch": 6.3654399999999995,
+      "grad_norm": 0.7385578155517578,
+      "learning_rate": 4.007002801120449e-05,
+      "loss": 0.6024,
+      "step": 4973
+    },
+    {
+      "epoch": 6.36672,
+      "grad_norm": 0.7962034344673157,
+      "learning_rate": 4.006802721088436e-05,
+      "loss": 0.6887,
+      "step": 4974
+    },
+    {
+      "epoch": 6.368,
+      "grad_norm": 0.7819473147392273,
+      "learning_rate": 4.0066026410564225e-05,
+      "loss": 0.6156,
+      "step": 4975
+    },
+    {
+      "epoch": 6.36928,
+      "grad_norm": 0.7631332278251648,
+      "learning_rate": 4.00640256102441e-05,
+      "loss": 0.6126,
+      "step": 4976
+    },
+    {
+      "epoch": 6.37056,
+      "grad_norm": 0.7788332104682922,
+      "learning_rate": 4.006202480992397e-05,
+      "loss": 0.6718,
+      "step": 4977
+    },
+    {
+      "epoch": 6.37184,
+      "grad_norm": 0.7646488547325134,
+      "learning_rate": 4.006002400960384e-05,
+      "loss": 0.6265,
+      "step": 4978
+    },
+    {
+      "epoch": 6.37312,
+      "grad_norm": 0.7828549742698669,
+      "learning_rate": 4.005802320928371e-05,
+      "loss": 0.6718,
+      "step": 4979
+    },
+    {
+      "epoch": 6.3744,
+      "grad_norm": 0.7134932279586792,
+      "learning_rate": 4.005602240896359e-05,
+      "loss": 0.5275,
+      "step": 4980
+    },
+    {
+      "epoch": 6.37568,
+      "grad_norm": 0.7558833360671997,
+      "learning_rate": 4.005402160864346e-05,
+      "loss": 0.6531,
+      "step": 4981
+    },
+    {
+      "epoch": 6.37696,
+      "grad_norm": 0.7762933373451233,
+      "learning_rate": 4.0052020808323335e-05,
+      "loss": 0.5951,
+      "step": 4982
+    },
+    {
+      "epoch": 6.37824,
+      "grad_norm": 0.7445576786994934,
+      "learning_rate": 4.00500200080032e-05,
+      "loss": 0.5966,
+      "step": 4983
+    },
+    {
+      "epoch": 6.37952,
+      "grad_norm": 0.7995836734771729,
+      "learning_rate": 4.004801920768307e-05,
+      "loss": 0.6623,
+      "step": 4984
+    },
+    {
+      "epoch": 6.3808,
+      "grad_norm": 0.7452435493469238,
+      "learning_rate": 4.0046018407362944e-05,
+      "loss": 0.5924,
+      "step": 4985
+    },
+    {
+      "epoch": 6.38208,
+      "grad_norm": 0.8057966828346252,
+      "learning_rate": 4.0044017607042816e-05,
+      "loss": 0.6282,
+      "step": 4986
+    },
+    {
+      "epoch": 6.38336,
+      "grad_norm": 0.8227620124816895,
+      "learning_rate": 4.0042016806722694e-05,
+      "loss": 0.6227,
+      "step": 4987
+    },
+    {
+      "epoch": 6.38464,
+      "grad_norm": 0.8170056343078613,
+      "learning_rate": 4.0040016006402566e-05,
+      "loss": 0.6289,
+      "step": 4988
+    },
+    {
+      "epoch": 6.38592,
+      "grad_norm": 0.7389938831329346,
+      "learning_rate": 4.003801520608244e-05,
+      "loss": 0.639,
+      "step": 4989
+    },
+    {
+      "epoch": 6.3872,
+      "grad_norm": 0.7633137106895447,
+      "learning_rate": 4.003601440576231e-05,
+      "loss": 0.6226,
+      "step": 4990
+    },
+    {
+      "epoch": 6.38848,
+      "grad_norm": 0.7744930982589722,
+      "learning_rate": 4.0034013605442175e-05,
+      "loss": 0.6296,
+      "step": 4991
+    },
+    {
+      "epoch": 6.38976,
+      "grad_norm": 0.7420474886894226,
+      "learning_rate": 4.003201280512205e-05,
+      "loss": 0.6408,
+      "step": 4992
+    },
+    {
+      "epoch": 6.39104,
+      "grad_norm": 0.7508412003517151,
+      "learning_rate": 4.003001200480192e-05,
+      "loss": 0.5994,
+      "step": 4993
+    },
+    {
+      "epoch": 6.39232,
+      "grad_norm": 0.783890962600708,
+      "learning_rate": 4.00280112044818e-05,
+      "loss": 0.6633,
+      "step": 4994
+    },
+    {
+      "epoch": 6.3936,
+      "grad_norm": 0.7705927491188049,
+      "learning_rate": 4.002601040416167e-05,
+      "loss": 0.6623,
+      "step": 4995
+    },
+    {
+      "epoch": 6.39488,
+      "grad_norm": 0.7970103621482849,
+      "learning_rate": 4.002400960384154e-05,
+      "loss": 0.6653,
+      "step": 4996
+    },
+    {
+      "epoch": 6.39616,
+      "grad_norm": 0.8254691958427429,
+      "learning_rate": 4.002200880352141e-05,
+      "loss": 0.6138,
+      "step": 4997
+    },
+    {
+      "epoch": 6.39744,
+      "grad_norm": 0.7935686707496643,
+      "learning_rate": 4.0020008003201285e-05,
+      "loss": 0.6288,
+      "step": 4998
+    },
+    {
+      "epoch": 6.39872,
+      "grad_norm": 0.7537709474563599,
+      "learning_rate": 4.001800720288115e-05,
+      "loss": 0.5739,
+      "step": 4999
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.7888927459716797,
+      "learning_rate": 4.001600640256102e-05,
+      "loss": 0.6501,
+      "step": 5000
+    },
+    {
+      "epoch": 6.40128,
+      "grad_norm": 0.7110080718994141,
+      "learning_rate": 4.00140056022409e-05,
+      "loss": 0.5607,
+      "step": 5001
+    },
+    {
+      "epoch": 6.40256,
+      "grad_norm": 0.7363867163658142,
+      "learning_rate": 4.001200480192077e-05,
+      "loss": 0.6134,
+      "step": 5002
+    },
+    {
+      "epoch": 6.40384,
+      "grad_norm": 0.7256652116775513,
+      "learning_rate": 4.0010004001600644e-05,
+      "loss": 0.5868,
+      "step": 5003
+    },
+    {
+      "epoch": 6.40512,
+      "grad_norm": 0.7670110464096069,
+      "learning_rate": 4.0008003201280516e-05,
+      "loss": 0.5823,
+      "step": 5004
+    },
+    {
+      "epoch": 6.4064,
+      "grad_norm": 0.778918445110321,
+      "learning_rate": 4.000600240096039e-05,
+      "loss": 0.6436,
+      "step": 5005
+    },
+    {
+      "epoch": 6.40768,
+      "grad_norm": 0.7200184464454651,
+      "learning_rate": 4.000400160064026e-05,
+      "loss": 0.5635,
+      "step": 5006
+    },
+    {
+      "epoch": 6.40896,
+      "grad_norm": 0.7693450450897217,
+      "learning_rate": 4.0002000800320125e-05,
+      "loss": 0.6088,
+      "step": 5007
+    },
+    {
+      "epoch": 6.41024,
+      "grad_norm": 0.7648208737373352,
+      "learning_rate": 4e-05,
+      "loss": 0.6234,
+      "step": 5008
+    },
+    {
+      "epoch": 6.41152,
+      "grad_norm": 0.784186840057373,
+      "learning_rate": 3.9997999199679875e-05,
+      "loss": 0.6424,
+      "step": 5009
+    },
+    {
+      "epoch": 6.4128,
+      "grad_norm": 0.7577515244483948,
+      "learning_rate": 3.999599839935975e-05,
+      "loss": 0.6087,
+      "step": 5010
+    },
+    {
+      "epoch": 6.41408,
+      "grad_norm": 0.7444120645523071,
+      "learning_rate": 3.999399759903962e-05,
+      "loss": 0.5809,
+      "step": 5011
+    },
+    {
+      "epoch": 6.41536,
+      "grad_norm": 0.7742969393730164,
+      "learning_rate": 3.999199679871949e-05,
+      "loss": 0.6226,
+      "step": 5012
+    },
+    {
+      "epoch": 6.41664,
+      "grad_norm": 0.7459928393363953,
+      "learning_rate": 3.998999599839936e-05,
+      "loss": 0.5572,
+      "step": 5013
+    },
+    {
+      "epoch": 6.41792,
+      "grad_norm": 0.7630308270454407,
+      "learning_rate": 3.9987995198079234e-05,
+      "loss": 0.6129,
+      "step": 5014
+    },
+    {
+      "epoch": 6.4192,
+      "grad_norm": 0.7499691843986511,
+      "learning_rate": 3.9985994397759106e-05,
+      "loss": 0.5652,
+      "step": 5015
+    },
+    {
+      "epoch": 6.42048,
+      "grad_norm": 0.7551511526107788,
+      "learning_rate": 3.998399359743898e-05,
+      "loss": 0.6122,
+      "step": 5016
+    },
+    {
+      "epoch": 6.42176,
+      "grad_norm": 0.8000392913818359,
+      "learning_rate": 3.998199279711885e-05,
+      "loss": 0.6467,
+      "step": 5017
+    },
+    {
+      "epoch": 6.42304,
+      "grad_norm": 0.7956501245498657,
+      "learning_rate": 3.997999199679872e-05,
+      "loss": 0.6459,
+      "step": 5018
+    },
+    {
+      "epoch": 6.42432,
+      "grad_norm": 0.7600113153457642,
+      "learning_rate": 3.9977991196478594e-05,
+      "loss": 0.6714,
+      "step": 5019
+    },
+    {
+      "epoch": 6.4256,
+      "grad_norm": 0.7289563417434692,
+      "learning_rate": 3.9975990396158466e-05,
+      "loss": 0.5878,
+      "step": 5020
+    },
+    {
+      "epoch": 6.42688,
+      "grad_norm": 0.7440637350082397,
+      "learning_rate": 3.997398959583834e-05,
+      "loss": 0.6087,
+      "step": 5021
+    },
+    {
+      "epoch": 6.42816,
+      "grad_norm": 0.8092576265335083,
+      "learning_rate": 3.997198879551821e-05,
+      "loss": 0.6633,
+      "step": 5022
+    },
+    {
+      "epoch": 6.42944,
+      "grad_norm": 0.7994837164878845,
+      "learning_rate": 3.996998799519808e-05,
+      "loss": 0.5933,
+      "step": 5023
+    },
+    {
+      "epoch": 6.43072,
+      "grad_norm": 0.7767707109451294,
+      "learning_rate": 3.996798719487795e-05,
+      "loss": 0.653,
+      "step": 5024
+    },
+    {
+      "epoch": 6.432,
+      "grad_norm": 0.780889093875885,
+      "learning_rate": 3.9965986394557825e-05,
+      "loss": 0.7127,
+      "step": 5025
+    },
+    {
+      "epoch": 6.43328,
+      "grad_norm": 0.7731380462646484,
+      "learning_rate": 3.99639855942377e-05,
+      "loss": 0.6146,
+      "step": 5026
+    },
+    {
+      "epoch": 6.43456,
+      "grad_norm": 0.7348530292510986,
+      "learning_rate": 3.996198479391757e-05,
+      "loss": 0.5598,
+      "step": 5027
+    },
+    {
+      "epoch": 6.43584,
+      "grad_norm": 0.7755183577537537,
+      "learning_rate": 3.995998399359744e-05,
+      "loss": 0.6158,
+      "step": 5028
+    },
+    {
+      "epoch": 6.43712,
+      "grad_norm": 0.7527555227279663,
+      "learning_rate": 3.995798319327731e-05,
+      "loss": 0.6186,
+      "step": 5029
+    },
+    {
+      "epoch": 6.4384,
+      "grad_norm": 0.7150756120681763,
+      "learning_rate": 3.9955982392957184e-05,
+      "loss": 0.5473,
+      "step": 5030
+    },
+    {
+      "epoch": 6.43968,
+      "grad_norm": 0.7636802792549133,
+      "learning_rate": 3.9953981592637056e-05,
+      "loss": 0.6463,
+      "step": 5031
+    },
+    {
+      "epoch": 6.4409600000000005,
+      "grad_norm": 0.8033157587051392,
+      "learning_rate": 3.995198079231693e-05,
+      "loss": 0.6095,
+      "step": 5032
+    },
+    {
+      "epoch": 6.44224,
+      "grad_norm": 0.7744354009628296,
+      "learning_rate": 3.99499799919968e-05,
+      "loss": 0.6034,
+      "step": 5033
+    },
+    {
+      "epoch": 6.44352,
+      "grad_norm": 0.7521213293075562,
+      "learning_rate": 3.994797919167667e-05,
+      "loss": 0.6075,
+      "step": 5034
+    },
+    {
+      "epoch": 6.4448,
+      "grad_norm": 0.7580350041389465,
+      "learning_rate": 3.9945978391356543e-05,
+      "loss": 0.6328,
+      "step": 5035
+    },
+    {
+      "epoch": 6.44608,
+      "grad_norm": 0.7581676244735718,
+      "learning_rate": 3.994397759103642e-05,
+      "loss": 0.5721,
+      "step": 5036
+    },
+    {
+      "epoch": 6.44736,
+      "grad_norm": 0.7432277202606201,
+      "learning_rate": 3.994197679071629e-05,
+      "loss": 0.5626,
+      "step": 5037
+    },
+    {
+      "epoch": 6.44864,
+      "grad_norm": 0.7694210410118103,
+      "learning_rate": 3.993997599039616e-05,
+      "loss": 0.6688,
+      "step": 5038
+    },
+    {
+      "epoch": 6.44992,
+      "grad_norm": 0.7114573121070862,
+      "learning_rate": 3.993797519007603e-05,
+      "loss": 0.5948,
+      "step": 5039
+    },
+    {
+      "epoch": 6.4512,
+      "grad_norm": 0.7279742360115051,
+      "learning_rate": 3.99359743897559e-05,
+      "loss": 0.5615,
+      "step": 5040
+    },
+    {
+      "epoch": 6.4524799999999995,
+      "grad_norm": 0.7480107545852661,
+      "learning_rate": 3.9933973589435775e-05,
+      "loss": 0.605,
+      "step": 5041
+    },
+    {
+      "epoch": 6.45376,
+      "grad_norm": 0.7534326314926147,
+      "learning_rate": 3.9931972789115646e-05,
+      "loss": 0.5966,
+      "step": 5042
+    },
+    {
+      "epoch": 6.45504,
+      "grad_norm": 0.7341785430908203,
+      "learning_rate": 3.9929971988795525e-05,
+      "loss": 0.5677,
+      "step": 5043
+    },
+    {
+      "epoch": 6.45632,
+      "grad_norm": 0.7698660492897034,
+      "learning_rate": 3.99279711884754e-05,
+      "loss": 0.6088,
+      "step": 5044
+    },
+    {
+      "epoch": 6.4576,
+      "grad_norm": 0.777713418006897,
+      "learning_rate": 3.992597038815526e-05,
+      "loss": 0.6248,
+      "step": 5045
+    },
+    {
+      "epoch": 6.45888,
+      "grad_norm": 0.7565256357192993,
+      "learning_rate": 3.9923969587835134e-05,
+      "loss": 0.6141,
+      "step": 5046
+    },
+    {
+      "epoch": 6.46016,
+      "grad_norm": 0.7639655470848083,
+      "learning_rate": 3.9921968787515006e-05,
+      "loss": 0.6114,
+      "step": 5047
+    },
+    {
+      "epoch": 6.46144,
+      "grad_norm": 0.7447236776351929,
+      "learning_rate": 3.991996798719488e-05,
+      "loss": 0.6114,
+      "step": 5048
+    },
+    {
+      "epoch": 6.46272,
+      "grad_norm": 0.7928009033203125,
+      "learning_rate": 3.991796718687475e-05,
+      "loss": 0.6378,
+      "step": 5049
+    },
+    {
+      "epoch": 6.464,
+      "grad_norm": 0.8144562244415283,
+      "learning_rate": 3.991596638655463e-05,
+      "loss": 0.6235,
+      "step": 5050
+    },
+    {
+      "epoch": 6.46528,
+      "grad_norm": 0.6955716609954834,
+      "learning_rate": 3.99139655862345e-05,
+      "loss": 0.5668,
+      "step": 5051
+    },
+    {
+      "epoch": 6.46656,
+      "grad_norm": 0.756885290145874,
+      "learning_rate": 3.991196478591437e-05,
+      "loss": 0.6575,
+      "step": 5052
+    },
+    {
+      "epoch": 6.46784,
+      "grad_norm": 0.7933226227760315,
+      "learning_rate": 3.990996398559424e-05,
+      "loss": 0.6359,
+      "step": 5053
+    },
+    {
+      "epoch": 6.46912,
+      "grad_norm": 0.764266848564148,
+      "learning_rate": 3.990796318527411e-05,
+      "loss": 0.6044,
+      "step": 5054
+    },
+    {
+      "epoch": 6.4704,
+      "grad_norm": 0.770219087600708,
+      "learning_rate": 3.990596238495398e-05,
+      "loss": 0.6511,
+      "step": 5055
+    },
+    {
+      "epoch": 6.47168,
+      "grad_norm": 0.8092508912086487,
+      "learning_rate": 3.990396158463385e-05,
+      "loss": 0.6976,
+      "step": 5056
+    },
+    {
+      "epoch": 6.47296,
+      "grad_norm": 0.7517983913421631,
+      "learning_rate": 3.990196078431373e-05,
+      "loss": 0.5846,
+      "step": 5057
+    },
+    {
+      "epoch": 6.47424,
+      "grad_norm": 0.7917564511299133,
+      "learning_rate": 3.98999599839936e-05,
+      "loss": 0.652,
+      "step": 5058
+    },
+    {
+      "epoch": 6.47552,
+      "grad_norm": 0.7756854295730591,
+      "learning_rate": 3.9897959183673475e-05,
+      "loss": 0.5847,
+      "step": 5059
+    },
+    {
+      "epoch": 6.4768,
+      "grad_norm": 0.7966377139091492,
+      "learning_rate": 3.9895958383353347e-05,
+      "loss": 0.5864,
+      "step": 5060
+    },
+    {
+      "epoch": 6.47808,
+      "grad_norm": 0.793876051902771,
+      "learning_rate": 3.989395758303321e-05,
+      "loss": 0.5858,
+      "step": 5061
+    },
+    {
+      "epoch": 6.47936,
+      "grad_norm": 0.7717666029930115,
+      "learning_rate": 3.9891956782713084e-05,
+      "loss": 0.6202,
+      "step": 5062
+    },
+    {
+      "epoch": 6.48064,
+      "grad_norm": 0.7384968400001526,
+      "learning_rate": 3.9889955982392955e-05,
+      "loss": 0.6262,
+      "step": 5063
+    },
+    {
+      "epoch": 6.48192,
+      "grad_norm": 0.7273770570755005,
+      "learning_rate": 3.9887955182072834e-05,
+      "loss": 0.5539,
+      "step": 5064
+    },
+    {
+      "epoch": 6.4832,
+      "grad_norm": 0.7145727276802063,
+      "learning_rate": 3.9885954381752706e-05,
+      "loss": 0.5845,
+      "step": 5065
+    },
+    {
+      "epoch": 6.48448,
+      "grad_norm": 0.7780152559280396,
+      "learning_rate": 3.988395358143258e-05,
+      "loss": 0.596,
+      "step": 5066
+    },
+    {
+      "epoch": 6.48576,
+      "grad_norm": 0.8011901378631592,
+      "learning_rate": 3.988195278111245e-05,
+      "loss": 0.6526,
+      "step": 5067
+    },
+    {
+      "epoch": 6.48704,
+      "grad_norm": 0.780985414981842,
+      "learning_rate": 3.987995198079232e-05,
+      "loss": 0.5757,
+      "step": 5068
+    },
+    {
+      "epoch": 6.48832,
+      "grad_norm": 0.7850869297981262,
+      "learning_rate": 3.9877951180472187e-05,
+      "loss": 0.639,
+      "step": 5069
+    },
+    {
+      "epoch": 6.4896,
+      "grad_norm": 0.7978724837303162,
+      "learning_rate": 3.987595038015206e-05,
+      "loss": 0.5953,
+      "step": 5070
+    },
+    {
+      "epoch": 6.49088,
+      "grad_norm": 0.7666551470756531,
+      "learning_rate": 3.987394957983194e-05,
+      "loss": 0.5911,
+      "step": 5071
+    },
+    {
+      "epoch": 6.49216,
+      "grad_norm": 0.8040878176689148,
+      "learning_rate": 3.987194877951181e-05,
+      "loss": 0.6198,
+      "step": 5072
+    },
+    {
+      "epoch": 6.49344,
+      "grad_norm": 0.7172399759292603,
+      "learning_rate": 3.986994797919168e-05,
+      "loss": 0.6161,
+      "step": 5073
+    },
+    {
+      "epoch": 6.49472,
+      "grad_norm": 0.7347850203514099,
+      "learning_rate": 3.986794717887155e-05,
+      "loss": 0.6082,
+      "step": 5074
+    },
+    {
+      "epoch": 6.496,
+      "grad_norm": 0.7927334904670715,
+      "learning_rate": 3.9865946378551424e-05,
+      "loss": 0.6269,
+      "step": 5075
+    },
+    {
+      "epoch": 6.49728,
+      "grad_norm": 0.7884413003921509,
+      "learning_rate": 3.9863945578231296e-05,
+      "loss": 0.6417,
+      "step": 5076
+    },
+    {
+      "epoch": 6.49856,
+      "grad_norm": 0.7497519850730896,
+      "learning_rate": 3.986194477791116e-05,
+      "loss": 0.5945,
+      "step": 5077
+    },
+    {
+      "epoch": 6.49984,
+      "grad_norm": 0.8451282382011414,
+      "learning_rate": 3.985994397759104e-05,
+      "loss": 0.6606,
+      "step": 5078
+    },
+    {
+      "epoch": 6.50112,
+      "grad_norm": 0.8098703026771545,
+      "learning_rate": 3.985794317727091e-05,
+      "loss": 0.6289,
+      "step": 5079
+    },
+    {
+      "epoch": 6.5024,
+      "grad_norm": 0.8050368428230286,
+      "learning_rate": 3.9855942376950784e-05,
+      "loss": 0.6251,
+      "step": 5080
+    },
+    {
+      "epoch": 6.50368,
+      "grad_norm": 0.7607022523880005,
+      "learning_rate": 3.9853941576630656e-05,
+      "loss": 0.5769,
+      "step": 5081
+    },
+    {
+      "epoch": 6.50496,
+      "grad_norm": 0.7871085405349731,
+      "learning_rate": 3.985194077631053e-05,
+      "loss": 0.6361,
+      "step": 5082
+    },
+    {
+      "epoch": 6.50624,
+      "grad_norm": 0.7968281507492065,
+      "learning_rate": 3.98499399759904e-05,
+      "loss": 0.6802,
+      "step": 5083
+    },
+    {
+      "epoch": 6.5075199999999995,
+      "grad_norm": 0.7657204270362854,
+      "learning_rate": 3.984793917567027e-05,
+      "loss": 0.6098,
+      "step": 5084
+    },
+    {
+      "epoch": 6.5088,
+      "grad_norm": 0.7545149922370911,
+      "learning_rate": 3.984593837535014e-05,
+      "loss": 0.6096,
+      "step": 5085
+    },
+    {
+      "epoch": 6.51008,
+      "grad_norm": 0.7787616848945618,
+      "learning_rate": 3.9843937575030015e-05,
+      "loss": 0.6258,
+      "step": 5086
+    },
+    {
+      "epoch": 6.51136,
+      "grad_norm": 0.795444905757904,
+      "learning_rate": 3.984193677470989e-05,
+      "loss": 0.6293,
+      "step": 5087
+    },
+    {
+      "epoch": 6.51264,
+      "grad_norm": 0.7523403763771057,
+      "learning_rate": 3.983993597438976e-05,
+      "loss": 0.6098,
+      "step": 5088
+    },
+    {
+      "epoch": 6.51392,
+      "grad_norm": 0.7614408731460571,
+      "learning_rate": 3.983793517406963e-05,
+      "loss": 0.5827,
+      "step": 5089
+    },
+    {
+      "epoch": 6.5152,
+      "grad_norm": 0.7404950857162476,
+      "learning_rate": 3.98359343737495e-05,
+      "loss": 0.6255,
+      "step": 5090
+    },
+    {
+      "epoch": 6.51648,
+      "grad_norm": 0.7416658997535706,
+      "learning_rate": 3.9833933573429374e-05,
+      "loss": 0.6169,
+      "step": 5091
+    },
+    {
+      "epoch": 6.51776,
+      "grad_norm": 0.7457707524299622,
+      "learning_rate": 3.9831932773109246e-05,
+      "loss": 0.5626,
+      "step": 5092
+    },
+    {
+      "epoch": 6.51904,
+      "grad_norm": 0.7906910181045532,
+      "learning_rate": 3.982993197278912e-05,
+      "loss": 0.6424,
+      "step": 5093
+    },
+    {
+      "epoch": 6.52032,
+      "grad_norm": 0.798050045967102,
+      "learning_rate": 3.982793117246899e-05,
+      "loss": 0.6044,
+      "step": 5094
+    },
+    {
+      "epoch": 6.5216,
+      "grad_norm": 0.7886822819709778,
+      "learning_rate": 3.982593037214886e-05,
+      "loss": 0.637,
+      "step": 5095
+    },
+    {
+      "epoch": 6.52288,
+      "grad_norm": 0.7528558373451233,
+      "learning_rate": 3.9823929571828733e-05,
+      "loss": 0.5701,
+      "step": 5096
+    },
+    {
+      "epoch": 6.52416,
+      "grad_norm": 0.78963303565979,
+      "learning_rate": 3.9821928771508605e-05,
+      "loss": 0.6763,
+      "step": 5097
+    },
+    {
+      "epoch": 6.52544,
+      "grad_norm": 0.7659367918968201,
+      "learning_rate": 3.981992797118848e-05,
+      "loss": 0.6181,
+      "step": 5098
+    },
+    {
+      "epoch": 6.52672,
+      "grad_norm": 0.7585764527320862,
+      "learning_rate": 3.981792717086835e-05,
+      "loss": 0.609,
+      "step": 5099
+    },
+    {
+      "epoch": 6.5280000000000005,
+      "grad_norm": 0.7747802138328552,
+      "learning_rate": 3.981592637054822e-05,
+      "loss": 0.6033,
+      "step": 5100
+    },
+    {
+      "epoch": 6.52928,
+      "grad_norm": 0.7253979444503784,
+      "learning_rate": 3.981392557022809e-05,
+      "loss": 0.5512,
+      "step": 5101
+    },
+    {
+      "epoch": 6.53056,
+      "grad_norm": 0.7542280554771423,
+      "learning_rate": 3.9811924769907965e-05,
+      "loss": 0.6486,
+      "step": 5102
+    },
+    {
+      "epoch": 6.53184,
+      "grad_norm": 0.7517480254173279,
+      "learning_rate": 3.9809923969587836e-05,
+      "loss": 0.6327,
+      "step": 5103
+    },
+    {
+      "epoch": 6.53312,
+      "grad_norm": 0.7435446381568909,
+      "learning_rate": 3.980792316926771e-05,
+      "loss": 0.6079,
+      "step": 5104
+    },
+    {
+      "epoch": 6.5344,
+      "grad_norm": 0.7595201730728149,
+      "learning_rate": 3.980592236894758e-05,
+      "loss": 0.604,
+      "step": 5105
+    },
+    {
+      "epoch": 6.53568,
+      "grad_norm": 0.7278428673744202,
+      "learning_rate": 3.980392156862745e-05,
+      "loss": 0.6111,
+      "step": 5106
+    },
+    {
+      "epoch": 6.53696,
+      "grad_norm": 0.7786182165145874,
+      "learning_rate": 3.9801920768307324e-05,
+      "loss": 0.6558,
+      "step": 5107
+    },
+    {
+      "epoch": 6.53824,
+      "grad_norm": 0.8050516247749329,
+      "learning_rate": 3.9799919967987196e-05,
+      "loss": 0.6563,
+      "step": 5108
+    },
+    {
+      "epoch": 6.5395199999999996,
+      "grad_norm": 0.7350043058395386,
+      "learning_rate": 3.979791916766707e-05,
+      "loss": 0.6049,
+      "step": 5109
+    },
+    {
+      "epoch": 6.5408,
+      "grad_norm": 0.7202991843223572,
+      "learning_rate": 3.979591836734694e-05,
+      "loss": 0.5954,
+      "step": 5110
+    },
+    {
+      "epoch": 6.54208,
+      "grad_norm": 0.7836583256721497,
+      "learning_rate": 3.979391756702681e-05,
+      "loss": 0.6103,
+      "step": 5111
+    },
+    {
+      "epoch": 6.54336,
+      "grad_norm": 0.7699684500694275,
+      "learning_rate": 3.979191676670668e-05,
+      "loss": 0.6243,
+      "step": 5112
+    },
+    {
+      "epoch": 6.54464,
+      "grad_norm": 0.8052793145179749,
+      "learning_rate": 3.9789915966386555e-05,
+      "loss": 0.6234,
+      "step": 5113
+    },
+    {
+      "epoch": 6.54592,
+      "grad_norm": 0.7986263632774353,
+      "learning_rate": 3.9787915166066434e-05,
+      "loss": 0.6337,
+      "step": 5114
+    },
+    {
+      "epoch": 6.5472,
+      "grad_norm": 0.7898523807525635,
+      "learning_rate": 3.97859143657463e-05,
+      "loss": 0.6222,
+      "step": 5115
+    },
+    {
+      "epoch": 6.54848,
+      "grad_norm": 0.7434815764427185,
+      "learning_rate": 3.978391356542617e-05,
+      "loss": 0.6167,
+      "step": 5116
+    },
+    {
+      "epoch": 6.54976,
+      "grad_norm": 0.7909148931503296,
+      "learning_rate": 3.978191276510604e-05,
+      "loss": 0.6231,
+      "step": 5117
+    },
+    {
+      "epoch": 6.55104,
+      "grad_norm": 0.7848829030990601,
+      "learning_rate": 3.9779911964785914e-05,
+      "loss": 0.6086,
+      "step": 5118
+    },
+    {
+      "epoch": 6.55232,
+      "grad_norm": 0.7340389490127563,
+      "learning_rate": 3.9777911164465786e-05,
+      "loss": 0.5672,
+      "step": 5119
+    },
+    {
+      "epoch": 6.5536,
+      "grad_norm": 0.8030784726142883,
+      "learning_rate": 3.977591036414566e-05,
+      "loss": 0.5965,
+      "step": 5120
+    },
+    {
+      "epoch": 6.55488,
+      "grad_norm": 0.783450186252594,
+      "learning_rate": 3.977390956382554e-05,
+      "loss": 0.6128,
+      "step": 5121
+    },
+    {
+      "epoch": 6.55616,
+      "grad_norm": 0.7899596095085144,
+      "learning_rate": 3.977190876350541e-05,
+      "loss": 0.6482,
+      "step": 5122
+    },
+    {
+      "epoch": 6.55744,
+      "grad_norm": 0.7937018275260925,
+      "learning_rate": 3.9769907963185274e-05,
+      "loss": 0.6245,
+      "step": 5123
+    },
+    {
+      "epoch": 6.55872,
+      "grad_norm": 0.7522284388542175,
+      "learning_rate": 3.9767907162865145e-05,
+      "loss": 0.5998,
+      "step": 5124
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 0.7983690500259399,
+      "learning_rate": 3.976590636254502e-05,
+      "loss": 0.6299,
+      "step": 5125
+    },
+    {
+      "epoch": 6.56128,
+      "grad_norm": 0.7660322785377502,
+      "learning_rate": 3.976390556222489e-05,
+      "loss": 0.6138,
+      "step": 5126
+    },
+    {
+      "epoch": 6.5625599999999995,
+      "grad_norm": 0.8127145767211914,
+      "learning_rate": 3.976190476190476e-05,
+      "loss": 0.5622,
+      "step": 5127
+    },
+    {
+      "epoch": 6.56384,
+      "grad_norm": 0.8016258478164673,
+      "learning_rate": 3.975990396158464e-05,
+      "loss": 0.6375,
+      "step": 5128
+    },
+    {
+      "epoch": 6.56512,
+      "grad_norm": 0.756418764591217,
+      "learning_rate": 3.975790316126451e-05,
+      "loss": 0.6423,
+      "step": 5129
+    },
+    {
+      "epoch": 6.5664,
+      "grad_norm": 0.7691940069198608,
+      "learning_rate": 3.975590236094438e-05,
+      "loss": 0.6323,
+      "step": 5130
+    },
+    {
+      "epoch": 6.56768,
+      "grad_norm": 0.7534066438674927,
+      "learning_rate": 3.975390156062425e-05,
+      "loss": 0.6174,
+      "step": 5131
+    },
+    {
+      "epoch": 6.56896,
+      "grad_norm": 0.8069169521331787,
+      "learning_rate": 3.975190076030412e-05,
+      "loss": 0.6324,
+      "step": 5132
+    },
+    {
+      "epoch": 6.57024,
+      "grad_norm": 0.7579085230827332,
+      "learning_rate": 3.974989995998399e-05,
+      "loss": 0.6137,
+      "step": 5133
+    },
+    {
+      "epoch": 6.57152,
+      "grad_norm": 0.759647011756897,
+      "learning_rate": 3.9747899159663864e-05,
+      "loss": 0.597,
+      "step": 5134
+    },
+    {
+      "epoch": 6.5728,
+      "grad_norm": 0.7449114918708801,
+      "learning_rate": 3.974589835934374e-05,
+      "loss": 0.5865,
+      "step": 5135
+    },
+    {
+      "epoch": 6.57408,
+      "grad_norm": 0.7468611001968384,
+      "learning_rate": 3.9743897559023614e-05,
+      "loss": 0.5255,
+      "step": 5136
+    },
+    {
+      "epoch": 6.57536,
+      "grad_norm": 0.8116476535797119,
+      "learning_rate": 3.9741896758703486e-05,
+      "loss": 0.6977,
+      "step": 5137
+    },
+    {
+      "epoch": 6.57664,
+      "grad_norm": 0.7768375873565674,
+      "learning_rate": 3.973989595838336e-05,
+      "loss": 0.6375,
+      "step": 5138
+    },
+    {
+      "epoch": 6.57792,
+      "grad_norm": 0.751890242099762,
+      "learning_rate": 3.973789515806322e-05,
+      "loss": 0.5899,
+      "step": 5139
+    },
+    {
+      "epoch": 6.5792,
+      "grad_norm": 0.7473777532577515,
+      "learning_rate": 3.9735894357743095e-05,
+      "loss": 0.5912,
+      "step": 5140
+    },
+    {
+      "epoch": 6.58048,
+      "grad_norm": 0.7794525623321533,
+      "learning_rate": 3.973389355742297e-05,
+      "loss": 0.6331,
+      "step": 5141
+    },
+    {
+      "epoch": 6.58176,
+      "grad_norm": 0.7751368880271912,
+      "learning_rate": 3.9731892757102846e-05,
+      "loss": 0.627,
+      "step": 5142
+    },
+    {
+      "epoch": 6.5830400000000004,
+      "grad_norm": 0.7366233468055725,
+      "learning_rate": 3.972989195678272e-05,
+      "loss": 0.6061,
+      "step": 5143
+    },
+    {
+      "epoch": 6.58432,
+      "grad_norm": 0.7520456910133362,
+      "learning_rate": 3.972789115646259e-05,
+      "loss": 0.6021,
+      "step": 5144
+    },
+    {
+      "epoch": 6.5856,
+      "grad_norm": 0.7733432650566101,
+      "learning_rate": 3.972589035614246e-05,
+      "loss": 0.5873,
+      "step": 5145
+    },
+    {
+      "epoch": 6.58688,
+      "grad_norm": 0.7415776252746582,
+      "learning_rate": 3.972388955582233e-05,
+      "loss": 0.5845,
+      "step": 5146
+    },
+    {
+      "epoch": 6.58816,
+      "grad_norm": 0.7746995091438293,
+      "learning_rate": 3.97218887555022e-05,
+      "loss": 0.5892,
+      "step": 5147
+    },
+    {
+      "epoch": 6.58944,
+      "grad_norm": 0.736659049987793,
+      "learning_rate": 3.971988795518207e-05,
+      "loss": 0.5841,
+      "step": 5148
+    },
+    {
+      "epoch": 6.59072,
+      "grad_norm": 0.7600561380386353,
+      "learning_rate": 3.971788715486195e-05,
+      "loss": 0.5765,
+      "step": 5149
+    },
+    {
+      "epoch": 6.592,
+      "grad_norm": 0.7675446271896362,
+      "learning_rate": 3.971588635454182e-05,
+      "loss": 0.6319,
+      "step": 5150
+    },
+    {
+      "epoch": 6.59328,
+      "grad_norm": 0.7852813601493835,
+      "learning_rate": 3.971388555422169e-05,
+      "loss": 0.6026,
+      "step": 5151
+    },
+    {
+      "epoch": 6.5945599999999995,
+      "grad_norm": 0.7649344801902771,
+      "learning_rate": 3.9711884753901564e-05,
+      "loss": 0.6173,
+      "step": 5152
+    },
+    {
+      "epoch": 6.59584,
+      "grad_norm": 0.7568708062171936,
+      "learning_rate": 3.9709883953581436e-05,
+      "loss": 0.6124,
+      "step": 5153
+    },
+    {
+      "epoch": 6.59712,
+      "grad_norm": 0.7136748433113098,
+      "learning_rate": 3.970788315326131e-05,
+      "loss": 0.6117,
+      "step": 5154
+    },
+    {
+      "epoch": 6.5984,
+      "grad_norm": 0.8341229557991028,
+      "learning_rate": 3.970588235294117e-05,
+      "loss": 0.6245,
+      "step": 5155
+    },
+    {
+      "epoch": 6.59968,
+      "grad_norm": 0.8061261773109436,
+      "learning_rate": 3.970388155262105e-05,
+      "loss": 0.6325,
+      "step": 5156
+    },
+    {
+      "epoch": 6.60096,
+      "grad_norm": 0.7360426187515259,
+      "learning_rate": 3.9701880752300923e-05,
+      "loss": 0.6087,
+      "step": 5157
+    },
+    {
+      "epoch": 6.60224,
+      "grad_norm": 0.7481448650360107,
+      "learning_rate": 3.9699879951980795e-05,
+      "loss": 0.5674,
+      "step": 5158
+    },
+    {
+      "epoch": 6.60352,
+      "grad_norm": 0.7844505310058594,
+      "learning_rate": 3.969787915166067e-05,
+      "loss": 0.6471,
+      "step": 5159
+    },
+    {
+      "epoch": 6.6048,
+      "grad_norm": 0.8189229965209961,
+      "learning_rate": 3.969587835134054e-05,
+      "loss": 0.6488,
+      "step": 5160
+    },
+    {
+      "epoch": 6.60608,
+      "grad_norm": 0.7833296656608582,
+      "learning_rate": 3.969387755102041e-05,
+      "loss": 0.6595,
+      "step": 5161
+    },
+    {
+      "epoch": 6.60736,
+      "grad_norm": 0.7521294951438904,
+      "learning_rate": 3.969187675070028e-05,
+      "loss": 0.6137,
+      "step": 5162
+    },
+    {
+      "epoch": 6.60864,
+      "grad_norm": 0.7876133918762207,
+      "learning_rate": 3.9689875950380155e-05,
+      "loss": 0.6154,
+      "step": 5163
+    },
+    {
+      "epoch": 6.60992,
+      "grad_norm": 0.7582443952560425,
+      "learning_rate": 3.9687875150060026e-05,
+      "loss": 0.5877,
+      "step": 5164
+    },
+    {
+      "epoch": 6.6112,
+      "grad_norm": 0.7658904790878296,
+      "learning_rate": 3.96858743497399e-05,
+      "loss": 0.647,
+      "step": 5165
+    },
+    {
+      "epoch": 6.61248,
+      "grad_norm": 0.7374891638755798,
+      "learning_rate": 3.968387354941977e-05,
+      "loss": 0.6219,
+      "step": 5166
+    },
+    {
+      "epoch": 6.61376,
+      "grad_norm": 0.7132728695869446,
+      "learning_rate": 3.968187274909964e-05,
+      "loss": 0.5317,
+      "step": 5167
+    },
+    {
+      "epoch": 6.6150400000000005,
+      "grad_norm": 0.7678626775741577,
+      "learning_rate": 3.9679871948779514e-05,
+      "loss": 0.5931,
+      "step": 5168
+    },
+    {
+      "epoch": 6.61632,
+      "grad_norm": 0.8119401335716248,
+      "learning_rate": 3.9677871148459386e-05,
+      "loss": 0.6636,
+      "step": 5169
+    },
+    {
+      "epoch": 6.6176,
+      "grad_norm": 0.7630581259727478,
+      "learning_rate": 3.967587034813926e-05,
+      "loss": 0.6401,
+      "step": 5170
+    },
+    {
+      "epoch": 6.61888,
+      "grad_norm": 0.7518944144248962,
+      "learning_rate": 3.967386954781913e-05,
+      "loss": 0.6174,
+      "step": 5171
+    },
+    {
+      "epoch": 6.62016,
+      "grad_norm": 0.7786079049110413,
+      "learning_rate": 3.9671868747499e-05,
+      "loss": 0.6526,
+      "step": 5172
+    },
+    {
+      "epoch": 6.62144,
+      "grad_norm": 0.8287822604179382,
+      "learning_rate": 3.966986794717887e-05,
+      "loss": 0.713,
+      "step": 5173
+    },
+    {
+      "epoch": 6.62272,
+      "grad_norm": 0.7201914191246033,
+      "learning_rate": 3.9667867146858745e-05,
+      "loss": 0.5696,
+      "step": 5174
+    },
+    {
+      "epoch": 6.624,
+      "grad_norm": 0.7631100416183472,
+      "learning_rate": 3.966586634653862e-05,
+      "loss": 0.6302,
+      "step": 5175
+    },
+    {
+      "epoch": 6.62528,
+      "grad_norm": 0.769008994102478,
+      "learning_rate": 3.966386554621849e-05,
+      "loss": 0.5921,
+      "step": 5176
+    },
+    {
+      "epoch": 6.62656,
+      "grad_norm": 0.8017969727516174,
+      "learning_rate": 3.966186474589836e-05,
+      "loss": 0.6579,
+      "step": 5177
+    },
+    {
+      "epoch": 6.62784,
+      "grad_norm": 0.7824108600616455,
+      "learning_rate": 3.965986394557823e-05,
+      "loss": 0.6594,
+      "step": 5178
+    },
+    {
+      "epoch": 6.62912,
+      "grad_norm": 0.7433846592903137,
+      "learning_rate": 3.9657863145258104e-05,
+      "loss": 0.624,
+      "step": 5179
+    },
+    {
+      "epoch": 6.6304,
+      "grad_norm": 0.7518701553344727,
+      "learning_rate": 3.9655862344937976e-05,
+      "loss": 0.6058,
+      "step": 5180
+    },
+    {
+      "epoch": 6.63168,
+      "grad_norm": 0.7567682266235352,
+      "learning_rate": 3.965386154461785e-05,
+      "loss": 0.6081,
+      "step": 5181
+    },
+    {
+      "epoch": 6.63296,
+      "grad_norm": 0.7529020309448242,
+      "learning_rate": 3.965186074429772e-05,
+      "loss": 0.6016,
+      "step": 5182
+    },
+    {
+      "epoch": 6.63424,
+      "grad_norm": 0.7778772115707397,
+      "learning_rate": 3.964985994397759e-05,
+      "loss": 0.6077,
+      "step": 5183
+    },
+    {
+      "epoch": 6.63552,
+      "grad_norm": 0.7256116271018982,
+      "learning_rate": 3.964785914365747e-05,
+      "loss": 0.5214,
+      "step": 5184
+    },
+    {
+      "epoch": 6.6368,
+      "grad_norm": 0.820444643497467,
+      "learning_rate": 3.9645858343337335e-05,
+      "loss": 0.6515,
+      "step": 5185
+    },
+    {
+      "epoch": 6.63808,
+      "grad_norm": 0.7796686291694641,
+      "learning_rate": 3.964385754301721e-05,
+      "loss": 0.6436,
+      "step": 5186
+    },
+    {
+      "epoch": 6.63936,
+      "grad_norm": 0.7803254127502441,
+      "learning_rate": 3.964185674269708e-05,
+      "loss": 0.6102,
+      "step": 5187
+    },
+    {
+      "epoch": 6.64064,
+      "grad_norm": 0.7486667037010193,
+      "learning_rate": 3.963985594237695e-05,
+      "loss": 0.65,
+      "step": 5188
+    },
+    {
+      "epoch": 6.64192,
+      "grad_norm": 0.7695333361625671,
+      "learning_rate": 3.963785514205682e-05,
+      "loss": 0.6523,
+      "step": 5189
+    },
+    {
+      "epoch": 6.6432,
+      "grad_norm": 0.7668389678001404,
+      "learning_rate": 3.9635854341736695e-05,
+      "loss": 0.6391,
+      "step": 5190
+    },
+    {
+      "epoch": 6.64448,
+      "grad_norm": 0.7639618515968323,
+      "learning_rate": 3.963385354141657e-05,
+      "loss": 0.6011,
+      "step": 5191
+    },
+    {
+      "epoch": 6.64576,
+      "grad_norm": 0.7803028225898743,
+      "learning_rate": 3.9631852741096445e-05,
+      "loss": 0.6203,
+      "step": 5192
+    },
+    {
+      "epoch": 6.64704,
+      "grad_norm": 0.7438964247703552,
+      "learning_rate": 3.962985194077631e-05,
+      "loss": 0.5911,
+      "step": 5193
+    },
+    {
+      "epoch": 6.64832,
+      "grad_norm": 0.7312869429588318,
+      "learning_rate": 3.962785114045618e-05,
+      "loss": 0.5731,
+      "step": 5194
+    },
+    {
+      "epoch": 6.6495999999999995,
+      "grad_norm": 0.7577333450317383,
+      "learning_rate": 3.9625850340136054e-05,
+      "loss": 0.6012,
+      "step": 5195
+    },
+    {
+      "epoch": 6.65088,
+      "grad_norm": 0.7712530493736267,
+      "learning_rate": 3.9623849539815926e-05,
+      "loss": 0.6512,
+      "step": 5196
+    },
+    {
+      "epoch": 6.65216,
+      "grad_norm": 0.7147469520568848,
+      "learning_rate": 3.96218487394958e-05,
+      "loss": 0.6153,
+      "step": 5197
+    },
+    {
+      "epoch": 6.65344,
+      "grad_norm": 0.7817697525024414,
+      "learning_rate": 3.961984793917567e-05,
+      "loss": 0.6256,
+      "step": 5198
+    },
+    {
+      "epoch": 6.65472,
+      "grad_norm": 0.7692977786064148,
+      "learning_rate": 3.961784713885555e-05,
+      "loss": 0.5991,
+      "step": 5199
+    },
+    {
+      "epoch": 6.656,
+      "grad_norm": 0.7339797616004944,
+      "learning_rate": 3.961584633853542e-05,
+      "loss": 0.5563,
+      "step": 5200
+    },
+    {
+      "epoch": 6.65728,
+      "grad_norm": 0.7555981278419495,
+      "learning_rate": 3.9613845538215285e-05,
+      "loss": 0.6133,
+      "step": 5201
+    },
+    {
+      "epoch": 6.65856,
+      "grad_norm": 0.7478959560394287,
+      "learning_rate": 3.961184473789516e-05,
+      "loss": 0.614,
+      "step": 5202
+    },
+    {
+      "epoch": 6.65984,
+      "grad_norm": 0.7598313689231873,
+      "learning_rate": 3.960984393757503e-05,
+      "loss": 0.6197,
+      "step": 5203
+    },
+    {
+      "epoch": 6.66112,
+      "grad_norm": 0.7408673763275146,
+      "learning_rate": 3.96078431372549e-05,
+      "loss": 0.5736,
+      "step": 5204
+    },
+    {
+      "epoch": 6.6624,
+      "grad_norm": 0.8013281226158142,
+      "learning_rate": 3.960584233693477e-05,
+      "loss": 0.6783,
+      "step": 5205
+    },
+    {
+      "epoch": 6.66368,
+      "grad_norm": 0.7539663910865784,
+      "learning_rate": 3.960384153661465e-05,
+      "loss": 0.5824,
+      "step": 5206
+    },
+    {
+      "epoch": 6.66496,
+      "grad_norm": 0.7687118649482727,
+      "learning_rate": 3.960184073629452e-05,
+      "loss": 0.6185,
+      "step": 5207
+    },
+    {
+      "epoch": 6.66624,
+      "grad_norm": 0.803480327129364,
+      "learning_rate": 3.9599839935974395e-05,
+      "loss": 0.5955,
+      "step": 5208
+    },
+    {
+      "epoch": 6.66752,
+      "grad_norm": 0.7527867555618286,
+      "learning_rate": 3.959783913565426e-05,
+      "loss": 0.6128,
+      "step": 5209
+    },
+    {
+      "epoch": 6.6688,
+      "grad_norm": 0.7492735981941223,
+      "learning_rate": 3.959583833533413e-05,
+      "loss": 0.6112,
+      "step": 5210
+    },
+    {
+      "epoch": 6.6700800000000005,
+      "grad_norm": 0.757946252822876,
+      "learning_rate": 3.9593837535014004e-05,
+      "loss": 0.57,
+      "step": 5211
+    },
+    {
+      "epoch": 6.67136,
+      "grad_norm": 0.7378002405166626,
+      "learning_rate": 3.9591836734693876e-05,
+      "loss": 0.5691,
+      "step": 5212
+    },
+    {
+      "epoch": 6.67264,
+      "grad_norm": 0.7848967909812927,
+      "learning_rate": 3.9589835934373754e-05,
+      "loss": 0.6498,
+      "step": 5213
+    },
+    {
+      "epoch": 6.67392,
+      "grad_norm": 0.7497094869613647,
+      "learning_rate": 3.9587835134053626e-05,
+      "loss": 0.5919,
+      "step": 5214
+    },
+    {
+      "epoch": 6.6752,
+      "grad_norm": 0.7710285782814026,
+      "learning_rate": 3.95858343337335e-05,
+      "loss": 0.623,
+      "step": 5215
+    },
+    {
+      "epoch": 6.67648,
+      "grad_norm": 0.7714881300926208,
+      "learning_rate": 3.958383353341337e-05,
+      "loss": 0.6167,
+      "step": 5216
+    },
+    {
+      "epoch": 6.67776,
+      "grad_norm": 0.8070691823959351,
+      "learning_rate": 3.9581832733093235e-05,
+      "loss": 0.627,
+      "step": 5217
+    },
+    {
+      "epoch": 6.67904,
+      "grad_norm": 0.7932981848716736,
+      "learning_rate": 3.957983193277311e-05,
+      "loss": 0.6538,
+      "step": 5218
+    },
+    {
+      "epoch": 6.68032,
+      "grad_norm": 0.7265870571136475,
+      "learning_rate": 3.957783113245298e-05,
+      "loss": 0.5791,
+      "step": 5219
+    },
+    {
+      "epoch": 6.6815999999999995,
+      "grad_norm": 0.7722101211547852,
+      "learning_rate": 3.957583033213286e-05,
+      "loss": 0.6188,
+      "step": 5220
+    },
+    {
+      "epoch": 6.68288,
+      "grad_norm": 0.7940667271614075,
+      "learning_rate": 3.957382953181273e-05,
+      "loss": 0.5663,
+      "step": 5221
+    },
+    {
+      "epoch": 6.68416,
+      "grad_norm": 0.7684674263000488,
+      "learning_rate": 3.95718287314926e-05,
+      "loss": 0.6121,
+      "step": 5222
+    },
+    {
+      "epoch": 6.68544,
+      "grad_norm": 0.7839207053184509,
+      "learning_rate": 3.956982793117247e-05,
+      "loss": 0.6378,
+      "step": 5223
+    },
+    {
+      "epoch": 6.68672,
+      "grad_norm": 0.6971110701560974,
+      "learning_rate": 3.9567827130852345e-05,
+      "loss": 0.5632,
+      "step": 5224
+    },
+    {
+      "epoch": 6.688,
+      "grad_norm": 0.7478116750717163,
+      "learning_rate": 3.956582633053221e-05,
+      "loss": 0.6001,
+      "step": 5225
+    },
+    {
+      "epoch": 6.68928,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 3.956382553021208e-05,
+      "loss": 0.647,
+      "step": 5226
+    },
+    {
+      "epoch": 6.69056,
+      "grad_norm": 0.7169869542121887,
+      "learning_rate": 3.956182472989196e-05,
+      "loss": 0.609,
+      "step": 5227
+    },
+    {
+      "epoch": 6.69184,
+      "grad_norm": 0.7155503034591675,
+      "learning_rate": 3.955982392957183e-05,
+      "loss": 0.5816,
+      "step": 5228
+    },
+    {
+      "epoch": 6.69312,
+      "grad_norm": 0.7647979855537415,
+      "learning_rate": 3.9557823129251704e-05,
+      "loss": 0.6386,
+      "step": 5229
+    },
+    {
+      "epoch": 6.6944,
+      "grad_norm": 0.717627227306366,
+      "learning_rate": 3.9555822328931576e-05,
+      "loss": 0.6124,
+      "step": 5230
+    },
+    {
+      "epoch": 6.69568,
+      "grad_norm": 0.7740793824195862,
+      "learning_rate": 3.955382152861145e-05,
+      "loss": 0.6819,
+      "step": 5231
+    },
+    {
+      "epoch": 6.69696,
+      "grad_norm": 0.7482873797416687,
+      "learning_rate": 3.955182072829132e-05,
+      "loss": 0.5816,
+      "step": 5232
+    },
+    {
+      "epoch": 6.69824,
+      "grad_norm": 0.8073453903198242,
+      "learning_rate": 3.9549819927971185e-05,
+      "loss": 0.6147,
+      "step": 5233
+    },
+    {
+      "epoch": 6.69952,
+      "grad_norm": 0.775709867477417,
+      "learning_rate": 3.954781912765106e-05,
+      "loss": 0.6217,
+      "step": 5234
+    },
+    {
+      "epoch": 6.7008,
+      "grad_norm": 0.8085676431655884,
+      "learning_rate": 3.9545818327330935e-05,
+      "loss": 0.6666,
+      "step": 5235
+    },
+    {
+      "epoch": 6.7020800000000005,
+      "grad_norm": 0.7595890164375305,
+      "learning_rate": 3.954381752701081e-05,
+      "loss": 0.64,
+      "step": 5236
+    },
+    {
+      "epoch": 6.70336,
+      "grad_norm": 0.7835769653320312,
+      "learning_rate": 3.954181672669068e-05,
+      "loss": 0.663,
+      "step": 5237
+    },
+    {
+      "epoch": 6.70464,
+      "grad_norm": 0.797363817691803,
+      "learning_rate": 3.953981592637055e-05,
+      "loss": 0.6874,
+      "step": 5238
+    },
+    {
+      "epoch": 6.70592,
+      "grad_norm": 0.7739813923835754,
+      "learning_rate": 3.953781512605042e-05,
+      "loss": 0.5998,
+      "step": 5239
+    },
+    {
+      "epoch": 6.7072,
+      "grad_norm": 0.7902897000312805,
+      "learning_rate": 3.9535814325730294e-05,
+      "loss": 0.6097,
+      "step": 5240
+    },
+    {
+      "epoch": 6.70848,
+      "grad_norm": 0.756549060344696,
+      "learning_rate": 3.9533813525410166e-05,
+      "loss": 0.6184,
+      "step": 5241
+    },
+    {
+      "epoch": 6.70976,
+      "grad_norm": 0.73802250623703,
+      "learning_rate": 3.953181272509004e-05,
+      "loss": 0.5682,
+      "step": 5242
+    },
+    {
+      "epoch": 6.71104,
+      "grad_norm": 0.8189078569412231,
+      "learning_rate": 3.952981192476991e-05,
+      "loss": 0.651,
+      "step": 5243
+    },
+    {
+      "epoch": 6.71232,
+      "grad_norm": 0.7824820280075073,
+      "learning_rate": 3.952781112444978e-05,
+      "loss": 0.6341,
+      "step": 5244
+    },
+    {
+      "epoch": 6.7136,
+      "grad_norm": 0.788250207901001,
+      "learning_rate": 3.9525810324129654e-05,
+      "loss": 0.5819,
+      "step": 5245
+    },
+    {
+      "epoch": 6.71488,
+      "grad_norm": 0.7493855953216553,
+      "learning_rate": 3.9523809523809526e-05,
+      "loss": 0.6157,
+      "step": 5246
+    },
+    {
+      "epoch": 6.71616,
+      "grad_norm": 0.7709637880325317,
+      "learning_rate": 3.95218087234894e-05,
+      "loss": 0.6511,
+      "step": 5247
+    },
+    {
+      "epoch": 6.71744,
+      "grad_norm": 0.7652736306190491,
+      "learning_rate": 3.951980792316927e-05,
+      "loss": 0.5601,
+      "step": 5248
+    },
+    {
+      "epoch": 6.71872,
+      "grad_norm": 0.8294394612312317,
+      "learning_rate": 3.951780712284914e-05,
+      "loss": 0.6263,
+      "step": 5249
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.7636884450912476,
+      "learning_rate": 3.951580632252901e-05,
+      "loss": 0.5633,
+      "step": 5250
+    },
+    {
+      "epoch": 6.72128,
+      "grad_norm": 0.7744940519332886,
+      "learning_rate": 3.9513805522208885e-05,
+      "loss": 0.6129,
+      "step": 5251
+    },
+    {
+      "epoch": 6.72256,
+      "grad_norm": 0.8224226236343384,
+      "learning_rate": 3.951180472188876e-05,
+      "loss": 0.6453,
+      "step": 5252
+    },
+    {
+      "epoch": 6.72384,
+      "grad_norm": 0.7555708885192871,
+      "learning_rate": 3.950980392156863e-05,
+      "loss": 0.6153,
+      "step": 5253
+    },
+    {
+      "epoch": 6.72512,
+      "grad_norm": 0.7619531750679016,
+      "learning_rate": 3.95078031212485e-05,
+      "loss": 0.6237,
+      "step": 5254
+    },
+    {
+      "epoch": 6.7264,
+      "grad_norm": 0.7484665513038635,
+      "learning_rate": 3.950580232092838e-05,
+      "loss": 0.5713,
+      "step": 5255
+    },
+    {
+      "epoch": 6.72768,
+      "grad_norm": 0.8010841608047485,
+      "learning_rate": 3.9503801520608244e-05,
+      "loss": 0.6829,
+      "step": 5256
+    },
+    {
+      "epoch": 6.72896,
+      "grad_norm": 0.7087874412536621,
+      "learning_rate": 3.9501800720288116e-05,
+      "loss": 0.6001,
+      "step": 5257
+    },
+    {
+      "epoch": 6.73024,
+      "grad_norm": 0.7901442050933838,
+      "learning_rate": 3.949979991996799e-05,
+      "loss": 0.6637,
+      "step": 5258
+    },
+    {
+      "epoch": 6.73152,
+      "grad_norm": 0.78165602684021,
+      "learning_rate": 3.949779911964786e-05,
+      "loss": 0.654,
+      "step": 5259
+    },
+    {
+      "epoch": 6.7328,
+      "grad_norm": 0.768568754196167,
+      "learning_rate": 3.949579831932773e-05,
+      "loss": 0.6469,
+      "step": 5260
+    },
+    {
+      "epoch": 6.73408,
+      "grad_norm": 0.7904379963874817,
+      "learning_rate": 3.94937975190076e-05,
+      "loss": 0.6348,
+      "step": 5261
+    },
+    {
+      "epoch": 6.73536,
+      "grad_norm": 0.7367870211601257,
+      "learning_rate": 3.949179671868748e-05,
+      "loss": 0.5939,
+      "step": 5262
+    },
+    {
+      "epoch": 6.7366399999999995,
+      "grad_norm": 0.7719533443450928,
+      "learning_rate": 3.9489795918367354e-05,
+      "loss": 0.6251,
+      "step": 5263
+    },
+    {
+      "epoch": 6.73792,
+      "grad_norm": 0.7805905938148499,
+      "learning_rate": 3.948779511804722e-05,
+      "loss": 0.5785,
+      "step": 5264
+    },
+    {
+      "epoch": 6.7392,
+      "grad_norm": 0.7767636775970459,
+      "learning_rate": 3.948579431772709e-05,
+      "loss": 0.5854,
+      "step": 5265
+    },
+    {
+      "epoch": 6.74048,
+      "grad_norm": 0.7528958916664124,
+      "learning_rate": 3.948379351740696e-05,
+      "loss": 0.5972,
+      "step": 5266
+    },
+    {
+      "epoch": 6.74176,
+      "grad_norm": 0.7544768452644348,
+      "learning_rate": 3.9481792717086835e-05,
+      "loss": 0.6143,
+      "step": 5267
+    },
+    {
+      "epoch": 6.74304,
+      "grad_norm": 0.8174731731414795,
+      "learning_rate": 3.9479791916766706e-05,
+      "loss": 0.6503,
+      "step": 5268
+    },
+    {
+      "epoch": 6.74432,
+      "grad_norm": 0.7854768633842468,
+      "learning_rate": 3.9477791116446585e-05,
+      "loss": 0.6503,
+      "step": 5269
+    },
+    {
+      "epoch": 6.7456,
+      "grad_norm": 0.7965456247329712,
+      "learning_rate": 3.947579031612646e-05,
+      "loss": 0.6077,
+      "step": 5270
+    },
+    {
+      "epoch": 6.74688,
+      "grad_norm": 0.7633908987045288,
+      "learning_rate": 3.947378951580633e-05,
+      "loss": 0.6433,
+      "step": 5271
+    },
+    {
+      "epoch": 6.74816,
+      "grad_norm": 0.7727089524269104,
+      "learning_rate": 3.9471788715486194e-05,
+      "loss": 0.6058,
+      "step": 5272
+    },
+    {
+      "epoch": 6.74944,
+      "grad_norm": 0.7607753276824951,
+      "learning_rate": 3.9469787915166066e-05,
+      "loss": 0.6032,
+      "step": 5273
+    },
+    {
+      "epoch": 6.75072,
+      "grad_norm": 0.7452098727226257,
+      "learning_rate": 3.946778711484594e-05,
+      "loss": 0.581,
+      "step": 5274
+    },
+    {
+      "epoch": 6.752,
+      "grad_norm": 0.7961545586585999,
+      "learning_rate": 3.946578631452581e-05,
+      "loss": 0.6433,
+      "step": 5275
+    },
+    {
+      "epoch": 6.75328,
+      "grad_norm": 0.7878684401512146,
+      "learning_rate": 3.946378551420569e-05,
+      "loss": 0.6343,
+      "step": 5276
+    },
+    {
+      "epoch": 6.75456,
+      "grad_norm": 0.7796621322631836,
+      "learning_rate": 3.946178471388556e-05,
+      "loss": 0.6007,
+      "step": 5277
+    },
+    {
+      "epoch": 6.75584,
+      "grad_norm": 0.7304415106773376,
+      "learning_rate": 3.945978391356543e-05,
+      "loss": 0.5637,
+      "step": 5278
+    },
+    {
+      "epoch": 6.7571200000000005,
+      "grad_norm": 0.8092875480651855,
+      "learning_rate": 3.9457783113245304e-05,
+      "loss": 0.6487,
+      "step": 5279
+    },
+    {
+      "epoch": 6.7584,
+      "grad_norm": 0.7983285188674927,
+      "learning_rate": 3.945578231292517e-05,
+      "loss": 0.6291,
+      "step": 5280
+    },
+    {
+      "epoch": 6.75968,
+      "grad_norm": 0.7523446679115295,
+      "learning_rate": 3.945378151260504e-05,
+      "loss": 0.5886,
+      "step": 5281
+    },
+    {
+      "epoch": 6.76096,
+      "grad_norm": 0.7880832552909851,
+      "learning_rate": 3.945178071228491e-05,
+      "loss": 0.6386,
+      "step": 5282
+    },
+    {
+      "epoch": 6.76224,
+      "grad_norm": 0.7898958921432495,
+      "learning_rate": 3.944977991196479e-05,
+      "loss": 0.616,
+      "step": 5283
+    },
+    {
+      "epoch": 6.76352,
+      "grad_norm": 0.7464529275894165,
+      "learning_rate": 3.944777911164466e-05,
+      "loss": 0.5919,
+      "step": 5284
+    },
+    {
+      "epoch": 6.7648,
+      "grad_norm": 0.7837912440299988,
+      "learning_rate": 3.9445778311324535e-05,
+      "loss": 0.6378,
+      "step": 5285
+    },
+    {
+      "epoch": 6.76608,
+      "grad_norm": 0.7559735774993896,
+      "learning_rate": 3.9443777511004407e-05,
+      "loss": 0.6033,
+      "step": 5286
+    },
+    {
+      "epoch": 6.76736,
+      "grad_norm": 0.7761905193328857,
+      "learning_rate": 3.944177671068428e-05,
+      "loss": 0.6288,
+      "step": 5287
+    },
+    {
+      "epoch": 6.7686399999999995,
+      "grad_norm": 0.7161477208137512,
+      "learning_rate": 3.9439775910364143e-05,
+      "loss": 0.595,
+      "step": 5288
+    },
+    {
+      "epoch": 6.76992,
+      "grad_norm": 0.7817723155021667,
+      "learning_rate": 3.9437775110044015e-05,
+      "loss": 0.6474,
+      "step": 5289
+    },
+    {
+      "epoch": 6.7712,
+      "grad_norm": 0.7308269739151001,
+      "learning_rate": 3.9435774309723894e-05,
+      "loss": 0.6068,
+      "step": 5290
+    },
+    {
+      "epoch": 6.77248,
+      "grad_norm": 0.8274186253547668,
+      "learning_rate": 3.9433773509403766e-05,
+      "loss": 0.6928,
+      "step": 5291
+    },
+    {
+      "epoch": 6.77376,
+      "grad_norm": 0.7699143886566162,
+      "learning_rate": 3.943177270908364e-05,
+      "loss": 0.636,
+      "step": 5292
+    },
+    {
+      "epoch": 6.77504,
+      "grad_norm": 0.7616050243377686,
+      "learning_rate": 3.942977190876351e-05,
+      "loss": 0.6021,
+      "step": 5293
+    },
+    {
+      "epoch": 6.77632,
+      "grad_norm": 0.7531759142875671,
+      "learning_rate": 3.942777110844338e-05,
+      "loss": 0.5899,
+      "step": 5294
+    },
+    {
+      "epoch": 6.7776,
+      "grad_norm": 0.7871730327606201,
+      "learning_rate": 3.942577030812325e-05,
+      "loss": 0.6186,
+      "step": 5295
+    },
+    {
+      "epoch": 6.77888,
+      "grad_norm": 0.7265611290931702,
+      "learning_rate": 3.942376950780312e-05,
+      "loss": 0.5459,
+      "step": 5296
+    },
+    {
+      "epoch": 6.78016,
+      "grad_norm": 0.7594027519226074,
+      "learning_rate": 3.9421768707483e-05,
+      "loss": 0.6046,
+      "step": 5297
+    },
+    {
+      "epoch": 6.78144,
+      "grad_norm": 0.792248010635376,
+      "learning_rate": 3.941976790716287e-05,
+      "loss": 0.6421,
+      "step": 5298
+    },
+    {
+      "epoch": 6.78272,
+      "grad_norm": 0.8279687166213989,
+      "learning_rate": 3.941776710684274e-05,
+      "loss": 0.6959,
+      "step": 5299
+    },
+    {
+      "epoch": 6.784,
+      "grad_norm": 0.764121413230896,
+      "learning_rate": 3.941576630652261e-05,
+      "loss": 0.6368,
+      "step": 5300
+    },
+    {
+      "epoch": 6.78528,
+      "grad_norm": 0.7647132873535156,
+      "learning_rate": 3.9413765506202484e-05,
+      "loss": 0.6331,
+      "step": 5301
+    },
+    {
+      "epoch": 6.78656,
+      "grad_norm": 0.7814575433731079,
+      "learning_rate": 3.9411764705882356e-05,
+      "loss": 0.647,
+      "step": 5302
+    },
+    {
+      "epoch": 6.78784,
+      "grad_norm": 0.7337337732315063,
+      "learning_rate": 3.940976390556223e-05,
+      "loss": 0.5798,
+      "step": 5303
+    },
+    {
+      "epoch": 6.7891200000000005,
+      "grad_norm": 0.7567817568778992,
+      "learning_rate": 3.94077631052421e-05,
+      "loss": 0.6858,
+      "step": 5304
+    },
+    {
+      "epoch": 6.7904,
+      "grad_norm": 0.7290568351745605,
+      "learning_rate": 3.940576230492197e-05,
+      "loss": 0.6001,
+      "step": 5305
+    },
+    {
+      "epoch": 6.79168,
+      "grad_norm": 0.7724208235740662,
+      "learning_rate": 3.9403761504601844e-05,
+      "loss": 0.6107,
+      "step": 5306
+    },
+    {
+      "epoch": 6.79296,
+      "grad_norm": 0.7930365800857544,
+      "learning_rate": 3.9401760704281716e-05,
+      "loss": 0.6773,
+      "step": 5307
+    },
+    {
+      "epoch": 6.79424,
+      "grad_norm": 0.7742063999176025,
+      "learning_rate": 3.939975990396159e-05,
+      "loss": 0.6405,
+      "step": 5308
+    },
+    {
+      "epoch": 6.79552,
+      "grad_norm": 0.808784008026123,
+      "learning_rate": 3.939775910364146e-05,
+      "loss": 0.6853,
+      "step": 5309
+    },
+    {
+      "epoch": 6.7968,
+      "grad_norm": 0.7806289196014404,
+      "learning_rate": 3.939575830332133e-05,
+      "loss": 0.6544,
+      "step": 5310
+    },
+    {
+      "epoch": 6.79808,
+      "grad_norm": 0.7165932655334473,
+      "learning_rate": 3.93937575030012e-05,
+      "loss": 0.5988,
+      "step": 5311
+    },
+    {
+      "epoch": 6.79936,
+      "grad_norm": 0.7418404221534729,
+      "learning_rate": 3.9391756702681075e-05,
+      "loss": 0.6256,
+      "step": 5312
+    },
+    {
+      "epoch": 6.80064,
+      "grad_norm": 0.7737796902656555,
+      "learning_rate": 3.938975590236095e-05,
+      "loss": 0.5891,
+      "step": 5313
+    },
+    {
+      "epoch": 6.80192,
+      "grad_norm": 0.7775099873542786,
+      "learning_rate": 3.938775510204082e-05,
+      "loss": 0.6219,
+      "step": 5314
+    },
+    {
+      "epoch": 6.8032,
+      "grad_norm": 0.7535067796707153,
+      "learning_rate": 3.938575430172069e-05,
+      "loss": 0.5918,
+      "step": 5315
+    },
+    {
+      "epoch": 6.80448,
+      "grad_norm": 0.7628635764122009,
+      "learning_rate": 3.938375350140056e-05,
+      "loss": 0.5637,
+      "step": 5316
+    },
+    {
+      "epoch": 6.80576,
+      "grad_norm": 0.7356747984886169,
+      "learning_rate": 3.9381752701080434e-05,
+      "loss": 0.5817,
+      "step": 5317
+    },
+    {
+      "epoch": 6.80704,
+      "grad_norm": 0.7762408256530762,
+      "learning_rate": 3.9379751900760306e-05,
+      "loss": 0.6193,
+      "step": 5318
+    },
+    {
+      "epoch": 6.80832,
+      "grad_norm": 0.74337238073349,
+      "learning_rate": 3.937775110044018e-05,
+      "loss": 0.6156,
+      "step": 5319
+    },
+    {
+      "epoch": 6.8096,
+      "grad_norm": 0.7793139815330505,
+      "learning_rate": 3.937575030012005e-05,
+      "loss": 0.6768,
+      "step": 5320
+    },
+    {
+      "epoch": 6.81088,
+      "grad_norm": 0.7772572040557861,
+      "learning_rate": 3.937374949979992e-05,
+      "loss": 0.6463,
+      "step": 5321
+    },
+    {
+      "epoch": 6.81216,
+      "grad_norm": 0.7629466652870178,
+      "learning_rate": 3.9371748699479793e-05,
+      "loss": 0.6277,
+      "step": 5322
+    },
+    {
+      "epoch": 6.81344,
+      "grad_norm": 0.7419812083244324,
+      "learning_rate": 3.9369747899159665e-05,
+      "loss": 0.5655,
+      "step": 5323
+    },
+    {
+      "epoch": 6.81472,
+      "grad_norm": 0.7378270030021667,
+      "learning_rate": 3.936774709883954e-05,
+      "loss": 0.5853,
+      "step": 5324
+    },
+    {
+      "epoch": 6.816,
+      "grad_norm": 0.7700802683830261,
+      "learning_rate": 3.936574629851941e-05,
+      "loss": 0.6331,
+      "step": 5325
+    },
+    {
+      "epoch": 6.81728,
+      "grad_norm": 0.7285279631614685,
+      "learning_rate": 3.936374549819928e-05,
+      "loss": 0.6299,
+      "step": 5326
+    },
+    {
+      "epoch": 6.81856,
+      "grad_norm": 0.7254111766815186,
+      "learning_rate": 3.936174469787915e-05,
+      "loss": 0.6142,
+      "step": 5327
+    },
+    {
+      "epoch": 6.81984,
+      "grad_norm": 0.7612648606300354,
+      "learning_rate": 3.9359743897559025e-05,
+      "loss": 0.6056,
+      "step": 5328
+    },
+    {
+      "epoch": 6.82112,
+      "grad_norm": 0.7452360391616821,
+      "learning_rate": 3.9357743097238896e-05,
+      "loss": 0.6107,
+      "step": 5329
+    },
+    {
+      "epoch": 6.8224,
+      "grad_norm": 0.7449943423271179,
+      "learning_rate": 3.935574229691877e-05,
+      "loss": 0.6078,
+      "step": 5330
+    },
+    {
+      "epoch": 6.8236799999999995,
+      "grad_norm": 0.728469729423523,
+      "learning_rate": 3.935374149659864e-05,
+      "loss": 0.593,
+      "step": 5331
+    },
+    {
+      "epoch": 6.82496,
+      "grad_norm": 0.8360829949378967,
+      "learning_rate": 3.935174069627851e-05,
+      "loss": 0.6543,
+      "step": 5332
+    },
+    {
+      "epoch": 6.82624,
+      "grad_norm": 0.8084598779678345,
+      "learning_rate": 3.934973989595839e-05,
+      "loss": 0.662,
+      "step": 5333
+    },
+    {
+      "epoch": 6.82752,
+      "grad_norm": 0.758673369884491,
+      "learning_rate": 3.9347739095638256e-05,
+      "loss": 0.6426,
+      "step": 5334
+    },
+    {
+      "epoch": 6.8288,
+      "grad_norm": 0.7647375464439392,
+      "learning_rate": 3.934573829531813e-05,
+      "loss": 0.5502,
+      "step": 5335
+    },
+    {
+      "epoch": 6.83008,
+      "grad_norm": 0.7250184416770935,
+      "learning_rate": 3.9343737494998e-05,
+      "loss": 0.6173,
+      "step": 5336
+    },
+    {
+      "epoch": 6.83136,
+      "grad_norm": 0.7183948159217834,
+      "learning_rate": 3.934173669467787e-05,
+      "loss": 0.5485,
+      "step": 5337
+    },
+    {
+      "epoch": 6.83264,
+      "grad_norm": 0.7605459690093994,
+      "learning_rate": 3.933973589435774e-05,
+      "loss": 0.5859,
+      "step": 5338
+    },
+    {
+      "epoch": 6.83392,
+      "grad_norm": 0.7369337677955627,
+      "learning_rate": 3.9337735094037615e-05,
+      "loss": 0.6079,
+      "step": 5339
+    },
+    {
+      "epoch": 6.8352,
+      "grad_norm": 0.7497186660766602,
+      "learning_rate": 3.9335734293717494e-05,
+      "loss": 0.5758,
+      "step": 5340
+    },
+    {
+      "epoch": 6.83648,
+      "grad_norm": 0.76474928855896,
+      "learning_rate": 3.9333733493397365e-05,
+      "loss": 0.645,
+      "step": 5341
+    },
+    {
+      "epoch": 6.83776,
+      "grad_norm": 0.7687287926673889,
+      "learning_rate": 3.933173269307723e-05,
+      "loss": 0.6172,
+      "step": 5342
+    },
+    {
+      "epoch": 6.83904,
+      "grad_norm": 0.7535450458526611,
+      "learning_rate": 3.93297318927571e-05,
+      "loss": 0.5974,
+      "step": 5343
+    },
+    {
+      "epoch": 6.84032,
+      "grad_norm": 0.7823262214660645,
+      "learning_rate": 3.9327731092436974e-05,
+      "loss": 0.5917,
+      "step": 5344
+    },
+    {
+      "epoch": 6.8416,
+      "grad_norm": 0.765173077583313,
+      "learning_rate": 3.9325730292116846e-05,
+      "loss": 0.67,
+      "step": 5345
+    },
+    {
+      "epoch": 6.84288,
+      "grad_norm": 0.7360157370567322,
+      "learning_rate": 3.932372949179672e-05,
+      "loss": 0.6097,
+      "step": 5346
+    },
+    {
+      "epoch": 6.8441600000000005,
+      "grad_norm": 0.7717430591583252,
+      "learning_rate": 3.9321728691476597e-05,
+      "loss": 0.6088,
+      "step": 5347
+    },
+    {
+      "epoch": 6.84544,
+      "grad_norm": 0.7661362886428833,
+      "learning_rate": 3.931972789115647e-05,
+      "loss": 0.6418,
+      "step": 5348
+    },
+    {
+      "epoch": 6.84672,
+      "grad_norm": 0.7785105109214783,
+      "learning_rate": 3.931772709083634e-05,
+      "loss": 0.603,
+      "step": 5349
+    },
+    {
+      "epoch": 6.848,
+      "grad_norm": 0.7660150527954102,
+      "learning_rate": 3.9315726290516205e-05,
+      "loss": 0.6095,
+      "step": 5350
+    },
+    {
+      "epoch": 6.84928,
+      "grad_norm": 0.7388876676559448,
+      "learning_rate": 3.931372549019608e-05,
+      "loss": 0.5756,
+      "step": 5351
+    },
+    {
+      "epoch": 6.85056,
+      "grad_norm": 0.7814899682998657,
+      "learning_rate": 3.931172468987595e-05,
+      "loss": 0.6265,
+      "step": 5352
+    },
+    {
+      "epoch": 6.85184,
+      "grad_norm": 0.7103585600852966,
+      "learning_rate": 3.930972388955582e-05,
+      "loss": 0.5408,
+      "step": 5353
+    },
+    {
+      "epoch": 6.85312,
+      "grad_norm": 0.7650427222251892,
+      "learning_rate": 3.93077230892357e-05,
+      "loss": 0.6405,
+      "step": 5354
+    },
+    {
+      "epoch": 6.8544,
+      "grad_norm": 0.7595013380050659,
+      "learning_rate": 3.930572228891557e-05,
+      "loss": 0.6052,
+      "step": 5355
+    },
+    {
+      "epoch": 6.8556799999999996,
+      "grad_norm": 0.7787649631500244,
+      "learning_rate": 3.930372148859544e-05,
+      "loss": 0.6362,
+      "step": 5356
+    },
+    {
+      "epoch": 6.85696,
+      "grad_norm": 0.7371984720230103,
+      "learning_rate": 3.9301720688275315e-05,
+      "loss": 0.6212,
+      "step": 5357
+    },
+    {
+      "epoch": 6.85824,
+      "grad_norm": 0.755862295627594,
+      "learning_rate": 3.929971988795518e-05,
+      "loss": 0.5907,
+      "step": 5358
+    },
+    {
+      "epoch": 6.85952,
+      "grad_norm": 0.7287248969078064,
+      "learning_rate": 3.929771908763505e-05,
+      "loss": 0.5912,
+      "step": 5359
+    },
+    {
+      "epoch": 6.8608,
+      "grad_norm": 0.7712188363075256,
+      "learning_rate": 3.9295718287314924e-05,
+      "loss": 0.6261,
+      "step": 5360
+    },
+    {
+      "epoch": 6.86208,
+      "grad_norm": 0.8140777945518494,
+      "learning_rate": 3.92937174869948e-05,
+      "loss": 0.6052,
+      "step": 5361
+    },
+    {
+      "epoch": 6.86336,
+      "grad_norm": 0.7377832531929016,
+      "learning_rate": 3.9291716686674674e-05,
+      "loss": 0.5631,
+      "step": 5362
+    },
+    {
+      "epoch": 6.86464,
+      "grad_norm": 0.75379878282547,
+      "learning_rate": 3.9289715886354546e-05,
+      "loss": 0.6071,
+      "step": 5363
+    },
+    {
+      "epoch": 6.86592,
+      "grad_norm": 0.7823106646537781,
+      "learning_rate": 3.928771508603442e-05,
+      "loss": 0.6651,
+      "step": 5364
+    },
+    {
+      "epoch": 6.8672,
+      "grad_norm": 0.7541218400001526,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 0.6412,
+      "step": 5365
+    },
+    {
+      "epoch": 6.86848,
+      "grad_norm": 0.7208373546600342,
+      "learning_rate": 3.9283713485394155e-05,
+      "loss": 0.605,
+      "step": 5366
+    },
+    {
+      "epoch": 6.86976,
+      "grad_norm": 0.7709495425224304,
+      "learning_rate": 3.928171268507403e-05,
+      "loss": 0.6041,
+      "step": 5367
+    },
+    {
+      "epoch": 6.87104,
+      "grad_norm": 0.8122734427452087,
+      "learning_rate": 3.9279711884753906e-05,
+      "loss": 0.6647,
+      "step": 5368
+    },
+    {
+      "epoch": 6.87232,
+      "grad_norm": 0.7816949486732483,
+      "learning_rate": 3.927771108443378e-05,
+      "loss": 0.6798,
+      "step": 5369
+    },
+    {
+      "epoch": 6.8736,
+      "grad_norm": 0.7514051198959351,
+      "learning_rate": 3.927571028411365e-05,
+      "loss": 0.5785,
+      "step": 5370
+    },
+    {
+      "epoch": 6.87488,
+      "grad_norm": 0.8315825462341309,
+      "learning_rate": 3.927370948379352e-05,
+      "loss": 0.6399,
+      "step": 5371
+    },
+    {
+      "epoch": 6.8761600000000005,
+      "grad_norm": 0.7676405310630798,
+      "learning_rate": 3.927170868347339e-05,
+      "loss": 0.6142,
+      "step": 5372
+    },
+    {
+      "epoch": 6.87744,
+      "grad_norm": 0.7865943312644958,
+      "learning_rate": 3.9269707883153265e-05,
+      "loss": 0.6319,
+      "step": 5373
+    },
+    {
+      "epoch": 6.8787199999999995,
+      "grad_norm": 0.7144403457641602,
+      "learning_rate": 3.926770708283313e-05,
+      "loss": 0.5566,
+      "step": 5374
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.8031983971595764,
+      "learning_rate": 3.926570628251301e-05,
+      "loss": 0.6461,
+      "step": 5375
+    },
+    {
+      "epoch": 6.88128,
+      "grad_norm": 0.761948823928833,
+      "learning_rate": 3.926370548219288e-05,
+      "loss": 0.6207,
+      "step": 5376
+    },
+    {
+      "epoch": 6.88256,
+      "grad_norm": 0.7399935126304626,
+      "learning_rate": 3.926170468187275e-05,
+      "loss": 0.6265,
+      "step": 5377
+    },
+    {
+      "epoch": 6.88384,
+      "grad_norm": 0.7543573975563049,
+      "learning_rate": 3.9259703881552624e-05,
+      "loss": 0.6648,
+      "step": 5378
+    },
+    {
+      "epoch": 6.88512,
+      "grad_norm": 0.7618049383163452,
+      "learning_rate": 3.9257703081232496e-05,
+      "loss": 0.6135,
+      "step": 5379
+    },
+    {
+      "epoch": 6.8864,
+      "grad_norm": 0.7822875380516052,
+      "learning_rate": 3.925570228091237e-05,
+      "loss": 0.6238,
+      "step": 5380
+    },
+    {
+      "epoch": 6.88768,
+      "grad_norm": 0.7895389795303345,
+      "learning_rate": 3.925370148059224e-05,
+      "loss": 0.613,
+      "step": 5381
+    },
+    {
+      "epoch": 6.88896,
+      "grad_norm": 0.7635522484779358,
+      "learning_rate": 3.925170068027211e-05,
+      "loss": 0.581,
+      "step": 5382
+    },
+    {
+      "epoch": 6.89024,
+      "grad_norm": 0.761661171913147,
+      "learning_rate": 3.9249699879951983e-05,
+      "loss": 0.5799,
+      "step": 5383
+    },
+    {
+      "epoch": 6.89152,
+      "grad_norm": 0.7741929888725281,
+      "learning_rate": 3.9247699079631855e-05,
+      "loss": 0.6283,
+      "step": 5384
+    },
+    {
+      "epoch": 6.8928,
+      "grad_norm": 0.7595140933990479,
+      "learning_rate": 3.924569827931173e-05,
+      "loss": 0.6398,
+      "step": 5385
+    },
+    {
+      "epoch": 6.89408,
+      "grad_norm": 0.7731903195381165,
+      "learning_rate": 3.92436974789916e-05,
+      "loss": 0.5852,
+      "step": 5386
+    },
+    {
+      "epoch": 6.89536,
+      "grad_norm": 0.7807347774505615,
+      "learning_rate": 3.924169667867147e-05,
+      "loss": 0.6452,
+      "step": 5387
+    },
+    {
+      "epoch": 6.89664,
+      "grad_norm": 0.7398995757102966,
+      "learning_rate": 3.923969587835134e-05,
+      "loss": 0.6283,
+      "step": 5388
+    },
+    {
+      "epoch": 6.89792,
+      "grad_norm": 0.7703590393066406,
+      "learning_rate": 3.9237695078031215e-05,
+      "loss": 0.5941,
+      "step": 5389
+    },
+    {
+      "epoch": 6.8992,
+      "grad_norm": 0.8060065507888794,
+      "learning_rate": 3.9235694277711086e-05,
+      "loss": 0.6347,
+      "step": 5390
+    },
+    {
+      "epoch": 6.90048,
+      "grad_norm": 0.7907054424285889,
+      "learning_rate": 3.923369347739096e-05,
+      "loss": 0.6734,
+      "step": 5391
+    },
+    {
+      "epoch": 6.90176,
+      "grad_norm": 0.7742048501968384,
+      "learning_rate": 3.923169267707083e-05,
+      "loss": 0.6362,
+      "step": 5392
+    },
+    {
+      "epoch": 6.90304,
+      "grad_norm": 0.7606136798858643,
+      "learning_rate": 3.92296918767507e-05,
+      "loss": 0.6501,
+      "step": 5393
+    },
+    {
+      "epoch": 6.90432,
+      "grad_norm": 0.7519256472587585,
+      "learning_rate": 3.9227691076430574e-05,
+      "loss": 0.6411,
+      "step": 5394
+    },
+    {
+      "epoch": 6.9056,
+      "grad_norm": 0.7524474859237671,
+      "learning_rate": 3.9225690276110446e-05,
+      "loss": 0.6171,
+      "step": 5395
+    },
+    {
+      "epoch": 6.90688,
+      "grad_norm": 0.8051052093505859,
+      "learning_rate": 3.922368947579032e-05,
+      "loss": 0.6251,
+      "step": 5396
+    },
+    {
+      "epoch": 6.90816,
+      "grad_norm": 0.7711687684059143,
+      "learning_rate": 3.922168867547019e-05,
+      "loss": 0.6428,
+      "step": 5397
+    },
+    {
+      "epoch": 6.90944,
+      "grad_norm": 0.7528000473976135,
+      "learning_rate": 3.921968787515006e-05,
+      "loss": 0.6054,
+      "step": 5398
+    },
+    {
+      "epoch": 6.9107199999999995,
+      "grad_norm": 0.7490524649620056,
+      "learning_rate": 3.921768707482993e-05,
+      "loss": 0.5886,
+      "step": 5399
+    },
+    {
+      "epoch": 6.912,
+      "grad_norm": 0.7919133901596069,
+      "learning_rate": 3.9215686274509805e-05,
+      "loss": 0.5964,
+      "step": 5400
+    },
+    {
+      "epoch": 6.91328,
+      "grad_norm": 0.8205950856208801,
+      "learning_rate": 3.921368547418968e-05,
+      "loss": 0.6063,
+      "step": 5401
+    },
+    {
+      "epoch": 6.91456,
+      "grad_norm": 0.7822876572608948,
+      "learning_rate": 3.921168467386955e-05,
+      "loss": 0.5971,
+      "step": 5402
+    },
+    {
+      "epoch": 6.91584,
+      "grad_norm": 0.7439231276512146,
+      "learning_rate": 3.920968387354943e-05,
+      "loss": 0.5844,
+      "step": 5403
+    },
+    {
+      "epoch": 6.91712,
+      "grad_norm": 0.7648040056228638,
+      "learning_rate": 3.920768307322929e-05,
+      "loss": 0.6627,
+      "step": 5404
+    },
+    {
+      "epoch": 6.9184,
+      "grad_norm": 0.7816339135169983,
+      "learning_rate": 3.9205682272909164e-05,
+      "loss": 0.6154,
+      "step": 5405
+    },
+    {
+      "epoch": 6.91968,
+      "grad_norm": 0.7725468873977661,
+      "learning_rate": 3.9203681472589036e-05,
+      "loss": 0.6032,
+      "step": 5406
+    },
+    {
+      "epoch": 6.92096,
+      "grad_norm": 0.7589234709739685,
+      "learning_rate": 3.920168067226891e-05,
+      "loss": 0.6349,
+      "step": 5407
+    },
+    {
+      "epoch": 6.92224,
+      "grad_norm": 0.7441187500953674,
+      "learning_rate": 3.919967987194878e-05,
+      "loss": 0.5901,
+      "step": 5408
+    },
+    {
+      "epoch": 6.92352,
+      "grad_norm": 0.8013135194778442,
+      "learning_rate": 3.919767907162865e-05,
+      "loss": 0.6455,
+      "step": 5409
+    },
+    {
+      "epoch": 6.9248,
+      "grad_norm": 0.7458381652832031,
+      "learning_rate": 3.919567827130853e-05,
+      "loss": 0.589,
+      "step": 5410
+    },
+    {
+      "epoch": 6.92608,
+      "grad_norm": 0.7533714771270752,
+      "learning_rate": 3.91936774709884e-05,
+      "loss": 0.5634,
+      "step": 5411
+    },
+    {
+      "epoch": 6.92736,
+      "grad_norm": 0.7798910140991211,
+      "learning_rate": 3.919167667066827e-05,
+      "loss": 0.6227,
+      "step": 5412
+    },
+    {
+      "epoch": 6.92864,
+      "grad_norm": 0.7700549960136414,
+      "learning_rate": 3.918967587034814e-05,
+      "loss": 0.6561,
+      "step": 5413
+    },
+    {
+      "epoch": 6.92992,
+      "grad_norm": 0.7472692728042603,
+      "learning_rate": 3.918767507002801e-05,
+      "loss": 0.6014,
+      "step": 5414
+    },
+    {
+      "epoch": 6.9312000000000005,
+      "grad_norm": 0.7239599227905273,
+      "learning_rate": 3.918567426970788e-05,
+      "loss": 0.5716,
+      "step": 5415
+    },
+    {
+      "epoch": 6.93248,
+      "grad_norm": 0.7691959738731384,
+      "learning_rate": 3.9183673469387755e-05,
+      "loss": 0.6603,
+      "step": 5416
+    },
+    {
+      "epoch": 6.93376,
+      "grad_norm": 0.8300781846046448,
+      "learning_rate": 3.918167266906763e-05,
+      "loss": 0.6178,
+      "step": 5417
+    },
+    {
+      "epoch": 6.93504,
+      "grad_norm": 0.7766384482383728,
+      "learning_rate": 3.9179671868747505e-05,
+      "loss": 0.6278,
+      "step": 5418
+    },
+    {
+      "epoch": 6.93632,
+      "grad_norm": 0.7828888297080994,
+      "learning_rate": 3.917767106842738e-05,
+      "loss": 0.6175,
+      "step": 5419
+    },
+    {
+      "epoch": 6.9376,
+      "grad_norm": 0.7322559356689453,
+      "learning_rate": 3.917567026810724e-05,
+      "loss": 0.5805,
+      "step": 5420
+    },
+    {
+      "epoch": 6.93888,
+      "grad_norm": 0.7959531545639038,
+      "learning_rate": 3.9173669467787114e-05,
+      "loss": 0.6813,
+      "step": 5421
+    },
+    {
+      "epoch": 6.94016,
+      "grad_norm": 0.7419981360435486,
+      "learning_rate": 3.9171668667466986e-05,
+      "loss": 0.6288,
+      "step": 5422
+    },
+    {
+      "epoch": 6.94144,
+      "grad_norm": 0.7409440875053406,
+      "learning_rate": 3.916966786714686e-05,
+      "loss": 0.5965,
+      "step": 5423
+    },
+    {
+      "epoch": 6.94272,
+      "grad_norm": 0.7452877759933472,
+      "learning_rate": 3.916766706682673e-05,
+      "loss": 0.6342,
+      "step": 5424
+    },
+    {
+      "epoch": 6.944,
+      "grad_norm": 0.7688354849815369,
+      "learning_rate": 3.916566626650661e-05,
+      "loss": 0.611,
+      "step": 5425
+    },
+    {
+      "epoch": 6.94528,
+      "grad_norm": 0.7577010989189148,
+      "learning_rate": 3.916366546618648e-05,
+      "loss": 0.6594,
+      "step": 5426
+    },
+    {
+      "epoch": 6.94656,
+      "grad_norm": 0.7698047161102295,
+      "learning_rate": 3.916166466586635e-05,
+      "loss": 0.6134,
+      "step": 5427
+    },
+    {
+      "epoch": 6.94784,
+      "grad_norm": 0.7376213073730469,
+      "learning_rate": 3.915966386554622e-05,
+      "loss": 0.6034,
+      "step": 5428
+    },
+    {
+      "epoch": 6.94912,
+      "grad_norm": 0.7267906665802002,
+      "learning_rate": 3.915766306522609e-05,
+      "loss": 0.5529,
+      "step": 5429
+    },
+    {
+      "epoch": 6.9504,
+      "grad_norm": 0.7915275692939758,
+      "learning_rate": 3.915566226490596e-05,
+      "loss": 0.6388,
+      "step": 5430
+    },
+    {
+      "epoch": 6.95168,
+      "grad_norm": 0.8539730906486511,
+      "learning_rate": 3.915366146458583e-05,
+      "loss": 0.6788,
+      "step": 5431
+    },
+    {
+      "epoch": 6.95296,
+      "grad_norm": 0.808468222618103,
+      "learning_rate": 3.915166066426571e-05,
+      "loss": 0.6558,
+      "step": 5432
+    },
+    {
+      "epoch": 6.95424,
+      "grad_norm": 0.8144415616989136,
+      "learning_rate": 3.914965986394558e-05,
+      "loss": 0.6735,
+      "step": 5433
+    },
+    {
+      "epoch": 6.95552,
+      "grad_norm": 0.7447432279586792,
+      "learning_rate": 3.9147659063625455e-05,
+      "loss": 0.5727,
+      "step": 5434
+    },
+    {
+      "epoch": 6.9568,
+      "grad_norm": 0.7698988914489746,
+      "learning_rate": 3.914565826330533e-05,
+      "loss": 0.6138,
+      "step": 5435
+    },
+    {
+      "epoch": 6.95808,
+      "grad_norm": 0.8165004849433899,
+      "learning_rate": 3.914365746298519e-05,
+      "loss": 0.6303,
+      "step": 5436
+    },
+    {
+      "epoch": 6.95936,
+      "grad_norm": 0.771243155002594,
+      "learning_rate": 3.9141656662665064e-05,
+      "loss": 0.6554,
+      "step": 5437
+    },
+    {
+      "epoch": 6.96064,
+      "grad_norm": 0.7581053376197815,
+      "learning_rate": 3.9139655862344936e-05,
+      "loss": 0.6171,
+      "step": 5438
+    },
+    {
+      "epoch": 6.96192,
+      "grad_norm": 0.7608379125595093,
+      "learning_rate": 3.9137655062024814e-05,
+      "loss": 0.5761,
+      "step": 5439
+    },
+    {
+      "epoch": 6.9632,
+      "grad_norm": 0.7522656321525574,
+      "learning_rate": 3.9135654261704686e-05,
+      "loss": 0.5646,
+      "step": 5440
+    },
+    {
+      "epoch": 6.96448,
+      "grad_norm": 0.7819010019302368,
+      "learning_rate": 3.913365346138456e-05,
+      "loss": 0.6509,
+      "step": 5441
+    },
+    {
+      "epoch": 6.9657599999999995,
+      "grad_norm": 0.783852756023407,
+      "learning_rate": 3.913165266106443e-05,
+      "loss": 0.6482,
+      "step": 5442
+    },
+    {
+      "epoch": 6.96704,
+      "grad_norm": 0.7638618350028992,
+      "learning_rate": 3.91296518607443e-05,
+      "loss": 0.6488,
+      "step": 5443
+    },
+    {
+      "epoch": 6.96832,
+      "grad_norm": 0.7872533798217773,
+      "learning_rate": 3.912765106042417e-05,
+      "loss": 0.6425,
+      "step": 5444
+    },
+    {
+      "epoch": 6.9696,
+      "grad_norm": 0.8017598986625671,
+      "learning_rate": 3.912565026010404e-05,
+      "loss": 0.6332,
+      "step": 5445
+    },
+    {
+      "epoch": 6.97088,
+      "grad_norm": 0.784069836139679,
+      "learning_rate": 3.912364945978392e-05,
+      "loss": 0.6307,
+      "step": 5446
+    },
+    {
+      "epoch": 6.97216,
+      "grad_norm": 0.8013083338737488,
+      "learning_rate": 3.912164865946379e-05,
+      "loss": 0.6808,
+      "step": 5447
+    },
+    {
+      "epoch": 6.97344,
+      "grad_norm": 0.7447139024734497,
+      "learning_rate": 3.911964785914366e-05,
+      "loss": 0.6048,
+      "step": 5448
+    },
+    {
+      "epoch": 6.97472,
+      "grad_norm": 0.7573011517524719,
+      "learning_rate": 3.911764705882353e-05,
+      "loss": 0.5808,
+      "step": 5449
+    },
+    {
+      "epoch": 6.976,
+      "grad_norm": 0.7706106901168823,
+      "learning_rate": 3.9115646258503405e-05,
+      "loss": 0.5773,
+      "step": 5450
+    },
+    {
+      "epoch": 6.97728,
+      "grad_norm": 0.8079164624214172,
+      "learning_rate": 3.9113645458183276e-05,
+      "loss": 0.6569,
+      "step": 5451
+    },
+    {
+      "epoch": 6.97856,
+      "grad_norm": 0.720331609249115,
+      "learning_rate": 3.911164465786314e-05,
+      "loss": 0.6154,
+      "step": 5452
+    },
+    {
+      "epoch": 6.97984,
+      "grad_norm": 0.7962987422943115,
+      "learning_rate": 3.910964385754302e-05,
+      "loss": 0.6097,
+      "step": 5453
+    },
+    {
+      "epoch": 6.98112,
+      "grad_norm": 0.7449235916137695,
+      "learning_rate": 3.910764305722289e-05,
+      "loss": 0.585,
+      "step": 5454
+    },
+    {
+      "epoch": 6.9824,
+      "grad_norm": 0.7487616539001465,
+      "learning_rate": 3.9105642256902764e-05,
+      "loss": 0.6123,
+      "step": 5455
+    },
+    {
+      "epoch": 6.98368,
+      "grad_norm": 0.775963544845581,
+      "learning_rate": 3.9103641456582636e-05,
+      "loss": 0.6661,
+      "step": 5456
+    },
+    {
+      "epoch": 6.98496,
+      "grad_norm": 0.710671603679657,
+      "learning_rate": 3.910164065626251e-05,
+      "loss": 0.5817,
+      "step": 5457
+    },
+    {
+      "epoch": 6.9862400000000004,
+      "grad_norm": 0.7331653237342834,
+      "learning_rate": 3.909963985594238e-05,
+      "loss": 0.5741,
+      "step": 5458
+    },
+    {
+      "epoch": 6.98752,
+      "grad_norm": 0.7193625569343567,
+      "learning_rate": 3.909763905562225e-05,
+      "loss": 0.5515,
+      "step": 5459
+    },
+    {
+      "epoch": 6.9888,
+      "grad_norm": 0.7541481852531433,
+      "learning_rate": 3.909563825530212e-05,
+      "loss": 0.595,
+      "step": 5460
+    },
+    {
+      "epoch": 6.99008,
+      "grad_norm": 0.7617225646972656,
+      "learning_rate": 3.9093637454981995e-05,
+      "loss": 0.6112,
+      "step": 5461
+    },
+    {
+      "epoch": 6.99136,
+      "grad_norm": 0.7736883163452148,
+      "learning_rate": 3.909163665466187e-05,
+      "loss": 0.6292,
+      "step": 5462
+    },
+    {
+      "epoch": 6.99264,
+      "grad_norm": 0.796480655670166,
+      "learning_rate": 3.908963585434174e-05,
+      "loss": 0.6122,
+      "step": 5463
+    },
+    {
+      "epoch": 6.99392,
+      "grad_norm": 0.7853350043296814,
+      "learning_rate": 3.908763505402161e-05,
+      "loss": 0.6188,
+      "step": 5464
+    },
+    {
+      "epoch": 6.9952,
+      "grad_norm": 0.7956370115280151,
+      "learning_rate": 3.908563425370148e-05,
+      "loss": 0.6699,
+      "step": 5465
+    },
+    {
+      "epoch": 6.99648,
+      "grad_norm": 0.7397189736366272,
+      "learning_rate": 3.9083633453381354e-05,
+      "loss": 0.647,
+      "step": 5466
+    },
+    {
+      "epoch": 6.9977599999999995,
+      "grad_norm": 0.6985254287719727,
+      "learning_rate": 3.9081632653061226e-05,
+      "loss": 0.5586,
+      "step": 5467
+    },
+    {
+      "epoch": 6.99904,
+      "grad_norm": 0.8187921047210693,
+      "learning_rate": 3.90796318527411e-05,
+      "loss": 0.6705,
+      "step": 5468
+    },
+    {
+      "epoch": 7.00032,
+      "grad_norm": 1.61716628074646,
+      "learning_rate": 3.907763105242097e-05,
+      "loss": 1.0948,
+      "step": 5469
+    },
+    {
+      "epoch": 7.0016,
+      "grad_norm": 0.7572510838508606,
+      "learning_rate": 3.907563025210084e-05,
+      "loss": 0.5787,
+      "step": 5470
+    },
+    {
+      "epoch": 7.00288,
+      "grad_norm": 0.7337214350700378,
+      "learning_rate": 3.9073629451780714e-05,
+      "loss": 0.579,
+      "step": 5471
+    },
+    {
+      "epoch": 7.00416,
+      "grad_norm": 0.7118316888809204,
+      "learning_rate": 3.9071628651460585e-05,
+      "loss": 0.5866,
+      "step": 5472
+    },
+    {
+      "epoch": 7.00544,
+      "grad_norm": 0.7408485412597656,
+      "learning_rate": 3.906962785114046e-05,
+      "loss": 0.6048,
+      "step": 5473
+    },
+    {
+      "epoch": 7.00672,
+      "grad_norm": 0.7442348003387451,
+      "learning_rate": 3.906762705082033e-05,
+      "loss": 0.5916,
+      "step": 5474
+    },
+    {
+      "epoch": 7.008,
+      "grad_norm": 0.7323760390281677,
+      "learning_rate": 3.90656262505002e-05,
+      "loss": 0.6146,
+      "step": 5475
+    },
+    {
+      "epoch": 7.00928,
+      "grad_norm": 0.7773112654685974,
+      "learning_rate": 3.906362545018007e-05,
+      "loss": 0.6156,
+      "step": 5476
+    },
+    {
+      "epoch": 7.01056,
+      "grad_norm": 0.7394047975540161,
+      "learning_rate": 3.9061624649859945e-05,
+      "loss": 0.5846,
+      "step": 5477
+    },
+    {
+      "epoch": 7.01184,
+      "grad_norm": 0.7753980755805969,
+      "learning_rate": 3.9059623849539817e-05,
+      "loss": 0.6151,
+      "step": 5478
+    },
+    {
+      "epoch": 7.01312,
+      "grad_norm": 0.7161932587623596,
+      "learning_rate": 3.905762304921969e-05,
+      "loss": 0.5535,
+      "step": 5479
+    },
+    {
+      "epoch": 7.0144,
+      "grad_norm": 0.7652511596679688,
+      "learning_rate": 3.905562224889956e-05,
+      "loss": 0.6409,
+      "step": 5480
+    },
+    {
+      "epoch": 7.01568,
+      "grad_norm": 0.7910155653953552,
+      "learning_rate": 3.905362144857944e-05,
+      "loss": 0.6257,
+      "step": 5481
+    },
+    {
+      "epoch": 7.01696,
+      "grad_norm": 0.7763440012931824,
+      "learning_rate": 3.9051620648259304e-05,
+      "loss": 0.5843,
+      "step": 5482
+    },
+    {
+      "epoch": 7.01824,
+      "grad_norm": 0.806786298751831,
+      "learning_rate": 3.9049619847939176e-05,
+      "loss": 0.5773,
+      "step": 5483
+    },
+    {
+      "epoch": 7.01952,
+      "grad_norm": 0.7779539227485657,
+      "learning_rate": 3.904761904761905e-05,
+      "loss": 0.6031,
+      "step": 5484
+    },
+    {
+      "epoch": 7.0208,
+      "grad_norm": 0.7639070153236389,
+      "learning_rate": 3.904561824729892e-05,
+      "loss": 0.5415,
+      "step": 5485
+    },
+    {
+      "epoch": 7.02208,
+      "grad_norm": 0.7557491064071655,
+      "learning_rate": 3.904361744697879e-05,
+      "loss": 0.5762,
+      "step": 5486
+    },
+    {
+      "epoch": 7.02336,
+      "grad_norm": 0.7613850235939026,
+      "learning_rate": 3.904161664665866e-05,
+      "loss": 0.6029,
+      "step": 5487
+    },
+    {
+      "epoch": 7.02464,
+      "grad_norm": 0.8083841800689697,
+      "learning_rate": 3.903961584633854e-05,
+      "loss": 0.6372,
+      "step": 5488
+    },
+    {
+      "epoch": 7.02592,
+      "grad_norm": 0.7833201885223389,
+      "learning_rate": 3.9037615046018414e-05,
+      "loss": 0.6152,
+      "step": 5489
+    },
+    {
+      "epoch": 7.0272,
+      "grad_norm": 0.7935197949409485,
+      "learning_rate": 3.903561424569828e-05,
+      "loss": 0.5762,
+      "step": 5490
+    },
+    {
+      "epoch": 7.02848,
+      "grad_norm": 0.7885215878486633,
+      "learning_rate": 3.903361344537815e-05,
+      "loss": 0.6361,
+      "step": 5491
+    },
+    {
+      "epoch": 7.02976,
+      "grad_norm": 0.770488440990448,
+      "learning_rate": 3.903161264505802e-05,
+      "loss": 0.6102,
+      "step": 5492
+    },
+    {
+      "epoch": 7.03104,
+      "grad_norm": 0.7559887170791626,
+      "learning_rate": 3.9029611844737894e-05,
+      "loss": 0.601,
+      "step": 5493
+    },
+    {
+      "epoch": 7.03232,
+      "grad_norm": 0.8430263996124268,
+      "learning_rate": 3.9027611044417766e-05,
+      "loss": 0.6871,
+      "step": 5494
+    },
+    {
+      "epoch": 7.0336,
+      "grad_norm": 0.729887068271637,
+      "learning_rate": 3.9025610244097645e-05,
+      "loss": 0.5806,
+      "step": 5495
+    },
+    {
+      "epoch": 7.03488,
+      "grad_norm": 0.7851121425628662,
+      "learning_rate": 3.902360944377752e-05,
+      "loss": 0.5748,
+      "step": 5496
+    },
+    {
+      "epoch": 7.03616,
+      "grad_norm": 0.7651329040527344,
+      "learning_rate": 3.902160864345739e-05,
+      "loss": 0.6004,
+      "step": 5497
+    },
+    {
+      "epoch": 7.03744,
+      "grad_norm": 0.7575582265853882,
+      "learning_rate": 3.9019607843137254e-05,
+      "loss": 0.6509,
+      "step": 5498
+    },
+    {
+      "epoch": 7.03872,
+      "grad_norm": 0.7376059293746948,
+      "learning_rate": 3.9017607042817126e-05,
+      "loss": 0.5799,
+      "step": 5499
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.723619282245636,
+      "learning_rate": 3.9015606242497e-05,
+      "loss": 0.6004,
+      "step": 5500
+    },
+    {
+      "epoch": 7.04128,
+      "grad_norm": 0.803107738494873,
+      "learning_rate": 3.901360544217687e-05,
+      "loss": 0.5998,
+      "step": 5501
+    },
+    {
+      "epoch": 7.04256,
+      "grad_norm": 0.8073145151138306,
+      "learning_rate": 3.901160464185675e-05,
+      "loss": 0.5843,
+      "step": 5502
+    },
+    {
+      "epoch": 7.04384,
+      "grad_norm": 0.8166995644569397,
+      "learning_rate": 3.900960384153662e-05,
+      "loss": 0.6536,
+      "step": 5503
+    },
+    {
+      "epoch": 7.04512,
+      "grad_norm": 0.7663634419441223,
+      "learning_rate": 3.900760304121649e-05,
+      "loss": 0.6081,
+      "step": 5504
+    },
+    {
+      "epoch": 7.0464,
+      "grad_norm": 0.8213377594947815,
+      "learning_rate": 3.9005602240896364e-05,
+      "loss": 0.5922,
+      "step": 5505
+    },
+    {
+      "epoch": 7.04768,
+      "grad_norm": 0.7828342318534851,
+      "learning_rate": 3.900360144057623e-05,
+      "loss": 0.6146,
+      "step": 5506
+    },
+    {
+      "epoch": 7.04896,
+      "grad_norm": 0.8070378303527832,
+      "learning_rate": 3.90016006402561e-05,
+      "loss": 0.6231,
+      "step": 5507
+    },
+    {
+      "epoch": 7.05024,
+      "grad_norm": 0.7503546476364136,
+      "learning_rate": 3.899959983993597e-05,
+      "loss": 0.5776,
+      "step": 5508
+    },
+    {
+      "epoch": 7.05152,
+      "grad_norm": 0.8065857291221619,
+      "learning_rate": 3.899759903961585e-05,
+      "loss": 0.6244,
+      "step": 5509
+    },
+    {
+      "epoch": 7.0528,
+      "grad_norm": 0.7999621629714966,
+      "learning_rate": 3.899559823929572e-05,
+      "loss": 0.5803,
+      "step": 5510
+    },
+    {
+      "epoch": 7.05408,
+      "grad_norm": 0.7979943752288818,
+      "learning_rate": 3.8993597438975595e-05,
+      "loss": 0.5895,
+      "step": 5511
+    },
+    {
+      "epoch": 7.05536,
+      "grad_norm": 0.8131271004676819,
+      "learning_rate": 3.8991596638655467e-05,
+      "loss": 0.615,
+      "step": 5512
+    },
+    {
+      "epoch": 7.05664,
+      "grad_norm": 0.7797995209693909,
+      "learning_rate": 3.898959583833534e-05,
+      "loss": 0.5712,
+      "step": 5513
+    },
+    {
+      "epoch": 7.05792,
+      "grad_norm": 0.8302449584007263,
+      "learning_rate": 3.8987595038015203e-05,
+      "loss": 0.6598,
+      "step": 5514
+    },
+    {
+      "epoch": 7.0592,
+      "grad_norm": 0.817923367023468,
+      "learning_rate": 3.8985594237695075e-05,
+      "loss": 0.627,
+      "step": 5515
+    },
+    {
+      "epoch": 7.06048,
+      "grad_norm": 0.8435928821563721,
+      "learning_rate": 3.8983593437374954e-05,
+      "loss": 0.6078,
+      "step": 5516
+    },
+    {
+      "epoch": 7.06176,
+      "grad_norm": 0.7841606736183167,
+      "learning_rate": 3.8981592637054826e-05,
+      "loss": 0.6069,
+      "step": 5517
+    },
+    {
+      "epoch": 7.06304,
+      "grad_norm": 0.7932061553001404,
+      "learning_rate": 3.89795918367347e-05,
+      "loss": 0.5993,
+      "step": 5518
+    },
+    {
+      "epoch": 7.06432,
+      "grad_norm": 0.8158665895462036,
+      "learning_rate": 3.897759103641457e-05,
+      "loss": 0.6412,
+      "step": 5519
+    },
+    {
+      "epoch": 7.0656,
+      "grad_norm": 0.7837883830070496,
+      "learning_rate": 3.897559023609444e-05,
+      "loss": 0.6011,
+      "step": 5520
+    },
+    {
+      "epoch": 7.06688,
+      "grad_norm": 0.8409795761108398,
+      "learning_rate": 3.897358943577431e-05,
+      "loss": 0.5954,
+      "step": 5521
+    },
+    {
+      "epoch": 7.06816,
+      "grad_norm": 0.818547785282135,
+      "learning_rate": 3.897158863545418e-05,
+      "loss": 0.6266,
+      "step": 5522
+    },
+    {
+      "epoch": 7.06944,
+      "grad_norm": 0.8386947512626648,
+      "learning_rate": 3.896958783513406e-05,
+      "loss": 0.6724,
+      "step": 5523
+    },
+    {
+      "epoch": 7.07072,
+      "grad_norm": 0.7939981818199158,
+      "learning_rate": 3.896758703481393e-05,
+      "loss": 0.571,
+      "step": 5524
+    },
+    {
+      "epoch": 7.072,
+      "grad_norm": 0.8104544281959534,
+      "learning_rate": 3.89655862344938e-05,
+      "loss": 0.5906,
+      "step": 5525
+    },
+    {
+      "epoch": 7.07328,
+      "grad_norm": 0.805988609790802,
+      "learning_rate": 3.896358543417367e-05,
+      "loss": 0.6188,
+      "step": 5526
+    },
+    {
+      "epoch": 7.07456,
+      "grad_norm": 0.7390158176422119,
+      "learning_rate": 3.8961584633853544e-05,
+      "loss": 0.5718,
+      "step": 5527
+    },
+    {
+      "epoch": 7.07584,
+      "grad_norm": 0.798747181892395,
+      "learning_rate": 3.8959583833533416e-05,
+      "loss": 0.5949,
+      "step": 5528
+    },
+    {
+      "epoch": 7.07712,
+      "grad_norm": 0.8065986633300781,
+      "learning_rate": 3.895758303321329e-05,
+      "loss": 0.6211,
+      "step": 5529
+    },
+    {
+      "epoch": 7.0784,
+      "grad_norm": 0.8075457215309143,
+      "learning_rate": 3.895558223289316e-05,
+      "loss": 0.5692,
+      "step": 5530
+    },
+    {
+      "epoch": 7.07968,
+      "grad_norm": 0.8485049605369568,
+      "learning_rate": 3.895358143257303e-05,
+      "loss": 0.6845,
+      "step": 5531
+    },
+    {
+      "epoch": 7.08096,
+      "grad_norm": 0.7971506118774414,
+      "learning_rate": 3.8951580632252904e-05,
+      "loss": 0.615,
+      "step": 5532
+    },
+    {
+      "epoch": 7.08224,
+      "grad_norm": 0.7371413707733154,
+      "learning_rate": 3.8949579831932776e-05,
+      "loss": 0.5338,
+      "step": 5533
+    },
+    {
+      "epoch": 7.08352,
+      "grad_norm": 0.7684740424156189,
+      "learning_rate": 3.894757903161265e-05,
+      "loss": 0.5948,
+      "step": 5534
+    },
+    {
+      "epoch": 7.0848,
+      "grad_norm": 0.8026983737945557,
+      "learning_rate": 3.894557823129252e-05,
+      "loss": 0.5749,
+      "step": 5535
+    },
+    {
+      "epoch": 7.08608,
+      "grad_norm": 0.8087561726570129,
+      "learning_rate": 3.894357743097239e-05,
+      "loss": 0.604,
+      "step": 5536
+    },
+    {
+      "epoch": 7.08736,
+      "grad_norm": 0.7562788724899292,
+      "learning_rate": 3.894157663065226e-05,
+      "loss": 0.5552,
+      "step": 5537
+    },
+    {
+      "epoch": 7.08864,
+      "grad_norm": 0.8130525946617126,
+      "learning_rate": 3.8939575830332135e-05,
+      "loss": 0.6516,
+      "step": 5538
+    },
+    {
+      "epoch": 7.08992,
+      "grad_norm": 0.7703153491020203,
+      "learning_rate": 3.893757503001201e-05,
+      "loss": 0.5952,
+      "step": 5539
+    },
+    {
+      "epoch": 7.0912,
+      "grad_norm": 0.8727670311927795,
+      "learning_rate": 3.893557422969188e-05,
+      "loss": 0.6897,
+      "step": 5540
+    },
+    {
+      "epoch": 7.09248,
+      "grad_norm": 0.7629865407943726,
+      "learning_rate": 3.893357342937175e-05,
+      "loss": 0.6019,
+      "step": 5541
+    },
+    {
+      "epoch": 7.09376,
+      "grad_norm": 0.7899230122566223,
+      "learning_rate": 3.893157262905162e-05,
+      "loss": 0.6197,
+      "step": 5542
+    },
+    {
+      "epoch": 7.09504,
+      "grad_norm": 0.7999109625816345,
+      "learning_rate": 3.8929571828731494e-05,
+      "loss": 0.6019,
+      "step": 5543
+    },
+    {
+      "epoch": 7.09632,
+      "grad_norm": 0.7331894040107727,
+      "learning_rate": 3.8927571028411366e-05,
+      "loss": 0.5198,
+      "step": 5544
+    },
+    {
+      "epoch": 7.0976,
+      "grad_norm": 0.8195477724075317,
+      "learning_rate": 3.892557022809124e-05,
+      "loss": 0.6371,
+      "step": 5545
+    },
+    {
+      "epoch": 7.09888,
+      "grad_norm": 0.8432417511940002,
+      "learning_rate": 3.892356942777111e-05,
+      "loss": 0.6299,
+      "step": 5546
+    },
+    {
+      "epoch": 7.10016,
+      "grad_norm": 0.8083561062812805,
+      "learning_rate": 3.892156862745098e-05,
+      "loss": 0.587,
+      "step": 5547
+    },
+    {
+      "epoch": 7.10144,
+      "grad_norm": 0.8173826932907104,
+      "learning_rate": 3.891956782713085e-05,
+      "loss": 0.6131,
+      "step": 5548
+    },
+    {
+      "epoch": 7.10272,
+      "grad_norm": 0.786679208278656,
+      "learning_rate": 3.8917567026810725e-05,
+      "loss": 0.5952,
+      "step": 5549
+    },
+    {
+      "epoch": 7.104,
+      "grad_norm": 0.7349452376365662,
+      "learning_rate": 3.89155662264906e-05,
+      "loss": 0.5611,
+      "step": 5550
+    },
+    {
+      "epoch": 7.10528,
+      "grad_norm": 0.7959557771682739,
+      "learning_rate": 3.891356542617047e-05,
+      "loss": 0.6117,
+      "step": 5551
+    },
+    {
+      "epoch": 7.10656,
+      "grad_norm": 0.7580567002296448,
+      "learning_rate": 3.891156462585034e-05,
+      "loss": 0.6021,
+      "step": 5552
+    },
+    {
+      "epoch": 7.10784,
+      "grad_norm": 0.797226071357727,
+      "learning_rate": 3.890956382553021e-05,
+      "loss": 0.6083,
+      "step": 5553
+    },
+    {
+      "epoch": 7.10912,
+      "grad_norm": 0.8558824062347412,
+      "learning_rate": 3.8907563025210084e-05,
+      "loss": 0.6534,
+      "step": 5554
+    },
+    {
+      "epoch": 7.1104,
+      "grad_norm": 0.7972661256790161,
+      "learning_rate": 3.8905562224889956e-05,
+      "loss": 0.6106,
+      "step": 5555
+    },
+    {
+      "epoch": 7.11168,
+      "grad_norm": 0.8266356587409973,
+      "learning_rate": 3.890356142456983e-05,
+      "loss": 0.6189,
+      "step": 5556
+    },
+    {
+      "epoch": 7.11296,
+      "grad_norm": 0.8113858103752136,
+      "learning_rate": 3.89015606242497e-05,
+      "loss": 0.5916,
+      "step": 5557
+    },
+    {
+      "epoch": 7.11424,
+      "grad_norm": 0.7940702438354492,
+      "learning_rate": 3.889955982392957e-05,
+      "loss": 0.6358,
+      "step": 5558
+    },
+    {
+      "epoch": 7.11552,
+      "grad_norm": 0.8410483598709106,
+      "learning_rate": 3.889755902360945e-05,
+      "loss": 0.6397,
+      "step": 5559
+    },
+    {
+      "epoch": 7.1168,
+      "grad_norm": 0.897203266620636,
+      "learning_rate": 3.8895558223289316e-05,
+      "loss": 0.6482,
+      "step": 5560
+    },
+    {
+      "epoch": 7.11808,
+      "grad_norm": 0.781025230884552,
+      "learning_rate": 3.889355742296919e-05,
+      "loss": 0.5471,
+      "step": 5561
+    },
+    {
+      "epoch": 7.11936,
+      "grad_norm": 0.7567470073699951,
+      "learning_rate": 3.889155662264906e-05,
+      "loss": 0.5391,
+      "step": 5562
+    },
+    {
+      "epoch": 7.12064,
+      "grad_norm": 0.8244337439537048,
+      "learning_rate": 3.888955582232893e-05,
+      "loss": 0.6146,
+      "step": 5563
+    },
+    {
+      "epoch": 7.12192,
+      "grad_norm": 0.8109177947044373,
+      "learning_rate": 3.88875550220088e-05,
+      "loss": 0.5691,
+      "step": 5564
+    },
+    {
+      "epoch": 7.1232,
+      "grad_norm": 0.7812331914901733,
+      "learning_rate": 3.8885554221688675e-05,
+      "loss": 0.6075,
+      "step": 5565
+    },
+    {
+      "epoch": 7.12448,
+      "grad_norm": 0.7723196148872375,
+      "learning_rate": 3.8883553421368554e-05,
+      "loss": 0.6153,
+      "step": 5566
+    },
+    {
+      "epoch": 7.12576,
+      "grad_norm": 0.790398120880127,
+      "learning_rate": 3.8881552621048425e-05,
+      "loss": 0.6043,
+      "step": 5567
+    },
+    {
+      "epoch": 7.12704,
+      "grad_norm": 0.7210956811904907,
+      "learning_rate": 3.887955182072829e-05,
+      "loss": 0.5861,
+      "step": 5568
+    },
+    {
+      "epoch": 7.12832,
+      "grad_norm": 0.7791348695755005,
+      "learning_rate": 3.887755102040816e-05,
+      "loss": 0.618,
+      "step": 5569
+    },
+    {
+      "epoch": 7.1296,
+      "grad_norm": 0.793732225894928,
+      "learning_rate": 3.8875550220088034e-05,
+      "loss": 0.6127,
+      "step": 5570
+    },
+    {
+      "epoch": 7.13088,
+      "grad_norm": 0.7834117412567139,
+      "learning_rate": 3.8873549419767906e-05,
+      "loss": 0.5885,
+      "step": 5571
+    },
+    {
+      "epoch": 7.13216,
+      "grad_norm": 0.8224747180938721,
+      "learning_rate": 3.887154861944778e-05,
+      "loss": 0.6073,
+      "step": 5572
+    },
+    {
+      "epoch": 7.13344,
+      "grad_norm": 0.7827395796775818,
+      "learning_rate": 3.8869547819127657e-05,
+      "loss": 0.5888,
+      "step": 5573
+    },
+    {
+      "epoch": 7.13472,
+      "grad_norm": 0.7737107872962952,
+      "learning_rate": 3.886754701880753e-05,
+      "loss": 0.6167,
+      "step": 5574
+    },
+    {
+      "epoch": 7.136,
+      "grad_norm": 0.8110948801040649,
+      "learning_rate": 3.88655462184874e-05,
+      "loss": 0.6296,
+      "step": 5575
+    },
+    {
+      "epoch": 7.13728,
+      "grad_norm": 0.7978016138076782,
+      "learning_rate": 3.8863545418167265e-05,
+      "loss": 0.6459,
+      "step": 5576
+    },
+    {
+      "epoch": 7.13856,
+      "grad_norm": 0.7949130535125732,
+      "learning_rate": 3.886154461784714e-05,
+      "loss": 0.622,
+      "step": 5577
+    },
+    {
+      "epoch": 7.13984,
+      "grad_norm": 0.7578536868095398,
+      "learning_rate": 3.885954381752701e-05,
+      "loss": 0.5487,
+      "step": 5578
+    },
+    {
+      "epoch": 7.14112,
+      "grad_norm": 0.778209388256073,
+      "learning_rate": 3.885754301720688e-05,
+      "loss": 0.6221,
+      "step": 5579
+    },
+    {
+      "epoch": 7.1424,
+      "grad_norm": 0.7351183891296387,
+      "learning_rate": 3.885554221688676e-05,
+      "loss": 0.5419,
+      "step": 5580
+    },
+    {
+      "epoch": 7.14368,
+      "grad_norm": 0.811617374420166,
+      "learning_rate": 3.885354141656663e-05,
+      "loss": 0.6545,
+      "step": 5581
+    },
+    {
+      "epoch": 7.14496,
+      "grad_norm": 0.7293868064880371,
+      "learning_rate": 3.88515406162465e-05,
+      "loss": 0.5763,
+      "step": 5582
+    },
+    {
+      "epoch": 7.14624,
+      "grad_norm": 0.7808999419212341,
+      "learning_rate": 3.8849539815926375e-05,
+      "loss": 0.5435,
+      "step": 5583
+    },
+    {
+      "epoch": 7.14752,
+      "grad_norm": 0.7816929221153259,
+      "learning_rate": 3.884753901560624e-05,
+      "loss": 0.6055,
+      "step": 5584
+    },
+    {
+      "epoch": 7.1488,
+      "grad_norm": 0.8008720278739929,
+      "learning_rate": 3.884553821528611e-05,
+      "loss": 0.6337,
+      "step": 5585
+    },
+    {
+      "epoch": 7.15008,
+      "grad_norm": 0.7841701507568359,
+      "learning_rate": 3.8843537414965984e-05,
+      "loss": 0.6371,
+      "step": 5586
+    },
+    {
+      "epoch": 7.15136,
+      "grad_norm": 0.7534654140472412,
+      "learning_rate": 3.884153661464586e-05,
+      "loss": 0.6057,
+      "step": 5587
+    },
+    {
+      "epoch": 7.15264,
+      "grad_norm": 0.7844194769859314,
+      "learning_rate": 3.8839535814325734e-05,
+      "loss": 0.6094,
+      "step": 5588
+    },
+    {
+      "epoch": 7.15392,
+      "grad_norm": 0.8300028443336487,
+      "learning_rate": 3.8837535014005606e-05,
+      "loss": 0.6388,
+      "step": 5589
+    },
+    {
+      "epoch": 7.1552,
+      "grad_norm": 0.7904089093208313,
+      "learning_rate": 3.883553421368548e-05,
+      "loss": 0.601,
+      "step": 5590
+    },
+    {
+      "epoch": 7.15648,
+      "grad_norm": 0.7863451838493347,
+      "learning_rate": 3.883353341336535e-05,
+      "loss": 0.6357,
+      "step": 5591
+    },
+    {
+      "epoch": 7.15776,
+      "grad_norm": 0.7628374695777893,
+      "learning_rate": 3.8831532613045215e-05,
+      "loss": 0.5687,
+      "step": 5592
+    },
+    {
+      "epoch": 7.15904,
+      "grad_norm": 0.8106046915054321,
+      "learning_rate": 3.882953181272509e-05,
+      "loss": 0.642,
+      "step": 5593
+    },
+    {
+      "epoch": 7.16032,
+      "grad_norm": 0.7651321887969971,
+      "learning_rate": 3.8827531012404966e-05,
+      "loss": 0.608,
+      "step": 5594
+    },
+    {
+      "epoch": 7.1616,
+      "grad_norm": 0.7789268493652344,
+      "learning_rate": 3.882553021208484e-05,
+      "loss": 0.6187,
+      "step": 5595
+    },
+    {
+      "epoch": 7.16288,
+      "grad_norm": 0.7903891801834106,
+      "learning_rate": 3.882352941176471e-05,
+      "loss": 0.5915,
+      "step": 5596
+    },
+    {
+      "epoch": 7.16416,
+      "grad_norm": 0.7863123416900635,
+      "learning_rate": 3.882152861144458e-05,
+      "loss": 0.6097,
+      "step": 5597
+    },
+    {
+      "epoch": 7.16544,
+      "grad_norm": 0.7845132350921631,
+      "learning_rate": 3.881952781112445e-05,
+      "loss": 0.5963,
+      "step": 5598
+    },
+    {
+      "epoch": 7.16672,
+      "grad_norm": 0.7682058811187744,
+      "learning_rate": 3.8817527010804325e-05,
+      "loss": 0.5804,
+      "step": 5599
+    },
+    {
+      "epoch": 7.168,
+      "grad_norm": 0.7563507556915283,
+      "learning_rate": 3.881552621048419e-05,
+      "loss": 0.5514,
+      "step": 5600
+    },
+    {
+      "epoch": 7.16928,
+      "grad_norm": 0.7621744275093079,
+      "learning_rate": 3.881352541016407e-05,
+      "loss": 0.5595,
+      "step": 5601
+    },
+    {
+      "epoch": 7.17056,
+      "grad_norm": 0.773902416229248,
+      "learning_rate": 3.881152460984394e-05,
+      "loss": 0.5483,
+      "step": 5602
+    },
+    {
+      "epoch": 7.1718399999999995,
+      "grad_norm": 0.7652880549430847,
+      "learning_rate": 3.880952380952381e-05,
+      "loss": 0.5439,
+      "step": 5603
+    },
+    {
+      "epoch": 7.17312,
+      "grad_norm": 0.8013029098510742,
+      "learning_rate": 3.8807523009203684e-05,
+      "loss": 0.583,
+      "step": 5604
+    },
+    {
+      "epoch": 7.1744,
+      "grad_norm": 0.7590806484222412,
+      "learning_rate": 3.8805522208883556e-05,
+      "loss": 0.569,
+      "step": 5605
+    },
+    {
+      "epoch": 7.17568,
+      "grad_norm": 0.8396809697151184,
+      "learning_rate": 3.880352140856343e-05,
+      "loss": 0.6549,
+      "step": 5606
+    },
+    {
+      "epoch": 7.17696,
+      "grad_norm": 0.796314537525177,
+      "learning_rate": 3.88015206082433e-05,
+      "loss": 0.6438,
+      "step": 5607
+    },
+    {
+      "epoch": 7.17824,
+      "grad_norm": 0.7749351859092712,
+      "learning_rate": 3.879951980792317e-05,
+      "loss": 0.5674,
+      "step": 5608
+    },
+    {
+      "epoch": 7.17952,
+      "grad_norm": 0.7739863395690918,
+      "learning_rate": 3.8797519007603043e-05,
+      "loss": 0.5771,
+      "step": 5609
+    },
+    {
+      "epoch": 7.1808,
+      "grad_norm": 0.808387279510498,
+      "learning_rate": 3.8795518207282915e-05,
+      "loss": 0.5839,
+      "step": 5610
+    },
+    {
+      "epoch": 7.18208,
+      "grad_norm": 0.7463328838348389,
+      "learning_rate": 3.879351740696279e-05,
+      "loss": 0.5557,
+      "step": 5611
+    },
+    {
+      "epoch": 7.18336,
+      "grad_norm": 0.7860255837440491,
+      "learning_rate": 3.879151660664266e-05,
+      "loss": 0.5787,
+      "step": 5612
+    },
+    {
+      "epoch": 7.18464,
+      "grad_norm": 0.841917097568512,
+      "learning_rate": 3.878951580632253e-05,
+      "loss": 0.6375,
+      "step": 5613
+    },
+    {
+      "epoch": 7.18592,
+      "grad_norm": 0.8427509665489197,
+      "learning_rate": 3.87875150060024e-05,
+      "loss": 0.6356,
+      "step": 5614
+    },
+    {
+      "epoch": 7.1872,
+      "grad_norm": 0.8131973147392273,
+      "learning_rate": 3.8785514205682275e-05,
+      "loss": 0.6143,
+      "step": 5615
+    },
+    {
+      "epoch": 7.18848,
+      "grad_norm": 0.8563538193702698,
+      "learning_rate": 3.8783513405362146e-05,
+      "loss": 0.6376,
+      "step": 5616
+    },
+    {
+      "epoch": 7.18976,
+      "grad_norm": 0.8263671398162842,
+      "learning_rate": 3.878151260504202e-05,
+      "loss": 0.5873,
+      "step": 5617
+    },
+    {
+      "epoch": 7.19104,
+      "grad_norm": 0.7478400468826294,
+      "learning_rate": 3.877951180472189e-05,
+      "loss": 0.5703,
+      "step": 5618
+    },
+    {
+      "epoch": 7.19232,
+      "grad_norm": 0.7577512860298157,
+      "learning_rate": 3.877751100440176e-05,
+      "loss": 0.5557,
+      "step": 5619
+    },
+    {
+      "epoch": 7.1936,
+      "grad_norm": 0.7820968627929688,
+      "learning_rate": 3.8775510204081634e-05,
+      "loss": 0.5891,
+      "step": 5620
+    },
+    {
+      "epoch": 7.19488,
+      "grad_norm": 0.7701216340065002,
+      "learning_rate": 3.8773509403761506e-05,
+      "loss": 0.6057,
+      "step": 5621
+    },
+    {
+      "epoch": 7.19616,
+      "grad_norm": 0.7920276522636414,
+      "learning_rate": 3.8771508603441384e-05,
+      "loss": 0.5919,
+      "step": 5622
+    },
+    {
+      "epoch": 7.19744,
+      "grad_norm": 0.8216814398765564,
+      "learning_rate": 3.876950780312125e-05,
+      "loss": 0.6177,
+      "step": 5623
+    },
+    {
+      "epoch": 7.19872,
+      "grad_norm": 0.8043842911720276,
+      "learning_rate": 3.876750700280112e-05,
+      "loss": 0.5935,
+      "step": 5624
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.7516394853591919,
+      "learning_rate": 3.876550620248099e-05,
+      "loss": 0.5912,
+      "step": 5625
+    },
+    {
+      "epoch": 7.20128,
+      "grad_norm": 0.7623422145843506,
+      "learning_rate": 3.8763505402160865e-05,
+      "loss": 0.5907,
+      "step": 5626
+    },
+    {
+      "epoch": 7.20256,
+      "grad_norm": 0.8073017597198486,
+      "learning_rate": 3.876150460184074e-05,
+      "loss": 0.6344,
+      "step": 5627
+    },
+    {
+      "epoch": 7.20384,
+      "grad_norm": 0.7998178005218506,
+      "learning_rate": 3.875950380152061e-05,
+      "loss": 0.6338,
+      "step": 5628
+    },
+    {
+      "epoch": 7.20512,
+      "grad_norm": 0.8303821086883545,
+      "learning_rate": 3.875750300120049e-05,
+      "loss": 0.6152,
+      "step": 5629
+    },
+    {
+      "epoch": 7.2064,
+      "grad_norm": 0.758705735206604,
+      "learning_rate": 3.875550220088036e-05,
+      "loss": 0.5706,
+      "step": 5630
+    },
+    {
+      "epoch": 7.20768,
+      "grad_norm": 0.7329617738723755,
+      "learning_rate": 3.8753501400560224e-05,
+      "loss": 0.5728,
+      "step": 5631
+    },
+    {
+      "epoch": 7.20896,
+      "grad_norm": 0.7714441418647766,
+      "learning_rate": 3.8751500600240096e-05,
+      "loss": 0.6087,
+      "step": 5632
+    },
+    {
+      "epoch": 7.21024,
+      "grad_norm": 0.8715619444847107,
+      "learning_rate": 3.874949979991997e-05,
+      "loss": 0.6635,
+      "step": 5633
+    },
+    {
+      "epoch": 7.21152,
+      "grad_norm": 0.7906916737556458,
+      "learning_rate": 3.874749899959984e-05,
+      "loss": 0.5903,
+      "step": 5634
+    },
+    {
+      "epoch": 7.2128,
+      "grad_norm": 0.7834184765815735,
+      "learning_rate": 3.874549819927971e-05,
+      "loss": 0.6234,
+      "step": 5635
+    },
+    {
+      "epoch": 7.21408,
+      "grad_norm": 0.7641972899436951,
+      "learning_rate": 3.874349739895959e-05,
+      "loss": 0.5475,
+      "step": 5636
+    },
+    {
+      "epoch": 7.21536,
+      "grad_norm": 0.7720842957496643,
+      "learning_rate": 3.874149659863946e-05,
+      "loss": 0.5907,
+      "step": 5637
+    },
+    {
+      "epoch": 7.21664,
+      "grad_norm": 0.8000057339668274,
+      "learning_rate": 3.8739495798319334e-05,
+      "loss": 0.6073,
+      "step": 5638
+    },
+    {
+      "epoch": 7.21792,
+      "grad_norm": 0.7984986901283264,
+      "learning_rate": 3.87374949979992e-05,
+      "loss": 0.5555,
+      "step": 5639
+    },
+    {
+      "epoch": 7.2192,
+      "grad_norm": 0.7842923998832703,
+      "learning_rate": 3.873549419767907e-05,
+      "loss": 0.5406,
+      "step": 5640
+    },
+    {
+      "epoch": 7.22048,
+      "grad_norm": 0.7952711582183838,
+      "learning_rate": 3.873349339735894e-05,
+      "loss": 0.6319,
+      "step": 5641
+    },
+    {
+      "epoch": 7.22176,
+      "grad_norm": 0.7931072115898132,
+      "learning_rate": 3.8731492597038815e-05,
+      "loss": 0.623,
+      "step": 5642
+    },
+    {
+      "epoch": 7.22304,
+      "grad_norm": 0.8495784997940063,
+      "learning_rate": 3.872949179671869e-05,
+      "loss": 0.6935,
+      "step": 5643
+    },
+    {
+      "epoch": 7.22432,
+      "grad_norm": 0.7872684001922607,
+      "learning_rate": 3.8727490996398565e-05,
+      "loss": 0.608,
+      "step": 5644
+    },
+    {
+      "epoch": 7.2256,
+      "grad_norm": 0.7937207221984863,
+      "learning_rate": 3.872549019607844e-05,
+      "loss": 0.6136,
+      "step": 5645
+    },
+    {
+      "epoch": 7.22688,
+      "grad_norm": 0.8002104163169861,
+      "learning_rate": 3.872348939575831e-05,
+      "loss": 0.6445,
+      "step": 5646
+    },
+    {
+      "epoch": 7.22816,
+      "grad_norm": 0.7604816555976868,
+      "learning_rate": 3.8721488595438174e-05,
+      "loss": 0.5837,
+      "step": 5647
+    },
+    {
+      "epoch": 7.22944,
+      "grad_norm": 0.804574728012085,
+      "learning_rate": 3.8719487795118046e-05,
+      "loss": 0.6667,
+      "step": 5648
+    },
+    {
+      "epoch": 7.23072,
+      "grad_norm": 0.7907571792602539,
+      "learning_rate": 3.871748699479792e-05,
+      "loss": 0.6271,
+      "step": 5649
+    },
+    {
+      "epoch": 7.232,
+      "grad_norm": 0.7569406628608704,
+      "learning_rate": 3.871548619447779e-05,
+      "loss": 0.5681,
+      "step": 5650
+    },
+    {
+      "epoch": 7.23328,
+      "grad_norm": 0.7395563721656799,
+      "learning_rate": 3.871348539415767e-05,
+      "loss": 0.5615,
+      "step": 5651
+    },
+    {
+      "epoch": 7.23456,
+      "grad_norm": 0.7570566534996033,
+      "learning_rate": 3.871148459383754e-05,
+      "loss": 0.5757,
+      "step": 5652
+    },
+    {
+      "epoch": 7.23584,
+      "grad_norm": 0.7983739376068115,
+      "learning_rate": 3.870948379351741e-05,
+      "loss": 0.5978,
+      "step": 5653
+    },
+    {
+      "epoch": 7.23712,
+      "grad_norm": 0.787769079208374,
+      "learning_rate": 3.8707482993197284e-05,
+      "loss": 0.5987,
+      "step": 5654
+    },
+    {
+      "epoch": 7.2384,
+      "grad_norm": 0.8369766473770142,
+      "learning_rate": 3.870548219287715e-05,
+      "loss": 0.6217,
+      "step": 5655
+    },
+    {
+      "epoch": 7.23968,
+      "grad_norm": 0.7803592681884766,
+      "learning_rate": 3.870348139255702e-05,
+      "loss": 0.5908,
+      "step": 5656
+    },
+    {
+      "epoch": 7.24096,
+      "grad_norm": 0.7458162307739258,
+      "learning_rate": 3.870148059223689e-05,
+      "loss": 0.5794,
+      "step": 5657
+    },
+    {
+      "epoch": 7.24224,
+      "grad_norm": 0.7700108885765076,
+      "learning_rate": 3.869947979191677e-05,
+      "loss": 0.6299,
+      "step": 5658
+    },
+    {
+      "epoch": 7.24352,
+      "grad_norm": 0.7748832702636719,
+      "learning_rate": 3.869747899159664e-05,
+      "loss": 0.5857,
+      "step": 5659
+    },
+    {
+      "epoch": 7.2448,
+      "grad_norm": 0.7574285864830017,
+      "learning_rate": 3.8695478191276515e-05,
+      "loss": 0.5838,
+      "step": 5660
+    },
+    {
+      "epoch": 7.24608,
+      "grad_norm": 0.8172944188117981,
+      "learning_rate": 3.869347739095639e-05,
+      "loss": 0.6803,
+      "step": 5661
+    },
+    {
+      "epoch": 7.24736,
+      "grad_norm": 0.79927659034729,
+      "learning_rate": 3.869147659063626e-05,
+      "loss": 0.5985,
+      "step": 5662
+    },
+    {
+      "epoch": 7.24864,
+      "grad_norm": 0.7518450617790222,
+      "learning_rate": 3.8689475790316124e-05,
+      "loss": 0.5637,
+      "step": 5663
+    },
+    {
+      "epoch": 7.24992,
+      "grad_norm": 0.7577493190765381,
+      "learning_rate": 3.8687474989995996e-05,
+      "loss": 0.5303,
+      "step": 5664
+    },
+    {
+      "epoch": 7.2512,
+      "grad_norm": 0.771763801574707,
+      "learning_rate": 3.8685474189675874e-05,
+      "loss": 0.5408,
+      "step": 5665
+    },
+    {
+      "epoch": 7.25248,
+      "grad_norm": 0.8028543591499329,
+      "learning_rate": 3.8683473389355746e-05,
+      "loss": 0.5607,
+      "step": 5666
+    },
+    {
+      "epoch": 7.25376,
+      "grad_norm": 0.8177222013473511,
+      "learning_rate": 3.868147258903562e-05,
+      "loss": 0.5871,
+      "step": 5667
+    },
+    {
+      "epoch": 7.25504,
+      "grad_norm": 0.772655189037323,
+      "learning_rate": 3.867947178871549e-05,
+      "loss": 0.5901,
+      "step": 5668
+    },
+    {
+      "epoch": 7.25632,
+      "grad_norm": 0.8063568472862244,
+      "learning_rate": 3.867747098839536e-05,
+      "loss": 0.5927,
+      "step": 5669
+    },
+    {
+      "epoch": 7.2576,
+      "grad_norm": 0.8081983327865601,
+      "learning_rate": 3.8675470188075233e-05,
+      "loss": 0.6183,
+      "step": 5670
+    },
+    {
+      "epoch": 7.2588799999999996,
+      "grad_norm": 0.7956660985946655,
+      "learning_rate": 3.86734693877551e-05,
+      "loss": 0.5789,
+      "step": 5671
+    },
+    {
+      "epoch": 7.26016,
+      "grad_norm": 0.7728344798088074,
+      "learning_rate": 3.867146858743498e-05,
+      "loss": 0.5848,
+      "step": 5672
+    },
+    {
+      "epoch": 7.26144,
+      "grad_norm": 0.7772632837295532,
+      "learning_rate": 3.866946778711485e-05,
+      "loss": 0.5777,
+      "step": 5673
+    },
+    {
+      "epoch": 7.26272,
+      "grad_norm": 0.7768100500106812,
+      "learning_rate": 3.866746698679472e-05,
+      "loss": 0.5562,
+      "step": 5674
+    },
+    {
+      "epoch": 7.264,
+      "grad_norm": 0.7781060934066772,
+      "learning_rate": 3.866546618647459e-05,
+      "loss": 0.5728,
+      "step": 5675
+    },
+    {
+      "epoch": 7.26528,
+      "grad_norm": 0.810567319393158,
+      "learning_rate": 3.8663465386154465e-05,
+      "loss": 0.5866,
+      "step": 5676
+    },
+    {
+      "epoch": 7.26656,
+      "grad_norm": 0.8131520748138428,
+      "learning_rate": 3.8661464585834336e-05,
+      "loss": 0.6461,
+      "step": 5677
+    },
+    {
+      "epoch": 7.26784,
+      "grad_norm": 0.8413848876953125,
+      "learning_rate": 3.865946378551421e-05,
+      "loss": 0.6273,
+      "step": 5678
+    },
+    {
+      "epoch": 7.26912,
+      "grad_norm": 0.7927761673927307,
+      "learning_rate": 3.865746298519408e-05,
+      "loss": 0.5788,
+      "step": 5679
+    },
+    {
+      "epoch": 7.2704,
+      "grad_norm": 0.7904192805290222,
+      "learning_rate": 3.865546218487395e-05,
+      "loss": 0.5934,
+      "step": 5680
+    },
+    {
+      "epoch": 7.27168,
+      "grad_norm": 0.8133504986763,
+      "learning_rate": 3.8653461384553824e-05,
+      "loss": 0.5641,
+      "step": 5681
+    },
+    {
+      "epoch": 7.27296,
+      "grad_norm": 0.8121780157089233,
+      "learning_rate": 3.8651460584233696e-05,
+      "loss": 0.6126,
+      "step": 5682
+    },
+    {
+      "epoch": 7.27424,
+      "grad_norm": 0.850824773311615,
+      "learning_rate": 3.864945978391357e-05,
+      "loss": 0.6447,
+      "step": 5683
+    },
+    {
+      "epoch": 7.27552,
+      "grad_norm": 0.7701993584632874,
+      "learning_rate": 3.864745898359344e-05,
+      "loss": 0.5702,
+      "step": 5684
+    },
+    {
+      "epoch": 7.2768,
+      "grad_norm": 0.8420401811599731,
+      "learning_rate": 3.864545818327331e-05,
+      "loss": 0.6527,
+      "step": 5685
+    },
+    {
+      "epoch": 7.27808,
+      "grad_norm": 0.7517603039741516,
+      "learning_rate": 3.864345738295318e-05,
+      "loss": 0.5792,
+      "step": 5686
+    },
+    {
+      "epoch": 7.27936,
+      "grad_norm": 0.877324640750885,
+      "learning_rate": 3.8641456582633055e-05,
+      "loss": 0.6342,
+      "step": 5687
+    },
+    {
+      "epoch": 7.28064,
+      "grad_norm": 0.7491464614868164,
+      "learning_rate": 3.863945578231293e-05,
+      "loss": 0.5504,
+      "step": 5688
+    },
+    {
+      "epoch": 7.28192,
+      "grad_norm": 0.7788039445877075,
+      "learning_rate": 3.86374549819928e-05,
+      "loss": 0.6107,
+      "step": 5689
+    },
+    {
+      "epoch": 7.2832,
+      "grad_norm": 0.8021331429481506,
+      "learning_rate": 3.863545418167267e-05,
+      "loss": 0.6289,
+      "step": 5690
+    },
+    {
+      "epoch": 7.28448,
+      "grad_norm": 0.7916176915168762,
+      "learning_rate": 3.863345338135254e-05,
+      "loss": 0.5962,
+      "step": 5691
+    },
+    {
+      "epoch": 7.28576,
+      "grad_norm": 0.7986366152763367,
+      "learning_rate": 3.8631452581032414e-05,
+      "loss": 0.5942,
+      "step": 5692
+    },
+    {
+      "epoch": 7.28704,
+      "grad_norm": 0.8029093742370605,
+      "learning_rate": 3.8629451780712286e-05,
+      "loss": 0.6185,
+      "step": 5693
+    },
+    {
+      "epoch": 7.28832,
+      "grad_norm": 0.7949596047401428,
+      "learning_rate": 3.862745098039216e-05,
+      "loss": 0.6337,
+      "step": 5694
+    },
+    {
+      "epoch": 7.2896,
+      "grad_norm": 0.7814388871192932,
+      "learning_rate": 3.862545018007203e-05,
+      "loss": 0.6188,
+      "step": 5695
+    },
+    {
+      "epoch": 7.29088,
+      "grad_norm": 0.8712118268013,
+      "learning_rate": 3.86234493797519e-05,
+      "loss": 0.6191,
+      "step": 5696
+    },
+    {
+      "epoch": 7.29216,
+      "grad_norm": 0.769758939743042,
+      "learning_rate": 3.8621448579431774e-05,
+      "loss": 0.5759,
+      "step": 5697
+    },
+    {
+      "epoch": 7.29344,
+      "grad_norm": 0.7669387459754944,
+      "learning_rate": 3.8619447779111645e-05,
+      "loss": 0.5509,
+      "step": 5698
+    },
+    {
+      "epoch": 7.29472,
+      "grad_norm": 0.7945777177810669,
+      "learning_rate": 3.861744697879152e-05,
+      "loss": 0.6417,
+      "step": 5699
+    },
+    {
+      "epoch": 7.296,
+      "grad_norm": 0.764313280582428,
+      "learning_rate": 3.8615446178471396e-05,
+      "loss": 0.5851,
+      "step": 5700
+    },
+    {
+      "epoch": 7.29728,
+      "grad_norm": 0.7636333107948303,
+      "learning_rate": 3.861344537815126e-05,
+      "loss": 0.5799,
+      "step": 5701
+    },
+    {
+      "epoch": 7.29856,
+      "grad_norm": 0.7744512557983398,
+      "learning_rate": 3.861144457783113e-05,
+      "loss": 0.5415,
+      "step": 5702
+    },
+    {
+      "epoch": 7.29984,
+      "grad_norm": 0.7718966007232666,
+      "learning_rate": 3.8609443777511005e-05,
+      "loss": 0.6007,
+      "step": 5703
+    },
+    {
+      "epoch": 7.30112,
+      "grad_norm": 0.7986489534378052,
+      "learning_rate": 3.8607442977190877e-05,
+      "loss": 0.6244,
+      "step": 5704
+    },
+    {
+      "epoch": 7.3024000000000004,
+      "grad_norm": 0.7697789072990417,
+      "learning_rate": 3.860544217687075e-05,
+      "loss": 0.6394,
+      "step": 5705
+    },
+    {
+      "epoch": 7.30368,
+      "grad_norm": 0.781538188457489,
+      "learning_rate": 3.860344137655062e-05,
+      "loss": 0.6007,
+      "step": 5706
+    },
+    {
+      "epoch": 7.30496,
+      "grad_norm": 0.7859370112419128,
+      "learning_rate": 3.86014405762305e-05,
+      "loss": 0.6103,
+      "step": 5707
+    },
+    {
+      "epoch": 7.30624,
+      "grad_norm": 0.8621463775634766,
+      "learning_rate": 3.859943977591037e-05,
+      "loss": 0.6401,
+      "step": 5708
+    },
+    {
+      "epoch": 7.30752,
+      "grad_norm": 0.8109675645828247,
+      "learning_rate": 3.8597438975590236e-05,
+      "loss": 0.611,
+      "step": 5709
+    },
+    {
+      "epoch": 7.3088,
+      "grad_norm": 0.7787231206893921,
+      "learning_rate": 3.859543817527011e-05,
+      "loss": 0.5673,
+      "step": 5710
+    },
+    {
+      "epoch": 7.31008,
+      "grad_norm": 0.7826237678527832,
+      "learning_rate": 3.859343737494998e-05,
+      "loss": 0.6092,
+      "step": 5711
+    },
+    {
+      "epoch": 7.31136,
+      "grad_norm": 0.8043519258499146,
+      "learning_rate": 3.859143657462985e-05,
+      "loss": 0.6405,
+      "step": 5712
+    },
+    {
+      "epoch": 7.31264,
+      "grad_norm": 0.7601256370544434,
+      "learning_rate": 3.858943577430972e-05,
+      "loss": 0.5687,
+      "step": 5713
+    },
+    {
+      "epoch": 7.3139199999999995,
+      "grad_norm": 0.8053227663040161,
+      "learning_rate": 3.85874349739896e-05,
+      "loss": 0.6097,
+      "step": 5714
+    },
+    {
+      "epoch": 7.3152,
+      "grad_norm": 0.7957972288131714,
+      "learning_rate": 3.8585434173669474e-05,
+      "loss": 0.6215,
+      "step": 5715
+    },
+    {
+      "epoch": 7.31648,
+      "grad_norm": 0.7682824730873108,
+      "learning_rate": 3.8583433373349346e-05,
+      "loss": 0.6002,
+      "step": 5716
+    },
+    {
+      "epoch": 7.31776,
+      "grad_norm": 0.7858462333679199,
+      "learning_rate": 3.858143257302921e-05,
+      "loss": 0.6155,
+      "step": 5717
+    },
+    {
+      "epoch": 7.31904,
+      "grad_norm": 0.7957653999328613,
+      "learning_rate": 3.857943177270908e-05,
+      "loss": 0.5728,
+      "step": 5718
+    },
+    {
+      "epoch": 7.32032,
+      "grad_norm": 0.7708908915519714,
+      "learning_rate": 3.8577430972388954e-05,
+      "loss": 0.6013,
+      "step": 5719
+    },
+    {
+      "epoch": 7.3216,
+      "grad_norm": 0.7898252010345459,
+      "learning_rate": 3.8575430172068826e-05,
+      "loss": 0.6068,
+      "step": 5720
+    },
+    {
+      "epoch": 7.32288,
+      "grad_norm": 0.8056557178497314,
+      "learning_rate": 3.8573429371748705e-05,
+      "loss": 0.6074,
+      "step": 5721
+    },
+    {
+      "epoch": 7.32416,
+      "grad_norm": 0.7734809517860413,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 0.6006,
+      "step": 5722
+    },
+    {
+      "epoch": 7.32544,
+      "grad_norm": 0.8016178607940674,
+      "learning_rate": 3.856942777110845e-05,
+      "loss": 0.6086,
+      "step": 5723
+    },
+    {
+      "epoch": 7.32672,
+      "grad_norm": 0.8026044964790344,
+      "learning_rate": 3.856742697078832e-05,
+      "loss": 0.6075,
+      "step": 5724
+    },
+    {
+      "epoch": 7.328,
+      "grad_norm": 0.7688097953796387,
+      "learning_rate": 3.8565426170468186e-05,
+      "loss": 0.5736,
+      "step": 5725
+    },
+    {
+      "epoch": 7.32928,
+      "grad_norm": 0.7733714580535889,
+      "learning_rate": 3.856342537014806e-05,
+      "loss": 0.5723,
+      "step": 5726
+    },
+    {
+      "epoch": 7.33056,
+      "grad_norm": 0.7920982837677002,
+      "learning_rate": 3.856142456982793e-05,
+      "loss": 0.6015,
+      "step": 5727
+    },
+    {
+      "epoch": 7.33184,
+      "grad_norm": 0.8239418864250183,
+      "learning_rate": 3.855942376950781e-05,
+      "loss": 0.5999,
+      "step": 5728
+    },
+    {
+      "epoch": 7.33312,
+      "grad_norm": 0.7626153826713562,
+      "learning_rate": 3.855742296918768e-05,
+      "loss": 0.5769,
+      "step": 5729
+    },
+    {
+      "epoch": 7.3344,
+      "grad_norm": 0.7836630344390869,
+      "learning_rate": 3.855542216886755e-05,
+      "loss": 0.5713,
+      "step": 5730
+    },
+    {
+      "epoch": 7.33568,
+      "grad_norm": 0.8354891538619995,
+      "learning_rate": 3.8553421368547423e-05,
+      "loss": 0.5959,
+      "step": 5731
+    },
+    {
+      "epoch": 7.33696,
+      "grad_norm": 0.8115443587303162,
+      "learning_rate": 3.8551420568227295e-05,
+      "loss": 0.581,
+      "step": 5732
+    },
+    {
+      "epoch": 7.33824,
+      "grad_norm": 0.787783682346344,
+      "learning_rate": 3.854941976790716e-05,
+      "loss": 0.6043,
+      "step": 5733
+    },
+    {
+      "epoch": 7.33952,
+      "grad_norm": 0.7985721230506897,
+      "learning_rate": 3.854741896758703e-05,
+      "loss": 0.6119,
+      "step": 5734
+    },
+    {
+      "epoch": 7.3408,
+      "grad_norm": 0.8209099173545837,
+      "learning_rate": 3.854541816726691e-05,
+      "loss": 0.6216,
+      "step": 5735
+    },
+    {
+      "epoch": 7.34208,
+      "grad_norm": 0.7716638445854187,
+      "learning_rate": 3.854341736694678e-05,
+      "loss": 0.5866,
+      "step": 5736
+    },
+    {
+      "epoch": 7.34336,
+      "grad_norm": 0.7463488578796387,
+      "learning_rate": 3.8541416566626655e-05,
+      "loss": 0.5584,
+      "step": 5737
+    },
+    {
+      "epoch": 7.34464,
+      "grad_norm": 0.7812283635139465,
+      "learning_rate": 3.8539415766306526e-05,
+      "loss": 0.6046,
+      "step": 5738
+    },
+    {
+      "epoch": 7.34592,
+      "grad_norm": 0.830808699131012,
+      "learning_rate": 3.85374149659864e-05,
+      "loss": 0.6134,
+      "step": 5739
+    },
+    {
+      "epoch": 7.3472,
+      "grad_norm": 0.802952766418457,
+      "learning_rate": 3.853541416566627e-05,
+      "loss": 0.6095,
+      "step": 5740
+    },
+    {
+      "epoch": 7.34848,
+      "grad_norm": 0.8335705995559692,
+      "learning_rate": 3.8533413365346135e-05,
+      "loss": 0.6473,
+      "step": 5741
+    },
+    {
+      "epoch": 7.34976,
+      "grad_norm": 0.8194312453269958,
+      "learning_rate": 3.8531412565026014e-05,
+      "loss": 0.6392,
+      "step": 5742
+    },
+    {
+      "epoch": 7.35104,
+      "grad_norm": 0.8065541386604309,
+      "learning_rate": 3.8529411764705886e-05,
+      "loss": 0.5786,
+      "step": 5743
+    },
+    {
+      "epoch": 7.35232,
+      "grad_norm": 0.8424248695373535,
+      "learning_rate": 3.852741096438576e-05,
+      "loss": 0.6275,
+      "step": 5744
+    },
+    {
+      "epoch": 7.3536,
+      "grad_norm": 0.8039935827255249,
+      "learning_rate": 3.852541016406563e-05,
+      "loss": 0.595,
+      "step": 5745
+    },
+    {
+      "epoch": 7.35488,
+      "grad_norm": 0.8151816129684448,
+      "learning_rate": 3.85234093637455e-05,
+      "loss": 0.6375,
+      "step": 5746
+    },
+    {
+      "epoch": 7.35616,
+      "grad_norm": 0.7974145412445068,
+      "learning_rate": 3.852140856342537e-05,
+      "loss": 0.605,
+      "step": 5747
+    },
+    {
+      "epoch": 7.35744,
+      "grad_norm": 0.8081220984458923,
+      "learning_rate": 3.8519407763105245e-05,
+      "loss": 0.6031,
+      "step": 5748
+    },
+    {
+      "epoch": 7.35872,
+      "grad_norm": 0.7766823768615723,
+      "learning_rate": 3.851740696278512e-05,
+      "loss": 0.6233,
+      "step": 5749
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.7977601885795593,
+      "learning_rate": 3.851540616246499e-05,
+      "loss": 0.6105,
+      "step": 5750
+    },
+    {
+      "epoch": 7.36128,
+      "grad_norm": 0.7957237958908081,
+      "learning_rate": 3.851340536214486e-05,
+      "loss": 0.6086,
+      "step": 5751
+    },
+    {
+      "epoch": 7.36256,
+      "grad_norm": 0.7621792554855347,
+      "learning_rate": 3.851140456182473e-05,
+      "loss": 0.6113,
+      "step": 5752
+    },
+    {
+      "epoch": 7.36384,
+      "grad_norm": 0.7617015242576599,
+      "learning_rate": 3.8509403761504604e-05,
+      "loss": 0.5529,
+      "step": 5753
+    },
+    {
+      "epoch": 7.36512,
+      "grad_norm": 0.7769597768783569,
+      "learning_rate": 3.8507402961184476e-05,
+      "loss": 0.6458,
+      "step": 5754
+    },
+    {
+      "epoch": 7.3664,
+      "grad_norm": 0.7622930407524109,
+      "learning_rate": 3.850540216086435e-05,
+      "loss": 0.5758,
+      "step": 5755
+    },
+    {
+      "epoch": 7.36768,
+      "grad_norm": 0.8526008129119873,
+      "learning_rate": 3.850340136054422e-05,
+      "loss": 0.6412,
+      "step": 5756
+    },
+    {
+      "epoch": 7.36896,
+      "grad_norm": 0.8075965046882629,
+      "learning_rate": 3.850140056022409e-05,
+      "loss": 0.598,
+      "step": 5757
+    },
+    {
+      "epoch": 7.37024,
+      "grad_norm": 0.859012246131897,
+      "learning_rate": 3.8499399759903964e-05,
+      "loss": 0.6844,
+      "step": 5758
+    },
+    {
+      "epoch": 7.37152,
+      "grad_norm": 0.7848984003067017,
+      "learning_rate": 3.8497398959583835e-05,
+      "loss": 0.5879,
+      "step": 5759
+    },
+    {
+      "epoch": 7.3728,
+      "grad_norm": 0.7643441557884216,
+      "learning_rate": 3.849539815926371e-05,
+      "loss": 0.5704,
+      "step": 5760
+    },
+    {
+      "epoch": 7.37408,
+      "grad_norm": 0.8113600015640259,
+      "learning_rate": 3.849339735894358e-05,
+      "loss": 0.6399,
+      "step": 5761
+    },
+    {
+      "epoch": 7.37536,
+      "grad_norm": 0.7590932846069336,
+      "learning_rate": 3.849139655862345e-05,
+      "loss": 0.5771,
+      "step": 5762
+    },
+    {
+      "epoch": 7.37664,
+      "grad_norm": 0.7780318260192871,
+      "learning_rate": 3.848939575830332e-05,
+      "loss": 0.6144,
+      "step": 5763
+    },
+    {
+      "epoch": 7.37792,
+      "grad_norm": 0.8140581846237183,
+      "learning_rate": 3.8487394957983195e-05,
+      "loss": 0.6571,
+      "step": 5764
+    },
+    {
+      "epoch": 7.3792,
+      "grad_norm": 0.7549552321434021,
+      "learning_rate": 3.8485394157663067e-05,
+      "loss": 0.5123,
+      "step": 5765
+    },
+    {
+      "epoch": 7.38048,
+      "grad_norm": 0.7963429093360901,
+      "learning_rate": 3.848339335734294e-05,
+      "loss": 0.5821,
+      "step": 5766
+    },
+    {
+      "epoch": 7.38176,
+      "grad_norm": 0.8283103704452515,
+      "learning_rate": 3.848139255702281e-05,
+      "loss": 0.627,
+      "step": 5767
+    },
+    {
+      "epoch": 7.38304,
+      "grad_norm": 0.7662546038627625,
+      "learning_rate": 3.847939175670268e-05,
+      "loss": 0.5702,
+      "step": 5768
+    },
+    {
+      "epoch": 7.38432,
+      "grad_norm": 0.780431866645813,
+      "learning_rate": 3.8477390956382554e-05,
+      "loss": 0.6072,
+      "step": 5769
+    },
+    {
+      "epoch": 7.3856,
+      "grad_norm": 0.8025130033493042,
+      "learning_rate": 3.8475390156062426e-05,
+      "loss": 0.5984,
+      "step": 5770
+    },
+    {
+      "epoch": 7.38688,
+      "grad_norm": 0.7883024215698242,
+      "learning_rate": 3.84733893557423e-05,
+      "loss": 0.5951,
+      "step": 5771
+    },
+    {
+      "epoch": 7.38816,
+      "grad_norm": 0.8061680197715759,
+      "learning_rate": 3.847138855542217e-05,
+      "loss": 0.6058,
+      "step": 5772
+    },
+    {
+      "epoch": 7.3894400000000005,
+      "grad_norm": 0.7774277925491333,
+      "learning_rate": 3.846938775510204e-05,
+      "loss": 0.5983,
+      "step": 5773
+    },
+    {
+      "epoch": 7.39072,
+      "grad_norm": 0.787751317024231,
+      "learning_rate": 3.846738695478191e-05,
+      "loss": 0.5822,
+      "step": 5774
+    },
+    {
+      "epoch": 7.392,
+      "grad_norm": 0.830544650554657,
+      "learning_rate": 3.8465386154461785e-05,
+      "loss": 0.6172,
+      "step": 5775
+    },
+    {
+      "epoch": 7.39328,
+      "grad_norm": 0.8001280426979065,
+      "learning_rate": 3.846338535414166e-05,
+      "loss": 0.6249,
+      "step": 5776
+    },
+    {
+      "epoch": 7.39456,
+      "grad_norm": 0.7862173914909363,
+      "learning_rate": 3.846138455382153e-05,
+      "loss": 0.5628,
+      "step": 5777
+    },
+    {
+      "epoch": 7.39584,
+      "grad_norm": 0.8081876039505005,
+      "learning_rate": 3.845938375350141e-05,
+      "loss": 0.6007,
+      "step": 5778
+    },
+    {
+      "epoch": 7.39712,
+      "grad_norm": 0.7967897057533264,
+      "learning_rate": 3.845738295318127e-05,
+      "loss": 0.5906,
+      "step": 5779
+    },
+    {
+      "epoch": 7.3984,
+      "grad_norm": 0.8136133551597595,
+      "learning_rate": 3.8455382152861144e-05,
+      "loss": 0.5849,
+      "step": 5780
+    },
+    {
+      "epoch": 7.39968,
+      "grad_norm": 0.7975330948829651,
+      "learning_rate": 3.8453381352541016e-05,
+      "loss": 0.6159,
+      "step": 5781
+    },
+    {
+      "epoch": 7.4009599999999995,
+      "grad_norm": 0.8437215089797974,
+      "learning_rate": 3.845138055222089e-05,
+      "loss": 0.6324,
+      "step": 5782
+    },
+    {
+      "epoch": 7.40224,
+      "grad_norm": 0.8381403684616089,
+      "learning_rate": 3.844937975190076e-05,
+      "loss": 0.6193,
+      "step": 5783
+    },
+    {
+      "epoch": 7.40352,
+      "grad_norm": 0.7790165543556213,
+      "learning_rate": 3.844737895158063e-05,
+      "loss": 0.6107,
+      "step": 5784
+    },
+    {
+      "epoch": 7.4048,
+      "grad_norm": 0.8007838726043701,
+      "learning_rate": 3.844537815126051e-05,
+      "loss": 0.6304,
+      "step": 5785
+    },
+    {
+      "epoch": 7.40608,
+      "grad_norm": 0.8209007978439331,
+      "learning_rate": 3.844337735094038e-05,
+      "loss": 0.6089,
+      "step": 5786
+    },
+    {
+      "epoch": 7.40736,
+      "grad_norm": 0.7682439684867859,
+      "learning_rate": 3.844137655062025e-05,
+      "loss": 0.5974,
+      "step": 5787
+    },
+    {
+      "epoch": 7.40864,
+      "grad_norm": 0.7677670121192932,
+      "learning_rate": 3.843937575030012e-05,
+      "loss": 0.5678,
+      "step": 5788
+    },
+    {
+      "epoch": 7.40992,
+      "grad_norm": 0.8111715912818909,
+      "learning_rate": 3.843737494997999e-05,
+      "loss": 0.5916,
+      "step": 5789
+    },
+    {
+      "epoch": 7.4112,
+      "grad_norm": 0.7806997895240784,
+      "learning_rate": 3.843537414965986e-05,
+      "loss": 0.6144,
+      "step": 5790
+    },
+    {
+      "epoch": 7.41248,
+      "grad_norm": 0.7777438759803772,
+      "learning_rate": 3.8433373349339735e-05,
+      "loss": 0.5683,
+      "step": 5791
+    },
+    {
+      "epoch": 7.41376,
+      "grad_norm": 0.751512885093689,
+      "learning_rate": 3.8431372549019614e-05,
+      "loss": 0.5775,
+      "step": 5792
+    },
+    {
+      "epoch": 7.41504,
+      "grad_norm": 0.7952166199684143,
+      "learning_rate": 3.8429371748699485e-05,
+      "loss": 0.5878,
+      "step": 5793
+    },
+    {
+      "epoch": 7.41632,
+      "grad_norm": 0.7829790711402893,
+      "learning_rate": 3.842737094837936e-05,
+      "loss": 0.6195,
+      "step": 5794
+    },
+    {
+      "epoch": 7.4176,
+      "grad_norm": 0.7563794851303101,
+      "learning_rate": 3.842537014805922e-05,
+      "loss": 0.5918,
+      "step": 5795
+    },
+    {
+      "epoch": 7.41888,
+      "grad_norm": 0.816536009311676,
+      "learning_rate": 3.8423369347739094e-05,
+      "loss": 0.6219,
+      "step": 5796
+    },
+    {
+      "epoch": 7.42016,
+      "grad_norm": 0.7779079079627991,
+      "learning_rate": 3.8421368547418966e-05,
+      "loss": 0.6168,
+      "step": 5797
+    },
+    {
+      "epoch": 7.42144,
+      "grad_norm": 0.8121761083602905,
+      "learning_rate": 3.841936774709884e-05,
+      "loss": 0.625,
+      "step": 5798
+    },
+    {
+      "epoch": 7.42272,
+      "grad_norm": 0.7565649151802063,
+      "learning_rate": 3.8417366946778717e-05,
+      "loss": 0.5407,
+      "step": 5799
+    },
+    {
+      "epoch": 7.424,
+      "grad_norm": 0.7441564798355103,
+      "learning_rate": 3.841536614645859e-05,
+      "loss": 0.5511,
+      "step": 5800
+    },
+    {
+      "epoch": 7.42528,
+      "grad_norm": 0.8287448883056641,
+      "learning_rate": 3.841336534613846e-05,
+      "loss": 0.6299,
+      "step": 5801
+    },
+    {
+      "epoch": 7.42656,
+      "grad_norm": 0.771754801273346,
+      "learning_rate": 3.841136454581833e-05,
+      "loss": 0.5718,
+      "step": 5802
+    },
+    {
+      "epoch": 7.42784,
+      "grad_norm": 0.7989311814308167,
+      "learning_rate": 3.84093637454982e-05,
+      "loss": 0.6068,
+      "step": 5803
+    },
+    {
+      "epoch": 7.42912,
+      "grad_norm": 0.7287749648094177,
+      "learning_rate": 3.840736294517807e-05,
+      "loss": 0.5552,
+      "step": 5804
+    },
+    {
+      "epoch": 7.4304,
+      "grad_norm": 0.8139730095863342,
+      "learning_rate": 3.840536214485794e-05,
+      "loss": 0.623,
+      "step": 5805
+    },
+    {
+      "epoch": 7.43168,
+      "grad_norm": 0.8090022206306458,
+      "learning_rate": 3.840336134453782e-05,
+      "loss": 0.5875,
+      "step": 5806
+    },
+    {
+      "epoch": 7.43296,
+      "grad_norm": 0.793850302696228,
+      "learning_rate": 3.840136054421769e-05,
+      "loss": 0.5948,
+      "step": 5807
+    },
+    {
+      "epoch": 7.43424,
+      "grad_norm": 0.8325986862182617,
+      "learning_rate": 3.839935974389756e-05,
+      "loss": 0.6497,
+      "step": 5808
+    },
+    {
+      "epoch": 7.43552,
+      "grad_norm": 0.7771223783493042,
+      "learning_rate": 3.8397358943577435e-05,
+      "loss": 0.584,
+      "step": 5809
+    },
+    {
+      "epoch": 7.4368,
+      "grad_norm": 0.8267020583152771,
+      "learning_rate": 3.839535814325731e-05,
+      "loss": 0.6556,
+      "step": 5810
+    },
+    {
+      "epoch": 7.43808,
+      "grad_norm": 0.8165448904037476,
+      "learning_rate": 3.839335734293717e-05,
+      "loss": 0.6132,
+      "step": 5811
+    },
+    {
+      "epoch": 7.43936,
+      "grad_norm": 0.8215618133544922,
+      "learning_rate": 3.8391356542617044e-05,
+      "loss": 0.6034,
+      "step": 5812
+    },
+    {
+      "epoch": 7.44064,
+      "grad_norm": 0.8251504302024841,
+      "learning_rate": 3.838935574229692e-05,
+      "loss": 0.6194,
+      "step": 5813
+    },
+    {
+      "epoch": 7.44192,
+      "grad_norm": 0.803627073764801,
+      "learning_rate": 3.8387354941976794e-05,
+      "loss": 0.5718,
+      "step": 5814
+    },
+    {
+      "epoch": 7.4432,
+      "grad_norm": 0.7754432559013367,
+      "learning_rate": 3.8385354141656666e-05,
+      "loss": 0.5652,
+      "step": 5815
+    },
+    {
+      "epoch": 7.44448,
+      "grad_norm": 0.7755488753318787,
+      "learning_rate": 3.838335334133654e-05,
+      "loss": 0.6237,
+      "step": 5816
+    },
+    {
+      "epoch": 7.44576,
+      "grad_norm": 0.8018819689750671,
+      "learning_rate": 3.838135254101641e-05,
+      "loss": 0.615,
+      "step": 5817
+    },
+    {
+      "epoch": 7.44704,
+      "grad_norm": 0.7852720618247986,
+      "learning_rate": 3.837935174069628e-05,
+      "loss": 0.6045,
+      "step": 5818
+    },
+    {
+      "epoch": 7.44832,
+      "grad_norm": 0.784041166305542,
+      "learning_rate": 3.837735094037615e-05,
+      "loss": 0.6095,
+      "step": 5819
+    },
+    {
+      "epoch": 7.4496,
+      "grad_norm": 0.8439889550209045,
+      "learning_rate": 3.8375350140056026e-05,
+      "loss": 0.7184,
+      "step": 5820
+    },
+    {
+      "epoch": 7.45088,
+      "grad_norm": 0.8502903580665588,
+      "learning_rate": 3.83733493397359e-05,
+      "loss": 0.6358,
+      "step": 5821
+    },
+    {
+      "epoch": 7.45216,
+      "grad_norm": 0.7948433756828308,
+      "learning_rate": 3.837134853941577e-05,
+      "loss": 0.5715,
+      "step": 5822
+    },
+    {
+      "epoch": 7.45344,
+      "grad_norm": 0.7365151047706604,
+      "learning_rate": 3.836934773909564e-05,
+      "loss": 0.567,
+      "step": 5823
+    },
+    {
+      "epoch": 7.45472,
+      "grad_norm": 0.8041279315948486,
+      "learning_rate": 3.836734693877551e-05,
+      "loss": 0.6025,
+      "step": 5824
+    },
+    {
+      "epoch": 7.456,
+      "grad_norm": 0.7834358215332031,
+      "learning_rate": 3.8365346138455385e-05,
+      "loss": 0.5941,
+      "step": 5825
+    },
+    {
+      "epoch": 7.45728,
+      "grad_norm": 0.7606781721115112,
+      "learning_rate": 3.836334533813526e-05,
+      "loss": 0.5979,
+      "step": 5826
+    },
+    {
+      "epoch": 7.45856,
+      "grad_norm": 0.7791493535041809,
+      "learning_rate": 3.836134453781513e-05,
+      "loss": 0.5739,
+      "step": 5827
+    },
+    {
+      "epoch": 7.45984,
+      "grad_norm": 0.8152048587799072,
+      "learning_rate": 3.8359343737495e-05,
+      "loss": 0.6035,
+      "step": 5828
+    },
+    {
+      "epoch": 7.46112,
+      "grad_norm": 0.8163161277770996,
+      "learning_rate": 3.835734293717487e-05,
+      "loss": 0.6061,
+      "step": 5829
+    },
+    {
+      "epoch": 7.4624,
+      "grad_norm": 0.7904027104377747,
+      "learning_rate": 3.8355342136854744e-05,
+      "loss": 0.6001,
+      "step": 5830
+    },
+    {
+      "epoch": 7.46368,
+      "grad_norm": 0.8345395922660828,
+      "learning_rate": 3.8353341336534616e-05,
+      "loss": 0.6474,
+      "step": 5831
+    },
+    {
+      "epoch": 7.46496,
+      "grad_norm": 0.764450192451477,
+      "learning_rate": 3.835134053621449e-05,
+      "loss": 0.5939,
+      "step": 5832
+    },
+    {
+      "epoch": 7.46624,
+      "grad_norm": 0.7856200933456421,
+      "learning_rate": 3.834933973589436e-05,
+      "loss": 0.5863,
+      "step": 5833
+    },
+    {
+      "epoch": 7.46752,
+      "grad_norm": 0.756513237953186,
+      "learning_rate": 3.834733893557423e-05,
+      "loss": 0.5699,
+      "step": 5834
+    },
+    {
+      "epoch": 7.4688,
+      "grad_norm": 0.8095502257347107,
+      "learning_rate": 3.83453381352541e-05,
+      "loss": 0.6145,
+      "step": 5835
+    },
+    {
+      "epoch": 7.47008,
+      "grad_norm": 0.7780006527900696,
+      "learning_rate": 3.8343337334933975e-05,
+      "loss": 0.622,
+      "step": 5836
+    },
+    {
+      "epoch": 7.47136,
+      "grad_norm": 0.7881510257720947,
+      "learning_rate": 3.834133653461385e-05,
+      "loss": 0.5873,
+      "step": 5837
+    },
+    {
+      "epoch": 7.47264,
+      "grad_norm": 0.7660511136054993,
+      "learning_rate": 3.833933573429372e-05,
+      "loss": 0.6096,
+      "step": 5838
+    },
+    {
+      "epoch": 7.47392,
+      "grad_norm": 0.7153475284576416,
+      "learning_rate": 3.833733493397359e-05,
+      "loss": 0.5051,
+      "step": 5839
+    },
+    {
+      "epoch": 7.4752,
+      "grad_norm": 0.7922013401985168,
+      "learning_rate": 3.833533413365346e-05,
+      "loss": 0.6159,
+      "step": 5840
+    },
+    {
+      "epoch": 7.4764800000000005,
+      "grad_norm": 0.7256450057029724,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 0.5558,
+      "step": 5841
+    },
+    {
+      "epoch": 7.47776,
+      "grad_norm": 0.7810547351837158,
+      "learning_rate": 3.8331332533013206e-05,
+      "loss": 0.6134,
+      "step": 5842
+    },
+    {
+      "epoch": 7.47904,
+      "grad_norm": 0.7948379516601562,
+      "learning_rate": 3.832933173269308e-05,
+      "loss": 0.6025,
+      "step": 5843
+    },
+    {
+      "epoch": 7.48032,
+      "grad_norm": 0.7824628353118896,
+      "learning_rate": 3.832733093237295e-05,
+      "loss": 0.5658,
+      "step": 5844
+    },
+    {
+      "epoch": 7.4816,
+      "grad_norm": 0.8361453413963318,
+      "learning_rate": 3.832533013205282e-05,
+      "loss": 0.6285,
+      "step": 5845
+    },
+    {
+      "epoch": 7.48288,
+      "grad_norm": 0.8043081164360046,
+      "learning_rate": 3.8323329331732694e-05,
+      "loss": 0.6019,
+      "step": 5846
+    },
+    {
+      "epoch": 7.48416,
+      "grad_norm": 0.7815614342689514,
+      "learning_rate": 3.8321328531412566e-05,
+      "loss": 0.5788,
+      "step": 5847
+    },
+    {
+      "epoch": 7.48544,
+      "grad_norm": 0.8125666379928589,
+      "learning_rate": 3.8319327731092444e-05,
+      "loss": 0.5843,
+      "step": 5848
+    },
+    {
+      "epoch": 7.48672,
+      "grad_norm": 0.8060123920440674,
+      "learning_rate": 3.831732693077231e-05,
+      "loss": 0.605,
+      "step": 5849
+    },
+    {
+      "epoch": 7.4879999999999995,
+      "grad_norm": 0.7951981425285339,
+      "learning_rate": 3.831532613045218e-05,
+      "loss": 0.5823,
+      "step": 5850
+    },
+    {
+      "epoch": 7.48928,
+      "grad_norm": 0.767137885093689,
+      "learning_rate": 3.831332533013205e-05,
+      "loss": 0.6463,
+      "step": 5851
+    },
+    {
+      "epoch": 7.49056,
+      "grad_norm": 0.7987186312675476,
+      "learning_rate": 3.8311324529811925e-05,
+      "loss": 0.5696,
+      "step": 5852
+    },
+    {
+      "epoch": 7.49184,
+      "grad_norm": 0.8147026896476746,
+      "learning_rate": 3.83093237294918e-05,
+      "loss": 0.6369,
+      "step": 5853
+    },
+    {
+      "epoch": 7.49312,
+      "grad_norm": 0.7695271968841553,
+      "learning_rate": 3.830732292917167e-05,
+      "loss": 0.6022,
+      "step": 5854
+    },
+    {
+      "epoch": 7.4944,
+      "grad_norm": 0.7904877066612244,
+      "learning_rate": 3.830532212885155e-05,
+      "loss": 0.5893,
+      "step": 5855
+    },
+    {
+      "epoch": 7.49568,
+      "grad_norm": 0.7861725687980652,
+      "learning_rate": 3.830332132853142e-05,
+      "loss": 0.6368,
+      "step": 5856
+    },
+    {
+      "epoch": 7.49696,
+      "grad_norm": 0.7913157343864441,
+      "learning_rate": 3.8301320528211284e-05,
+      "loss": 0.6459,
+      "step": 5857
+    },
+    {
+      "epoch": 7.49824,
+      "grad_norm": 0.7918976545333862,
+      "learning_rate": 3.8299319727891156e-05,
+      "loss": 0.5793,
+      "step": 5858
+    },
+    {
+      "epoch": 7.49952,
+      "grad_norm": 0.8110979199409485,
+      "learning_rate": 3.829731892757103e-05,
+      "loss": 0.589,
+      "step": 5859
+    },
+    {
+      "epoch": 7.5008,
+      "grad_norm": 0.8010942339897156,
+      "learning_rate": 3.82953181272509e-05,
+      "loss": 0.5936,
+      "step": 5860
+    },
+    {
+      "epoch": 7.50208,
+      "grad_norm": 0.8619551062583923,
+      "learning_rate": 3.829331732693077e-05,
+      "loss": 0.6526,
+      "step": 5861
+    },
+    {
+      "epoch": 7.50336,
+      "grad_norm": 0.7884719371795654,
+      "learning_rate": 3.829131652661065e-05,
+      "loss": 0.5931,
+      "step": 5862
+    },
+    {
+      "epoch": 7.50464,
+      "grad_norm": 0.7682931423187256,
+      "learning_rate": 3.828931572629052e-05,
+      "loss": 0.5891,
+      "step": 5863
+    },
+    {
+      "epoch": 7.50592,
+      "grad_norm": 0.7763561606407166,
+      "learning_rate": 3.8287314925970394e-05,
+      "loss": 0.5551,
+      "step": 5864
+    },
+    {
+      "epoch": 7.5072,
+      "grad_norm": 0.7990860342979431,
+      "learning_rate": 3.828531412565026e-05,
+      "loss": 0.586,
+      "step": 5865
+    },
+    {
+      "epoch": 7.5084800000000005,
+      "grad_norm": 0.7618961334228516,
+      "learning_rate": 3.828331332533013e-05,
+      "loss": 0.5892,
+      "step": 5866
+    },
+    {
+      "epoch": 7.50976,
+      "grad_norm": 0.8227397203445435,
+      "learning_rate": 3.828131252501e-05,
+      "loss": 0.5808,
+      "step": 5867
+    },
+    {
+      "epoch": 7.51104,
+      "grad_norm": 0.7972336411476135,
+      "learning_rate": 3.8279311724689875e-05,
+      "loss": 0.5937,
+      "step": 5868
+    },
+    {
+      "epoch": 7.51232,
+      "grad_norm": 0.7725977301597595,
+      "learning_rate": 3.8277310924369746e-05,
+      "loss": 0.5742,
+      "step": 5869
+    },
+    {
+      "epoch": 7.5136,
+      "grad_norm": 0.7537288665771484,
+      "learning_rate": 3.8275310124049625e-05,
+      "loss": 0.5535,
+      "step": 5870
+    },
+    {
+      "epoch": 7.51488,
+      "grad_norm": 0.8171656131744385,
+      "learning_rate": 3.82733093237295e-05,
+      "loss": 0.5987,
+      "step": 5871
+    },
+    {
+      "epoch": 7.51616,
+      "grad_norm": 0.8476085662841797,
+      "learning_rate": 3.827130852340937e-05,
+      "loss": 0.6341,
+      "step": 5872
+    },
+    {
+      "epoch": 7.51744,
+      "grad_norm": 0.7884910106658936,
+      "learning_rate": 3.8269307723089234e-05,
+      "loss": 0.5655,
+      "step": 5873
+    },
+    {
+      "epoch": 7.51872,
+      "grad_norm": 0.7917491793632507,
+      "learning_rate": 3.8267306922769106e-05,
+      "loss": 0.5747,
+      "step": 5874
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.7862417697906494,
+      "learning_rate": 3.826530612244898e-05,
+      "loss": 0.5739,
+      "step": 5875
+    },
+    {
+      "epoch": 7.52128,
+      "grad_norm": 0.7860782146453857,
+      "learning_rate": 3.826330532212885e-05,
+      "loss": 0.5958,
+      "step": 5876
+    },
+    {
+      "epoch": 7.52256,
+      "grad_norm": 0.7811517715454102,
+      "learning_rate": 3.826130452180873e-05,
+      "loss": 0.5845,
+      "step": 5877
+    },
+    {
+      "epoch": 7.52384,
+      "grad_norm": 0.7601850628852844,
+      "learning_rate": 3.82593037214886e-05,
+      "loss": 0.5989,
+      "step": 5878
+    },
+    {
+      "epoch": 7.52512,
+      "grad_norm": 0.727415919303894,
+      "learning_rate": 3.825730292116847e-05,
+      "loss": 0.5449,
+      "step": 5879
+    },
+    {
+      "epoch": 7.5264,
+      "grad_norm": 0.8012895584106445,
+      "learning_rate": 3.8255302120848344e-05,
+      "loss": 0.5999,
+      "step": 5880
+    },
+    {
+      "epoch": 7.52768,
+      "grad_norm": 0.7879264950752258,
+      "learning_rate": 3.825330132052821e-05,
+      "loss": 0.5992,
+      "step": 5881
+    },
+    {
+      "epoch": 7.52896,
+      "grad_norm": 0.7692047953605652,
+      "learning_rate": 3.825130052020808e-05,
+      "loss": 0.5393,
+      "step": 5882
+    },
+    {
+      "epoch": 7.53024,
+      "grad_norm": 0.8411558270454407,
+      "learning_rate": 3.824929971988795e-05,
+      "loss": 0.593,
+      "step": 5883
+    },
+    {
+      "epoch": 7.53152,
+      "grad_norm": 0.8363029956817627,
+      "learning_rate": 3.824729891956783e-05,
+      "loss": 0.6034,
+      "step": 5884
+    },
+    {
+      "epoch": 7.5328,
+      "grad_norm": 0.7915286421775818,
+      "learning_rate": 3.82452981192477e-05,
+      "loss": 0.5933,
+      "step": 5885
+    },
+    {
+      "epoch": 7.53408,
+      "grad_norm": 0.8054776787757874,
+      "learning_rate": 3.8243297318927575e-05,
+      "loss": 0.6009,
+      "step": 5886
+    },
+    {
+      "epoch": 7.53536,
+      "grad_norm": 0.7464303970336914,
+      "learning_rate": 3.824129651860745e-05,
+      "loss": 0.5387,
+      "step": 5887
+    },
+    {
+      "epoch": 7.53664,
+      "grad_norm": 0.7987693548202515,
+      "learning_rate": 3.823929571828732e-05,
+      "loss": 0.5608,
+      "step": 5888
+    },
+    {
+      "epoch": 7.53792,
+      "grad_norm": 0.7968281507492065,
+      "learning_rate": 3.8237294917967184e-05,
+      "loss": 0.605,
+      "step": 5889
+    },
+    {
+      "epoch": 7.5392,
+      "grad_norm": 0.7655385136604309,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.5751,
+      "step": 5890
+    },
+    {
+      "epoch": 7.54048,
+      "grad_norm": 0.8068759441375732,
+      "learning_rate": 3.8233293317326934e-05,
+      "loss": 0.6387,
+      "step": 5891
+    },
+    {
+      "epoch": 7.54176,
+      "grad_norm": 0.7618523240089417,
+      "learning_rate": 3.8231292517006806e-05,
+      "loss": 0.5853,
+      "step": 5892
+    },
+    {
+      "epoch": 7.5430399999999995,
+      "grad_norm": 0.7742576599121094,
+      "learning_rate": 3.822929171668668e-05,
+      "loss": 0.5768,
+      "step": 5893
+    },
+    {
+      "epoch": 7.54432,
+      "grad_norm": 0.7480475902557373,
+      "learning_rate": 3.822729091636655e-05,
+      "loss": 0.5423,
+      "step": 5894
+    },
+    {
+      "epoch": 7.5456,
+      "grad_norm": 0.780965268611908,
+      "learning_rate": 3.822529011604642e-05,
+      "loss": 0.6088,
+      "step": 5895
+    },
+    {
+      "epoch": 7.54688,
+      "grad_norm": 0.8063462376594543,
+      "learning_rate": 3.8223289315726293e-05,
+      "loss": 0.5745,
+      "step": 5896
+    },
+    {
+      "epoch": 7.54816,
+      "grad_norm": 0.8213686347007751,
+      "learning_rate": 3.822128851540616e-05,
+      "loss": 0.6227,
+      "step": 5897
+    },
+    {
+      "epoch": 7.54944,
+      "grad_norm": 0.8014224767684937,
+      "learning_rate": 3.821928771508604e-05,
+      "loss": 0.6174,
+      "step": 5898
+    },
+    {
+      "epoch": 7.55072,
+      "grad_norm": 0.8072211146354675,
+      "learning_rate": 3.821728691476591e-05,
+      "loss": 0.6356,
+      "step": 5899
+    },
+    {
+      "epoch": 7.552,
+      "grad_norm": 0.8300227522850037,
+      "learning_rate": 3.821528611444578e-05,
+      "loss": 0.62,
+      "step": 5900
+    },
+    {
+      "epoch": 7.55328,
+      "grad_norm": 0.7829582095146179,
+      "learning_rate": 3.821328531412565e-05,
+      "loss": 0.5797,
+      "step": 5901
+    },
+    {
+      "epoch": 7.55456,
+      "grad_norm": 0.8026710152626038,
+      "learning_rate": 3.8211284513805525e-05,
+      "loss": 0.573,
+      "step": 5902
+    },
+    {
+      "epoch": 7.55584,
+      "grad_norm": 0.8413091897964478,
+      "learning_rate": 3.8209283713485396e-05,
+      "loss": 0.6193,
+      "step": 5903
+    },
+    {
+      "epoch": 7.55712,
+      "grad_norm": 0.7783622741699219,
+      "learning_rate": 3.820728291316527e-05,
+      "loss": 0.6419,
+      "step": 5904
+    },
+    {
+      "epoch": 7.5584,
+      "grad_norm": 0.804766058921814,
+      "learning_rate": 3.820528211284514e-05,
+      "loss": 0.5698,
+      "step": 5905
+    },
+    {
+      "epoch": 7.55968,
+      "grad_norm": 0.7991184592247009,
+      "learning_rate": 3.820328131252501e-05,
+      "loss": 0.6626,
+      "step": 5906
+    },
+    {
+      "epoch": 7.56096,
+      "grad_norm": 0.8455173969268799,
+      "learning_rate": 3.8201280512204884e-05,
+      "loss": 0.6492,
+      "step": 5907
+    },
+    {
+      "epoch": 7.56224,
+      "grad_norm": 0.8010911345481873,
+      "learning_rate": 3.8199279711884756e-05,
+      "loss": 0.625,
+      "step": 5908
+    },
+    {
+      "epoch": 7.5635200000000005,
+      "grad_norm": 0.7779704928398132,
+      "learning_rate": 3.819727891156463e-05,
+      "loss": 0.5656,
+      "step": 5909
+    },
+    {
+      "epoch": 7.5648,
+      "grad_norm": 0.7678871750831604,
+      "learning_rate": 3.81952781112445e-05,
+      "loss": 0.5712,
+      "step": 5910
+    },
+    {
+      "epoch": 7.56608,
+      "grad_norm": 0.7485666871070862,
+      "learning_rate": 3.819327731092437e-05,
+      "loss": 0.5766,
+      "step": 5911
+    },
+    {
+      "epoch": 7.56736,
+      "grad_norm": 0.8102645874023438,
+      "learning_rate": 3.819127651060424e-05,
+      "loss": 0.6139,
+      "step": 5912
+    },
+    {
+      "epoch": 7.56864,
+      "grad_norm": 0.8365371227264404,
+      "learning_rate": 3.8189275710284115e-05,
+      "loss": 0.6378,
+      "step": 5913
+    },
+    {
+      "epoch": 7.56992,
+      "grad_norm": 0.7460724115371704,
+      "learning_rate": 3.818727490996399e-05,
+      "loss": 0.5879,
+      "step": 5914
+    },
+    {
+      "epoch": 7.5712,
+      "grad_norm": 0.8165127038955688,
+      "learning_rate": 3.818527410964386e-05,
+      "loss": 0.5859,
+      "step": 5915
+    },
+    {
+      "epoch": 7.57248,
+      "grad_norm": 0.8666216135025024,
+      "learning_rate": 3.818327330932373e-05,
+      "loss": 0.6287,
+      "step": 5916
+    },
+    {
+      "epoch": 7.57376,
+      "grad_norm": 0.8045700788497925,
+      "learning_rate": 3.81812725090036e-05,
+      "loss": 0.6372,
+      "step": 5917
+    },
+    {
+      "epoch": 7.5750399999999996,
+      "grad_norm": 0.7294300198554993,
+      "learning_rate": 3.8179271708683474e-05,
+      "loss": 0.5485,
+      "step": 5918
+    },
+    {
+      "epoch": 7.57632,
+      "grad_norm": 0.7684528827667236,
+      "learning_rate": 3.817727090836335e-05,
+      "loss": 0.5719,
+      "step": 5919
+    },
+    {
+      "epoch": 7.5776,
+      "grad_norm": 0.8162046670913696,
+      "learning_rate": 3.817527010804322e-05,
+      "loss": 0.6237,
+      "step": 5920
+    },
+    {
+      "epoch": 7.57888,
+      "grad_norm": 0.7951498031616211,
+      "learning_rate": 3.817326930772309e-05,
+      "loss": 0.6499,
+      "step": 5921
+    },
+    {
+      "epoch": 7.58016,
+      "grad_norm": 0.7112811803817749,
+      "learning_rate": 3.817126850740296e-05,
+      "loss": 0.5304,
+      "step": 5922
+    },
+    {
+      "epoch": 7.58144,
+      "grad_norm": 0.794484555721283,
+      "learning_rate": 3.8169267707082834e-05,
+      "loss": 0.6157,
+      "step": 5923
+    },
+    {
+      "epoch": 7.58272,
+      "grad_norm": 0.769756019115448,
+      "learning_rate": 3.8167266906762705e-05,
+      "loss": 0.5664,
+      "step": 5924
+    },
+    {
+      "epoch": 7.584,
+      "grad_norm": 0.8022031784057617,
+      "learning_rate": 3.816526610644258e-05,
+      "loss": 0.6011,
+      "step": 5925
+    },
+    {
+      "epoch": 7.58528,
+      "grad_norm": 0.8301845788955688,
+      "learning_rate": 3.8163265306122456e-05,
+      "loss": 0.6288,
+      "step": 5926
+    },
+    {
+      "epoch": 7.58656,
+      "grad_norm": 0.7498910427093506,
+      "learning_rate": 3.816126450580233e-05,
+      "loss": 0.5681,
+      "step": 5927
+    },
+    {
+      "epoch": 7.58784,
+      "grad_norm": 0.8285461068153381,
+      "learning_rate": 3.815926370548219e-05,
+      "loss": 0.6179,
+      "step": 5928
+    },
+    {
+      "epoch": 7.58912,
+      "grad_norm": 0.795591413974762,
+      "learning_rate": 3.8157262905162065e-05,
+      "loss": 0.6189,
+      "step": 5929
+    },
+    {
+      "epoch": 7.5904,
+      "grad_norm": 0.7645912170410156,
+      "learning_rate": 3.8155262104841937e-05,
+      "loss": 0.558,
+      "step": 5930
+    },
+    {
+      "epoch": 7.59168,
+      "grad_norm": 0.7344127297401428,
+      "learning_rate": 3.815326130452181e-05,
+      "loss": 0.5886,
+      "step": 5931
+    },
+    {
+      "epoch": 7.59296,
+      "grad_norm": 0.7792043685913086,
+      "learning_rate": 3.815126050420168e-05,
+      "loss": 0.5778,
+      "step": 5932
+    },
+    {
+      "epoch": 7.59424,
+      "grad_norm": 0.7682183980941772,
+      "learning_rate": 3.814925970388156e-05,
+      "loss": 0.6103,
+      "step": 5933
+    },
+    {
+      "epoch": 7.5955200000000005,
+      "grad_norm": 0.8293752670288086,
+      "learning_rate": 3.814725890356143e-05,
+      "loss": 0.6615,
+      "step": 5934
+    },
+    {
+      "epoch": 7.5968,
+      "grad_norm": 0.8193676471710205,
+      "learning_rate": 3.81452581032413e-05,
+      "loss": 0.6054,
+      "step": 5935
+    },
+    {
+      "epoch": 7.5980799999999995,
+      "grad_norm": 0.799968957901001,
+      "learning_rate": 3.814325730292117e-05,
+      "loss": 0.5967,
+      "step": 5936
+    },
+    {
+      "epoch": 7.59936,
+      "grad_norm": 0.7881608605384827,
+      "learning_rate": 3.814125650260104e-05,
+      "loss": 0.5734,
+      "step": 5937
+    },
+    {
+      "epoch": 7.60064,
+      "grad_norm": 0.7928145527839661,
+      "learning_rate": 3.813925570228091e-05,
+      "loss": 0.62,
+      "step": 5938
+    },
+    {
+      "epoch": 7.60192,
+      "grad_norm": 0.8052318096160889,
+      "learning_rate": 3.813725490196078e-05,
+      "loss": 0.6097,
+      "step": 5939
+    },
+    {
+      "epoch": 7.6032,
+      "grad_norm": 0.7783374786376953,
+      "learning_rate": 3.813525410164066e-05,
+      "loss": 0.624,
+      "step": 5940
+    },
+    {
+      "epoch": 7.60448,
+      "grad_norm": 0.8174735307693481,
+      "learning_rate": 3.8133253301320534e-05,
+      "loss": 0.5862,
+      "step": 5941
+    },
+    {
+      "epoch": 7.60576,
+      "grad_norm": 0.8126026391983032,
+      "learning_rate": 3.8131252501000406e-05,
+      "loss": 0.6185,
+      "step": 5942
+    },
+    {
+      "epoch": 7.60704,
+      "grad_norm": 0.796207070350647,
+      "learning_rate": 3.812925170068028e-05,
+      "loss": 0.5721,
+      "step": 5943
+    },
+    {
+      "epoch": 7.60832,
+      "grad_norm": 0.8204118013381958,
+      "learning_rate": 3.812725090036014e-05,
+      "loss": 0.6101,
+      "step": 5944
+    },
+    {
+      "epoch": 7.6096,
+      "grad_norm": 0.8275149464607239,
+      "learning_rate": 3.8125250100040014e-05,
+      "loss": 0.6486,
+      "step": 5945
+    },
+    {
+      "epoch": 7.61088,
+      "grad_norm": 0.8166235089302063,
+      "learning_rate": 3.8123249299719886e-05,
+      "loss": 0.5877,
+      "step": 5946
+    },
+    {
+      "epoch": 7.61216,
+      "grad_norm": 0.7479138970375061,
+      "learning_rate": 3.8121248499399765e-05,
+      "loss": 0.5492,
+      "step": 5947
+    },
+    {
+      "epoch": 7.61344,
+      "grad_norm": 0.8284672498703003,
+      "learning_rate": 3.811924769907964e-05,
+      "loss": 0.6257,
+      "step": 5948
+    },
+    {
+      "epoch": 7.61472,
+      "grad_norm": 0.7393089532852173,
+      "learning_rate": 3.811724689875951e-05,
+      "loss": 0.5608,
+      "step": 5949
+    },
+    {
+      "epoch": 7.616,
+      "grad_norm": 0.8117731213569641,
+      "learning_rate": 3.811524609843938e-05,
+      "loss": 0.6299,
+      "step": 5950
+    },
+    {
+      "epoch": 7.61728,
+      "grad_norm": 0.7665390968322754,
+      "learning_rate": 3.811324529811925e-05,
+      "loss": 0.5933,
+      "step": 5951
+    },
+    {
+      "epoch": 7.61856,
+      "grad_norm": 0.7984145879745483,
+      "learning_rate": 3.811124449779912e-05,
+      "loss": 0.6006,
+      "step": 5952
+    },
+    {
+      "epoch": 7.61984,
+      "grad_norm": 0.7704722881317139,
+      "learning_rate": 3.810924369747899e-05,
+      "loss": 0.5863,
+      "step": 5953
+    },
+    {
+      "epoch": 7.62112,
+      "grad_norm": 0.845035970211029,
+      "learning_rate": 3.810724289715887e-05,
+      "loss": 0.6554,
+      "step": 5954
+    },
+    {
+      "epoch": 7.6224,
+      "grad_norm": 0.8365921378135681,
+      "learning_rate": 3.810524209683874e-05,
+      "loss": 0.6329,
+      "step": 5955
+    },
+    {
+      "epoch": 7.62368,
+      "grad_norm": 0.8249934911727905,
+      "learning_rate": 3.810324129651861e-05,
+      "loss": 0.5811,
+      "step": 5956
+    },
+    {
+      "epoch": 7.62496,
+      "grad_norm": 0.7940043807029724,
+      "learning_rate": 3.8101240496198483e-05,
+      "loss": 0.5887,
+      "step": 5957
+    },
+    {
+      "epoch": 7.62624,
+      "grad_norm": 0.8075708746910095,
+      "learning_rate": 3.8099239695878355e-05,
+      "loss": 0.6427,
+      "step": 5958
+    },
+    {
+      "epoch": 7.62752,
+      "grad_norm": 0.7509844303131104,
+      "learning_rate": 3.809723889555823e-05,
+      "loss": 0.5248,
+      "step": 5959
+    },
+    {
+      "epoch": 7.6288,
+      "grad_norm": 0.8051660060882568,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.591,
+      "step": 5960
+    },
+    {
+      "epoch": 7.6300799999999995,
+      "grad_norm": 0.8326472640037537,
+      "learning_rate": 3.809323729491797e-05,
+      "loss": 0.6136,
+      "step": 5961
+    },
+    {
+      "epoch": 7.63136,
+      "grad_norm": 0.792871356010437,
+      "learning_rate": 3.809123649459784e-05,
+      "loss": 0.5931,
+      "step": 5962
+    },
+    {
+      "epoch": 7.63264,
+      "grad_norm": 0.7740362882614136,
+      "learning_rate": 3.8089235694277715e-05,
+      "loss": 0.5771,
+      "step": 5963
+    },
+    {
+      "epoch": 7.63392,
+      "grad_norm": 0.8323248624801636,
+      "learning_rate": 3.8087234893957586e-05,
+      "loss": 0.6383,
+      "step": 5964
+    },
+    {
+      "epoch": 7.6352,
+      "grad_norm": 0.8095336556434631,
+      "learning_rate": 3.808523409363746e-05,
+      "loss": 0.5937,
+      "step": 5965
+    },
+    {
+      "epoch": 7.63648,
+      "grad_norm": 0.8045695424079895,
+      "learning_rate": 3.808323329331733e-05,
+      "loss": 0.5883,
+      "step": 5966
+    },
+    {
+      "epoch": 7.63776,
+      "grad_norm": 0.825339138507843,
+      "learning_rate": 3.80812324929972e-05,
+      "loss": 0.5761,
+      "step": 5967
+    },
+    {
+      "epoch": 7.63904,
+      "grad_norm": 0.8067010641098022,
+      "learning_rate": 3.8079231692677074e-05,
+      "loss": 0.6219,
+      "step": 5968
+    },
+    {
+      "epoch": 7.64032,
+      "grad_norm": 0.7915658950805664,
+      "learning_rate": 3.8077230892356946e-05,
+      "loss": 0.5803,
+      "step": 5969
+    },
+    {
+      "epoch": 7.6416,
+      "grad_norm": 0.7723389267921448,
+      "learning_rate": 3.807523009203682e-05,
+      "loss": 0.5954,
+      "step": 5970
+    },
+    {
+      "epoch": 7.64288,
+      "grad_norm": 0.8201357126235962,
+      "learning_rate": 3.807322929171669e-05,
+      "loss": 0.627,
+      "step": 5971
+    },
+    {
+      "epoch": 7.64416,
+      "grad_norm": 0.8042327761650085,
+      "learning_rate": 3.807122849139656e-05,
+      "loss": 0.6212,
+      "step": 5972
+    },
+    {
+      "epoch": 7.64544,
+      "grad_norm": 0.8056513667106628,
+      "learning_rate": 3.806922769107643e-05,
+      "loss": 0.5762,
+      "step": 5973
+    },
+    {
+      "epoch": 7.64672,
+      "grad_norm": 0.7721224427223206,
+      "learning_rate": 3.8067226890756305e-05,
+      "loss": 0.6134,
+      "step": 5974
+    },
+    {
+      "epoch": 7.648,
+      "grad_norm": 0.8109544515609741,
+      "learning_rate": 3.806522609043618e-05,
+      "loss": 0.6003,
+      "step": 5975
+    },
+    {
+      "epoch": 7.64928,
+      "grad_norm": 0.7777767181396484,
+      "learning_rate": 3.806322529011605e-05,
+      "loss": 0.6004,
+      "step": 5976
+    },
+    {
+      "epoch": 7.6505600000000005,
+      "grad_norm": 0.8008784651756287,
+      "learning_rate": 3.806122448979592e-05,
+      "loss": 0.6076,
+      "step": 5977
+    },
+    {
+      "epoch": 7.65184,
+      "grad_norm": 0.8032917976379395,
+      "learning_rate": 3.805922368947579e-05,
+      "loss": 0.595,
+      "step": 5978
+    },
+    {
+      "epoch": 7.65312,
+      "grad_norm": 0.7758605480194092,
+      "learning_rate": 3.8057222889155664e-05,
+      "loss": 0.5175,
+      "step": 5979
+    },
+    {
+      "epoch": 7.6544,
+      "grad_norm": 0.7716826796531677,
+      "learning_rate": 3.8055222088835536e-05,
+      "loss": 0.603,
+      "step": 5980
+    },
+    {
+      "epoch": 7.65568,
+      "grad_norm": 0.7985824346542358,
+      "learning_rate": 3.805322128851541e-05,
+      "loss": 0.602,
+      "step": 5981
+    },
+    {
+      "epoch": 7.65696,
+      "grad_norm": 0.8194901347160339,
+      "learning_rate": 3.805122048819528e-05,
+      "loss": 0.6157,
+      "step": 5982
+    },
+    {
+      "epoch": 7.65824,
+      "grad_norm": 0.8596216440200806,
+      "learning_rate": 3.804921968787515e-05,
+      "loss": 0.6033,
+      "step": 5983
+    },
+    {
+      "epoch": 7.65952,
+      "grad_norm": 0.8513696193695068,
+      "learning_rate": 3.8047218887555024e-05,
+      "loss": 0.6309,
+      "step": 5984
+    },
+    {
+      "epoch": 7.6608,
+      "grad_norm": 0.8196579813957214,
+      "learning_rate": 3.8045218087234895e-05,
+      "loss": 0.6085,
+      "step": 5985
+    },
+    {
+      "epoch": 7.66208,
+      "grad_norm": 0.7808190584182739,
+      "learning_rate": 3.804321728691477e-05,
+      "loss": 0.6266,
+      "step": 5986
+    },
+    {
+      "epoch": 7.66336,
+      "grad_norm": 0.792535662651062,
+      "learning_rate": 3.804121648659464e-05,
+      "loss": 0.5974,
+      "step": 5987
+    },
+    {
+      "epoch": 7.66464,
+      "grad_norm": 0.7937912940979004,
+      "learning_rate": 3.803921568627451e-05,
+      "loss": 0.6269,
+      "step": 5988
+    },
+    {
+      "epoch": 7.66592,
+      "grad_norm": 0.7414507269859314,
+      "learning_rate": 3.803721488595438e-05,
+      "loss": 0.575,
+      "step": 5989
+    },
+    {
+      "epoch": 7.6672,
+      "grad_norm": 0.8128142952919006,
+      "learning_rate": 3.8035214085634255e-05,
+      "loss": 0.5785,
+      "step": 5990
+    },
+    {
+      "epoch": 7.66848,
+      "grad_norm": 0.8728705644607544,
+      "learning_rate": 3.8033213285314127e-05,
+      "loss": 0.6354,
+      "step": 5991
+    },
+    {
+      "epoch": 7.66976,
+      "grad_norm": 0.7969766855239868,
+      "learning_rate": 3.8031212484994e-05,
+      "loss": 0.6223,
+      "step": 5992
+    },
+    {
+      "epoch": 7.67104,
+      "grad_norm": 0.8199662566184998,
+      "learning_rate": 3.802921168467387e-05,
+      "loss": 0.6278,
+      "step": 5993
+    },
+    {
+      "epoch": 7.67232,
+      "grad_norm": 0.7900199890136719,
+      "learning_rate": 3.802721088435374e-05,
+      "loss": 0.5949,
+      "step": 5994
+    },
+    {
+      "epoch": 7.6736,
+      "grad_norm": 0.7966949939727783,
+      "learning_rate": 3.8025210084033614e-05,
+      "loss": 0.6385,
+      "step": 5995
+    },
+    {
+      "epoch": 7.67488,
+      "grad_norm": 0.7564438581466675,
+      "learning_rate": 3.8023209283713486e-05,
+      "loss": 0.5726,
+      "step": 5996
+    },
+    {
+      "epoch": 7.67616,
+      "grad_norm": 0.8227134943008423,
+      "learning_rate": 3.8021208483393364e-05,
+      "loss": 0.6368,
+      "step": 5997
+    },
+    {
+      "epoch": 7.67744,
+      "grad_norm": 0.8298145532608032,
+      "learning_rate": 3.801920768307323e-05,
+      "loss": 0.6483,
+      "step": 5998
+    },
+    {
+      "epoch": 7.67872,
+      "grad_norm": 0.8414880037307739,
+      "learning_rate": 3.80172068827531e-05,
+      "loss": 0.6564,
+      "step": 5999
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.7608658075332642,
+      "learning_rate": 3.801520608243297e-05,
+      "loss": 0.5326,
+      "step": 6000
+    },
+    {
+      "epoch": 7.68128,
+      "grad_norm": 0.816177487373352,
+      "learning_rate": 3.8013205282112845e-05,
+      "loss": 0.5995,
+      "step": 6001
+    },
+    {
+      "epoch": 7.68256,
+      "grad_norm": 0.7969012260437012,
+      "learning_rate": 3.801120448179272e-05,
+      "loss": 0.576,
+      "step": 6002
+    },
+    {
+      "epoch": 7.68384,
+      "grad_norm": 0.786906361579895,
+      "learning_rate": 3.800920368147259e-05,
+      "loss": 0.6084,
+      "step": 6003
+    },
+    {
+      "epoch": 7.6851199999999995,
+      "grad_norm": 0.764916718006134,
+      "learning_rate": 3.800720288115247e-05,
+      "loss": 0.6095,
+      "step": 6004
+    },
+    {
+      "epoch": 7.6864,
+      "grad_norm": 0.7754150629043579,
+      "learning_rate": 3.800520208083234e-05,
+      "loss": 0.6024,
+      "step": 6005
+    },
+    {
+      "epoch": 7.68768,
+      "grad_norm": 0.7677636742591858,
+      "learning_rate": 3.8003201280512204e-05,
+      "loss": 0.5886,
+      "step": 6006
+    },
+    {
+      "epoch": 7.68896,
+      "grad_norm": 0.8274985551834106,
+      "learning_rate": 3.8001200480192076e-05,
+      "loss": 0.6117,
+      "step": 6007
+    },
+    {
+      "epoch": 7.69024,
+      "grad_norm": 0.8422303199768066,
+      "learning_rate": 3.799919967987195e-05,
+      "loss": 0.6012,
+      "step": 6008
+    },
+    {
+      "epoch": 7.69152,
+      "grad_norm": 0.7773991227149963,
+      "learning_rate": 3.799719887955182e-05,
+      "loss": 0.6188,
+      "step": 6009
+    },
+    {
+      "epoch": 7.6928,
+      "grad_norm": 0.8241961002349854,
+      "learning_rate": 3.799519807923169e-05,
+      "loss": 0.6102,
+      "step": 6010
+    },
+    {
+      "epoch": 7.69408,
+      "grad_norm": 0.7824380397796631,
+      "learning_rate": 3.799319727891157e-05,
+      "loss": 0.6254,
+      "step": 6011
+    },
+    {
+      "epoch": 7.69536,
+      "grad_norm": 0.7920189499855042,
+      "learning_rate": 3.799119647859144e-05,
+      "loss": 0.687,
+      "step": 6012
+    },
+    {
+      "epoch": 7.69664,
+      "grad_norm": 0.7902247309684753,
+      "learning_rate": 3.7989195678271314e-05,
+      "loss": 0.5669,
+      "step": 6013
+    },
+    {
+      "epoch": 7.69792,
+      "grad_norm": 0.8157680630683899,
+      "learning_rate": 3.798719487795118e-05,
+      "loss": 0.6307,
+      "step": 6014
+    },
+    {
+      "epoch": 7.6992,
+      "grad_norm": 0.8133694529533386,
+      "learning_rate": 3.798519407763105e-05,
+      "loss": 0.6101,
+      "step": 6015
+    },
+    {
+      "epoch": 7.70048,
+      "grad_norm": 0.8677835464477539,
+      "learning_rate": 3.798319327731092e-05,
+      "loss": 0.6809,
+      "step": 6016
+    },
+    {
+      "epoch": 7.70176,
+      "grad_norm": 0.7828230261802673,
+      "learning_rate": 3.7981192476990795e-05,
+      "loss": 0.5994,
+      "step": 6017
+    },
+    {
+      "epoch": 7.70304,
+      "grad_norm": 0.8306964039802551,
+      "learning_rate": 3.7979191676670673e-05,
+      "loss": 0.6098,
+      "step": 6018
+    },
+    {
+      "epoch": 7.70432,
+      "grad_norm": 0.7977209687232971,
+      "learning_rate": 3.7977190876350545e-05,
+      "loss": 0.6301,
+      "step": 6019
+    },
+    {
+      "epoch": 7.7056000000000004,
+      "grad_norm": 0.7888930439949036,
+      "learning_rate": 3.797519007603042e-05,
+      "loss": 0.6073,
+      "step": 6020
+    },
+    {
+      "epoch": 7.70688,
+      "grad_norm": 0.813720166683197,
+      "learning_rate": 3.797318927571029e-05,
+      "loss": 0.6056,
+      "step": 6021
+    },
+    {
+      "epoch": 7.70816,
+      "grad_norm": 0.7712581157684326,
+      "learning_rate": 3.7971188475390154e-05,
+      "loss": 0.6198,
+      "step": 6022
+    },
+    {
+      "epoch": 7.70944,
+      "grad_norm": 0.8599300384521484,
+      "learning_rate": 3.7969187675070026e-05,
+      "loss": 0.6541,
+      "step": 6023
+    },
+    {
+      "epoch": 7.71072,
+      "grad_norm": 0.8106684684753418,
+      "learning_rate": 3.79671868747499e-05,
+      "loss": 0.625,
+      "step": 6024
+    },
+    {
+      "epoch": 7.712,
+      "grad_norm": 0.8189576268196106,
+      "learning_rate": 3.7965186074429776e-05,
+      "loss": 0.6379,
+      "step": 6025
+    },
+    {
+      "epoch": 7.71328,
+      "grad_norm": 0.8133044838905334,
+      "learning_rate": 3.796318527410965e-05,
+      "loss": 0.6289,
+      "step": 6026
+    },
+    {
+      "epoch": 7.71456,
+      "grad_norm": 0.7700027227401733,
+      "learning_rate": 3.796118447378952e-05,
+      "loss": 0.5445,
+      "step": 6027
+    },
+    {
+      "epoch": 7.71584,
+      "grad_norm": 0.7435338497161865,
+      "learning_rate": 3.795918367346939e-05,
+      "loss": 0.5761,
+      "step": 6028
+    },
+    {
+      "epoch": 7.7171199999999995,
+      "grad_norm": 0.7466886639595032,
+      "learning_rate": 3.7957182873149264e-05,
+      "loss": 0.563,
+      "step": 6029
+    },
+    {
+      "epoch": 7.7184,
+      "grad_norm": 0.7899182438850403,
+      "learning_rate": 3.795518207282913e-05,
+      "loss": 0.5442,
+      "step": 6030
+    },
+    {
+      "epoch": 7.71968,
+      "grad_norm": 0.8155211806297302,
+      "learning_rate": 3.7953181272509e-05,
+      "loss": 0.6445,
+      "step": 6031
+    },
+    {
+      "epoch": 7.72096,
+      "grad_norm": 0.8087602257728577,
+      "learning_rate": 3.795118047218888e-05,
+      "loss": 0.6062,
+      "step": 6032
+    },
+    {
+      "epoch": 7.72224,
+      "grad_norm": 0.7604132890701294,
+      "learning_rate": 3.794917967186875e-05,
+      "loss": 0.6147,
+      "step": 6033
+    },
+    {
+      "epoch": 7.72352,
+      "grad_norm": 0.8204428553581238,
+      "learning_rate": 3.794717887154862e-05,
+      "loss": 0.5717,
+      "step": 6034
+    },
+    {
+      "epoch": 7.7248,
+      "grad_norm": 0.7669458985328674,
+      "learning_rate": 3.7945178071228495e-05,
+      "loss": 0.5662,
+      "step": 6035
+    },
+    {
+      "epoch": 7.72608,
+      "grad_norm": 0.765567421913147,
+      "learning_rate": 3.794317727090837e-05,
+      "loss": 0.5558,
+      "step": 6036
+    },
+    {
+      "epoch": 7.72736,
+      "grad_norm": 0.8454002141952515,
+      "learning_rate": 3.794117647058824e-05,
+      "loss": 0.6204,
+      "step": 6037
+    },
+    {
+      "epoch": 7.72864,
+      "grad_norm": 0.7641090750694275,
+      "learning_rate": 3.7939175670268104e-05,
+      "loss": 0.5612,
+      "step": 6038
+    },
+    {
+      "epoch": 7.72992,
+      "grad_norm": 0.8416698575019836,
+      "learning_rate": 3.793717486994798e-05,
+      "loss": 0.63,
+      "step": 6039
+    },
+    {
+      "epoch": 7.7312,
+      "grad_norm": 0.8227651119232178,
+      "learning_rate": 3.7935174069627854e-05,
+      "loss": 0.6666,
+      "step": 6040
+    },
+    {
+      "epoch": 7.73248,
+      "grad_norm": 0.8543378114700317,
+      "learning_rate": 3.7933173269307726e-05,
+      "loss": 0.6023,
+      "step": 6041
+    },
+    {
+      "epoch": 7.73376,
+      "grad_norm": 0.761553943157196,
+      "learning_rate": 3.79311724689876e-05,
+      "loss": 0.583,
+      "step": 6042
+    },
+    {
+      "epoch": 7.73504,
+      "grad_norm": 0.8009664416313171,
+      "learning_rate": 3.792917166866747e-05,
+      "loss": 0.6192,
+      "step": 6043
+    },
+    {
+      "epoch": 7.73632,
+      "grad_norm": 0.7856703400611877,
+      "learning_rate": 3.792717086834734e-05,
+      "loss": 0.5523,
+      "step": 6044
+    },
+    {
+      "epoch": 7.7376000000000005,
+      "grad_norm": 0.7818490862846375,
+      "learning_rate": 3.7925170068027214e-05,
+      "loss": 0.6144,
+      "step": 6045
+    },
+    {
+      "epoch": 7.73888,
+      "grad_norm": 0.7923960089683533,
+      "learning_rate": 3.7923169267707085e-05,
+      "loss": 0.6022,
+      "step": 6046
+    },
+    {
+      "epoch": 7.74016,
+      "grad_norm": 0.769059956073761,
+      "learning_rate": 3.792116846738696e-05,
+      "loss": 0.5452,
+      "step": 6047
+    },
+    {
+      "epoch": 7.74144,
+      "grad_norm": 0.7857373952865601,
+      "learning_rate": 3.791916766706683e-05,
+      "loss": 0.6226,
+      "step": 6048
+    },
+    {
+      "epoch": 7.74272,
+      "grad_norm": 0.7994586825370789,
+      "learning_rate": 3.79171668667467e-05,
+      "loss": 0.5875,
+      "step": 6049
+    },
+    {
+      "epoch": 7.744,
+      "grad_norm": 0.7892863154411316,
+      "learning_rate": 3.791516606642657e-05,
+      "loss": 0.6333,
+      "step": 6050
+    },
+    {
+      "epoch": 7.74528,
+      "grad_norm": 0.8037270903587341,
+      "learning_rate": 3.7913165266106445e-05,
+      "loss": 0.5936,
+      "step": 6051
+    },
+    {
+      "epoch": 7.74656,
+      "grad_norm": 0.7748731374740601,
+      "learning_rate": 3.7911164465786317e-05,
+      "loss": 0.5943,
+      "step": 6052
+    },
+    {
+      "epoch": 7.74784,
+      "grad_norm": 0.7597405910491943,
+      "learning_rate": 3.790916366546619e-05,
+      "loss": 0.5837,
+      "step": 6053
+    },
+    {
+      "epoch": 7.74912,
+      "grad_norm": 0.8196899890899658,
+      "learning_rate": 3.790716286514606e-05,
+      "loss": 0.6103,
+      "step": 6054
+    },
+    {
+      "epoch": 7.7504,
+      "grad_norm": 0.7854442000389099,
+      "learning_rate": 3.790516206482593e-05,
+      "loss": 0.6141,
+      "step": 6055
+    },
+    {
+      "epoch": 7.75168,
+      "grad_norm": 0.7270790934562683,
+      "learning_rate": 3.7903161264505804e-05,
+      "loss": 0.4887,
+      "step": 6056
+    },
+    {
+      "epoch": 7.75296,
+      "grad_norm": 0.75218266248703,
+      "learning_rate": 3.7901160464185676e-05,
+      "loss": 0.5902,
+      "step": 6057
+    },
+    {
+      "epoch": 7.75424,
+      "grad_norm": 0.8306524753570557,
+      "learning_rate": 3.789915966386555e-05,
+      "loss": 0.6816,
+      "step": 6058
+    },
+    {
+      "epoch": 7.75552,
+      "grad_norm": 0.7545936703681946,
+      "learning_rate": 3.789715886354542e-05,
+      "loss": 0.5668,
+      "step": 6059
+    },
+    {
+      "epoch": 7.7568,
+      "grad_norm": 0.7960179448127747,
+      "learning_rate": 3.789515806322529e-05,
+      "loss": 0.5391,
+      "step": 6060
+    },
+    {
+      "epoch": 7.75808,
+      "grad_norm": 0.755833625793457,
+      "learning_rate": 3.789315726290516e-05,
+      "loss": 0.5984,
+      "step": 6061
+    },
+    {
+      "epoch": 7.75936,
+      "grad_norm": 0.8086370229721069,
+      "learning_rate": 3.7891156462585035e-05,
+      "loss": 0.6226,
+      "step": 6062
+    },
+    {
+      "epoch": 7.76064,
+      "grad_norm": 0.8106130361557007,
+      "learning_rate": 3.788915566226491e-05,
+      "loss": 0.6215,
+      "step": 6063
+    },
+    {
+      "epoch": 7.76192,
+      "grad_norm": 0.7895246744155884,
+      "learning_rate": 3.788715486194478e-05,
+      "loss": 0.6468,
+      "step": 6064
+    },
+    {
+      "epoch": 7.7632,
+      "grad_norm": 0.8080568313598633,
+      "learning_rate": 3.788515406162465e-05,
+      "loss": 0.6082,
+      "step": 6065
+    },
+    {
+      "epoch": 7.76448,
+      "grad_norm": 0.7731059193611145,
+      "learning_rate": 3.788315326130452e-05,
+      "loss": 0.6086,
+      "step": 6066
+    },
+    {
+      "epoch": 7.76576,
+      "grad_norm": 0.8263415694236755,
+      "learning_rate": 3.78811524609844e-05,
+      "loss": 0.6548,
+      "step": 6067
+    },
+    {
+      "epoch": 7.76704,
+      "grad_norm": 0.7245672345161438,
+      "learning_rate": 3.7879151660664266e-05,
+      "loss": 0.5606,
+      "step": 6068
+    },
+    {
+      "epoch": 7.76832,
+      "grad_norm": 0.764879047870636,
+      "learning_rate": 3.787715086034414e-05,
+      "loss": 0.6432,
+      "step": 6069
+    },
+    {
+      "epoch": 7.7696,
+      "grad_norm": 0.7973184585571289,
+      "learning_rate": 3.787515006002401e-05,
+      "loss": 0.6108,
+      "step": 6070
+    },
+    {
+      "epoch": 7.77088,
+      "grad_norm": 0.7751659750938416,
+      "learning_rate": 3.787314925970388e-05,
+      "loss": 0.6245,
+      "step": 6071
+    },
+    {
+      "epoch": 7.7721599999999995,
+      "grad_norm": 0.76836758852005,
+      "learning_rate": 3.7871148459383754e-05,
+      "loss": 0.5633,
+      "step": 6072
+    },
+    {
+      "epoch": 7.77344,
+      "grad_norm": 0.7604205012321472,
+      "learning_rate": 3.7869147659063626e-05,
+      "loss": 0.6017,
+      "step": 6073
+    },
+    {
+      "epoch": 7.77472,
+      "grad_norm": 0.8240510821342468,
+      "learning_rate": 3.7867146858743504e-05,
+      "loss": 0.5819,
+      "step": 6074
+    },
+    {
+      "epoch": 7.776,
+      "grad_norm": 0.7578047513961792,
+      "learning_rate": 3.7865146058423376e-05,
+      "loss": 0.562,
+      "step": 6075
+    },
+    {
+      "epoch": 7.77728,
+      "grad_norm": 0.7977784872055054,
+      "learning_rate": 3.786314525810324e-05,
+      "loss": 0.6071,
+      "step": 6076
+    },
+    {
+      "epoch": 7.77856,
+      "grad_norm": 0.7755118608474731,
+      "learning_rate": 3.786114445778311e-05,
+      "loss": 0.5861,
+      "step": 6077
+    },
+    {
+      "epoch": 7.77984,
+      "grad_norm": 0.7646934390068054,
+      "learning_rate": 3.7859143657462985e-05,
+      "loss": 0.6142,
+      "step": 6078
+    },
+    {
+      "epoch": 7.78112,
+      "grad_norm": 0.8070874214172363,
+      "learning_rate": 3.785714285714286e-05,
+      "loss": 0.6089,
+      "step": 6079
+    },
+    {
+      "epoch": 7.7824,
+      "grad_norm": 0.7948157787322998,
+      "learning_rate": 3.785514205682273e-05,
+      "loss": 0.6112,
+      "step": 6080
+    },
+    {
+      "epoch": 7.78368,
+      "grad_norm": 0.7674957513809204,
+      "learning_rate": 3.785314125650261e-05,
+      "loss": 0.5461,
+      "step": 6081
+    },
+    {
+      "epoch": 7.78496,
+      "grad_norm": 0.7721199989318848,
+      "learning_rate": 3.785114045618248e-05,
+      "loss": 0.5872,
+      "step": 6082
+    },
+    {
+      "epoch": 7.78624,
+      "grad_norm": 0.8199625015258789,
+      "learning_rate": 3.784913965586235e-05,
+      "loss": 0.6066,
+      "step": 6083
+    },
+    {
+      "epoch": 7.78752,
+      "grad_norm": 0.8024269342422485,
+      "learning_rate": 3.7847138855542216e-05,
+      "loss": 0.5541,
+      "step": 6084
+    },
+    {
+      "epoch": 7.7888,
+      "grad_norm": 0.7629589438438416,
+      "learning_rate": 3.784513805522209e-05,
+      "loss": 0.5699,
+      "step": 6085
+    },
+    {
+      "epoch": 7.79008,
+      "grad_norm": 0.7849573493003845,
+      "learning_rate": 3.784313725490196e-05,
+      "loss": 0.5958,
+      "step": 6086
+    },
+    {
+      "epoch": 7.79136,
+      "grad_norm": 0.8116906881332397,
+      "learning_rate": 3.784113645458183e-05,
+      "loss": 0.6552,
+      "step": 6087
+    },
+    {
+      "epoch": 7.7926400000000005,
+      "grad_norm": 0.7979661822319031,
+      "learning_rate": 3.783913565426171e-05,
+      "loss": 0.6307,
+      "step": 6088
+    },
+    {
+      "epoch": 7.79392,
+      "grad_norm": 0.8299638628959656,
+      "learning_rate": 3.783713485394158e-05,
+      "loss": 0.6451,
+      "step": 6089
+    },
+    {
+      "epoch": 7.7952,
+      "grad_norm": 0.787670910358429,
+      "learning_rate": 3.7835134053621454e-05,
+      "loss": 0.6323,
+      "step": 6090
+    },
+    {
+      "epoch": 7.79648,
+      "grad_norm": 0.8025368452072144,
+      "learning_rate": 3.7833133253301326e-05,
+      "loss": 0.6199,
+      "step": 6091
+    },
+    {
+      "epoch": 7.79776,
+      "grad_norm": 0.7752863168716431,
+      "learning_rate": 3.783113245298119e-05,
+      "loss": 0.5975,
+      "step": 6092
+    },
+    {
+      "epoch": 7.79904,
+      "grad_norm": 0.7841124534606934,
+      "learning_rate": 3.782913165266106e-05,
+      "loss": 0.5705,
+      "step": 6093
+    },
+    {
+      "epoch": 7.80032,
+      "grad_norm": 0.7986131310462952,
+      "learning_rate": 3.7827130852340935e-05,
+      "loss": 0.5626,
+      "step": 6094
+    },
+    {
+      "epoch": 7.8016,
+      "grad_norm": 0.7659592032432556,
+      "learning_rate": 3.7825130052020806e-05,
+      "loss": 0.582,
+      "step": 6095
+    },
+    {
+      "epoch": 7.80288,
+      "grad_norm": 0.7751227617263794,
+      "learning_rate": 3.7823129251700685e-05,
+      "loss": 0.589,
+      "step": 6096
+    },
+    {
+      "epoch": 7.8041599999999995,
+      "grad_norm": 0.810896635055542,
+      "learning_rate": 3.782112845138056e-05,
+      "loss": 0.5888,
+      "step": 6097
+    },
+    {
+      "epoch": 7.80544,
+      "grad_norm": 0.8425770998001099,
+      "learning_rate": 3.781912765106043e-05,
+      "loss": 0.6118,
+      "step": 6098
+    },
+    {
+      "epoch": 7.80672,
+      "grad_norm": 0.81880784034729,
+      "learning_rate": 3.78171268507403e-05,
+      "loss": 0.6555,
+      "step": 6099
+    },
+    {
+      "epoch": 7.808,
+      "grad_norm": 0.8141629099845886,
+      "learning_rate": 3.7815126050420166e-05,
+      "loss": 0.6628,
+      "step": 6100
+    },
+    {
+      "epoch": 7.80928,
+      "grad_norm": 0.8063963651657104,
+      "learning_rate": 3.781312525010004e-05,
+      "loss": 0.5981,
+      "step": 6101
+    },
+    {
+      "epoch": 7.81056,
+      "grad_norm": 0.8052300214767456,
+      "learning_rate": 3.781112444977991e-05,
+      "loss": 0.5995,
+      "step": 6102
+    },
+    {
+      "epoch": 7.81184,
+      "grad_norm": 0.804269552230835,
+      "learning_rate": 3.780912364945979e-05,
+      "loss": 0.6108,
+      "step": 6103
+    },
+    {
+      "epoch": 7.81312,
+      "grad_norm": 0.7562627196311951,
+      "learning_rate": 3.780712284913966e-05,
+      "loss": 0.5164,
+      "step": 6104
+    },
+    {
+      "epoch": 7.8144,
+      "grad_norm": 0.800515353679657,
+      "learning_rate": 3.780512204881953e-05,
+      "loss": 0.5988,
+      "step": 6105
+    },
+    {
+      "epoch": 7.81568,
+      "grad_norm": 0.8121452331542969,
+      "learning_rate": 3.7803121248499404e-05,
+      "loss": 0.632,
+      "step": 6106
+    },
+    {
+      "epoch": 7.81696,
+      "grad_norm": 0.8428505063056946,
+      "learning_rate": 3.7801120448179275e-05,
+      "loss": 0.6044,
+      "step": 6107
+    },
+    {
+      "epoch": 7.81824,
+      "grad_norm": 0.8227953314781189,
+      "learning_rate": 3.779911964785914e-05,
+      "loss": 0.6519,
+      "step": 6108
+    },
+    {
+      "epoch": 7.81952,
+      "grad_norm": 0.7917014360427856,
+      "learning_rate": 3.779711884753901e-05,
+      "loss": 0.5866,
+      "step": 6109
+    },
+    {
+      "epoch": 7.8208,
+      "grad_norm": 0.8045591115951538,
+      "learning_rate": 3.779511804721889e-05,
+      "loss": 0.6472,
+      "step": 6110
+    },
+    {
+      "epoch": 7.82208,
+      "grad_norm": 0.772802472114563,
+      "learning_rate": 3.779311724689876e-05,
+      "loss": 0.5495,
+      "step": 6111
+    },
+    {
+      "epoch": 7.82336,
+      "grad_norm": 0.8186236023902893,
+      "learning_rate": 3.7791116446578635e-05,
+      "loss": 0.664,
+      "step": 6112
+    },
+    {
+      "epoch": 7.8246400000000005,
+      "grad_norm": 0.806842029094696,
+      "learning_rate": 3.778911564625851e-05,
+      "loss": 0.6216,
+      "step": 6113
+    },
+    {
+      "epoch": 7.82592,
+      "grad_norm": 0.7762079238891602,
+      "learning_rate": 3.778711484593838e-05,
+      "loss": 0.5853,
+      "step": 6114
+    },
+    {
+      "epoch": 7.8272,
+      "grad_norm": 0.8283485770225525,
+      "learning_rate": 3.778511404561825e-05,
+      "loss": 0.5947,
+      "step": 6115
+    },
+    {
+      "epoch": 7.82848,
+      "grad_norm": 0.8360929489135742,
+      "learning_rate": 3.7783113245298115e-05,
+      "loss": 0.6088,
+      "step": 6116
+    },
+    {
+      "epoch": 7.82976,
+      "grad_norm": 0.7732667922973633,
+      "learning_rate": 3.7781112444977994e-05,
+      "loss": 0.6161,
+      "step": 6117
+    },
+    {
+      "epoch": 7.83104,
+      "grad_norm": 0.8080599308013916,
+      "learning_rate": 3.7779111644657866e-05,
+      "loss": 0.6314,
+      "step": 6118
+    },
+    {
+      "epoch": 7.83232,
+      "grad_norm": 0.8000563979148865,
+      "learning_rate": 3.777711084433774e-05,
+      "loss": 0.6548,
+      "step": 6119
+    },
+    {
+      "epoch": 7.8336,
+      "grad_norm": 0.8152042031288147,
+      "learning_rate": 3.777511004401761e-05,
+      "loss": 0.6637,
+      "step": 6120
+    },
+    {
+      "epoch": 7.83488,
+      "grad_norm": 0.8046169877052307,
+      "learning_rate": 3.777310924369748e-05,
+      "loss": 0.6306,
+      "step": 6121
+    },
+    {
+      "epoch": 7.83616,
+      "grad_norm": 0.7880290746688843,
+      "learning_rate": 3.777110844337735e-05,
+      "loss": 0.6418,
+      "step": 6122
+    },
+    {
+      "epoch": 7.83744,
+      "grad_norm": 0.8029178380966187,
+      "learning_rate": 3.7769107643057225e-05,
+      "loss": 0.6183,
+      "step": 6123
+    },
+    {
+      "epoch": 7.83872,
+      "grad_norm": 0.7654139995574951,
+      "learning_rate": 3.77671068427371e-05,
+      "loss": 0.5793,
+      "step": 6124
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.8131774067878723,
+      "learning_rate": 3.776510604241697e-05,
+      "loss": 0.6485,
+      "step": 6125
+    },
+    {
+      "epoch": 7.84128,
+      "grad_norm": 0.7340885996818542,
+      "learning_rate": 3.776310524209684e-05,
+      "loss": 0.5605,
+      "step": 6126
+    },
+    {
+      "epoch": 7.84256,
+      "grad_norm": 0.7948605418205261,
+      "learning_rate": 3.776110444177671e-05,
+      "loss": 0.588,
+      "step": 6127
+    },
+    {
+      "epoch": 7.84384,
+      "grad_norm": 0.823373019695282,
+      "learning_rate": 3.7759103641456584e-05,
+      "loss": 0.5925,
+      "step": 6128
+    },
+    {
+      "epoch": 7.84512,
+      "grad_norm": 0.8136553764343262,
+      "learning_rate": 3.7757102841136456e-05,
+      "loss": 0.6336,
+      "step": 6129
+    },
+    {
+      "epoch": 7.8464,
+      "grad_norm": 0.7955484390258789,
+      "learning_rate": 3.775510204081633e-05,
+      "loss": 0.5678,
+      "step": 6130
+    },
+    {
+      "epoch": 7.84768,
+      "grad_norm": 0.7093358039855957,
+      "learning_rate": 3.77531012404962e-05,
+      "loss": 0.5829,
+      "step": 6131
+    },
+    {
+      "epoch": 7.84896,
+      "grad_norm": 0.7424338459968567,
+      "learning_rate": 3.775110044017607e-05,
+      "loss": 0.5677,
+      "step": 6132
+    },
+    {
+      "epoch": 7.85024,
+      "grad_norm": 0.7559502720832825,
+      "learning_rate": 3.7749099639855944e-05,
+      "loss": 0.5777,
+      "step": 6133
+    },
+    {
+      "epoch": 7.85152,
+      "grad_norm": 0.8359054923057556,
+      "learning_rate": 3.7747098839535816e-05,
+      "loss": 0.5884,
+      "step": 6134
+    },
+    {
+      "epoch": 7.8528,
+      "grad_norm": 0.8114975094795227,
+      "learning_rate": 3.774509803921569e-05,
+      "loss": 0.5916,
+      "step": 6135
+    },
+    {
+      "epoch": 7.85408,
+      "grad_norm": 0.8038936853408813,
+      "learning_rate": 3.774309723889556e-05,
+      "loss": 0.6401,
+      "step": 6136
+    },
+    {
+      "epoch": 7.85536,
+      "grad_norm": 0.7915340065956116,
+      "learning_rate": 3.774109643857543e-05,
+      "loss": 0.574,
+      "step": 6137
+    },
+    {
+      "epoch": 7.85664,
+      "grad_norm": 0.8028757572174072,
+      "learning_rate": 3.77390956382553e-05,
+      "loss": 0.5954,
+      "step": 6138
+    },
+    {
+      "epoch": 7.85792,
+      "grad_norm": 0.7789997458457947,
+      "learning_rate": 3.7737094837935175e-05,
+      "loss": 0.5468,
+      "step": 6139
+    },
+    {
+      "epoch": 7.8591999999999995,
+      "grad_norm": 0.8444523215293884,
+      "learning_rate": 3.773509403761505e-05,
+      "loss": 0.6584,
+      "step": 6140
+    },
+    {
+      "epoch": 7.86048,
+      "grad_norm": 0.7982633709907532,
+      "learning_rate": 3.773309323729492e-05,
+      "loss": 0.6439,
+      "step": 6141
+    },
+    {
+      "epoch": 7.86176,
+      "grad_norm": 0.8487026691436768,
+      "learning_rate": 3.773109243697479e-05,
+      "loss": 0.6307,
+      "step": 6142
+    },
+    {
+      "epoch": 7.86304,
+      "grad_norm": 0.8426850438117981,
+      "learning_rate": 3.772909163665466e-05,
+      "loss": 0.5936,
+      "step": 6143
+    },
+    {
+      "epoch": 7.86432,
+      "grad_norm": 0.82231205701828,
+      "learning_rate": 3.7727090836334534e-05,
+      "loss": 0.5931,
+      "step": 6144
+    },
+    {
+      "epoch": 7.8656,
+      "grad_norm": 0.8080248832702637,
+      "learning_rate": 3.772509003601441e-05,
+      "loss": 0.5993,
+      "step": 6145
+    },
+    {
+      "epoch": 7.86688,
+      "grad_norm": 0.7820031046867371,
+      "learning_rate": 3.772308923569428e-05,
+      "loss": 0.5974,
+      "step": 6146
+    },
+    {
+      "epoch": 7.86816,
+      "grad_norm": 0.7480020523071289,
+      "learning_rate": 3.772108843537415e-05,
+      "loss": 0.5778,
+      "step": 6147
+    },
+    {
+      "epoch": 7.86944,
+      "grad_norm": 0.7739498019218445,
+      "learning_rate": 3.771908763505402e-05,
+      "loss": 0.6069,
+      "step": 6148
+    },
+    {
+      "epoch": 7.87072,
+      "grad_norm": 0.79982590675354,
+      "learning_rate": 3.7717086834733893e-05,
+      "loss": 0.6673,
+      "step": 6149
+    },
+    {
+      "epoch": 7.872,
+      "grad_norm": 0.7504255175590515,
+      "learning_rate": 3.7715086034413765e-05,
+      "loss": 0.5946,
+      "step": 6150
+    },
+    {
+      "epoch": 7.87328,
+      "grad_norm": 0.7930676341056824,
+      "learning_rate": 3.771308523409364e-05,
+      "loss": 0.6112,
+      "step": 6151
+    },
+    {
+      "epoch": 7.87456,
+      "grad_norm": 0.7735934853553772,
+      "learning_rate": 3.7711084433773516e-05,
+      "loss": 0.5914,
+      "step": 6152
+    },
+    {
+      "epoch": 7.87584,
+      "grad_norm": 0.8146105408668518,
+      "learning_rate": 3.770908363345339e-05,
+      "loss": 0.6164,
+      "step": 6153
+    },
+    {
+      "epoch": 7.87712,
+      "grad_norm": 0.7409805655479431,
+      "learning_rate": 3.770708283313325e-05,
+      "loss": 0.4994,
+      "step": 6154
+    },
+    {
+      "epoch": 7.8784,
+      "grad_norm": 0.8767794370651245,
+      "learning_rate": 3.7705082032813125e-05,
+      "loss": 0.6723,
+      "step": 6155
+    },
+    {
+      "epoch": 7.8796800000000005,
+      "grad_norm": 0.7842835187911987,
+      "learning_rate": 3.7703081232492996e-05,
+      "loss": 0.5496,
+      "step": 6156
+    },
+    {
+      "epoch": 7.88096,
+      "grad_norm": 0.8183816075325012,
+      "learning_rate": 3.770108043217287e-05,
+      "loss": 0.5986,
+      "step": 6157
+    },
+    {
+      "epoch": 7.88224,
+      "grad_norm": 0.7943567037582397,
+      "learning_rate": 3.769907963185274e-05,
+      "loss": 0.6196,
+      "step": 6158
+    },
+    {
+      "epoch": 7.88352,
+      "grad_norm": 0.7995247840881348,
+      "learning_rate": 3.769707883153262e-05,
+      "loss": 0.6269,
+      "step": 6159
+    },
+    {
+      "epoch": 7.8848,
+      "grad_norm": 0.798160970211029,
+      "learning_rate": 3.769507803121249e-05,
+      "loss": 0.5612,
+      "step": 6160
+    },
+    {
+      "epoch": 7.88608,
+      "grad_norm": 0.7991229295730591,
+      "learning_rate": 3.769307723089236e-05,
+      "loss": 0.5785,
+      "step": 6161
+    },
+    {
+      "epoch": 7.88736,
+      "grad_norm": 0.7615501880645752,
+      "learning_rate": 3.769107643057223e-05,
+      "loss": 0.5869,
+      "step": 6162
+    },
+    {
+      "epoch": 7.88864,
+      "grad_norm": 0.7592645287513733,
+      "learning_rate": 3.76890756302521e-05,
+      "loss": 0.5602,
+      "step": 6163
+    },
+    {
+      "epoch": 7.88992,
+      "grad_norm": 0.8296216726303101,
+      "learning_rate": 3.768707482993197e-05,
+      "loss": 0.6587,
+      "step": 6164
+    },
+    {
+      "epoch": 7.8911999999999995,
+      "grad_norm": 0.7715473175048828,
+      "learning_rate": 3.768507402961184e-05,
+      "loss": 0.6282,
+      "step": 6165
+    },
+    {
+      "epoch": 7.89248,
+      "grad_norm": 0.786847710609436,
+      "learning_rate": 3.768307322929172e-05,
+      "loss": 0.6401,
+      "step": 6166
+    },
+    {
+      "epoch": 7.89376,
+      "grad_norm": 0.7830649018287659,
+      "learning_rate": 3.7681072428971594e-05,
+      "loss": 0.6306,
+      "step": 6167
+    },
+    {
+      "epoch": 7.89504,
+      "grad_norm": 0.8743398189544678,
+      "learning_rate": 3.7679071628651466e-05,
+      "loss": 0.6588,
+      "step": 6168
+    },
+    {
+      "epoch": 7.89632,
+      "grad_norm": 0.8453099131584167,
+      "learning_rate": 3.767707082833134e-05,
+      "loss": 0.6599,
+      "step": 6169
+    },
+    {
+      "epoch": 7.8976,
+      "grad_norm": 0.8115950226783752,
+      "learning_rate": 3.76750700280112e-05,
+      "loss": 0.6126,
+      "step": 6170
+    },
+    {
+      "epoch": 7.89888,
+      "grad_norm": 0.8021664619445801,
+      "learning_rate": 3.7673069227691074e-05,
+      "loss": 0.5736,
+      "step": 6171
+    },
+    {
+      "epoch": 7.90016,
+      "grad_norm": 0.7566850781440735,
+      "learning_rate": 3.7671068427370946e-05,
+      "loss": 0.5289,
+      "step": 6172
+    },
+    {
+      "epoch": 7.90144,
+      "grad_norm": 0.8024584054946899,
+      "learning_rate": 3.7669067627050825e-05,
+      "loss": 0.6521,
+      "step": 6173
+    },
+    {
+      "epoch": 7.90272,
+      "grad_norm": 0.7810555100440979,
+      "learning_rate": 3.76670668267307e-05,
+      "loss": 0.6171,
+      "step": 6174
+    },
+    {
+      "epoch": 7.904,
+      "grad_norm": 0.839141309261322,
+      "learning_rate": 3.766506602641057e-05,
+      "loss": 0.6384,
+      "step": 6175
+    },
+    {
+      "epoch": 7.90528,
+      "grad_norm": 0.8140023350715637,
+      "learning_rate": 3.766306522609044e-05,
+      "loss": 0.6113,
+      "step": 6176
+    },
+    {
+      "epoch": 7.90656,
+      "grad_norm": 0.7578498721122742,
+      "learning_rate": 3.766106442577031e-05,
+      "loss": 0.5981,
+      "step": 6177
+    },
+    {
+      "epoch": 7.90784,
+      "grad_norm": 0.850180447101593,
+      "learning_rate": 3.765906362545018e-05,
+      "loss": 0.6304,
+      "step": 6178
+    },
+    {
+      "epoch": 7.90912,
+      "grad_norm": 0.7836924195289612,
+      "learning_rate": 3.765706282513005e-05,
+      "loss": 0.6016,
+      "step": 6179
+    },
+    {
+      "epoch": 7.9104,
+      "grad_norm": 0.7858523726463318,
+      "learning_rate": 3.765506202480993e-05,
+      "loss": 0.5774,
+      "step": 6180
+    },
+    {
+      "epoch": 7.9116800000000005,
+      "grad_norm": 0.8169428706169128,
+      "learning_rate": 3.76530612244898e-05,
+      "loss": 0.6141,
+      "step": 6181
+    },
+    {
+      "epoch": 7.91296,
+      "grad_norm": 0.779033362865448,
+      "learning_rate": 3.765106042416967e-05,
+      "loss": 0.5847,
+      "step": 6182
+    },
+    {
+      "epoch": 7.91424,
+      "grad_norm": 0.8012449741363525,
+      "learning_rate": 3.764905962384954e-05,
+      "loss": 0.6393,
+      "step": 6183
+    },
+    {
+      "epoch": 7.91552,
+      "grad_norm": 0.7956286668777466,
+      "learning_rate": 3.7647058823529415e-05,
+      "loss": 0.5944,
+      "step": 6184
+    },
+    {
+      "epoch": 7.9168,
+      "grad_norm": 0.79158616065979,
+      "learning_rate": 3.764505802320929e-05,
+      "loss": 0.5891,
+      "step": 6185
+    },
+    {
+      "epoch": 7.91808,
+      "grad_norm": 0.7679545879364014,
+      "learning_rate": 3.764305722288915e-05,
+      "loss": 0.5964,
+      "step": 6186
+    },
+    {
+      "epoch": 7.91936,
+      "grad_norm": 0.7634429931640625,
+      "learning_rate": 3.764105642256903e-05,
+      "loss": 0.6025,
+      "step": 6187
+    },
+    {
+      "epoch": 7.92064,
+      "grad_norm": 0.840979278087616,
+      "learning_rate": 3.76390556222489e-05,
+      "loss": 0.6005,
+      "step": 6188
+    },
+    {
+      "epoch": 7.92192,
+      "grad_norm": 0.8021059632301331,
+      "learning_rate": 3.7637054821928775e-05,
+      "loss": 0.6507,
+      "step": 6189
+    },
+    {
+      "epoch": 7.9232,
+      "grad_norm": 0.7973845601081848,
+      "learning_rate": 3.7635054021608646e-05,
+      "loss": 0.6245,
+      "step": 6190
+    },
+    {
+      "epoch": 7.92448,
+      "grad_norm": 0.8019742369651794,
+      "learning_rate": 3.763305322128852e-05,
+      "loss": 0.6052,
+      "step": 6191
+    },
+    {
+      "epoch": 7.92576,
+      "grad_norm": 0.8176249265670776,
+      "learning_rate": 3.763105242096839e-05,
+      "loss": 0.6816,
+      "step": 6192
+    },
+    {
+      "epoch": 7.92704,
+      "grad_norm": 0.799094557762146,
+      "learning_rate": 3.762905162064826e-05,
+      "loss": 0.5972,
+      "step": 6193
+    },
+    {
+      "epoch": 7.92832,
+      "grad_norm": 0.7592592835426331,
+      "learning_rate": 3.7627050820328134e-05,
+      "loss": 0.5614,
+      "step": 6194
+    },
+    {
+      "epoch": 7.9296,
+      "grad_norm": 0.7643021941184998,
+      "learning_rate": 3.7625050020008006e-05,
+      "loss": 0.618,
+      "step": 6195
+    },
+    {
+      "epoch": 7.93088,
+      "grad_norm": 0.8433973789215088,
+      "learning_rate": 3.762304921968788e-05,
+      "loss": 0.6484,
+      "step": 6196
+    },
+    {
+      "epoch": 7.93216,
+      "grad_norm": 0.8006365299224854,
+      "learning_rate": 3.762104841936775e-05,
+      "loss": 0.5503,
+      "step": 6197
+    },
+    {
+      "epoch": 7.93344,
+      "grad_norm": 0.8090243935585022,
+      "learning_rate": 3.761904761904762e-05,
+      "loss": 0.6048,
+      "step": 6198
+    },
+    {
+      "epoch": 7.93472,
+      "grad_norm": 0.8174999356269836,
+      "learning_rate": 3.761704681872749e-05,
+      "loss": 0.6092,
+      "step": 6199
+    },
+    {
+      "epoch": 7.936,
+      "grad_norm": 0.7761614322662354,
+      "learning_rate": 3.7615046018407365e-05,
+      "loss": 0.5421,
+      "step": 6200
+    },
+    {
+      "epoch": 7.93728,
+      "grad_norm": 0.7439830303192139,
+      "learning_rate": 3.761304521808724e-05,
+      "loss": 0.6109,
+      "step": 6201
+    },
+    {
+      "epoch": 7.93856,
+      "grad_norm": 0.8538080453872681,
+      "learning_rate": 3.761104441776711e-05,
+      "loss": 0.6092,
+      "step": 6202
+    },
+    {
+      "epoch": 7.93984,
+      "grad_norm": 0.7983587980270386,
+      "learning_rate": 3.760904361744698e-05,
+      "loss": 0.6157,
+      "step": 6203
+    },
+    {
+      "epoch": 7.94112,
+      "grad_norm": 0.8033997416496277,
+      "learning_rate": 3.760704281712685e-05,
+      "loss": 0.6363,
+      "step": 6204
+    },
+    {
+      "epoch": 7.9424,
+      "grad_norm": 0.8110709190368652,
+      "learning_rate": 3.7605042016806724e-05,
+      "loss": 0.5945,
+      "step": 6205
+    },
+    {
+      "epoch": 7.94368,
+      "grad_norm": 0.8630302548408508,
+      "learning_rate": 3.7603041216486596e-05,
+      "loss": 0.5869,
+      "step": 6206
+    },
+    {
+      "epoch": 7.94496,
+      "grad_norm": 0.865012526512146,
+      "learning_rate": 3.760104041616647e-05,
+      "loss": 0.6861,
+      "step": 6207
+    },
+    {
+      "epoch": 7.9462399999999995,
+      "grad_norm": 0.8047817945480347,
+      "learning_rate": 3.759903961584634e-05,
+      "loss": 0.6484,
+      "step": 6208
+    },
+    {
+      "epoch": 7.94752,
+      "grad_norm": 0.7712967395782471,
+      "learning_rate": 3.759703881552621e-05,
+      "loss": 0.6387,
+      "step": 6209
+    },
+    {
+      "epoch": 7.9488,
+      "grad_norm": 0.7860550284385681,
+      "learning_rate": 3.7595038015206084e-05,
+      "loss": 0.6384,
+      "step": 6210
+    },
+    {
+      "epoch": 7.95008,
+      "grad_norm": 0.7760273218154907,
+      "learning_rate": 3.7593037214885955e-05,
+      "loss": 0.602,
+      "step": 6211
+    },
+    {
+      "epoch": 7.95136,
+      "grad_norm": 0.7923306822776794,
+      "learning_rate": 3.759103641456583e-05,
+      "loss": 0.6,
+      "step": 6212
+    },
+    {
+      "epoch": 7.95264,
+      "grad_norm": 0.7732347846031189,
+      "learning_rate": 3.75890356142457e-05,
+      "loss": 0.5843,
+      "step": 6213
+    },
+    {
+      "epoch": 7.95392,
+      "grad_norm": 0.7786732912063599,
+      "learning_rate": 3.758703481392557e-05,
+      "loss": 0.6319,
+      "step": 6214
+    },
+    {
+      "epoch": 7.9552,
+      "grad_norm": 0.822283148765564,
+      "learning_rate": 3.758503401360544e-05,
+      "loss": 0.638,
+      "step": 6215
+    },
+    {
+      "epoch": 7.95648,
+      "grad_norm": 0.8353981375694275,
+      "learning_rate": 3.7583033213285315e-05,
+      "loss": 0.6462,
+      "step": 6216
+    },
+    {
+      "epoch": 7.95776,
+      "grad_norm": 0.7998156547546387,
+      "learning_rate": 3.7581032412965187e-05,
+      "loss": 0.6271,
+      "step": 6217
+    },
+    {
+      "epoch": 7.95904,
+      "grad_norm": 0.8036489486694336,
+      "learning_rate": 3.757903161264506e-05,
+      "loss": 0.6486,
+      "step": 6218
+    },
+    {
+      "epoch": 7.96032,
+      "grad_norm": 0.8148389458656311,
+      "learning_rate": 3.757703081232493e-05,
+      "loss": 0.5962,
+      "step": 6219
+    },
+    {
+      "epoch": 7.9616,
+      "grad_norm": 0.7913573980331421,
+      "learning_rate": 3.75750300120048e-05,
+      "loss": 0.5983,
+      "step": 6220
+    },
+    {
+      "epoch": 7.96288,
+      "grad_norm": 0.7523571848869324,
+      "learning_rate": 3.7573029211684674e-05,
+      "loss": 0.6503,
+      "step": 6221
+    },
+    {
+      "epoch": 7.96416,
+      "grad_norm": 0.7958295941352844,
+      "learning_rate": 3.7571028411364546e-05,
+      "loss": 0.5914,
+      "step": 6222
+    },
+    {
+      "epoch": 7.96544,
+      "grad_norm": 0.7768007516860962,
+      "learning_rate": 3.7569027611044424e-05,
+      "loss": 0.617,
+      "step": 6223
+    },
+    {
+      "epoch": 7.9667200000000005,
+      "grad_norm": 0.7701081037521362,
+      "learning_rate": 3.756702681072429e-05,
+      "loss": 0.5931,
+      "step": 6224
+    },
+    {
+      "epoch": 7.968,
+      "grad_norm": 0.7834358811378479,
+      "learning_rate": 3.756502601040416e-05,
+      "loss": 0.5555,
+      "step": 6225
+    },
+    {
+      "epoch": 7.96928,
+      "grad_norm": 0.7959411144256592,
+      "learning_rate": 3.756302521008403e-05,
+      "loss": 0.63,
+      "step": 6226
+    },
+    {
+      "epoch": 7.97056,
+      "grad_norm": 0.7836350202560425,
+      "learning_rate": 3.7561024409763905e-05,
+      "loss": 0.5953,
+      "step": 6227
+    },
+    {
+      "epoch": 7.97184,
+      "grad_norm": 0.7769935131072998,
+      "learning_rate": 3.755902360944378e-05,
+      "loss": 0.6272,
+      "step": 6228
+    },
+    {
+      "epoch": 7.97312,
+      "grad_norm": 0.7422361373901367,
+      "learning_rate": 3.755702280912365e-05,
+      "loss": 0.5554,
+      "step": 6229
+    },
+    {
+      "epoch": 7.9744,
+      "grad_norm": 0.8061965703964233,
+      "learning_rate": 3.755502200880353e-05,
+      "loss": 0.5511,
+      "step": 6230
+    },
+    {
+      "epoch": 7.97568,
+      "grad_norm": 0.7815845608711243,
+      "learning_rate": 3.75530212084834e-05,
+      "loss": 0.5608,
+      "step": 6231
+    },
+    {
+      "epoch": 7.97696,
+      "grad_norm": 0.7946850657463074,
+      "learning_rate": 3.7551020408163264e-05,
+      "loss": 0.5856,
+      "step": 6232
+    },
+    {
+      "epoch": 7.9782399999999996,
+      "grad_norm": 0.738746702671051,
+      "learning_rate": 3.7549019607843136e-05,
+      "loss": 0.5789,
+      "step": 6233
+    },
+    {
+      "epoch": 7.97952,
+      "grad_norm": 0.7725175619125366,
+      "learning_rate": 3.754701880752301e-05,
+      "loss": 0.6236,
+      "step": 6234
+    },
+    {
+      "epoch": 7.9808,
+      "grad_norm": 0.8103924989700317,
+      "learning_rate": 3.754501800720288e-05,
+      "loss": 0.6221,
+      "step": 6235
+    },
+    {
+      "epoch": 7.98208,
+      "grad_norm": 0.8287813067436218,
+      "learning_rate": 3.754301720688275e-05,
+      "loss": 0.6418,
+      "step": 6236
+    },
+    {
+      "epoch": 7.98336,
+      "grad_norm": 0.8044477701187134,
+      "learning_rate": 3.754101640656263e-05,
+      "loss": 0.5968,
+      "step": 6237
+    },
+    {
+      "epoch": 7.98464,
+      "grad_norm": 0.7894521951675415,
+      "learning_rate": 3.75390156062425e-05,
+      "loss": 0.5945,
+      "step": 6238
+    },
+    {
+      "epoch": 7.98592,
+      "grad_norm": 0.8011317253112793,
+      "learning_rate": 3.7537014805922374e-05,
+      "loss": 0.6376,
+      "step": 6239
+    },
+    {
+      "epoch": 7.9872,
+      "grad_norm": 0.7666401863098145,
+      "learning_rate": 3.753501400560224e-05,
+      "loss": 0.5611,
+      "step": 6240
+    },
+    {
+      "epoch": 7.98848,
+      "grad_norm": 0.8295350670814514,
+      "learning_rate": 3.753301320528211e-05,
+      "loss": 0.6115,
+      "step": 6241
+    },
+    {
+      "epoch": 7.98976,
+      "grad_norm": 0.7832409739494324,
+      "learning_rate": 3.753101240496198e-05,
+      "loss": 0.599,
+      "step": 6242
+    },
+    {
+      "epoch": 7.99104,
+      "grad_norm": 0.8104324340820312,
+      "learning_rate": 3.7529011604641855e-05,
+      "loss": 0.6137,
+      "step": 6243
+    },
+    {
+      "epoch": 7.99232,
+      "grad_norm": 0.7831165790557861,
+      "learning_rate": 3.7527010804321733e-05,
+      "loss": 0.5828,
+      "step": 6244
+    },
+    {
+      "epoch": 7.9936,
+      "grad_norm": 0.8100852370262146,
+      "learning_rate": 3.7525010004001605e-05,
+      "loss": 0.5893,
+      "step": 6245
+    },
+    {
+      "epoch": 7.99488,
+      "grad_norm": 0.8356633186340332,
+      "learning_rate": 3.752300920368148e-05,
+      "loss": 0.6308,
+      "step": 6246
+    },
+    {
+      "epoch": 7.99616,
+      "grad_norm": 0.85598224401474,
+      "learning_rate": 3.752100840336135e-05,
+      "loss": 0.6509,
+      "step": 6247
+    },
+    {
+      "epoch": 7.99744,
+      "grad_norm": 0.8315650224685669,
+      "learning_rate": 3.7519007603041214e-05,
+      "loss": 0.6497,
+      "step": 6248
+    },
+    {
+      "epoch": 7.9987200000000005,
+      "grad_norm": 0.7793347835540771,
+      "learning_rate": 3.7517006802721086e-05,
+      "loss": 0.5806,
+      "step": 6249
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.704264521598816,
+      "learning_rate": 3.751500600240096e-05,
+      "loss": 1.0745,
+      "step": 6250
+    },
+    {
+      "epoch": 8.00128,
+      "grad_norm": 0.7346014976501465,
+      "learning_rate": 3.7513005202080836e-05,
+      "loss": 0.5634,
+      "step": 6251
+    },
+    {
+      "epoch": 8.00256,
+      "grad_norm": 0.7524464130401611,
+      "learning_rate": 3.751100440176071e-05,
+      "loss": 0.5499,
+      "step": 6252
+    },
+    {
+      "epoch": 8.00384,
+      "grad_norm": 0.8172950148582458,
+      "learning_rate": 3.750900360144058e-05,
+      "loss": 0.5984,
+      "step": 6253
+    },
+    {
+      "epoch": 8.00512,
+      "grad_norm": 0.8188893795013428,
+      "learning_rate": 3.750700280112045e-05,
+      "loss": 0.6099,
+      "step": 6254
+    },
+    {
+      "epoch": 8.0064,
+      "grad_norm": 0.825077474117279,
+      "learning_rate": 3.7505002000800324e-05,
+      "loss": 0.6207,
+      "step": 6255
+    },
+    {
+      "epoch": 8.00768,
+      "grad_norm": 0.7807846069335938,
+      "learning_rate": 3.750300120048019e-05,
+      "loss": 0.5735,
+      "step": 6256
+    },
+    {
+      "epoch": 8.00896,
+      "grad_norm": 0.7946782112121582,
+      "learning_rate": 3.750100040016006e-05,
+      "loss": 0.6036,
+      "step": 6257
+    },
+    {
+      "epoch": 8.01024,
+      "grad_norm": 0.8482866883277893,
+      "learning_rate": 3.749899959983994e-05,
+      "loss": 0.5998,
+      "step": 6258
+    },
+    {
+      "epoch": 8.01152,
+      "grad_norm": 0.7960243821144104,
+      "learning_rate": 3.749699879951981e-05,
+      "loss": 0.6042,
+      "step": 6259
+    },
+    {
+      "epoch": 8.0128,
+      "grad_norm": 0.830877959728241,
+      "learning_rate": 3.749499799919968e-05,
+      "loss": 0.586,
+      "step": 6260
+    },
+    {
+      "epoch": 8.01408,
+      "grad_norm": 0.782637357711792,
+      "learning_rate": 3.7492997198879555e-05,
+      "loss": 0.5895,
+      "step": 6261
+    },
+    {
+      "epoch": 8.01536,
+      "grad_norm": 0.8307430148124695,
+      "learning_rate": 3.749099639855943e-05,
+      "loss": 0.5734,
+      "step": 6262
+    },
+    {
+      "epoch": 8.01664,
+      "grad_norm": 0.832324206829071,
+      "learning_rate": 3.74889955982393e-05,
+      "loss": 0.5546,
+      "step": 6263
+    },
+    {
+      "epoch": 8.01792,
+      "grad_norm": 0.8269922733306885,
+      "learning_rate": 3.7486994797919164e-05,
+      "loss": 0.5858,
+      "step": 6264
+    },
+    {
+      "epoch": 8.0192,
+      "grad_norm": 0.8416856527328491,
+      "learning_rate": 3.748499399759904e-05,
+      "loss": 0.6297,
+      "step": 6265
+    },
+    {
+      "epoch": 8.02048,
+      "grad_norm": 0.806425929069519,
+      "learning_rate": 3.7482993197278914e-05,
+      "loss": 0.5737,
+      "step": 6266
+    },
+    {
+      "epoch": 8.02176,
+      "grad_norm": 0.8288019895553589,
+      "learning_rate": 3.7480992396958786e-05,
+      "loss": 0.5856,
+      "step": 6267
+    },
+    {
+      "epoch": 8.02304,
+      "grad_norm": 0.8416863679885864,
+      "learning_rate": 3.747899159663866e-05,
+      "loss": 0.589,
+      "step": 6268
+    },
+    {
+      "epoch": 8.02432,
+      "grad_norm": 0.8179052472114563,
+      "learning_rate": 3.747699079631853e-05,
+      "loss": 0.6037,
+      "step": 6269
+    },
+    {
+      "epoch": 8.0256,
+      "grad_norm": 0.7979174256324768,
+      "learning_rate": 3.74749899959984e-05,
+      "loss": 0.5805,
+      "step": 6270
+    },
+    {
+      "epoch": 8.02688,
+      "grad_norm": 0.7354505658149719,
+      "learning_rate": 3.7472989195678274e-05,
+      "loss": 0.5747,
+      "step": 6271
+    },
+    {
+      "epoch": 8.02816,
+      "grad_norm": 0.8041462898254395,
+      "learning_rate": 3.7470988395358145e-05,
+      "loss": 0.5907,
+      "step": 6272
+    },
+    {
+      "epoch": 8.02944,
+      "grad_norm": 0.7784780263900757,
+      "learning_rate": 3.746898759503802e-05,
+      "loss": 0.5735,
+      "step": 6273
+    },
+    {
+      "epoch": 8.03072,
+      "grad_norm": 0.8290993571281433,
+      "learning_rate": 3.746698679471789e-05,
+      "loss": 0.5509,
+      "step": 6274
+    },
+    {
+      "epoch": 8.032,
+      "grad_norm": 0.7944415807723999,
+      "learning_rate": 3.746498599439776e-05,
+      "loss": 0.5914,
+      "step": 6275
+    },
+    {
+      "epoch": 8.03328,
+      "grad_norm": 0.7763034701347351,
+      "learning_rate": 3.746298519407763e-05,
+      "loss": 0.5719,
+      "step": 6276
+    },
+    {
+      "epoch": 8.03456,
+      "grad_norm": 0.8510763645172119,
+      "learning_rate": 3.7460984393757505e-05,
+      "loss": 0.6297,
+      "step": 6277
+    },
+    {
+      "epoch": 8.03584,
+      "grad_norm": 0.7973195314407349,
+      "learning_rate": 3.7458983593437377e-05,
+      "loss": 0.5668,
+      "step": 6278
+    },
+    {
+      "epoch": 8.03712,
+      "grad_norm": 0.7961564064025879,
+      "learning_rate": 3.745698279311725e-05,
+      "loss": 0.5721,
+      "step": 6279
+    },
+    {
+      "epoch": 8.0384,
+      "grad_norm": 0.7879326343536377,
+      "learning_rate": 3.745498199279712e-05,
+      "loss": 0.5706,
+      "step": 6280
+    },
+    {
+      "epoch": 8.03968,
+      "grad_norm": 0.8085858225822449,
+      "learning_rate": 3.745298119247699e-05,
+      "loss": 0.5476,
+      "step": 6281
+    },
+    {
+      "epoch": 8.04096,
+      "grad_norm": 0.8712663054466248,
+      "learning_rate": 3.7450980392156864e-05,
+      "loss": 0.6085,
+      "step": 6282
+    },
+    {
+      "epoch": 8.04224,
+      "grad_norm": 0.7817235589027405,
+      "learning_rate": 3.7448979591836736e-05,
+      "loss": 0.5623,
+      "step": 6283
+    },
+    {
+      "epoch": 8.043520000000001,
+      "grad_norm": 0.8824650049209595,
+      "learning_rate": 3.744697879151661e-05,
+      "loss": 0.612,
+      "step": 6284
+    },
+    {
+      "epoch": 8.0448,
+      "grad_norm": 0.8118883371353149,
+      "learning_rate": 3.744497799119648e-05,
+      "loss": 0.5937,
+      "step": 6285
+    },
+    {
+      "epoch": 8.04608,
+      "grad_norm": 0.8654536604881287,
+      "learning_rate": 3.744297719087636e-05,
+      "loss": 0.6412,
+      "step": 6286
+    },
+    {
+      "epoch": 8.04736,
+      "grad_norm": 0.8161172866821289,
+      "learning_rate": 3.744097639055622e-05,
+      "loss": 0.6114,
+      "step": 6287
+    },
+    {
+      "epoch": 8.04864,
+      "grad_norm": 0.8604544997215271,
+      "learning_rate": 3.7438975590236095e-05,
+      "loss": 0.6085,
+      "step": 6288
+    },
+    {
+      "epoch": 8.04992,
+      "grad_norm": 0.816491425037384,
+      "learning_rate": 3.743697478991597e-05,
+      "loss": 0.6371,
+      "step": 6289
+    },
+    {
+      "epoch": 8.0512,
+      "grad_norm": 0.8112639784812927,
+      "learning_rate": 3.743497398959584e-05,
+      "loss": 0.5633,
+      "step": 6290
+    },
+    {
+      "epoch": 8.05248,
+      "grad_norm": 0.8095521330833435,
+      "learning_rate": 3.743297318927571e-05,
+      "loss": 0.6115,
+      "step": 6291
+    },
+    {
+      "epoch": 8.05376,
+      "grad_norm": 0.8249873518943787,
+      "learning_rate": 3.743097238895558e-05,
+      "loss": 0.5975,
+      "step": 6292
+    },
+    {
+      "epoch": 8.05504,
+      "grad_norm": 0.7858548164367676,
+      "learning_rate": 3.742897158863546e-05,
+      "loss": 0.5539,
+      "step": 6293
+    },
+    {
+      "epoch": 8.05632,
+      "grad_norm": 0.8200925588607788,
+      "learning_rate": 3.742697078831533e-05,
+      "loss": 0.5892,
+      "step": 6294
+    },
+    {
+      "epoch": 8.0576,
+      "grad_norm": 0.7779000997543335,
+      "learning_rate": 3.74249699879952e-05,
+      "loss": 0.5372,
+      "step": 6295
+    },
+    {
+      "epoch": 8.05888,
+      "grad_norm": 0.7932630181312561,
+      "learning_rate": 3.742296918767507e-05,
+      "loss": 0.5852,
+      "step": 6296
+    },
+    {
+      "epoch": 8.06016,
+      "grad_norm": 0.8097670078277588,
+      "learning_rate": 3.742096838735494e-05,
+      "loss": 0.6018,
+      "step": 6297
+    },
+    {
+      "epoch": 8.06144,
+      "grad_norm": 0.7957705855369568,
+      "learning_rate": 3.7418967587034814e-05,
+      "loss": 0.5866,
+      "step": 6298
+    },
+    {
+      "epoch": 8.06272,
+      "grad_norm": 0.7967826128005981,
+      "learning_rate": 3.7416966786714686e-05,
+      "loss": 0.5785,
+      "step": 6299
+    },
+    {
+      "epoch": 8.064,
+      "grad_norm": 0.7749722003936768,
+      "learning_rate": 3.7414965986394564e-05,
+      "loss": 0.5758,
+      "step": 6300
+    },
+    {
+      "epoch": 8.06528,
+      "grad_norm": 0.8013430237770081,
+      "learning_rate": 3.7412965186074436e-05,
+      "loss": 0.5362,
+      "step": 6301
+    },
+    {
+      "epoch": 8.06656,
+      "grad_norm": 0.8903826475143433,
+      "learning_rate": 3.741096438575431e-05,
+      "loss": 0.6127,
+      "step": 6302
+    },
+    {
+      "epoch": 8.06784,
+      "grad_norm": 0.7794708013534546,
+      "learning_rate": 3.740896358543417e-05,
+      "loss": 0.5785,
+      "step": 6303
+    },
+    {
+      "epoch": 8.06912,
+      "grad_norm": 0.8395883440971375,
+      "learning_rate": 3.7406962785114045e-05,
+      "loss": 0.612,
+      "step": 6304
+    },
+    {
+      "epoch": 8.0704,
+      "grad_norm": 0.7724868655204773,
+      "learning_rate": 3.740496198479392e-05,
+      "loss": 0.5437,
+      "step": 6305
+    },
+    {
+      "epoch": 8.07168,
+      "grad_norm": 0.8095210790634155,
+      "learning_rate": 3.740296118447379e-05,
+      "loss": 0.6093,
+      "step": 6306
+    },
+    {
+      "epoch": 8.07296,
+      "grad_norm": 0.7821049690246582,
+      "learning_rate": 3.740096038415367e-05,
+      "loss": 0.5682,
+      "step": 6307
+    },
+    {
+      "epoch": 8.07424,
+      "grad_norm": 0.8228663802146912,
+      "learning_rate": 3.739895958383354e-05,
+      "loss": 0.5655,
+      "step": 6308
+    },
+    {
+      "epoch": 8.07552,
+      "grad_norm": 0.7802860140800476,
+      "learning_rate": 3.739695878351341e-05,
+      "loss": 0.5646,
+      "step": 6309
+    },
+    {
+      "epoch": 8.0768,
+      "grad_norm": 0.8123772740364075,
+      "learning_rate": 3.739495798319328e-05,
+      "loss": 0.5666,
+      "step": 6310
+    },
+    {
+      "epoch": 8.07808,
+      "grad_norm": 0.8669785857200623,
+      "learning_rate": 3.739295718287315e-05,
+      "loss": 0.5712,
+      "step": 6311
+    },
+    {
+      "epoch": 8.07936,
+      "grad_norm": 0.798360288143158,
+      "learning_rate": 3.739095638255302e-05,
+      "loss": 0.5568,
+      "step": 6312
+    },
+    {
+      "epoch": 8.08064,
+      "grad_norm": 0.8681429624557495,
+      "learning_rate": 3.738895558223289e-05,
+      "loss": 0.6515,
+      "step": 6313
+    },
+    {
+      "epoch": 8.08192,
+      "grad_norm": 0.8483299612998962,
+      "learning_rate": 3.7386954781912763e-05,
+      "loss": 0.6093,
+      "step": 6314
+    },
+    {
+      "epoch": 8.0832,
+      "grad_norm": 0.8004496693611145,
+      "learning_rate": 3.738495398159264e-05,
+      "loss": 0.5589,
+      "step": 6315
+    },
+    {
+      "epoch": 8.08448,
+      "grad_norm": 0.8437052369117737,
+      "learning_rate": 3.7382953181272514e-05,
+      "loss": 0.6215,
+      "step": 6316
+    },
+    {
+      "epoch": 8.08576,
+      "grad_norm": 0.7914602756500244,
+      "learning_rate": 3.7380952380952386e-05,
+      "loss": 0.5765,
+      "step": 6317
+    },
+    {
+      "epoch": 8.08704,
+      "grad_norm": 0.8129740357398987,
+      "learning_rate": 3.737895158063226e-05,
+      "loss": 0.5463,
+      "step": 6318
+    },
+    {
+      "epoch": 8.08832,
+      "grad_norm": 0.7785910964012146,
+      "learning_rate": 3.737695078031212e-05,
+      "loss": 0.5845,
+      "step": 6319
+    },
+    {
+      "epoch": 8.0896,
+      "grad_norm": 0.781987726688385,
+      "learning_rate": 3.7374949979991995e-05,
+      "loss": 0.5644,
+      "step": 6320
+    },
+    {
+      "epoch": 8.09088,
+      "grad_norm": 0.7904685139656067,
+      "learning_rate": 3.7372949179671866e-05,
+      "loss": 0.5591,
+      "step": 6321
+    },
+    {
+      "epoch": 8.09216,
+      "grad_norm": 0.7658714652061462,
+      "learning_rate": 3.7370948379351745e-05,
+      "loss": 0.5768,
+      "step": 6322
+    },
+    {
+      "epoch": 8.09344,
+      "grad_norm": 0.8255518674850464,
+      "learning_rate": 3.736894757903162e-05,
+      "loss": 0.6142,
+      "step": 6323
+    },
+    {
+      "epoch": 8.09472,
+      "grad_norm": 0.7947357296943665,
+      "learning_rate": 3.736694677871149e-05,
+      "loss": 0.5057,
+      "step": 6324
+    },
+    {
+      "epoch": 8.096,
+      "grad_norm": 0.8527635335922241,
+      "learning_rate": 3.736494597839136e-05,
+      "loss": 0.5636,
+      "step": 6325
+    },
+    {
+      "epoch": 8.09728,
+      "grad_norm": 0.8318933248519897,
+      "learning_rate": 3.736294517807123e-05,
+      "loss": 0.6075,
+      "step": 6326
+    },
+    {
+      "epoch": 8.09856,
+      "grad_norm": 0.8209018707275391,
+      "learning_rate": 3.73609443777511e-05,
+      "loss": 0.5532,
+      "step": 6327
+    },
+    {
+      "epoch": 8.09984,
+      "grad_norm": 0.7837770581245422,
+      "learning_rate": 3.735894357743097e-05,
+      "loss": 0.576,
+      "step": 6328
+    },
+    {
+      "epoch": 8.10112,
+      "grad_norm": 0.7998737096786499,
+      "learning_rate": 3.735694277711085e-05,
+      "loss": 0.5571,
+      "step": 6329
+    },
+    {
+      "epoch": 8.1024,
+      "grad_norm": 0.7599712014198303,
+      "learning_rate": 3.735494197679072e-05,
+      "loss": 0.5275,
+      "step": 6330
+    },
+    {
+      "epoch": 8.10368,
+      "grad_norm": 0.8159705400466919,
+      "learning_rate": 3.735294117647059e-05,
+      "loss": 0.5565,
+      "step": 6331
+    },
+    {
+      "epoch": 8.10496,
+      "grad_norm": 0.8160362243652344,
+      "learning_rate": 3.7350940376150464e-05,
+      "loss": 0.5994,
+      "step": 6332
+    },
+    {
+      "epoch": 8.10624,
+      "grad_norm": 0.7920660972595215,
+      "learning_rate": 3.7348939575830335e-05,
+      "loss": 0.6277,
+      "step": 6333
+    },
+    {
+      "epoch": 8.10752,
+      "grad_norm": 0.8129349946975708,
+      "learning_rate": 3.734693877551021e-05,
+      "loss": 0.5953,
+      "step": 6334
+    },
+    {
+      "epoch": 8.1088,
+      "grad_norm": 0.8045061826705933,
+      "learning_rate": 3.734493797519007e-05,
+      "loss": 0.5593,
+      "step": 6335
+    },
+    {
+      "epoch": 8.11008,
+      "grad_norm": 0.8104987144470215,
+      "learning_rate": 3.734293717486995e-05,
+      "loss": 0.587,
+      "step": 6336
+    },
+    {
+      "epoch": 8.11136,
+      "grad_norm": 0.8352326154708862,
+      "learning_rate": 3.734093637454982e-05,
+      "loss": 0.6288,
+      "step": 6337
+    },
+    {
+      "epoch": 8.11264,
+      "grad_norm": 0.7948821187019348,
+      "learning_rate": 3.7338935574229695e-05,
+      "loss": 0.5477,
+      "step": 6338
+    },
+    {
+      "epoch": 8.11392,
+      "grad_norm": 0.8230301141738892,
+      "learning_rate": 3.7336934773909567e-05,
+      "loss": 0.5805,
+      "step": 6339
+    },
+    {
+      "epoch": 8.1152,
+      "grad_norm": 0.7984503507614136,
+      "learning_rate": 3.733493397358944e-05,
+      "loss": 0.5385,
+      "step": 6340
+    },
+    {
+      "epoch": 8.11648,
+      "grad_norm": 0.817011833190918,
+      "learning_rate": 3.733293317326931e-05,
+      "loss": 0.6042,
+      "step": 6341
+    },
+    {
+      "epoch": 8.11776,
+      "grad_norm": 0.7629779577255249,
+      "learning_rate": 3.733093237294918e-05,
+      "loss": 0.5647,
+      "step": 6342
+    },
+    {
+      "epoch": 8.11904,
+      "grad_norm": 0.7868363261222839,
+      "learning_rate": 3.7328931572629054e-05,
+      "loss": 0.551,
+      "step": 6343
+    },
+    {
+      "epoch": 8.12032,
+      "grad_norm": 0.8329203128814697,
+      "learning_rate": 3.7326930772308926e-05,
+      "loss": 0.6125,
+      "step": 6344
+    },
+    {
+      "epoch": 8.1216,
+      "grad_norm": 0.835198700428009,
+      "learning_rate": 3.73249299719888e-05,
+      "loss": 0.6188,
+      "step": 6345
+    },
+    {
+      "epoch": 8.12288,
+      "grad_norm": 0.8320775032043457,
+      "learning_rate": 3.732292917166867e-05,
+      "loss": 0.5952,
+      "step": 6346
+    },
+    {
+      "epoch": 8.12416,
+      "grad_norm": 0.8890385627746582,
+      "learning_rate": 3.732092837134854e-05,
+      "loss": 0.5802,
+      "step": 6347
+    },
+    {
+      "epoch": 8.12544,
+      "grad_norm": 0.9071861505508423,
+      "learning_rate": 3.731892757102841e-05,
+      "loss": 0.6534,
+      "step": 6348
+    },
+    {
+      "epoch": 8.12672,
+      "grad_norm": 0.8464532494544983,
+      "learning_rate": 3.7316926770708285e-05,
+      "loss": 0.5867,
+      "step": 6349
+    },
+    {
+      "epoch": 8.128,
+      "grad_norm": 0.8280791640281677,
+      "learning_rate": 3.731492597038816e-05,
+      "loss": 0.5995,
+      "step": 6350
+    },
+    {
+      "epoch": 8.12928,
+      "grad_norm": 0.7893093228340149,
+      "learning_rate": 3.731292517006803e-05,
+      "loss": 0.5685,
+      "step": 6351
+    },
+    {
+      "epoch": 8.13056,
+      "grad_norm": 0.8106672167778015,
+      "learning_rate": 3.73109243697479e-05,
+      "loss": 0.5615,
+      "step": 6352
+    },
+    {
+      "epoch": 8.13184,
+      "grad_norm": 0.8697341084480286,
+      "learning_rate": 3.730892356942777e-05,
+      "loss": 0.6196,
+      "step": 6353
+    },
+    {
+      "epoch": 8.13312,
+      "grad_norm": 0.837616503238678,
+      "learning_rate": 3.7306922769107644e-05,
+      "loss": 0.5308,
+      "step": 6354
+    },
+    {
+      "epoch": 8.1344,
+      "grad_norm": 0.7826257944107056,
+      "learning_rate": 3.7304921968787516e-05,
+      "loss": 0.5694,
+      "step": 6355
+    },
+    {
+      "epoch": 8.13568,
+      "grad_norm": 0.8352533578872681,
+      "learning_rate": 3.730292116846739e-05,
+      "loss": 0.6239,
+      "step": 6356
+    },
+    {
+      "epoch": 8.13696,
+      "grad_norm": 0.835431694984436,
+      "learning_rate": 3.730092036814726e-05,
+      "loss": 0.6214,
+      "step": 6357
+    },
+    {
+      "epoch": 8.13824,
+      "grad_norm": 0.8025604486465454,
+      "learning_rate": 3.729891956782713e-05,
+      "loss": 0.579,
+      "step": 6358
+    },
+    {
+      "epoch": 8.13952,
+      "grad_norm": 0.7842774987220764,
+      "learning_rate": 3.7296918767507004e-05,
+      "loss": 0.5335,
+      "step": 6359
+    },
+    {
+      "epoch": 8.1408,
+      "grad_norm": 0.8423367142677307,
+      "learning_rate": 3.7294917967186876e-05,
+      "loss": 0.6108,
+      "step": 6360
+    },
+    {
+      "epoch": 8.14208,
+      "grad_norm": 0.8245922327041626,
+      "learning_rate": 3.729291716686675e-05,
+      "loss": 0.5721,
+      "step": 6361
+    },
+    {
+      "epoch": 8.14336,
+      "grad_norm": 0.8257554173469543,
+      "learning_rate": 3.729091636654662e-05,
+      "loss": 0.5574,
+      "step": 6362
+    },
+    {
+      "epoch": 8.14464,
+      "grad_norm": 0.8311666250228882,
+      "learning_rate": 3.728891556622649e-05,
+      "loss": 0.5973,
+      "step": 6363
+    },
+    {
+      "epoch": 8.14592,
+      "grad_norm": 0.814927339553833,
+      "learning_rate": 3.728691476590637e-05,
+      "loss": 0.6139,
+      "step": 6364
+    },
+    {
+      "epoch": 8.1472,
+      "grad_norm": 0.8726287484169006,
+      "learning_rate": 3.7284913965586235e-05,
+      "loss": 0.6423,
+      "step": 6365
+    },
+    {
+      "epoch": 8.14848,
+      "grad_norm": 0.8697568774223328,
+      "learning_rate": 3.728291316526611e-05,
+      "loss": 0.556,
+      "step": 6366
+    },
+    {
+      "epoch": 8.14976,
+      "grad_norm": 0.7923082113265991,
+      "learning_rate": 3.728091236494598e-05,
+      "loss": 0.5776,
+      "step": 6367
+    },
+    {
+      "epoch": 8.15104,
+      "grad_norm": 0.8061732053756714,
+      "learning_rate": 3.727891156462585e-05,
+      "loss": 0.6374,
+      "step": 6368
+    },
+    {
+      "epoch": 8.15232,
+      "grad_norm": 0.8172518610954285,
+      "learning_rate": 3.727691076430572e-05,
+      "loss": 0.5489,
+      "step": 6369
+    },
+    {
+      "epoch": 8.1536,
+      "grad_norm": 0.7999323606491089,
+      "learning_rate": 3.7274909963985594e-05,
+      "loss": 0.556,
+      "step": 6370
+    },
+    {
+      "epoch": 8.15488,
+      "grad_norm": 0.8566176891326904,
+      "learning_rate": 3.727290916366547e-05,
+      "loss": 0.6124,
+      "step": 6371
+    },
+    {
+      "epoch": 8.15616,
+      "grad_norm": 0.8322476148605347,
+      "learning_rate": 3.7270908363345345e-05,
+      "loss": 0.5665,
+      "step": 6372
+    },
+    {
+      "epoch": 8.15744,
+      "grad_norm": 0.8234203457832336,
+      "learning_rate": 3.726890756302521e-05,
+      "loss": 0.5859,
+      "step": 6373
+    },
+    {
+      "epoch": 8.15872,
+      "grad_norm": 0.7753047347068787,
+      "learning_rate": 3.726690676270508e-05,
+      "loss": 0.5526,
+      "step": 6374
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.8442295789718628,
+      "learning_rate": 3.7264905962384953e-05,
+      "loss": 0.5812,
+      "step": 6375
+    },
+    {
+      "epoch": 8.16128,
+      "grad_norm": 0.7989637851715088,
+      "learning_rate": 3.7262905162064825e-05,
+      "loss": 0.5454,
+      "step": 6376
+    },
+    {
+      "epoch": 8.16256,
+      "grad_norm": 0.8231421113014221,
+      "learning_rate": 3.72609043617447e-05,
+      "loss": 0.5522,
+      "step": 6377
+    },
+    {
+      "epoch": 8.16384,
+      "grad_norm": 0.8021591901779175,
+      "learning_rate": 3.7258903561424576e-05,
+      "loss": 0.5667,
+      "step": 6378
+    },
+    {
+      "epoch": 8.16512,
+      "grad_norm": 0.8049442768096924,
+      "learning_rate": 3.725690276110445e-05,
+      "loss": 0.5194,
+      "step": 6379
+    },
+    {
+      "epoch": 8.1664,
+      "grad_norm": 0.8534220457077026,
+      "learning_rate": 3.725490196078432e-05,
+      "loss": 0.6475,
+      "step": 6380
+    },
+    {
+      "epoch": 8.16768,
+      "grad_norm": 0.7750808596611023,
+      "learning_rate": 3.7252901160464185e-05,
+      "loss": 0.5171,
+      "step": 6381
+    },
+    {
+      "epoch": 8.16896,
+      "grad_norm": 0.7859979867935181,
+      "learning_rate": 3.7250900360144056e-05,
+      "loss": 0.5755,
+      "step": 6382
+    },
+    {
+      "epoch": 8.17024,
+      "grad_norm": 0.9099715948104858,
+      "learning_rate": 3.724889955982393e-05,
+      "loss": 0.5664,
+      "step": 6383
+    },
+    {
+      "epoch": 8.17152,
+      "grad_norm": 0.7696498036384583,
+      "learning_rate": 3.72468987595038e-05,
+      "loss": 0.5575,
+      "step": 6384
+    },
+    {
+      "epoch": 8.1728,
+      "grad_norm": 0.8566391468048096,
+      "learning_rate": 3.724489795918368e-05,
+      "loss": 0.6486,
+      "step": 6385
+    },
+    {
+      "epoch": 8.17408,
+      "grad_norm": 0.7947507500648499,
+      "learning_rate": 3.724289715886355e-05,
+      "loss": 0.5506,
+      "step": 6386
+    },
+    {
+      "epoch": 8.17536,
+      "grad_norm": 0.8360593318939209,
+      "learning_rate": 3.724089635854342e-05,
+      "loss": 0.6098,
+      "step": 6387
+    },
+    {
+      "epoch": 8.17664,
+      "grad_norm": 0.8018707036972046,
+      "learning_rate": 3.7238895558223294e-05,
+      "loss": 0.585,
+      "step": 6388
+    },
+    {
+      "epoch": 8.17792,
+      "grad_norm": 0.8566110134124756,
+      "learning_rate": 3.723689475790316e-05,
+      "loss": 0.5698,
+      "step": 6389
+    },
+    {
+      "epoch": 8.1792,
+      "grad_norm": 0.7939445376396179,
+      "learning_rate": 3.723489395758303e-05,
+      "loss": 0.5424,
+      "step": 6390
+    },
+    {
+      "epoch": 8.18048,
+      "grad_norm": 0.816756010055542,
+      "learning_rate": 3.72328931572629e-05,
+      "loss": 0.563,
+      "step": 6391
+    },
+    {
+      "epoch": 8.18176,
+      "grad_norm": 0.8309502601623535,
+      "learning_rate": 3.723089235694278e-05,
+      "loss": 0.5777,
+      "step": 6392
+    },
+    {
+      "epoch": 8.18304,
+      "grad_norm": 0.8525409698486328,
+      "learning_rate": 3.7228891556622654e-05,
+      "loss": 0.6131,
+      "step": 6393
+    },
+    {
+      "epoch": 8.18432,
+      "grad_norm": 0.8380247354507446,
+      "learning_rate": 3.7226890756302525e-05,
+      "loss": 0.5809,
+      "step": 6394
+    },
+    {
+      "epoch": 8.1856,
+      "grad_norm": 0.7954679131507874,
+      "learning_rate": 3.72248899559824e-05,
+      "loss": 0.5846,
+      "step": 6395
+    },
+    {
+      "epoch": 8.18688,
+      "grad_norm": 0.8677021265029907,
+      "learning_rate": 3.722288915566227e-05,
+      "loss": 0.6068,
+      "step": 6396
+    },
+    {
+      "epoch": 8.18816,
+      "grad_norm": 0.8796229362487793,
+      "learning_rate": 3.7220888355342134e-05,
+      "loss": 0.637,
+      "step": 6397
+    },
+    {
+      "epoch": 8.18944,
+      "grad_norm": 0.7766791582107544,
+      "learning_rate": 3.7218887555022006e-05,
+      "loss": 0.5808,
+      "step": 6398
+    },
+    {
+      "epoch": 8.19072,
+      "grad_norm": 0.8018509149551392,
+      "learning_rate": 3.7216886754701885e-05,
+      "loss": 0.5721,
+      "step": 6399
+    },
+    {
+      "epoch": 8.192,
+      "grad_norm": 0.8333677053451538,
+      "learning_rate": 3.721488595438176e-05,
+      "loss": 0.5966,
+      "step": 6400
+    },
+    {
+      "epoch": 8.19328,
+      "grad_norm": 0.8443406224250793,
+      "learning_rate": 3.721288515406163e-05,
+      "loss": 0.5973,
+      "step": 6401
+    },
+    {
+      "epoch": 8.19456,
+      "grad_norm": 0.8139026761054993,
+      "learning_rate": 3.72108843537415e-05,
+      "loss": 0.5496,
+      "step": 6402
+    },
+    {
+      "epoch": 8.19584,
+      "grad_norm": 0.846799373626709,
+      "learning_rate": 3.720888355342137e-05,
+      "loss": 0.5896,
+      "step": 6403
+    },
+    {
+      "epoch": 8.19712,
+      "grad_norm": 0.8184740543365479,
+      "learning_rate": 3.7206882753101244e-05,
+      "loss": 0.5586,
+      "step": 6404
+    },
+    {
+      "epoch": 8.1984,
+      "grad_norm": 0.8557119965553284,
+      "learning_rate": 3.720488195278111e-05,
+      "loss": 0.5687,
+      "step": 6405
+    },
+    {
+      "epoch": 8.19968,
+      "grad_norm": 0.7910146713256836,
+      "learning_rate": 3.720288115246099e-05,
+      "loss": 0.5474,
+      "step": 6406
+    },
+    {
+      "epoch": 8.20096,
+      "grad_norm": 0.8577878475189209,
+      "learning_rate": 3.720088035214086e-05,
+      "loss": 0.6,
+      "step": 6407
+    },
+    {
+      "epoch": 8.20224,
+      "grad_norm": 0.8686122894287109,
+      "learning_rate": 3.719887955182073e-05,
+      "loss": 0.6026,
+      "step": 6408
+    },
+    {
+      "epoch": 8.20352,
+      "grad_norm": 0.7906977534294128,
+      "learning_rate": 3.71968787515006e-05,
+      "loss": 0.5427,
+      "step": 6409
+    },
+    {
+      "epoch": 8.2048,
+      "grad_norm": 0.8151067495346069,
+      "learning_rate": 3.7194877951180475e-05,
+      "loss": 0.5795,
+      "step": 6410
+    },
+    {
+      "epoch": 8.20608,
+      "grad_norm": 0.8321256041526794,
+      "learning_rate": 3.719287715086035e-05,
+      "loss": 0.6022,
+      "step": 6411
+    },
+    {
+      "epoch": 8.20736,
+      "grad_norm": 0.7891202569007874,
+      "learning_rate": 3.719087635054022e-05,
+      "loss": 0.5335,
+      "step": 6412
+    },
+    {
+      "epoch": 8.20864,
+      "grad_norm": 0.836283802986145,
+      "learning_rate": 3.718887555022009e-05,
+      "loss": 0.6082,
+      "step": 6413
+    },
+    {
+      "epoch": 8.20992,
+      "grad_norm": 0.7881807684898376,
+      "learning_rate": 3.718687474989996e-05,
+      "loss": 0.562,
+      "step": 6414
+    },
+    {
+      "epoch": 8.2112,
+      "grad_norm": 0.7665523886680603,
+      "learning_rate": 3.7184873949579834e-05,
+      "loss": 0.571,
+      "step": 6415
+    },
+    {
+      "epoch": 8.21248,
+      "grad_norm": 0.8602042198181152,
+      "learning_rate": 3.7182873149259706e-05,
+      "loss": 0.6157,
+      "step": 6416
+    },
+    {
+      "epoch": 8.21376,
+      "grad_norm": 0.8061181902885437,
+      "learning_rate": 3.718087234893958e-05,
+      "loss": 0.5852,
+      "step": 6417
+    },
+    {
+      "epoch": 8.21504,
+      "grad_norm": 0.7839178442955017,
+      "learning_rate": 3.717887154861945e-05,
+      "loss": 0.5376,
+      "step": 6418
+    },
+    {
+      "epoch": 8.21632,
+      "grad_norm": 0.8473336696624756,
+      "learning_rate": 3.717687074829932e-05,
+      "loss": 0.6365,
+      "step": 6419
+    },
+    {
+      "epoch": 8.2176,
+      "grad_norm": 0.8579205870628357,
+      "learning_rate": 3.7174869947979194e-05,
+      "loss": 0.6029,
+      "step": 6420
+    },
+    {
+      "epoch": 8.21888,
+      "grad_norm": 0.8451679348945618,
+      "learning_rate": 3.7172869147659066e-05,
+      "loss": 0.6307,
+      "step": 6421
+    },
+    {
+      "epoch": 8.22016,
+      "grad_norm": 0.7627133131027222,
+      "learning_rate": 3.717086834733894e-05,
+      "loss": 0.5673,
+      "step": 6422
+    },
+    {
+      "epoch": 8.22144,
+      "grad_norm": 0.8411123156547546,
+      "learning_rate": 3.716886754701881e-05,
+      "loss": 0.5622,
+      "step": 6423
+    },
+    {
+      "epoch": 8.22272,
+      "grad_norm": 0.8919776678085327,
+      "learning_rate": 3.716686674669868e-05,
+      "loss": 0.6082,
+      "step": 6424
+    },
+    {
+      "epoch": 8.224,
+      "grad_norm": 0.7615383267402649,
+      "learning_rate": 3.716486594637855e-05,
+      "loss": 0.5266,
+      "step": 6425
+    },
+    {
+      "epoch": 8.22528,
+      "grad_norm": 0.8244317770004272,
+      "learning_rate": 3.7162865146058425e-05,
+      "loss": 0.5582,
+      "step": 6426
+    },
+    {
+      "epoch": 8.22656,
+      "grad_norm": 0.8274043202400208,
+      "learning_rate": 3.71608643457383e-05,
+      "loss": 0.5784,
+      "step": 6427
+    },
+    {
+      "epoch": 8.22784,
+      "grad_norm": 0.8274574279785156,
+      "learning_rate": 3.715886354541817e-05,
+      "loss": 0.5993,
+      "step": 6428
+    },
+    {
+      "epoch": 8.22912,
+      "grad_norm": 0.7799541354179382,
+      "learning_rate": 3.715686274509804e-05,
+      "loss": 0.6209,
+      "step": 6429
+    },
+    {
+      "epoch": 8.2304,
+      "grad_norm": 0.8560000658035278,
+      "learning_rate": 3.715486194477791e-05,
+      "loss": 0.6404,
+      "step": 6430
+    },
+    {
+      "epoch": 8.23168,
+      "grad_norm": 0.8321433663368225,
+      "learning_rate": 3.7152861144457784e-05,
+      "loss": 0.6115,
+      "step": 6431
+    },
+    {
+      "epoch": 8.23296,
+      "grad_norm": 0.7989788055419922,
+      "learning_rate": 3.7150860344137656e-05,
+      "loss": 0.5221,
+      "step": 6432
+    },
+    {
+      "epoch": 8.23424,
+      "grad_norm": 0.8330096006393433,
+      "learning_rate": 3.714885954381753e-05,
+      "loss": 0.6045,
+      "step": 6433
+    },
+    {
+      "epoch": 8.23552,
+      "grad_norm": 0.7904120683670044,
+      "learning_rate": 3.71468587434974e-05,
+      "loss": 0.5978,
+      "step": 6434
+    },
+    {
+      "epoch": 8.2368,
+      "grad_norm": 0.8202242255210876,
+      "learning_rate": 3.714485794317727e-05,
+      "loss": 0.5845,
+      "step": 6435
+    },
+    {
+      "epoch": 8.23808,
+      "grad_norm": 0.8055812120437622,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.5653,
+      "step": 6436
+    },
+    {
+      "epoch": 8.23936,
+      "grad_norm": 0.8505423665046692,
+      "learning_rate": 3.7140856342537015e-05,
+      "loss": 0.6106,
+      "step": 6437
+    },
+    {
+      "epoch": 8.24064,
+      "grad_norm": 0.8569924235343933,
+      "learning_rate": 3.713885554221689e-05,
+      "loss": 0.5948,
+      "step": 6438
+    },
+    {
+      "epoch": 8.24192,
+      "grad_norm": 0.785232663154602,
+      "learning_rate": 3.713685474189676e-05,
+      "loss": 0.5768,
+      "step": 6439
+    },
+    {
+      "epoch": 8.2432,
+      "grad_norm": 0.8176567554473877,
+      "learning_rate": 3.713485394157663e-05,
+      "loss": 0.5744,
+      "step": 6440
+    },
+    {
+      "epoch": 8.24448,
+      "grad_norm": 0.8164046406745911,
+      "learning_rate": 3.71328531412565e-05,
+      "loss": 0.5559,
+      "step": 6441
+    },
+    {
+      "epoch": 8.24576,
+      "grad_norm": 0.7764423489570618,
+      "learning_rate": 3.713085234093638e-05,
+      "loss": 0.5796,
+      "step": 6442
+    },
+    {
+      "epoch": 8.24704,
+      "grad_norm": 0.8372679352760315,
+      "learning_rate": 3.7128851540616246e-05,
+      "loss": 0.5921,
+      "step": 6443
+    },
+    {
+      "epoch": 8.24832,
+      "grad_norm": 0.8438205122947693,
+      "learning_rate": 3.712685074029612e-05,
+      "loss": 0.6257,
+      "step": 6444
+    },
+    {
+      "epoch": 8.2496,
+      "grad_norm": 0.810208261013031,
+      "learning_rate": 3.712484993997599e-05,
+      "loss": 0.5474,
+      "step": 6445
+    },
+    {
+      "epoch": 8.25088,
+      "grad_norm": 0.866226315498352,
+      "learning_rate": 3.712284913965586e-05,
+      "loss": 0.5989,
+      "step": 6446
+    },
+    {
+      "epoch": 8.25216,
+      "grad_norm": 0.8516311645507812,
+      "learning_rate": 3.7120848339335734e-05,
+      "loss": 0.6025,
+      "step": 6447
+    },
+    {
+      "epoch": 8.25344,
+      "grad_norm": 0.8606278896331787,
+      "learning_rate": 3.7118847539015606e-05,
+      "loss": 0.5967,
+      "step": 6448
+    },
+    {
+      "epoch": 8.25472,
+      "grad_norm": 0.8150187134742737,
+      "learning_rate": 3.7116846738695484e-05,
+      "loss": 0.5981,
+      "step": 6449
+    },
+    {
+      "epoch": 8.256,
+      "grad_norm": 0.7826385498046875,
+      "learning_rate": 3.7114845938375356e-05,
+      "loss": 0.5887,
+      "step": 6450
+    },
+    {
+      "epoch": 8.25728,
+      "grad_norm": 0.8195633888244629,
+      "learning_rate": 3.711284513805522e-05,
+      "loss": 0.587,
+      "step": 6451
+    },
+    {
+      "epoch": 8.25856,
+      "grad_norm": 0.9077224135398865,
+      "learning_rate": 3.711084433773509e-05,
+      "loss": 0.6068,
+      "step": 6452
+    },
+    {
+      "epoch": 8.25984,
+      "grad_norm": 0.8052554130554199,
+      "learning_rate": 3.7108843537414965e-05,
+      "loss": 0.5859,
+      "step": 6453
+    },
+    {
+      "epoch": 8.26112,
+      "grad_norm": 0.8076302409172058,
+      "learning_rate": 3.710684273709484e-05,
+      "loss": 0.576,
+      "step": 6454
+    },
+    {
+      "epoch": 8.2624,
+      "grad_norm": 0.8576489090919495,
+      "learning_rate": 3.710484193677471e-05,
+      "loss": 0.6078,
+      "step": 6455
+    },
+    {
+      "epoch": 8.26368,
+      "grad_norm": 0.8286393880844116,
+      "learning_rate": 3.710284113645459e-05,
+      "loss": 0.6022,
+      "step": 6456
+    },
+    {
+      "epoch": 8.26496,
+      "grad_norm": 0.8481602072715759,
+      "learning_rate": 3.710084033613446e-05,
+      "loss": 0.5954,
+      "step": 6457
+    },
+    {
+      "epoch": 8.26624,
+      "grad_norm": 0.8346137404441833,
+      "learning_rate": 3.709883953581433e-05,
+      "loss": 0.5913,
+      "step": 6458
+    },
+    {
+      "epoch": 8.26752,
+      "grad_norm": 0.8080378770828247,
+      "learning_rate": 3.7096838735494196e-05,
+      "loss": 0.6009,
+      "step": 6459
+    },
+    {
+      "epoch": 8.2688,
+      "grad_norm": 0.8473795652389526,
+      "learning_rate": 3.709483793517407e-05,
+      "loss": 0.6306,
+      "step": 6460
+    },
+    {
+      "epoch": 8.27008,
+      "grad_norm": 0.8970816731452942,
+      "learning_rate": 3.709283713485394e-05,
+      "loss": 0.5936,
+      "step": 6461
+    },
+    {
+      "epoch": 8.27136,
+      "grad_norm": 0.8006930947303772,
+      "learning_rate": 3.709083633453381e-05,
+      "loss": 0.5833,
+      "step": 6462
+    },
+    {
+      "epoch": 8.272639999999999,
+      "grad_norm": 0.800212562084198,
+      "learning_rate": 3.708883553421369e-05,
+      "loss": 0.5658,
+      "step": 6463
+    },
+    {
+      "epoch": 8.27392,
+      "grad_norm": 0.8020644783973694,
+      "learning_rate": 3.708683473389356e-05,
+      "loss": 0.6084,
+      "step": 6464
+    },
+    {
+      "epoch": 8.2752,
+      "grad_norm": 0.8160932660102844,
+      "learning_rate": 3.7084833933573434e-05,
+      "loss": 0.5927,
+      "step": 6465
+    },
+    {
+      "epoch": 8.27648,
+      "grad_norm": 0.8528776168823242,
+      "learning_rate": 3.7082833133253306e-05,
+      "loss": 0.6092,
+      "step": 6466
+    },
+    {
+      "epoch": 8.27776,
+      "grad_norm": 0.8029847145080566,
+      "learning_rate": 3.708083233293317e-05,
+      "loss": 0.6283,
+      "step": 6467
+    },
+    {
+      "epoch": 8.27904,
+      "grad_norm": 0.7798078060150146,
+      "learning_rate": 3.707883153261304e-05,
+      "loss": 0.5935,
+      "step": 6468
+    },
+    {
+      "epoch": 8.28032,
+      "grad_norm": 0.7831089496612549,
+      "learning_rate": 3.7076830732292915e-05,
+      "loss": 0.5904,
+      "step": 6469
+    },
+    {
+      "epoch": 8.2816,
+      "grad_norm": 0.9106951951980591,
+      "learning_rate": 3.707482993197279e-05,
+      "loss": 0.6896,
+      "step": 6470
+    },
+    {
+      "epoch": 8.28288,
+      "grad_norm": 0.8110149502754211,
+      "learning_rate": 3.7072829131652665e-05,
+      "loss": 0.5925,
+      "step": 6471
+    },
+    {
+      "epoch": 8.28416,
+      "grad_norm": 0.7867901921272278,
+      "learning_rate": 3.707082833133254e-05,
+      "loss": 0.5726,
+      "step": 6472
+    },
+    {
+      "epoch": 8.28544,
+      "grad_norm": 0.7895092368125916,
+      "learning_rate": 3.706882753101241e-05,
+      "loss": 0.5712,
+      "step": 6473
+    },
+    {
+      "epoch": 8.28672,
+      "grad_norm": 0.8702477216720581,
+      "learning_rate": 3.706682673069228e-05,
+      "loss": 0.6259,
+      "step": 6474
+    },
+    {
+      "epoch": 8.288,
+      "grad_norm": 0.8143934011459351,
+      "learning_rate": 3.7064825930372146e-05,
+      "loss": 0.5903,
+      "step": 6475
+    },
+    {
+      "epoch": 8.28928,
+      "grad_norm": 0.8529263138771057,
+      "learning_rate": 3.706282513005202e-05,
+      "loss": 0.6155,
+      "step": 6476
+    },
+    {
+      "epoch": 8.29056,
+      "grad_norm": 0.7962032556533813,
+      "learning_rate": 3.7060824329731896e-05,
+      "loss": 0.5951,
+      "step": 6477
+    },
+    {
+      "epoch": 8.29184,
+      "grad_norm": 0.7777064442634583,
+      "learning_rate": 3.705882352941177e-05,
+      "loss": 0.5678,
+      "step": 6478
+    },
+    {
+      "epoch": 8.29312,
+      "grad_norm": 0.8295832872390747,
+      "learning_rate": 3.705682272909164e-05,
+      "loss": 0.5486,
+      "step": 6479
+    },
+    {
+      "epoch": 8.2944,
+      "grad_norm": 0.7815911769866943,
+      "learning_rate": 3.705482192877151e-05,
+      "loss": 0.5732,
+      "step": 6480
+    },
+    {
+      "epoch": 8.29568,
+      "grad_norm": 0.7912229895591736,
+      "learning_rate": 3.7052821128451384e-05,
+      "loss": 0.5611,
+      "step": 6481
+    },
+    {
+      "epoch": 8.29696,
+      "grad_norm": 0.8012306690216064,
+      "learning_rate": 3.7050820328131256e-05,
+      "loss": 0.5652,
+      "step": 6482
+    },
+    {
+      "epoch": 8.29824,
+      "grad_norm": 0.7731996178627014,
+      "learning_rate": 3.704881952781112e-05,
+      "loss": 0.511,
+      "step": 6483
+    },
+    {
+      "epoch": 8.29952,
+      "grad_norm": 0.858241081237793,
+      "learning_rate": 3.7046818727491e-05,
+      "loss": 0.5636,
+      "step": 6484
+    },
+    {
+      "epoch": 8.3008,
+      "grad_norm": 0.816175639629364,
+      "learning_rate": 3.704481792717087e-05,
+      "loss": 0.5677,
+      "step": 6485
+    },
+    {
+      "epoch": 8.30208,
+      "grad_norm": 0.7947445511817932,
+      "learning_rate": 3.704281712685074e-05,
+      "loss": 0.5523,
+      "step": 6486
+    },
+    {
+      "epoch": 8.30336,
+      "grad_norm": 0.8075773119926453,
+      "learning_rate": 3.7040816326530615e-05,
+      "loss": 0.566,
+      "step": 6487
+    },
+    {
+      "epoch": 8.30464,
+      "grad_norm": 0.8670647740364075,
+      "learning_rate": 3.703881552621049e-05,
+      "loss": 0.6355,
+      "step": 6488
+    },
+    {
+      "epoch": 8.30592,
+      "grad_norm": 0.8332636952400208,
+      "learning_rate": 3.703681472589036e-05,
+      "loss": 0.6226,
+      "step": 6489
+    },
+    {
+      "epoch": 8.3072,
+      "grad_norm": 0.8159798979759216,
+      "learning_rate": 3.703481392557023e-05,
+      "loss": 0.5558,
+      "step": 6490
+    },
+    {
+      "epoch": 8.30848,
+      "grad_norm": 0.8338848948478699,
+      "learning_rate": 3.70328131252501e-05,
+      "loss": 0.5839,
+      "step": 6491
+    },
+    {
+      "epoch": 8.30976,
+      "grad_norm": 0.8652037382125854,
+      "learning_rate": 3.7030812324929974e-05,
+      "loss": 0.6033,
+      "step": 6492
+    },
+    {
+      "epoch": 8.31104,
+      "grad_norm": 0.8078513145446777,
+      "learning_rate": 3.7028811524609846e-05,
+      "loss": 0.5618,
+      "step": 6493
+    },
+    {
+      "epoch": 8.31232,
+      "grad_norm": 0.8094210028648376,
+      "learning_rate": 3.702681072428972e-05,
+      "loss": 0.5781,
+      "step": 6494
+    },
+    {
+      "epoch": 8.3136,
+      "grad_norm": 0.7953526973724365,
+      "learning_rate": 3.702480992396959e-05,
+      "loss": 0.6069,
+      "step": 6495
+    },
+    {
+      "epoch": 8.31488,
+      "grad_norm": 0.8320873379707336,
+      "learning_rate": 3.702280912364946e-05,
+      "loss": 0.5769,
+      "step": 6496
+    },
+    {
+      "epoch": 8.31616,
+      "grad_norm": 0.8492526412010193,
+      "learning_rate": 3.7020808323329334e-05,
+      "loss": 0.6435,
+      "step": 6497
+    },
+    {
+      "epoch": 8.31744,
+      "grad_norm": 0.8726949691772461,
+      "learning_rate": 3.7018807523009205e-05,
+      "loss": 0.5645,
+      "step": 6498
+    },
+    {
+      "epoch": 8.31872,
+      "grad_norm": 0.8547096252441406,
+      "learning_rate": 3.701680672268908e-05,
+      "loss": 0.5968,
+      "step": 6499
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.8142231702804565,
+      "learning_rate": 3.701480592236895e-05,
+      "loss": 0.6064,
+      "step": 6500
+    },
+    {
+      "epoch": 8.32128,
+      "grad_norm": 0.8083314895629883,
+      "learning_rate": 3.701280512204882e-05,
+      "loss": 0.5691,
+      "step": 6501
+    },
+    {
+      "epoch": 8.32256,
+      "grad_norm": 0.7994301319122314,
+      "learning_rate": 3.701080432172869e-05,
+      "loss": 0.6027,
+      "step": 6502
+    },
+    {
+      "epoch": 8.32384,
+      "grad_norm": 0.8659411072731018,
+      "learning_rate": 3.7008803521408565e-05,
+      "loss": 0.6085,
+      "step": 6503
+    },
+    {
+      "epoch": 8.32512,
+      "grad_norm": 0.8234634399414062,
+      "learning_rate": 3.7006802721088437e-05,
+      "loss": 0.593,
+      "step": 6504
+    },
+    {
+      "epoch": 8.3264,
+      "grad_norm": 0.8921228051185608,
+      "learning_rate": 3.700480192076831e-05,
+      "loss": 0.6614,
+      "step": 6505
+    },
+    {
+      "epoch": 8.32768,
+      "grad_norm": 0.7711195945739746,
+      "learning_rate": 3.700280112044818e-05,
+      "loss": 0.5886,
+      "step": 6506
+    },
+    {
+      "epoch": 8.32896,
+      "grad_norm": 0.8125150203704834,
+      "learning_rate": 3.700080032012805e-05,
+      "loss": 0.58,
+      "step": 6507
+    },
+    {
+      "epoch": 8.33024,
+      "grad_norm": 0.8450069427490234,
+      "learning_rate": 3.6998799519807924e-05,
+      "loss": 0.6523,
+      "step": 6508
+    },
+    {
+      "epoch": 8.33152,
+      "grad_norm": 0.8026819229125977,
+      "learning_rate": 3.6996798719487796e-05,
+      "loss": 0.5938,
+      "step": 6509
+    },
+    {
+      "epoch": 8.3328,
+      "grad_norm": 0.7969563603401184,
+      "learning_rate": 3.699479791916767e-05,
+      "loss": 0.591,
+      "step": 6510
+    },
+    {
+      "epoch": 8.33408,
+      "grad_norm": 0.8721278309822083,
+      "learning_rate": 3.699279711884754e-05,
+      "loss": 0.636,
+      "step": 6511
+    },
+    {
+      "epoch": 8.33536,
+      "grad_norm": 0.81974196434021,
+      "learning_rate": 3.699079631852742e-05,
+      "loss": 0.6409,
+      "step": 6512
+    },
+    {
+      "epoch": 8.33664,
+      "grad_norm": 0.7834908366203308,
+      "learning_rate": 3.698879551820728e-05,
+      "loss": 0.5799,
+      "step": 6513
+    },
+    {
+      "epoch": 8.33792,
+      "grad_norm": 0.815528929233551,
+      "learning_rate": 3.6986794717887155e-05,
+      "loss": 0.5999,
+      "step": 6514
+    },
+    {
+      "epoch": 8.3392,
+      "grad_norm": 0.8319604992866516,
+      "learning_rate": 3.698479391756703e-05,
+      "loss": 0.6098,
+      "step": 6515
+    },
+    {
+      "epoch": 8.34048,
+      "grad_norm": 0.7863122820854187,
+      "learning_rate": 3.69827931172469e-05,
+      "loss": 0.5897,
+      "step": 6516
+    },
+    {
+      "epoch": 8.34176,
+      "grad_norm": 0.8292329907417297,
+      "learning_rate": 3.698079231692677e-05,
+      "loss": 0.5612,
+      "step": 6517
+    },
+    {
+      "epoch": 8.34304,
+      "grad_norm": 0.8444699645042419,
+      "learning_rate": 3.697879151660664e-05,
+      "loss": 0.6062,
+      "step": 6518
+    },
+    {
+      "epoch": 8.34432,
+      "grad_norm": 0.8435462117195129,
+      "learning_rate": 3.697679071628652e-05,
+      "loss": 0.5952,
+      "step": 6519
+    },
+    {
+      "epoch": 8.3456,
+      "grad_norm": 0.8049376606941223,
+      "learning_rate": 3.697478991596639e-05,
+      "loss": 0.5783,
+      "step": 6520
+    },
+    {
+      "epoch": 8.34688,
+      "grad_norm": 0.8144257068634033,
+      "learning_rate": 3.697278911564626e-05,
+      "loss": 0.6279,
+      "step": 6521
+    },
+    {
+      "epoch": 8.34816,
+      "grad_norm": 0.8180856108665466,
+      "learning_rate": 3.697078831532613e-05,
+      "loss": 0.5911,
+      "step": 6522
+    },
+    {
+      "epoch": 8.34944,
+      "grad_norm": 0.8317684531211853,
+      "learning_rate": 3.6968787515006e-05,
+      "loss": 0.6043,
+      "step": 6523
+    },
+    {
+      "epoch": 8.35072,
+      "grad_norm": 0.8124802112579346,
+      "learning_rate": 3.6966786714685874e-05,
+      "loss": 0.6082,
+      "step": 6524
+    },
+    {
+      "epoch": 8.352,
+      "grad_norm": 0.8340176939964294,
+      "learning_rate": 3.6964785914365746e-05,
+      "loss": 0.6172,
+      "step": 6525
+    },
+    {
+      "epoch": 8.35328,
+      "grad_norm": 0.8047380447387695,
+      "learning_rate": 3.6962785114045624e-05,
+      "loss": 0.5722,
+      "step": 6526
+    },
+    {
+      "epoch": 8.35456,
+      "grad_norm": 0.8233848810195923,
+      "learning_rate": 3.6960784313725496e-05,
+      "loss": 0.6046,
+      "step": 6527
+    },
+    {
+      "epoch": 8.35584,
+      "grad_norm": 0.7907252311706543,
+      "learning_rate": 3.695878351340537e-05,
+      "loss": 0.5304,
+      "step": 6528
+    },
+    {
+      "epoch": 8.35712,
+      "grad_norm": 0.8269102573394775,
+      "learning_rate": 3.695678271308523e-05,
+      "loss": 0.5938,
+      "step": 6529
+    },
+    {
+      "epoch": 8.3584,
+      "grad_norm": 0.8250780701637268,
+      "learning_rate": 3.6954781912765105e-05,
+      "loss": 0.5838,
+      "step": 6530
+    },
+    {
+      "epoch": 8.35968,
+      "grad_norm": 0.7899469137191772,
+      "learning_rate": 3.695278111244498e-05,
+      "loss": 0.5508,
+      "step": 6531
+    },
+    {
+      "epoch": 8.36096,
+      "grad_norm": 0.8494967818260193,
+      "learning_rate": 3.695078031212485e-05,
+      "loss": 0.6586,
+      "step": 6532
+    },
+    {
+      "epoch": 8.36224,
+      "grad_norm": 0.775540828704834,
+      "learning_rate": 3.694877951180473e-05,
+      "loss": 0.5521,
+      "step": 6533
+    },
+    {
+      "epoch": 8.36352,
+      "grad_norm": 0.7813871502876282,
+      "learning_rate": 3.69467787114846e-05,
+      "loss": 0.5531,
+      "step": 6534
+    },
+    {
+      "epoch": 8.3648,
+      "grad_norm": 0.7743489742279053,
+      "learning_rate": 3.694477791116447e-05,
+      "loss": 0.5521,
+      "step": 6535
+    },
+    {
+      "epoch": 8.36608,
+      "grad_norm": 0.8797667026519775,
+      "learning_rate": 3.694277711084434e-05,
+      "loss": 0.5972,
+      "step": 6536
+    },
+    {
+      "epoch": 8.36736,
+      "grad_norm": 0.8481217622756958,
+      "learning_rate": 3.694077631052421e-05,
+      "loss": 0.6185,
+      "step": 6537
+    },
+    {
+      "epoch": 8.36864,
+      "grad_norm": 0.8754853010177612,
+      "learning_rate": 3.693877551020408e-05,
+      "loss": 0.6194,
+      "step": 6538
+    },
+    {
+      "epoch": 8.36992,
+      "grad_norm": 0.8245212435722351,
+      "learning_rate": 3.693677470988395e-05,
+      "loss": 0.5823,
+      "step": 6539
+    },
+    {
+      "epoch": 8.3712,
+      "grad_norm": 0.8101980090141296,
+      "learning_rate": 3.693477390956382e-05,
+      "loss": 0.5779,
+      "step": 6540
+    },
+    {
+      "epoch": 8.37248,
+      "grad_norm": 0.8348204493522644,
+      "learning_rate": 3.69327731092437e-05,
+      "loss": 0.5794,
+      "step": 6541
+    },
+    {
+      "epoch": 8.37376,
+      "grad_norm": 0.84105384349823,
+      "learning_rate": 3.6930772308923574e-05,
+      "loss": 0.5751,
+      "step": 6542
+    },
+    {
+      "epoch": 8.37504,
+      "grad_norm": 0.8668518662452698,
+      "learning_rate": 3.6928771508603446e-05,
+      "loss": 0.5879,
+      "step": 6543
+    },
+    {
+      "epoch": 8.37632,
+      "grad_norm": 0.8224269151687622,
+      "learning_rate": 3.692677070828332e-05,
+      "loss": 0.5835,
+      "step": 6544
+    },
+    {
+      "epoch": 8.3776,
+      "grad_norm": 0.8665353059768677,
+      "learning_rate": 3.692476990796318e-05,
+      "loss": 0.5756,
+      "step": 6545
+    },
+    {
+      "epoch": 8.37888,
+      "grad_norm": 0.8154611587524414,
+      "learning_rate": 3.6922769107643054e-05,
+      "loss": 0.6244,
+      "step": 6546
+    },
+    {
+      "epoch": 8.38016,
+      "grad_norm": 0.7815151810646057,
+      "learning_rate": 3.6920768307322926e-05,
+      "loss": 0.5823,
+      "step": 6547
+    },
+    {
+      "epoch": 8.38144,
+      "grad_norm": 0.8650956749916077,
+      "learning_rate": 3.6918767507002805e-05,
+      "loss": 0.5881,
+      "step": 6548
+    },
+    {
+      "epoch": 8.38272,
+      "grad_norm": 0.8099015355110168,
+      "learning_rate": 3.691676670668268e-05,
+      "loss": 0.6132,
+      "step": 6549
+    },
+    {
+      "epoch": 8.384,
+      "grad_norm": 0.8544344902038574,
+      "learning_rate": 3.691476590636255e-05,
+      "loss": 0.5674,
+      "step": 6550
+    },
+    {
+      "epoch": 8.38528,
+      "grad_norm": 0.8150829672813416,
+      "learning_rate": 3.691276510604242e-05,
+      "loss": 0.6013,
+      "step": 6551
+    },
+    {
+      "epoch": 8.38656,
+      "grad_norm": 0.8252568244934082,
+      "learning_rate": 3.691076430572229e-05,
+      "loss": 0.5952,
+      "step": 6552
+    },
+    {
+      "epoch": 8.38784,
+      "grad_norm": 0.8224411010742188,
+      "learning_rate": 3.690876350540216e-05,
+      "loss": 0.613,
+      "step": 6553
+    },
+    {
+      "epoch": 8.38912,
+      "grad_norm": 0.8442636728286743,
+      "learning_rate": 3.690676270508203e-05,
+      "loss": 0.6177,
+      "step": 6554
+    },
+    {
+      "epoch": 8.3904,
+      "grad_norm": 0.8492255210876465,
+      "learning_rate": 3.690476190476191e-05,
+      "loss": 0.5842,
+      "step": 6555
+    },
+    {
+      "epoch": 8.39168,
+      "grad_norm": 0.8563710451126099,
+      "learning_rate": 3.690276110444178e-05,
+      "loss": 0.6015,
+      "step": 6556
+    },
+    {
+      "epoch": 8.39296,
+      "grad_norm": 0.8145362138748169,
+      "learning_rate": 3.690076030412165e-05,
+      "loss": 0.5714,
+      "step": 6557
+    },
+    {
+      "epoch": 8.39424,
+      "grad_norm": 0.8347020745277405,
+      "learning_rate": 3.6898759503801524e-05,
+      "loss": 0.6105,
+      "step": 6558
+    },
+    {
+      "epoch": 8.39552,
+      "grad_norm": 0.8055539131164551,
+      "learning_rate": 3.6896758703481395e-05,
+      "loss": 0.6214,
+      "step": 6559
+    },
+    {
+      "epoch": 8.3968,
+      "grad_norm": 0.7828962802886963,
+      "learning_rate": 3.689475790316127e-05,
+      "loss": 0.6206,
+      "step": 6560
+    },
+    {
+      "epoch": 8.39808,
+      "grad_norm": 0.8174715042114258,
+      "learning_rate": 3.689275710284113e-05,
+      "loss": 0.536,
+      "step": 6561
+    },
+    {
+      "epoch": 8.39936,
+      "grad_norm": 0.7967883944511414,
+      "learning_rate": 3.689075630252101e-05,
+      "loss": 0.6092,
+      "step": 6562
+    },
+    {
+      "epoch": 8.40064,
+      "grad_norm": 0.8208482265472412,
+      "learning_rate": 3.688875550220088e-05,
+      "loss": 0.5818,
+      "step": 6563
+    },
+    {
+      "epoch": 8.40192,
+      "grad_norm": 0.7751407623291016,
+      "learning_rate": 3.6886754701880755e-05,
+      "loss": 0.5362,
+      "step": 6564
+    },
+    {
+      "epoch": 8.4032,
+      "grad_norm": 0.8539714217185974,
+      "learning_rate": 3.6884753901560627e-05,
+      "loss": 0.5734,
+      "step": 6565
+    },
+    {
+      "epoch": 8.40448,
+      "grad_norm": 0.8336576819419861,
+      "learning_rate": 3.68827531012405e-05,
+      "loss": 0.5931,
+      "step": 6566
+    },
+    {
+      "epoch": 8.40576,
+      "grad_norm": 0.7995145320892334,
+      "learning_rate": 3.688075230092037e-05,
+      "loss": 0.6048,
+      "step": 6567
+    },
+    {
+      "epoch": 8.40704,
+      "grad_norm": 0.8343474268913269,
+      "learning_rate": 3.687875150060024e-05,
+      "loss": 0.5621,
+      "step": 6568
+    },
+    {
+      "epoch": 8.40832,
+      "grad_norm": 0.8455119132995605,
+      "learning_rate": 3.6876750700280114e-05,
+      "loss": 0.597,
+      "step": 6569
+    },
+    {
+      "epoch": 8.4096,
+      "grad_norm": 0.839838445186615,
+      "learning_rate": 3.6874749899959986e-05,
+      "loss": 0.5993,
+      "step": 6570
+    },
+    {
+      "epoch": 8.41088,
+      "grad_norm": 0.8005172610282898,
+      "learning_rate": 3.687274909963986e-05,
+      "loss": 0.5688,
+      "step": 6571
+    },
+    {
+      "epoch": 8.41216,
+      "grad_norm": 0.8667073249816895,
+      "learning_rate": 3.687074829931973e-05,
+      "loss": 0.5971,
+      "step": 6572
+    },
+    {
+      "epoch": 8.41344,
+      "grad_norm": 0.7542425990104675,
+      "learning_rate": 3.68687474989996e-05,
+      "loss": 0.5329,
+      "step": 6573
+    },
+    {
+      "epoch": 8.414719999999999,
+      "grad_norm": 0.8643616437911987,
+      "learning_rate": 3.686674669867947e-05,
+      "loss": 0.6388,
+      "step": 6574
+    },
+    {
+      "epoch": 8.416,
+      "grad_norm": 0.807478666305542,
+      "learning_rate": 3.6864745898359345e-05,
+      "loss": 0.6263,
+      "step": 6575
+    },
+    {
+      "epoch": 8.41728,
+      "grad_norm": 0.8404189944267273,
+      "learning_rate": 3.686274509803922e-05,
+      "loss": 0.6011,
+      "step": 6576
+    },
+    {
+      "epoch": 8.41856,
+      "grad_norm": 0.8093534111976624,
+      "learning_rate": 3.686074429771909e-05,
+      "loss": 0.5659,
+      "step": 6577
+    },
+    {
+      "epoch": 8.41984,
+      "grad_norm": 0.8560016751289368,
+      "learning_rate": 3.685874349739896e-05,
+      "loss": 0.6221,
+      "step": 6578
+    },
+    {
+      "epoch": 8.42112,
+      "grad_norm": 0.836397647857666,
+      "learning_rate": 3.685674269707883e-05,
+      "loss": 0.6274,
+      "step": 6579
+    },
+    {
+      "epoch": 8.4224,
+      "grad_norm": 0.8117139935493469,
+      "learning_rate": 3.6854741896758704e-05,
+      "loss": 0.6296,
+      "step": 6580
+    },
+    {
+      "epoch": 8.42368,
+      "grad_norm": 0.8044874668121338,
+      "learning_rate": 3.6852741096438576e-05,
+      "loss": 0.5879,
+      "step": 6581
+    },
+    {
+      "epoch": 8.42496,
+      "grad_norm": 0.7988526225090027,
+      "learning_rate": 3.685074029611845e-05,
+      "loss": 0.5576,
+      "step": 6582
+    },
+    {
+      "epoch": 8.42624,
+      "grad_norm": 0.7902975678443909,
+      "learning_rate": 3.684873949579833e-05,
+      "loss": 0.5694,
+      "step": 6583
+    },
+    {
+      "epoch": 8.42752,
+      "grad_norm": 0.8768497109413147,
+      "learning_rate": 3.684673869547819e-05,
+      "loss": 0.6101,
+      "step": 6584
+    },
+    {
+      "epoch": 8.4288,
+      "grad_norm": 0.806703507900238,
+      "learning_rate": 3.6844737895158064e-05,
+      "loss": 0.576,
+      "step": 6585
+    },
+    {
+      "epoch": 8.43008,
+      "grad_norm": 0.8013177514076233,
+      "learning_rate": 3.6842737094837936e-05,
+      "loss": 0.6118,
+      "step": 6586
+    },
+    {
+      "epoch": 8.43136,
+      "grad_norm": 0.8201990723609924,
+      "learning_rate": 3.684073629451781e-05,
+      "loss": 0.5902,
+      "step": 6587
+    },
+    {
+      "epoch": 8.43264,
+      "grad_norm": 0.8329010009765625,
+      "learning_rate": 3.683873549419768e-05,
+      "loss": 0.5995,
+      "step": 6588
+    },
+    {
+      "epoch": 8.43392,
+      "grad_norm": 0.7939520478248596,
+      "learning_rate": 3.683673469387755e-05,
+      "loss": 0.5849,
+      "step": 6589
+    },
+    {
+      "epoch": 8.4352,
+      "grad_norm": 0.8510729670524597,
+      "learning_rate": 3.683473389355743e-05,
+      "loss": 0.6381,
+      "step": 6590
+    },
+    {
+      "epoch": 8.43648,
+      "grad_norm": 0.8639024496078491,
+      "learning_rate": 3.68327330932373e-05,
+      "loss": 0.6025,
+      "step": 6591
+    },
+    {
+      "epoch": 8.43776,
+      "grad_norm": 0.81855708360672,
+      "learning_rate": 3.683073229291717e-05,
+      "loss": 0.5719,
+      "step": 6592
+    },
+    {
+      "epoch": 8.43904,
+      "grad_norm": 0.8242548108100891,
+      "learning_rate": 3.682873149259704e-05,
+      "loss": 0.5777,
+      "step": 6593
+    },
+    {
+      "epoch": 8.44032,
+      "grad_norm": 0.8355400562286377,
+      "learning_rate": 3.682673069227691e-05,
+      "loss": 0.5496,
+      "step": 6594
+    },
+    {
+      "epoch": 8.4416,
+      "grad_norm": 0.8287314772605896,
+      "learning_rate": 3.682472989195678e-05,
+      "loss": 0.6298,
+      "step": 6595
+    },
+    {
+      "epoch": 8.44288,
+      "grad_norm": 0.8493402600288391,
+      "learning_rate": 3.6822729091636654e-05,
+      "loss": 0.624,
+      "step": 6596
+    },
+    {
+      "epoch": 8.44416,
+      "grad_norm": 0.786516547203064,
+      "learning_rate": 3.682072829131653e-05,
+      "loss": 0.5663,
+      "step": 6597
+    },
+    {
+      "epoch": 8.44544,
+      "grad_norm": 0.7651713490486145,
+      "learning_rate": 3.6818727490996405e-05,
+      "loss": 0.5248,
+      "step": 6598
+    },
+    {
+      "epoch": 8.44672,
+      "grad_norm": 0.8335928320884705,
+      "learning_rate": 3.6816726690676276e-05,
+      "loss": 0.5722,
+      "step": 6599
+    },
+    {
+      "epoch": 8.448,
+      "grad_norm": 0.8091786503791809,
+      "learning_rate": 3.681472589035614e-05,
+      "loss": 0.554,
+      "step": 6600
+    },
+    {
+      "epoch": 8.44928,
+      "grad_norm": 0.846765398979187,
+      "learning_rate": 3.6812725090036013e-05,
+      "loss": 0.663,
+      "step": 6601
+    },
+    {
+      "epoch": 8.45056,
+      "grad_norm": 0.8616847991943359,
+      "learning_rate": 3.6810724289715885e-05,
+      "loss": 0.6325,
+      "step": 6602
+    },
+    {
+      "epoch": 8.45184,
+      "grad_norm": 0.7922678589820862,
+      "learning_rate": 3.680872348939576e-05,
+      "loss": 0.5701,
+      "step": 6603
+    },
+    {
+      "epoch": 8.45312,
+      "grad_norm": 0.8290618658065796,
+      "learning_rate": 3.6806722689075636e-05,
+      "loss": 0.6524,
+      "step": 6604
+    },
+    {
+      "epoch": 8.4544,
+      "grad_norm": 0.8202276825904846,
+      "learning_rate": 3.680472188875551e-05,
+      "loss": 0.5703,
+      "step": 6605
+    },
+    {
+      "epoch": 8.45568,
+      "grad_norm": 0.8039751052856445,
+      "learning_rate": 3.680272108843538e-05,
+      "loss": 0.5763,
+      "step": 6606
+    },
+    {
+      "epoch": 8.45696,
+      "grad_norm": 0.7898495197296143,
+      "learning_rate": 3.680072028811525e-05,
+      "loss": 0.5615,
+      "step": 6607
+    },
+    {
+      "epoch": 8.45824,
+      "grad_norm": 0.833078145980835,
+      "learning_rate": 3.6798719487795116e-05,
+      "loss": 0.5745,
+      "step": 6608
+    },
+    {
+      "epoch": 8.45952,
+      "grad_norm": 0.908738911151886,
+      "learning_rate": 3.679671868747499e-05,
+      "loss": 0.5853,
+      "step": 6609
+    },
+    {
+      "epoch": 8.4608,
+      "grad_norm": 0.8550048470497131,
+      "learning_rate": 3.679471788715486e-05,
+      "loss": 0.5834,
+      "step": 6610
+    },
+    {
+      "epoch": 8.46208,
+      "grad_norm": 0.8882399201393127,
+      "learning_rate": 3.679271708683474e-05,
+      "loss": 0.6226,
+      "step": 6611
+    },
+    {
+      "epoch": 8.46336,
+      "grad_norm": 0.832423746585846,
+      "learning_rate": 3.679071628651461e-05,
+      "loss": 0.5573,
+      "step": 6612
+    },
+    {
+      "epoch": 8.46464,
+      "grad_norm": 0.842068076133728,
+      "learning_rate": 3.678871548619448e-05,
+      "loss": 0.6053,
+      "step": 6613
+    },
+    {
+      "epoch": 8.46592,
+      "grad_norm": 0.8376506567001343,
+      "learning_rate": 3.6786714685874354e-05,
+      "loss": 0.5527,
+      "step": 6614
+    },
+    {
+      "epoch": 8.4672,
+      "grad_norm": 0.8572778701782227,
+      "learning_rate": 3.6784713885554226e-05,
+      "loss": 0.6259,
+      "step": 6615
+    },
+    {
+      "epoch": 8.46848,
+      "grad_norm": 0.817713737487793,
+      "learning_rate": 3.678271308523409e-05,
+      "loss": 0.5689,
+      "step": 6616
+    },
+    {
+      "epoch": 8.46976,
+      "grad_norm": 0.8326448202133179,
+      "learning_rate": 3.678071228491396e-05,
+      "loss": 0.5906,
+      "step": 6617
+    },
+    {
+      "epoch": 8.47104,
+      "grad_norm": 0.8092232942581177,
+      "learning_rate": 3.677871148459384e-05,
+      "loss": 0.5749,
+      "step": 6618
+    },
+    {
+      "epoch": 8.47232,
+      "grad_norm": 0.8321169018745422,
+      "learning_rate": 3.6776710684273714e-05,
+      "loss": 0.5773,
+      "step": 6619
+    },
+    {
+      "epoch": 8.4736,
+      "grad_norm": 0.8475345373153687,
+      "learning_rate": 3.6774709883953585e-05,
+      "loss": 0.6028,
+      "step": 6620
+    },
+    {
+      "epoch": 8.47488,
+      "grad_norm": 0.8301557898521423,
+      "learning_rate": 3.677270908363346e-05,
+      "loss": 0.5961,
+      "step": 6621
+    },
+    {
+      "epoch": 8.47616,
+      "grad_norm": 0.8141891360282898,
+      "learning_rate": 3.677070828331333e-05,
+      "loss": 0.6512,
+      "step": 6622
+    },
+    {
+      "epoch": 8.47744,
+      "grad_norm": 0.7891356348991394,
+      "learning_rate": 3.67687074829932e-05,
+      "loss": 0.539,
+      "step": 6623
+    },
+    {
+      "epoch": 8.47872,
+      "grad_norm": 0.7641344666481018,
+      "learning_rate": 3.6766706682673066e-05,
+      "loss": 0.5497,
+      "step": 6624
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.766279935836792,
+      "learning_rate": 3.6764705882352945e-05,
+      "loss": 0.5496,
+      "step": 6625
+    },
+    {
+      "epoch": 8.48128,
+      "grad_norm": 0.836593508720398,
+      "learning_rate": 3.6762705082032817e-05,
+      "loss": 0.6149,
+      "step": 6626
+    },
+    {
+      "epoch": 8.48256,
+      "grad_norm": 0.8140498399734497,
+      "learning_rate": 3.676070428171269e-05,
+      "loss": 0.5873,
+      "step": 6627
+    },
+    {
+      "epoch": 8.48384,
+      "grad_norm": 0.8106237053871155,
+      "learning_rate": 3.675870348139256e-05,
+      "loss": 0.5868,
+      "step": 6628
+    },
+    {
+      "epoch": 8.48512,
+      "grad_norm": 0.7953768968582153,
+      "learning_rate": 3.675670268107243e-05,
+      "loss": 0.5578,
+      "step": 6629
+    },
+    {
+      "epoch": 8.4864,
+      "grad_norm": 0.8248597383499146,
+      "learning_rate": 3.6754701880752304e-05,
+      "loss": 0.6027,
+      "step": 6630
+    },
+    {
+      "epoch": 8.48768,
+      "grad_norm": 0.8090713024139404,
+      "learning_rate": 3.6752701080432176e-05,
+      "loss": 0.5848,
+      "step": 6631
+    },
+    {
+      "epoch": 8.48896,
+      "grad_norm": 0.8814820647239685,
+      "learning_rate": 3.675070028011205e-05,
+      "loss": 0.6117,
+      "step": 6632
+    },
+    {
+      "epoch": 8.49024,
+      "grad_norm": 0.8482004404067993,
+      "learning_rate": 3.674869947979192e-05,
+      "loss": 0.6126,
+      "step": 6633
+    },
+    {
+      "epoch": 8.49152,
+      "grad_norm": 0.8183587789535522,
+      "learning_rate": 3.674669867947179e-05,
+      "loss": 0.5719,
+      "step": 6634
+    },
+    {
+      "epoch": 8.4928,
+      "grad_norm": 0.8571783304214478,
+      "learning_rate": 3.674469787915166e-05,
+      "loss": 0.6063,
+      "step": 6635
+    },
+    {
+      "epoch": 8.49408,
+      "grad_norm": 0.802609384059906,
+      "learning_rate": 3.6742697078831535e-05,
+      "loss": 0.625,
+      "step": 6636
+    },
+    {
+      "epoch": 8.49536,
+      "grad_norm": 0.8341289758682251,
+      "learning_rate": 3.674069627851141e-05,
+      "loss": 0.5994,
+      "step": 6637
+    },
+    {
+      "epoch": 8.49664,
+      "grad_norm": 0.8001972436904907,
+      "learning_rate": 3.673869547819128e-05,
+      "loss": 0.5502,
+      "step": 6638
+    },
+    {
+      "epoch": 8.49792,
+      "grad_norm": 0.8287181258201599,
+      "learning_rate": 3.673669467787115e-05,
+      "loss": 0.602,
+      "step": 6639
+    },
+    {
+      "epoch": 8.4992,
+      "grad_norm": 0.7981618046760559,
+      "learning_rate": 3.673469387755102e-05,
+      "loss": 0.6023,
+      "step": 6640
+    },
+    {
+      "epoch": 8.50048,
+      "grad_norm": 0.8224805593490601,
+      "learning_rate": 3.6732693077230894e-05,
+      "loss": 0.5563,
+      "step": 6641
+    },
+    {
+      "epoch": 8.50176,
+      "grad_norm": 0.7999238967895508,
+      "learning_rate": 3.6730692276910766e-05,
+      "loss": 0.5549,
+      "step": 6642
+    },
+    {
+      "epoch": 8.50304,
+      "grad_norm": 0.8282129764556885,
+      "learning_rate": 3.672869147659064e-05,
+      "loss": 0.6004,
+      "step": 6643
+    },
+    {
+      "epoch": 8.50432,
+      "grad_norm": 0.8010920882225037,
+      "learning_rate": 3.672669067627051e-05,
+      "loss": 0.5956,
+      "step": 6644
+    },
+    {
+      "epoch": 8.5056,
+      "grad_norm": 0.7640748023986816,
+      "learning_rate": 3.672468987595038e-05,
+      "loss": 0.5638,
+      "step": 6645
+    },
+    {
+      "epoch": 8.50688,
+      "grad_norm": 0.8032206296920776,
+      "learning_rate": 3.6722689075630254e-05,
+      "loss": 0.5797,
+      "step": 6646
+    },
+    {
+      "epoch": 8.50816,
+      "grad_norm": 0.8864914774894714,
+      "learning_rate": 3.6720688275310126e-05,
+      "loss": 0.5798,
+      "step": 6647
+    },
+    {
+      "epoch": 8.50944,
+      "grad_norm": 0.8176901340484619,
+      "learning_rate": 3.671868747499e-05,
+      "loss": 0.6298,
+      "step": 6648
+    },
+    {
+      "epoch": 8.51072,
+      "grad_norm": 0.8001031279563904,
+      "learning_rate": 3.671668667466987e-05,
+      "loss": 0.6211,
+      "step": 6649
+    },
+    {
+      "epoch": 8.512,
+      "grad_norm": 0.8112642168998718,
+      "learning_rate": 3.671468587434974e-05,
+      "loss": 0.6003,
+      "step": 6650
+    },
+    {
+      "epoch": 8.51328,
+      "grad_norm": 0.8111798763275146,
+      "learning_rate": 3.671268507402961e-05,
+      "loss": 0.603,
+      "step": 6651
+    },
+    {
+      "epoch": 8.51456,
+      "grad_norm": 0.8160041570663452,
+      "learning_rate": 3.6710684273709485e-05,
+      "loss": 0.5565,
+      "step": 6652
+    },
+    {
+      "epoch": 8.51584,
+      "grad_norm": 0.7813338041305542,
+      "learning_rate": 3.670868347338936e-05,
+      "loss": 0.5783,
+      "step": 6653
+    },
+    {
+      "epoch": 8.51712,
+      "grad_norm": 0.8111140131950378,
+      "learning_rate": 3.670668267306923e-05,
+      "loss": 0.6126,
+      "step": 6654
+    },
+    {
+      "epoch": 8.5184,
+      "grad_norm": 0.8105128407478333,
+      "learning_rate": 3.67046818727491e-05,
+      "loss": 0.5711,
+      "step": 6655
+    },
+    {
+      "epoch": 8.51968,
+      "grad_norm": 0.8445085287094116,
+      "learning_rate": 3.670268107242897e-05,
+      "loss": 0.64,
+      "step": 6656
+    },
+    {
+      "epoch": 8.52096,
+      "grad_norm": 0.8055734634399414,
+      "learning_rate": 3.6700680272108844e-05,
+      "loss": 0.5938,
+      "step": 6657
+    },
+    {
+      "epoch": 8.52224,
+      "grad_norm": 0.8156753182411194,
+      "learning_rate": 3.6698679471788716e-05,
+      "loss": 0.5672,
+      "step": 6658
+    },
+    {
+      "epoch": 8.52352,
+      "grad_norm": 0.7987210154533386,
+      "learning_rate": 3.669667867146859e-05,
+      "loss": 0.5924,
+      "step": 6659
+    },
+    {
+      "epoch": 8.5248,
+      "grad_norm": 0.878676176071167,
+      "learning_rate": 3.669467787114846e-05,
+      "loss": 0.6441,
+      "step": 6660
+    },
+    {
+      "epoch": 8.52608,
+      "grad_norm": 0.8461551666259766,
+      "learning_rate": 3.669267707082834e-05,
+      "loss": 0.5701,
+      "step": 6661
+    },
+    {
+      "epoch": 8.52736,
+      "grad_norm": 0.7994030117988586,
+      "learning_rate": 3.6690676270508203e-05,
+      "loss": 0.5331,
+      "step": 6662
+    },
+    {
+      "epoch": 8.52864,
+      "grad_norm": 0.8361437320709229,
+      "learning_rate": 3.6688675470188075e-05,
+      "loss": 0.6438,
+      "step": 6663
+    },
+    {
+      "epoch": 8.52992,
+      "grad_norm": 0.8350645899772644,
+      "learning_rate": 3.668667466986795e-05,
+      "loss": 0.5887,
+      "step": 6664
+    },
+    {
+      "epoch": 8.5312,
+      "grad_norm": 0.8255120515823364,
+      "learning_rate": 3.668467386954782e-05,
+      "loss": 0.5805,
+      "step": 6665
+    },
+    {
+      "epoch": 8.53248,
+      "grad_norm": 0.8830844759941101,
+      "learning_rate": 3.668267306922769e-05,
+      "loss": 0.6237,
+      "step": 6666
+    },
+    {
+      "epoch": 8.533760000000001,
+      "grad_norm": 0.8089183568954468,
+      "learning_rate": 3.668067226890756e-05,
+      "loss": 0.5825,
+      "step": 6667
+    },
+    {
+      "epoch": 8.53504,
+      "grad_norm": 0.7940279245376587,
+      "learning_rate": 3.667867146858744e-05,
+      "loss": 0.5835,
+      "step": 6668
+    },
+    {
+      "epoch": 8.53632,
+      "grad_norm": 0.8011153340339661,
+      "learning_rate": 3.667667066826731e-05,
+      "loss": 0.5635,
+      "step": 6669
+    },
+    {
+      "epoch": 8.5376,
+      "grad_norm": 0.8229313492774963,
+      "learning_rate": 3.667466986794718e-05,
+      "loss": 0.553,
+      "step": 6670
+    },
+    {
+      "epoch": 8.53888,
+      "grad_norm": 0.8413478136062622,
+      "learning_rate": 3.667266906762705e-05,
+      "loss": 0.5636,
+      "step": 6671
+    },
+    {
+      "epoch": 8.54016,
+      "grad_norm": 0.8253154754638672,
+      "learning_rate": 3.667066826730692e-05,
+      "loss": 0.6305,
+      "step": 6672
+    },
+    {
+      "epoch": 8.54144,
+      "grad_norm": 0.8086980581283569,
+      "learning_rate": 3.6668667466986794e-05,
+      "loss": 0.6359,
+      "step": 6673
+    },
+    {
+      "epoch": 8.54272,
+      "grad_norm": 0.8422557711601257,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.5517,
+      "step": 6674
+    },
+    {
+      "epoch": 8.544,
+      "grad_norm": 0.7937127947807312,
+      "learning_rate": 3.6664665866346544e-05,
+      "loss": 0.6099,
+      "step": 6675
+    },
+    {
+      "epoch": 8.54528,
+      "grad_norm": 0.8104080557823181,
+      "learning_rate": 3.6662665066026416e-05,
+      "loss": 0.647,
+      "step": 6676
+    },
+    {
+      "epoch": 8.54656,
+      "grad_norm": 0.7860154509544373,
+      "learning_rate": 3.666066426570629e-05,
+      "loss": 0.5656,
+      "step": 6677
+    },
+    {
+      "epoch": 8.54784,
+      "grad_norm": 0.8154928684234619,
+      "learning_rate": 3.665866346538615e-05,
+      "loss": 0.5811,
+      "step": 6678
+    },
+    {
+      "epoch": 8.54912,
+      "grad_norm": 0.8200475573539734,
+      "learning_rate": 3.6656662665066025e-05,
+      "loss": 0.5985,
+      "step": 6679
+    },
+    {
+      "epoch": 8.5504,
+      "grad_norm": 0.7977676391601562,
+      "learning_rate": 3.66546618647459e-05,
+      "loss": 0.535,
+      "step": 6680
+    },
+    {
+      "epoch": 8.55168,
+      "grad_norm": 0.7876496315002441,
+      "learning_rate": 3.665266106442577e-05,
+      "loss": 0.582,
+      "step": 6681
+    },
+    {
+      "epoch": 8.55296,
+      "grad_norm": 0.7595927119255066,
+      "learning_rate": 3.665066026410565e-05,
+      "loss": 0.5749,
+      "step": 6682
+    },
+    {
+      "epoch": 8.55424,
+      "grad_norm": 0.8560596108436584,
+      "learning_rate": 3.664865946378552e-05,
+      "loss": 0.6061,
+      "step": 6683
+    },
+    {
+      "epoch": 8.55552,
+      "grad_norm": 0.7883579134941101,
+      "learning_rate": 3.664665866346539e-05,
+      "loss": 0.5673,
+      "step": 6684
+    },
+    {
+      "epoch": 8.556799999999999,
+      "grad_norm": 0.8544105887413025,
+      "learning_rate": 3.664465786314526e-05,
+      "loss": 0.644,
+      "step": 6685
+    },
+    {
+      "epoch": 8.55808,
+      "grad_norm": 0.8229860663414001,
+      "learning_rate": 3.664265706282513e-05,
+      "loss": 0.5651,
+      "step": 6686
+    },
+    {
+      "epoch": 8.55936,
+      "grad_norm": 0.8209177851676941,
+      "learning_rate": 3.6640656262505e-05,
+      "loss": 0.6064,
+      "step": 6687
+    },
+    {
+      "epoch": 8.56064,
+      "grad_norm": 0.867435097694397,
+      "learning_rate": 3.663865546218487e-05,
+      "loss": 0.568,
+      "step": 6688
+    },
+    {
+      "epoch": 8.56192,
+      "grad_norm": 0.8385619521141052,
+      "learning_rate": 3.663665466186475e-05,
+      "loss": 0.5947,
+      "step": 6689
+    },
+    {
+      "epoch": 8.5632,
+      "grad_norm": 0.7875993847846985,
+      "learning_rate": 3.663465386154462e-05,
+      "loss": 0.5768,
+      "step": 6690
+    },
+    {
+      "epoch": 8.56448,
+      "grad_norm": 0.8218605518341064,
+      "learning_rate": 3.6632653061224494e-05,
+      "loss": 0.5482,
+      "step": 6691
+    },
+    {
+      "epoch": 8.565760000000001,
+      "grad_norm": 0.7455148696899414,
+      "learning_rate": 3.6630652260904366e-05,
+      "loss": 0.5475,
+      "step": 6692
+    },
+    {
+      "epoch": 8.56704,
+      "grad_norm": 0.7991039156913757,
+      "learning_rate": 3.662865146058424e-05,
+      "loss": 0.6022,
+      "step": 6693
+    },
+    {
+      "epoch": 8.56832,
+      "grad_norm": 0.8934693932533264,
+      "learning_rate": 3.66266506602641e-05,
+      "loss": 0.6481,
+      "step": 6694
+    },
+    {
+      "epoch": 8.5696,
+      "grad_norm": 0.8251892328262329,
+      "learning_rate": 3.6624649859943975e-05,
+      "loss": 0.6115,
+      "step": 6695
+    },
+    {
+      "epoch": 8.57088,
+      "grad_norm": 0.7520719766616821,
+      "learning_rate": 3.662264905962385e-05,
+      "loss": 0.5618,
+      "step": 6696
+    },
+    {
+      "epoch": 8.57216,
+      "grad_norm": 0.7765500545501709,
+      "learning_rate": 3.6620648259303725e-05,
+      "loss": 0.5914,
+      "step": 6697
+    },
+    {
+      "epoch": 8.57344,
+      "grad_norm": 0.8249855637550354,
+      "learning_rate": 3.66186474589836e-05,
+      "loss": 0.5604,
+      "step": 6698
+    },
+    {
+      "epoch": 8.57472,
+      "grad_norm": 0.8163955211639404,
+      "learning_rate": 3.661664665866347e-05,
+      "loss": 0.5495,
+      "step": 6699
+    },
+    {
+      "epoch": 8.576,
+      "grad_norm": 0.7878378033638,
+      "learning_rate": 3.661464585834334e-05,
+      "loss": 0.5725,
+      "step": 6700
+    },
+    {
+      "epoch": 8.57728,
+      "grad_norm": 0.8135743141174316,
+      "learning_rate": 3.661264505802321e-05,
+      "loss": 0.5855,
+      "step": 6701
+    },
+    {
+      "epoch": 8.57856,
+      "grad_norm": 0.8749912977218628,
+      "learning_rate": 3.661064425770308e-05,
+      "loss": 0.6184,
+      "step": 6702
+    },
+    {
+      "epoch": 8.57984,
+      "grad_norm": 0.7891753315925598,
+      "learning_rate": 3.6608643457382956e-05,
+      "loss": 0.5873,
+      "step": 6703
+    },
+    {
+      "epoch": 8.58112,
+      "grad_norm": 0.8346560001373291,
+      "learning_rate": 3.660664265706283e-05,
+      "loss": 0.5899,
+      "step": 6704
+    },
+    {
+      "epoch": 8.5824,
+      "grad_norm": 0.874178946018219,
+      "learning_rate": 3.66046418567427e-05,
+      "loss": 0.6579,
+      "step": 6705
+    },
+    {
+      "epoch": 8.58368,
+      "grad_norm": 0.7875180244445801,
+      "learning_rate": 3.660264105642257e-05,
+      "loss": 0.5685,
+      "step": 6706
+    },
+    {
+      "epoch": 8.58496,
+      "grad_norm": 0.8282900452613831,
+      "learning_rate": 3.6600640256102444e-05,
+      "loss": 0.5997,
+      "step": 6707
+    },
+    {
+      "epoch": 8.58624,
+      "grad_norm": 0.8672759532928467,
+      "learning_rate": 3.6598639455782316e-05,
+      "loss": 0.5734,
+      "step": 6708
+    },
+    {
+      "epoch": 8.58752,
+      "grad_norm": 0.8106910586357117,
+      "learning_rate": 3.659663865546219e-05,
+      "loss": 0.6093,
+      "step": 6709
+    },
+    {
+      "epoch": 8.588799999999999,
+      "grad_norm": 0.8382178544998169,
+      "learning_rate": 3.659463785514206e-05,
+      "loss": 0.6188,
+      "step": 6710
+    },
+    {
+      "epoch": 8.59008,
+      "grad_norm": 0.8015230298042297,
+      "learning_rate": 3.659263705482193e-05,
+      "loss": 0.5521,
+      "step": 6711
+    },
+    {
+      "epoch": 8.59136,
+      "grad_norm": 0.8438718914985657,
+      "learning_rate": 3.65906362545018e-05,
+      "loss": 0.6076,
+      "step": 6712
+    },
+    {
+      "epoch": 8.59264,
+      "grad_norm": 0.8490085601806641,
+      "learning_rate": 3.6588635454181675e-05,
+      "loss": 0.6349,
+      "step": 6713
+    },
+    {
+      "epoch": 8.59392,
+      "grad_norm": 0.7863012552261353,
+      "learning_rate": 3.658663465386155e-05,
+      "loss": 0.5578,
+      "step": 6714
+    },
+    {
+      "epoch": 8.5952,
+      "grad_norm": 0.7872239947319031,
+      "learning_rate": 3.658463385354142e-05,
+      "loss": 0.5816,
+      "step": 6715
+    },
+    {
+      "epoch": 8.59648,
+      "grad_norm": 0.7982747554779053,
+      "learning_rate": 3.658263305322129e-05,
+      "loss": 0.5745,
+      "step": 6716
+    },
+    {
+      "epoch": 8.59776,
+      "grad_norm": 0.7870101928710938,
+      "learning_rate": 3.658063225290116e-05,
+      "loss": 0.5575,
+      "step": 6717
+    },
+    {
+      "epoch": 8.59904,
+      "grad_norm": 0.8285995125770569,
+      "learning_rate": 3.6578631452581034e-05,
+      "loss": 0.5857,
+      "step": 6718
+    },
+    {
+      "epoch": 8.60032,
+      "grad_norm": 0.8226214051246643,
+      "learning_rate": 3.6576630652260906e-05,
+      "loss": 0.5893,
+      "step": 6719
+    },
+    {
+      "epoch": 8.6016,
+      "grad_norm": 0.8384042382240295,
+      "learning_rate": 3.657462985194078e-05,
+      "loss": 0.5664,
+      "step": 6720
+    },
+    {
+      "epoch": 8.60288,
+      "grad_norm": 0.8193085789680481,
+      "learning_rate": 3.657262905162065e-05,
+      "loss": 0.5688,
+      "step": 6721
+    },
+    {
+      "epoch": 8.60416,
+      "grad_norm": 0.8433443307876587,
+      "learning_rate": 3.657062825130052e-05,
+      "loss": 0.6191,
+      "step": 6722
+    },
+    {
+      "epoch": 8.60544,
+      "grad_norm": 0.7860081791877747,
+      "learning_rate": 3.6568627450980393e-05,
+      "loss": 0.547,
+      "step": 6723
+    },
+    {
+      "epoch": 8.60672,
+      "grad_norm": 0.7903274893760681,
+      "learning_rate": 3.6566626650660265e-05,
+      "loss": 0.5653,
+      "step": 6724
+    },
+    {
+      "epoch": 8.608,
+      "grad_norm": 0.8736501336097717,
+      "learning_rate": 3.656462585034014e-05,
+      "loss": 0.5936,
+      "step": 6725
+    },
+    {
+      "epoch": 8.60928,
+      "grad_norm": 0.8456231355667114,
+      "learning_rate": 3.656262505002001e-05,
+      "loss": 0.5829,
+      "step": 6726
+    },
+    {
+      "epoch": 8.61056,
+      "grad_norm": 0.7985215187072754,
+      "learning_rate": 3.656062424969988e-05,
+      "loss": 0.5847,
+      "step": 6727
+    },
+    {
+      "epoch": 8.61184,
+      "grad_norm": 0.8238279223442078,
+      "learning_rate": 3.655862344937975e-05,
+      "loss": 0.5796,
+      "step": 6728
+    },
+    {
+      "epoch": 8.61312,
+      "grad_norm": 0.826208233833313,
+      "learning_rate": 3.6556622649059625e-05,
+      "loss": 0.5839,
+      "step": 6729
+    },
+    {
+      "epoch": 8.6144,
+      "grad_norm": 0.7756603956222534,
+      "learning_rate": 3.6554621848739496e-05,
+      "loss": 0.544,
+      "step": 6730
+    },
+    {
+      "epoch": 8.61568,
+      "grad_norm": 0.7803246974945068,
+      "learning_rate": 3.6552621048419375e-05,
+      "loss": 0.5523,
+      "step": 6731
+    },
+    {
+      "epoch": 8.61696,
+      "grad_norm": 0.8132696747779846,
+      "learning_rate": 3.655062024809924e-05,
+      "loss": 0.5963,
+      "step": 6732
+    },
+    {
+      "epoch": 8.61824,
+      "grad_norm": 0.8277651071548462,
+      "learning_rate": 3.654861944777911e-05,
+      "loss": 0.6203,
+      "step": 6733
+    },
+    {
+      "epoch": 8.61952,
+      "grad_norm": 0.8409605622291565,
+      "learning_rate": 3.6546618647458984e-05,
+      "loss": 0.6087,
+      "step": 6734
+    },
+    {
+      "epoch": 8.6208,
+      "grad_norm": 0.8315775990486145,
+      "learning_rate": 3.6544617847138856e-05,
+      "loss": 0.5775,
+      "step": 6735
+    },
+    {
+      "epoch": 8.62208,
+      "grad_norm": 0.8047201037406921,
+      "learning_rate": 3.654261704681873e-05,
+      "loss": 0.6142,
+      "step": 6736
+    },
+    {
+      "epoch": 8.62336,
+      "grad_norm": 0.7707719802856445,
+      "learning_rate": 3.65406162464986e-05,
+      "loss": 0.5723,
+      "step": 6737
+    },
+    {
+      "epoch": 8.62464,
+      "grad_norm": 0.7818657755851746,
+      "learning_rate": 3.653861544617848e-05,
+      "loss": 0.6173,
+      "step": 6738
+    },
+    {
+      "epoch": 8.62592,
+      "grad_norm": 0.8295494914054871,
+      "learning_rate": 3.653661464585835e-05,
+      "loss": 0.598,
+      "step": 6739
+    },
+    {
+      "epoch": 8.6272,
+      "grad_norm": 0.8243419528007507,
+      "learning_rate": 3.6534613845538215e-05,
+      "loss": 0.5859,
+      "step": 6740
+    },
+    {
+      "epoch": 8.62848,
+      "grad_norm": 0.8508108258247375,
+      "learning_rate": 3.653261304521809e-05,
+      "loss": 0.6116,
+      "step": 6741
+    },
+    {
+      "epoch": 8.62976,
+      "grad_norm": 0.8881233334541321,
+      "learning_rate": 3.653061224489796e-05,
+      "loss": 0.6642,
+      "step": 6742
+    },
+    {
+      "epoch": 8.63104,
+      "grad_norm": 0.8283036351203918,
+      "learning_rate": 3.652861144457783e-05,
+      "loss": 0.5926,
+      "step": 6743
+    },
+    {
+      "epoch": 8.63232,
+      "grad_norm": 0.7696643471717834,
+      "learning_rate": 3.65266106442577e-05,
+      "loss": 0.5894,
+      "step": 6744
+    },
+    {
+      "epoch": 8.6336,
+      "grad_norm": 0.8470257520675659,
+      "learning_rate": 3.652460984393758e-05,
+      "loss": 0.5801,
+      "step": 6745
+    },
+    {
+      "epoch": 8.63488,
+      "grad_norm": 0.8444271087646484,
+      "learning_rate": 3.652260904361745e-05,
+      "loss": 0.5986,
+      "step": 6746
+    },
+    {
+      "epoch": 8.63616,
+      "grad_norm": 0.7664961814880371,
+      "learning_rate": 3.6520608243297325e-05,
+      "loss": 0.553,
+      "step": 6747
+    },
+    {
+      "epoch": 8.63744,
+      "grad_norm": 0.8508731722831726,
+      "learning_rate": 3.651860744297719e-05,
+      "loss": 0.6113,
+      "step": 6748
+    },
+    {
+      "epoch": 8.63872,
+      "grad_norm": 0.7583442330360413,
+      "learning_rate": 3.651660664265706e-05,
+      "loss": 0.533,
+      "step": 6749
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.812756359577179,
+      "learning_rate": 3.6514605842336934e-05,
+      "loss": 0.5829,
+      "step": 6750
+    },
+    {
+      "epoch": 8.64128,
+      "grad_norm": 0.8092144727706909,
+      "learning_rate": 3.6512605042016805e-05,
+      "loss": 0.5675,
+      "step": 6751
+    },
+    {
+      "epoch": 8.64256,
+      "grad_norm": 0.8396515846252441,
+      "learning_rate": 3.6510604241696684e-05,
+      "loss": 0.6065,
+      "step": 6752
+    },
+    {
+      "epoch": 8.64384,
+      "grad_norm": 0.7741784453392029,
+      "learning_rate": 3.6508603441376556e-05,
+      "loss": 0.5881,
+      "step": 6753
+    },
+    {
+      "epoch": 8.64512,
+      "grad_norm": 0.8222891092300415,
+      "learning_rate": 3.650660264105643e-05,
+      "loss": 0.6125,
+      "step": 6754
+    },
+    {
+      "epoch": 8.6464,
+      "grad_norm": 0.8462512493133545,
+      "learning_rate": 3.65046018407363e-05,
+      "loss": 0.6372,
+      "step": 6755
+    },
+    {
+      "epoch": 8.64768,
+      "grad_norm": 0.8327202796936035,
+      "learning_rate": 3.6502601040416165e-05,
+      "loss": 0.6419,
+      "step": 6756
+    },
+    {
+      "epoch": 8.64896,
+      "grad_norm": 0.7695246338844299,
+      "learning_rate": 3.6500600240096037e-05,
+      "loss": 0.5385,
+      "step": 6757
+    },
+    {
+      "epoch": 8.65024,
+      "grad_norm": 0.8170539736747742,
+      "learning_rate": 3.649859943977591e-05,
+      "loss": 0.5818,
+      "step": 6758
+    },
+    {
+      "epoch": 8.65152,
+      "grad_norm": 0.8147903680801392,
+      "learning_rate": 3.649659863945579e-05,
+      "loss": 0.616,
+      "step": 6759
+    },
+    {
+      "epoch": 8.6528,
+      "grad_norm": 0.7922413945198059,
+      "learning_rate": 3.649459783913566e-05,
+      "loss": 0.6042,
+      "step": 6760
+    },
+    {
+      "epoch": 8.65408,
+      "grad_norm": 0.7604812979698181,
+      "learning_rate": 3.649259703881553e-05,
+      "loss": 0.5312,
+      "step": 6761
+    },
+    {
+      "epoch": 8.65536,
+      "grad_norm": 0.855582594871521,
+      "learning_rate": 3.64905962384954e-05,
+      "loss": 0.6038,
+      "step": 6762
+    },
+    {
+      "epoch": 8.65664,
+      "grad_norm": 0.8385626077651978,
+      "learning_rate": 3.6488595438175275e-05,
+      "loss": 0.569,
+      "step": 6763
+    },
+    {
+      "epoch": 8.65792,
+      "grad_norm": 0.8010948300361633,
+      "learning_rate": 3.648659463785514e-05,
+      "loss": 0.5846,
+      "step": 6764
+    },
+    {
+      "epoch": 8.6592,
+      "grad_norm": 0.8321779370307922,
+      "learning_rate": 3.648459383753501e-05,
+      "loss": 0.6176,
+      "step": 6765
+    },
+    {
+      "epoch": 8.66048,
+      "grad_norm": 0.8671482801437378,
+      "learning_rate": 3.648259303721488e-05,
+      "loss": 0.6062,
+      "step": 6766
+    },
+    {
+      "epoch": 8.66176,
+      "grad_norm": 0.8177146911621094,
+      "learning_rate": 3.648059223689476e-05,
+      "loss": 0.5906,
+      "step": 6767
+    },
+    {
+      "epoch": 8.66304,
+      "grad_norm": 0.8458163738250732,
+      "learning_rate": 3.6478591436574634e-05,
+      "loss": 0.6042,
+      "step": 6768
+    },
+    {
+      "epoch": 8.66432,
+      "grad_norm": 0.7744457125663757,
+      "learning_rate": 3.6476590636254506e-05,
+      "loss": 0.5768,
+      "step": 6769
+    },
+    {
+      "epoch": 8.6656,
+      "grad_norm": 0.8034748435020447,
+      "learning_rate": 3.647458983593438e-05,
+      "loss": 0.5837,
+      "step": 6770
+    },
+    {
+      "epoch": 8.66688,
+      "grad_norm": 0.7622728943824768,
+      "learning_rate": 3.647258903561425e-05,
+      "loss": 0.5629,
+      "step": 6771
+    },
+    {
+      "epoch": 8.66816,
+      "grad_norm": 0.8260266780853271,
+      "learning_rate": 3.6470588235294114e-05,
+      "loss": 0.5742,
+      "step": 6772
+    },
+    {
+      "epoch": 8.66944,
+      "grad_norm": 0.8134649395942688,
+      "learning_rate": 3.6468587434973986e-05,
+      "loss": 0.5833,
+      "step": 6773
+    },
+    {
+      "epoch": 8.67072,
+      "grad_norm": 0.827939510345459,
+      "learning_rate": 3.6466586634653865e-05,
+      "loss": 0.5361,
+      "step": 6774
+    },
+    {
+      "epoch": 8.672,
+      "grad_norm": 0.8437283635139465,
+      "learning_rate": 3.646458583433374e-05,
+      "loss": 0.6519,
+      "step": 6775
+    },
+    {
+      "epoch": 8.67328,
+      "grad_norm": 0.836875319480896,
+      "learning_rate": 3.646258503401361e-05,
+      "loss": 0.605,
+      "step": 6776
+    },
+    {
+      "epoch": 8.67456,
+      "grad_norm": 0.8690876364707947,
+      "learning_rate": 3.646058423369348e-05,
+      "loss": 0.6817,
+      "step": 6777
+    },
+    {
+      "epoch": 8.67584,
+      "grad_norm": 0.80400550365448,
+      "learning_rate": 3.645858343337335e-05,
+      "loss": 0.5805,
+      "step": 6778
+    },
+    {
+      "epoch": 8.67712,
+      "grad_norm": 0.7860267162322998,
+      "learning_rate": 3.6456582633053224e-05,
+      "loss": 0.5937,
+      "step": 6779
+    },
+    {
+      "epoch": 8.6784,
+      "grad_norm": 0.8096144199371338,
+      "learning_rate": 3.645458183273309e-05,
+      "loss": 0.5387,
+      "step": 6780
+    },
+    {
+      "epoch": 8.67968,
+      "grad_norm": 0.7942869067192078,
+      "learning_rate": 3.645258103241297e-05,
+      "loss": 0.5842,
+      "step": 6781
+    },
+    {
+      "epoch": 8.68096,
+      "grad_norm": 0.768929123878479,
+      "learning_rate": 3.645058023209284e-05,
+      "loss": 0.6049,
+      "step": 6782
+    },
+    {
+      "epoch": 8.68224,
+      "grad_norm": 0.8190042972564697,
+      "learning_rate": 3.644857943177271e-05,
+      "loss": 0.5947,
+      "step": 6783
+    },
+    {
+      "epoch": 8.68352,
+      "grad_norm": 0.847685694694519,
+      "learning_rate": 3.6446578631452584e-05,
+      "loss": 0.6109,
+      "step": 6784
+    },
+    {
+      "epoch": 8.6848,
+      "grad_norm": 0.7972142696380615,
+      "learning_rate": 3.6444577831132455e-05,
+      "loss": 0.5764,
+      "step": 6785
+    },
+    {
+      "epoch": 8.68608,
+      "grad_norm": 0.8239948153495789,
+      "learning_rate": 3.644257703081233e-05,
+      "loss": 0.6277,
+      "step": 6786
+    },
+    {
+      "epoch": 8.68736,
+      "grad_norm": 0.7374285459518433,
+      "learning_rate": 3.64405762304922e-05,
+      "loss": 0.5331,
+      "step": 6787
+    },
+    {
+      "epoch": 8.68864,
+      "grad_norm": 0.7998077273368835,
+      "learning_rate": 3.643857543017207e-05,
+      "loss": 0.5555,
+      "step": 6788
+    },
+    {
+      "epoch": 8.68992,
+      "grad_norm": 0.8306496739387512,
+      "learning_rate": 3.643657462985194e-05,
+      "loss": 0.5906,
+      "step": 6789
+    },
+    {
+      "epoch": 8.6912,
+      "grad_norm": 0.8808903694152832,
+      "learning_rate": 3.6434573829531815e-05,
+      "loss": 0.6105,
+      "step": 6790
+    },
+    {
+      "epoch": 8.69248,
+      "grad_norm": 0.7989248037338257,
+      "learning_rate": 3.6432573029211687e-05,
+      "loss": 0.5822,
+      "step": 6791
+    },
+    {
+      "epoch": 8.69376,
+      "grad_norm": 0.7795215845108032,
+      "learning_rate": 3.643057222889156e-05,
+      "loss": 0.5552,
+      "step": 6792
+    },
+    {
+      "epoch": 8.69504,
+      "grad_norm": 0.8287959098815918,
+      "learning_rate": 3.642857142857143e-05,
+      "loss": 0.6018,
+      "step": 6793
+    },
+    {
+      "epoch": 8.69632,
+      "grad_norm": 0.8223719000816345,
+      "learning_rate": 3.64265706282513e-05,
+      "loss": 0.5816,
+      "step": 6794
+    },
+    {
+      "epoch": 8.6976,
+      "grad_norm": 0.8280214667320251,
+      "learning_rate": 3.6424569827931174e-05,
+      "loss": 0.6346,
+      "step": 6795
+    },
+    {
+      "epoch": 8.698879999999999,
+      "grad_norm": 0.795192539691925,
+      "learning_rate": 3.6422569027611046e-05,
+      "loss": 0.6239,
+      "step": 6796
+    },
+    {
+      "epoch": 8.70016,
+      "grad_norm": 0.7667868137359619,
+      "learning_rate": 3.642056822729092e-05,
+      "loss": 0.5148,
+      "step": 6797
+    },
+    {
+      "epoch": 8.70144,
+      "grad_norm": 0.8211344480514526,
+      "learning_rate": 3.641856742697079e-05,
+      "loss": 0.5715,
+      "step": 6798
+    },
+    {
+      "epoch": 8.70272,
+      "grad_norm": 0.8203229904174805,
+      "learning_rate": 3.641656662665066e-05,
+      "loss": 0.6021,
+      "step": 6799
+    },
+    {
+      "epoch": 8.704,
+      "grad_norm": 0.8265122175216675,
+      "learning_rate": 3.641456582633053e-05,
+      "loss": 0.6179,
+      "step": 6800
+    },
+    {
+      "epoch": 8.70528,
+      "grad_norm": 0.9294067025184631,
+      "learning_rate": 3.6412565026010405e-05,
+      "loss": 0.6312,
+      "step": 6801
+    },
+    {
+      "epoch": 8.70656,
+      "grad_norm": 0.8427034020423889,
+      "learning_rate": 3.641056422569028e-05,
+      "loss": 0.5626,
+      "step": 6802
+    },
+    {
+      "epoch": 8.707840000000001,
+      "grad_norm": 0.8450182676315308,
+      "learning_rate": 3.640856342537015e-05,
+      "loss": 0.6435,
+      "step": 6803
+    },
+    {
+      "epoch": 8.70912,
+      "grad_norm": 0.8336526155471802,
+      "learning_rate": 3.640656262505002e-05,
+      "loss": 0.5936,
+      "step": 6804
+    },
+    {
+      "epoch": 8.7104,
+      "grad_norm": 0.8691603541374207,
+      "learning_rate": 3.640456182472989e-05,
+      "loss": 0.605,
+      "step": 6805
+    },
+    {
+      "epoch": 8.71168,
+      "grad_norm": 0.7679282426834106,
+      "learning_rate": 3.6402561024409764e-05,
+      "loss": 0.5439,
+      "step": 6806
+    },
+    {
+      "epoch": 8.71296,
+      "grad_norm": 0.8385738730430603,
+      "learning_rate": 3.6400560224089636e-05,
+      "loss": 0.6184,
+      "step": 6807
+    },
+    {
+      "epoch": 8.71424,
+      "grad_norm": 0.8129993677139282,
+      "learning_rate": 3.639855942376951e-05,
+      "loss": 0.5656,
+      "step": 6808
+    },
+    {
+      "epoch": 8.71552,
+      "grad_norm": 0.7901842594146729,
+      "learning_rate": 3.639655862344939e-05,
+      "loss": 0.573,
+      "step": 6809
+    },
+    {
+      "epoch": 8.7168,
+      "grad_norm": 0.8075759410858154,
+      "learning_rate": 3.639455782312925e-05,
+      "loss": 0.5685,
+      "step": 6810
+    },
+    {
+      "epoch": 8.71808,
+      "grad_norm": 0.7919057011604309,
+      "learning_rate": 3.6392557022809124e-05,
+      "loss": 0.58,
+      "step": 6811
+    },
+    {
+      "epoch": 8.71936,
+      "grad_norm": 0.8212895393371582,
+      "learning_rate": 3.6390556222488995e-05,
+      "loss": 0.5962,
+      "step": 6812
+    },
+    {
+      "epoch": 8.72064,
+      "grad_norm": 0.8056111931800842,
+      "learning_rate": 3.638855542216887e-05,
+      "loss": 0.5887,
+      "step": 6813
+    },
+    {
+      "epoch": 8.72192,
+      "grad_norm": 0.794323742389679,
+      "learning_rate": 3.638655462184874e-05,
+      "loss": 0.5535,
+      "step": 6814
+    },
+    {
+      "epoch": 8.7232,
+      "grad_norm": 0.8696872591972351,
+      "learning_rate": 3.638455382152861e-05,
+      "loss": 0.5984,
+      "step": 6815
+    },
+    {
+      "epoch": 8.72448,
+      "grad_norm": 0.8847552537918091,
+      "learning_rate": 3.638255302120849e-05,
+      "loss": 0.6501,
+      "step": 6816
+    },
+    {
+      "epoch": 8.72576,
+      "grad_norm": 0.8174288272857666,
+      "learning_rate": 3.638055222088836e-05,
+      "loss": 0.6027,
+      "step": 6817
+    },
+    {
+      "epoch": 8.72704,
+      "grad_norm": 0.8479911088943481,
+      "learning_rate": 3.637855142056823e-05,
+      "loss": 0.6216,
+      "step": 6818
+    },
+    {
+      "epoch": 8.72832,
+      "grad_norm": 0.8264278173446655,
+      "learning_rate": 3.63765506202481e-05,
+      "loss": 0.615,
+      "step": 6819
+    },
+    {
+      "epoch": 8.7296,
+      "grad_norm": 0.8488075137138367,
+      "learning_rate": 3.637454981992797e-05,
+      "loss": 0.5891,
+      "step": 6820
+    },
+    {
+      "epoch": 8.730879999999999,
+      "grad_norm": 0.8699356317520142,
+      "learning_rate": 3.637254901960784e-05,
+      "loss": 0.6251,
+      "step": 6821
+    },
+    {
+      "epoch": 8.73216,
+      "grad_norm": 0.8555402755737305,
+      "learning_rate": 3.6370548219287714e-05,
+      "loss": 0.635,
+      "step": 6822
+    },
+    {
+      "epoch": 8.73344,
+      "grad_norm": 0.7735991477966309,
+      "learning_rate": 3.636854741896759e-05,
+      "loss": 0.5566,
+      "step": 6823
+    },
+    {
+      "epoch": 8.73472,
+      "grad_norm": 0.8107903599739075,
+      "learning_rate": 3.6366546618647465e-05,
+      "loss": 0.5747,
+      "step": 6824
+    },
+    {
+      "epoch": 8.736,
+      "grad_norm": 0.7944985032081604,
+      "learning_rate": 3.6364545818327336e-05,
+      "loss": 0.5683,
+      "step": 6825
+    },
+    {
+      "epoch": 8.73728,
+      "grad_norm": 0.8392530083656311,
+      "learning_rate": 3.63625450180072e-05,
+      "loss": 0.5869,
+      "step": 6826
+    },
+    {
+      "epoch": 8.73856,
+      "grad_norm": 0.8549109101295471,
+      "learning_rate": 3.636054421768707e-05,
+      "loss": 0.5996,
+      "step": 6827
+    },
+    {
+      "epoch": 8.739840000000001,
+      "grad_norm": 0.8116100430488586,
+      "learning_rate": 3.6358543417366945e-05,
+      "loss": 0.5885,
+      "step": 6828
+    },
+    {
+      "epoch": 8.74112,
+      "grad_norm": 0.81255042552948,
+      "learning_rate": 3.635654261704682e-05,
+      "loss": 0.5747,
+      "step": 6829
+    },
+    {
+      "epoch": 8.7424,
+      "grad_norm": 0.8412413597106934,
+      "learning_rate": 3.6354541816726696e-05,
+      "loss": 0.595,
+      "step": 6830
+    },
+    {
+      "epoch": 8.74368,
+      "grad_norm": 0.8014543056488037,
+      "learning_rate": 3.635254101640657e-05,
+      "loss": 0.5669,
+      "step": 6831
+    },
+    {
+      "epoch": 8.74496,
+      "grad_norm": 0.7754011750221252,
+      "learning_rate": 3.635054021608644e-05,
+      "loss": 0.5427,
+      "step": 6832
+    },
+    {
+      "epoch": 8.74624,
+      "grad_norm": 0.8052383065223694,
+      "learning_rate": 3.634853941576631e-05,
+      "loss": 0.576,
+      "step": 6833
+    },
+    {
+      "epoch": 8.74752,
+      "grad_norm": 0.8288300633430481,
+      "learning_rate": 3.6346538615446176e-05,
+      "loss": 0.6057,
+      "step": 6834
+    },
+    {
+      "epoch": 8.7488,
+      "grad_norm": 0.8074236512184143,
+      "learning_rate": 3.634453781512605e-05,
+      "loss": 0.5821,
+      "step": 6835
+    },
+    {
+      "epoch": 8.75008,
+      "grad_norm": 0.7861261963844299,
+      "learning_rate": 3.634253701480592e-05,
+      "loss": 0.5665,
+      "step": 6836
+    },
+    {
+      "epoch": 8.75136,
+      "grad_norm": 0.8360334634780884,
+      "learning_rate": 3.63405362144858e-05,
+      "loss": 0.6376,
+      "step": 6837
+    },
+    {
+      "epoch": 8.75264,
+      "grad_norm": 0.7639442086219788,
+      "learning_rate": 3.633853541416567e-05,
+      "loss": 0.5622,
+      "step": 6838
+    },
+    {
+      "epoch": 8.75392,
+      "grad_norm": 0.7803179025650024,
+      "learning_rate": 3.633653461384554e-05,
+      "loss": 0.5281,
+      "step": 6839
+    },
+    {
+      "epoch": 8.7552,
+      "grad_norm": 0.7837743163108826,
+      "learning_rate": 3.6334533813525414e-05,
+      "loss": 0.5903,
+      "step": 6840
+    },
+    {
+      "epoch": 8.75648,
+      "grad_norm": 0.7992691397666931,
+      "learning_rate": 3.6332533013205286e-05,
+      "loss": 0.5786,
+      "step": 6841
+    },
+    {
+      "epoch": 8.75776,
+      "grad_norm": 0.8223332762718201,
+      "learning_rate": 3.633053221288515e-05,
+      "loss": 0.5554,
+      "step": 6842
+    },
+    {
+      "epoch": 8.75904,
+      "grad_norm": 0.8041775226593018,
+      "learning_rate": 3.632853141256502e-05,
+      "loss": 0.5898,
+      "step": 6843
+    },
+    {
+      "epoch": 8.76032,
+      "grad_norm": 0.8420320153236389,
+      "learning_rate": 3.63265306122449e-05,
+      "loss": 0.5924,
+      "step": 6844
+    },
+    {
+      "epoch": 8.7616,
+      "grad_norm": 0.7990514039993286,
+      "learning_rate": 3.6324529811924774e-05,
+      "loss": 0.6265,
+      "step": 6845
+    },
+    {
+      "epoch": 8.76288,
+      "grad_norm": 0.7773391008377075,
+      "learning_rate": 3.6322529011604645e-05,
+      "loss": 0.5407,
+      "step": 6846
+    },
+    {
+      "epoch": 8.76416,
+      "grad_norm": 0.8004972338676453,
+      "learning_rate": 3.632052821128452e-05,
+      "loss": 0.6135,
+      "step": 6847
+    },
+    {
+      "epoch": 8.76544,
+      "grad_norm": 0.7751970887184143,
+      "learning_rate": 3.631852741096439e-05,
+      "loss": 0.5591,
+      "step": 6848
+    },
+    {
+      "epoch": 8.76672,
+      "grad_norm": 0.7655860185623169,
+      "learning_rate": 3.631652661064426e-05,
+      "loss": 0.5736,
+      "step": 6849
+    },
+    {
+      "epoch": 8.768,
+      "grad_norm": 0.8234750628471375,
+      "learning_rate": 3.6314525810324126e-05,
+      "loss": 0.5844,
+      "step": 6850
+    },
+    {
+      "epoch": 8.76928,
+      "grad_norm": 0.8051683306694031,
+      "learning_rate": 3.6312525010004005e-05,
+      "loss": 0.5872,
+      "step": 6851
+    },
+    {
+      "epoch": 8.77056,
+      "grad_norm": 0.9070225954055786,
+      "learning_rate": 3.6310524209683877e-05,
+      "loss": 0.6276,
+      "step": 6852
+    },
+    {
+      "epoch": 8.77184,
+      "grad_norm": 0.8074741959571838,
+      "learning_rate": 3.630852340936375e-05,
+      "loss": 0.5962,
+      "step": 6853
+    },
+    {
+      "epoch": 8.77312,
+      "grad_norm": 0.8040851354598999,
+      "learning_rate": 3.630652260904362e-05,
+      "loss": 0.5674,
+      "step": 6854
+    },
+    {
+      "epoch": 8.7744,
+      "grad_norm": 0.8116294741630554,
+      "learning_rate": 3.630452180872349e-05,
+      "loss": 0.5863,
+      "step": 6855
+    },
+    {
+      "epoch": 8.77568,
+      "grad_norm": 0.9247410297393799,
+      "learning_rate": 3.6302521008403364e-05,
+      "loss": 0.6572,
+      "step": 6856
+    },
+    {
+      "epoch": 8.77696,
+      "grad_norm": 0.8132933378219604,
+      "learning_rate": 3.6300520208083236e-05,
+      "loss": 0.5931,
+      "step": 6857
+    },
+    {
+      "epoch": 8.77824,
+      "grad_norm": 0.8136230707168579,
+      "learning_rate": 3.629851940776311e-05,
+      "loss": 0.579,
+      "step": 6858
+    },
+    {
+      "epoch": 8.77952,
+      "grad_norm": 0.8221474885940552,
+      "learning_rate": 3.629651860744298e-05,
+      "loss": 0.6021,
+      "step": 6859
+    },
+    {
+      "epoch": 8.7808,
+      "grad_norm": 0.8838080763816833,
+      "learning_rate": 3.629451780712285e-05,
+      "loss": 0.5825,
+      "step": 6860
+    },
+    {
+      "epoch": 8.78208,
+      "grad_norm": 0.8534195423126221,
+      "learning_rate": 3.629251700680272e-05,
+      "loss": 0.6262,
+      "step": 6861
+    },
+    {
+      "epoch": 8.78336,
+      "grad_norm": 0.8286737203598022,
+      "learning_rate": 3.6290516206482595e-05,
+      "loss": 0.5569,
+      "step": 6862
+    },
+    {
+      "epoch": 8.78464,
+      "grad_norm": 0.7863860130310059,
+      "learning_rate": 3.628851540616247e-05,
+      "loss": 0.5827,
+      "step": 6863
+    },
+    {
+      "epoch": 8.78592,
+      "grad_norm": 0.8157966136932373,
+      "learning_rate": 3.628651460584234e-05,
+      "loss": 0.6109,
+      "step": 6864
+    },
+    {
+      "epoch": 8.7872,
+      "grad_norm": 0.7991543412208557,
+      "learning_rate": 3.628451380552221e-05,
+      "loss": 0.5703,
+      "step": 6865
+    },
+    {
+      "epoch": 8.78848,
+      "grad_norm": 0.7719929814338684,
+      "learning_rate": 3.628251300520208e-05,
+      "loss": 0.5925,
+      "step": 6866
+    },
+    {
+      "epoch": 8.78976,
+      "grad_norm": 0.8264296054840088,
+      "learning_rate": 3.6280512204881954e-05,
+      "loss": 0.584,
+      "step": 6867
+    },
+    {
+      "epoch": 8.79104,
+      "grad_norm": 0.8640742897987366,
+      "learning_rate": 3.6278511404561826e-05,
+      "loss": 0.6214,
+      "step": 6868
+    },
+    {
+      "epoch": 8.79232,
+      "grad_norm": 0.7543335556983948,
+      "learning_rate": 3.62765106042417e-05,
+      "loss": 0.5316,
+      "step": 6869
+    },
+    {
+      "epoch": 8.7936,
+      "grad_norm": 0.8252272605895996,
+      "learning_rate": 3.627450980392157e-05,
+      "loss": 0.6209,
+      "step": 6870
+    },
+    {
+      "epoch": 8.79488,
+      "grad_norm": 0.7842090129852295,
+      "learning_rate": 3.627250900360144e-05,
+      "loss": 0.5349,
+      "step": 6871
+    },
+    {
+      "epoch": 8.79616,
+      "grad_norm": 0.794221818447113,
+      "learning_rate": 3.6270508203281314e-05,
+      "loss": 0.6018,
+      "step": 6872
+    },
+    {
+      "epoch": 8.79744,
+      "grad_norm": 0.7998900413513184,
+      "learning_rate": 3.6268507402961186e-05,
+      "loss": 0.5694,
+      "step": 6873
+    },
+    {
+      "epoch": 8.79872,
+      "grad_norm": 0.8280250430107117,
+      "learning_rate": 3.626650660264106e-05,
+      "loss": 0.5735,
+      "step": 6874
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.8256487846374512,
+      "learning_rate": 3.626450580232093e-05,
+      "loss": 0.5985,
+      "step": 6875
+    },
+    {
+      "epoch": 8.80128,
+      "grad_norm": 0.86505526304245,
+      "learning_rate": 3.62625050020008e-05,
+      "loss": 0.6511,
+      "step": 6876
+    },
+    {
+      "epoch": 8.80256,
+      "grad_norm": 0.7898805141448975,
+      "learning_rate": 3.626050420168067e-05,
+      "loss": 0.5708,
+      "step": 6877
+    },
+    {
+      "epoch": 8.80384,
+      "grad_norm": 0.8345022797584534,
+      "learning_rate": 3.6258503401360545e-05,
+      "loss": 0.5572,
+      "step": 6878
+    },
+    {
+      "epoch": 8.80512,
+      "grad_norm": 0.8210886120796204,
+      "learning_rate": 3.625650260104042e-05,
+      "loss": 0.581,
+      "step": 6879
+    },
+    {
+      "epoch": 8.8064,
+      "grad_norm": 0.7753137350082397,
+      "learning_rate": 3.625450180072029e-05,
+      "loss": 0.6006,
+      "step": 6880
+    },
+    {
+      "epoch": 8.80768,
+      "grad_norm": 0.8361926078796387,
+      "learning_rate": 3.625250100040016e-05,
+      "loss": 0.5687,
+      "step": 6881
+    },
+    {
+      "epoch": 8.80896,
+      "grad_norm": 0.8121229410171509,
+      "learning_rate": 3.625050020008003e-05,
+      "loss": 0.5764,
+      "step": 6882
+    },
+    {
+      "epoch": 8.81024,
+      "grad_norm": 0.8940417766571045,
+      "learning_rate": 3.6248499399759904e-05,
+      "loss": 0.6222,
+      "step": 6883
+    },
+    {
+      "epoch": 8.81152,
+      "grad_norm": 0.7658796906471252,
+      "learning_rate": 3.6246498599439776e-05,
+      "loss": 0.5626,
+      "step": 6884
+    },
+    {
+      "epoch": 8.8128,
+      "grad_norm": 0.8226578831672668,
+      "learning_rate": 3.624449779911965e-05,
+      "loss": 0.5823,
+      "step": 6885
+    },
+    {
+      "epoch": 8.81408,
+      "grad_norm": 0.8413184881210327,
+      "learning_rate": 3.624249699879952e-05,
+      "loss": 0.6244,
+      "step": 6886
+    },
+    {
+      "epoch": 8.81536,
+      "grad_norm": 0.8740143775939941,
+      "learning_rate": 3.62404961984794e-05,
+      "loss": 0.6014,
+      "step": 6887
+    },
+    {
+      "epoch": 8.81664,
+      "grad_norm": 0.7891989946365356,
+      "learning_rate": 3.6238495398159263e-05,
+      "loss": 0.5616,
+      "step": 6888
+    },
+    {
+      "epoch": 8.81792,
+      "grad_norm": 0.8135474920272827,
+      "learning_rate": 3.6236494597839135e-05,
+      "loss": 0.5674,
+      "step": 6889
+    },
+    {
+      "epoch": 8.8192,
+      "grad_norm": 0.7830061316490173,
+      "learning_rate": 3.623449379751901e-05,
+      "loss": 0.5726,
+      "step": 6890
+    },
+    {
+      "epoch": 8.82048,
+      "grad_norm": 0.8667339086532593,
+      "learning_rate": 3.623249299719888e-05,
+      "loss": 0.5673,
+      "step": 6891
+    },
+    {
+      "epoch": 8.82176,
+      "grad_norm": 0.7781888246536255,
+      "learning_rate": 3.623049219687875e-05,
+      "loss": 0.5379,
+      "step": 6892
+    },
+    {
+      "epoch": 8.82304,
+      "grad_norm": 0.8315311670303345,
+      "learning_rate": 3.622849139655862e-05,
+      "loss": 0.5652,
+      "step": 6893
+    },
+    {
+      "epoch": 8.82432,
+      "grad_norm": 0.8577895164489746,
+      "learning_rate": 3.62264905962385e-05,
+      "loss": 0.631,
+      "step": 6894
+    },
+    {
+      "epoch": 8.8256,
+      "grad_norm": 0.8391826152801514,
+      "learning_rate": 3.622448979591837e-05,
+      "loss": 0.6021,
+      "step": 6895
+    },
+    {
+      "epoch": 8.82688,
+      "grad_norm": 0.8212698698043823,
+      "learning_rate": 3.622248899559824e-05,
+      "loss": 0.6771,
+      "step": 6896
+    },
+    {
+      "epoch": 8.82816,
+      "grad_norm": 0.7723982334136963,
+      "learning_rate": 3.622048819527811e-05,
+      "loss": 0.5674,
+      "step": 6897
+    },
+    {
+      "epoch": 8.82944,
+      "grad_norm": 0.807265043258667,
+      "learning_rate": 3.621848739495798e-05,
+      "loss": 0.6185,
+      "step": 6898
+    },
+    {
+      "epoch": 8.83072,
+      "grad_norm": 0.8213419318199158,
+      "learning_rate": 3.6216486594637854e-05,
+      "loss": 0.6491,
+      "step": 6899
+    },
+    {
+      "epoch": 8.832,
+      "grad_norm": 0.8153632283210754,
+      "learning_rate": 3.6214485794317726e-05,
+      "loss": 0.6082,
+      "step": 6900
+    },
+    {
+      "epoch": 8.83328,
+      "grad_norm": 0.80569988489151,
+      "learning_rate": 3.6212484993997604e-05,
+      "loss": 0.6263,
+      "step": 6901
+    },
+    {
+      "epoch": 8.83456,
+      "grad_norm": 0.8200774192810059,
+      "learning_rate": 3.6210484193677476e-05,
+      "loss": 0.6107,
+      "step": 6902
+    },
+    {
+      "epoch": 8.83584,
+      "grad_norm": 0.783087432384491,
+      "learning_rate": 3.620848339335735e-05,
+      "loss": 0.5873,
+      "step": 6903
+    },
+    {
+      "epoch": 8.83712,
+      "grad_norm": 0.8208640813827515,
+      "learning_rate": 3.620648259303721e-05,
+      "loss": 0.6105,
+      "step": 6904
+    },
+    {
+      "epoch": 8.8384,
+      "grad_norm": 0.8038390278816223,
+      "learning_rate": 3.6204481792717085e-05,
+      "loss": 0.6263,
+      "step": 6905
+    },
+    {
+      "epoch": 8.83968,
+      "grad_norm": 0.7997908592224121,
+      "learning_rate": 3.620248099239696e-05,
+      "loss": 0.631,
+      "step": 6906
+    },
+    {
+      "epoch": 8.84096,
+      "grad_norm": 0.8349906802177429,
+      "learning_rate": 3.620048019207683e-05,
+      "loss": 0.5978,
+      "step": 6907
+    },
+    {
+      "epoch": 8.84224,
+      "grad_norm": 0.8827357292175293,
+      "learning_rate": 3.619847939175671e-05,
+      "loss": 0.6295,
+      "step": 6908
+    },
+    {
+      "epoch": 8.84352,
+      "grad_norm": 0.8000686168670654,
+      "learning_rate": 3.619647859143658e-05,
+      "loss": 0.6055,
+      "step": 6909
+    },
+    {
+      "epoch": 8.8448,
+      "grad_norm": 0.8262300491333008,
+      "learning_rate": 3.619447779111645e-05,
+      "loss": 0.5929,
+      "step": 6910
+    },
+    {
+      "epoch": 8.84608,
+      "grad_norm": 0.8049215078353882,
+      "learning_rate": 3.619247699079632e-05,
+      "loss": 0.5946,
+      "step": 6911
+    },
+    {
+      "epoch": 8.84736,
+      "grad_norm": 0.8154301643371582,
+      "learning_rate": 3.619047619047619e-05,
+      "loss": 0.5801,
+      "step": 6912
+    },
+    {
+      "epoch": 8.84864,
+      "grad_norm": 0.8478993773460388,
+      "learning_rate": 3.618847539015606e-05,
+      "loss": 0.6509,
+      "step": 6913
+    },
+    {
+      "epoch": 8.849920000000001,
+      "grad_norm": 0.8522371649742126,
+      "learning_rate": 3.618647458983593e-05,
+      "loss": 0.5853,
+      "step": 6914
+    },
+    {
+      "epoch": 8.8512,
+      "grad_norm": 0.8328931331634521,
+      "learning_rate": 3.618447378951581e-05,
+      "loss": 0.6368,
+      "step": 6915
+    },
+    {
+      "epoch": 8.85248,
+      "grad_norm": 0.7752423882484436,
+      "learning_rate": 3.618247298919568e-05,
+      "loss": 0.558,
+      "step": 6916
+    },
+    {
+      "epoch": 8.85376,
+      "grad_norm": 0.7981879115104675,
+      "learning_rate": 3.6180472188875554e-05,
+      "loss": 0.5696,
+      "step": 6917
+    },
+    {
+      "epoch": 8.85504,
+      "grad_norm": 0.837726354598999,
+      "learning_rate": 3.6178471388555426e-05,
+      "loss": 0.6013,
+      "step": 6918
+    },
+    {
+      "epoch": 8.85632,
+      "grad_norm": 0.8261789679527283,
+      "learning_rate": 3.61764705882353e-05,
+      "loss": 0.6242,
+      "step": 6919
+    },
+    {
+      "epoch": 8.8576,
+      "grad_norm": 0.8679850697517395,
+      "learning_rate": 3.617446978791516e-05,
+      "loss": 0.6112,
+      "step": 6920
+    },
+    {
+      "epoch": 8.85888,
+      "grad_norm": 0.8656749129295349,
+      "learning_rate": 3.6172468987595035e-05,
+      "loss": 0.6052,
+      "step": 6921
+    },
+    {
+      "epoch": 8.86016,
+      "grad_norm": 0.7777904272079468,
+      "learning_rate": 3.617046818727491e-05,
+      "loss": 0.5814,
+      "step": 6922
+    },
+    {
+      "epoch": 8.86144,
+      "grad_norm": 0.8450097441673279,
+      "learning_rate": 3.6168467386954785e-05,
+      "loss": 0.5964,
+      "step": 6923
+    },
+    {
+      "epoch": 8.86272,
+      "grad_norm": 0.865833580493927,
+      "learning_rate": 3.616646658663466e-05,
+      "loss": 0.6279,
+      "step": 6924
+    },
+    {
+      "epoch": 8.864,
+      "grad_norm": 0.8091690540313721,
+      "learning_rate": 3.616446578631453e-05,
+      "loss": 0.5425,
+      "step": 6925
+    },
+    {
+      "epoch": 8.86528,
+      "grad_norm": 0.8185449838638306,
+      "learning_rate": 3.61624649859944e-05,
+      "loss": 0.606,
+      "step": 6926
+    },
+    {
+      "epoch": 8.86656,
+      "grad_norm": 0.8065361976623535,
+      "learning_rate": 3.616046418567427e-05,
+      "loss": 0.5938,
+      "step": 6927
+    },
+    {
+      "epoch": 8.86784,
+      "grad_norm": 0.8017386198043823,
+      "learning_rate": 3.615846338535414e-05,
+      "loss": 0.5727,
+      "step": 6928
+    },
+    {
+      "epoch": 8.86912,
+      "grad_norm": 0.8515564799308777,
+      "learning_rate": 3.6156462585034016e-05,
+      "loss": 0.6308,
+      "step": 6929
+    },
+    {
+      "epoch": 8.8704,
+      "grad_norm": 0.837009608745575,
+      "learning_rate": 3.615446178471389e-05,
+      "loss": 0.6016,
+      "step": 6930
+    },
+    {
+      "epoch": 8.87168,
+      "grad_norm": 0.8977416157722473,
+      "learning_rate": 3.615246098439376e-05,
+      "loss": 0.6323,
+      "step": 6931
+    },
+    {
+      "epoch": 8.872959999999999,
+      "grad_norm": 0.8525516390800476,
+      "learning_rate": 3.615046018407363e-05,
+      "loss": 0.627,
+      "step": 6932
+    },
+    {
+      "epoch": 8.87424,
+      "grad_norm": 0.838636577129364,
+      "learning_rate": 3.6148459383753504e-05,
+      "loss": 0.6056,
+      "step": 6933
+    },
+    {
+      "epoch": 8.87552,
+      "grad_norm": 0.8404485583305359,
+      "learning_rate": 3.6146458583433376e-05,
+      "loss": 0.6425,
+      "step": 6934
+    },
+    {
+      "epoch": 8.8768,
+      "grad_norm": 0.8780763745307922,
+      "learning_rate": 3.614445778311325e-05,
+      "loss": 0.677,
+      "step": 6935
+    },
+    {
+      "epoch": 8.87808,
+      "grad_norm": 0.7928383350372314,
+      "learning_rate": 3.614245698279312e-05,
+      "loss": 0.584,
+      "step": 6936
+    },
+    {
+      "epoch": 8.87936,
+      "grad_norm": 0.8395443558692932,
+      "learning_rate": 3.614045618247299e-05,
+      "loss": 0.6165,
+      "step": 6937
+    },
+    {
+      "epoch": 8.88064,
+      "grad_norm": 0.8272031545639038,
+      "learning_rate": 3.613845538215286e-05,
+      "loss": 0.5808,
+      "step": 6938
+    },
+    {
+      "epoch": 8.881920000000001,
+      "grad_norm": 0.8448715806007385,
+      "learning_rate": 3.6136454581832735e-05,
+      "loss": 0.6293,
+      "step": 6939
+    },
+    {
+      "epoch": 8.8832,
+      "grad_norm": 0.8742750287055969,
+      "learning_rate": 3.613445378151261e-05,
+      "loss": 0.5869,
+      "step": 6940
+    },
+    {
+      "epoch": 8.88448,
+      "grad_norm": 0.7730932235717773,
+      "learning_rate": 3.613245298119248e-05,
+      "loss": 0.5513,
+      "step": 6941
+    },
+    {
+      "epoch": 8.88576,
+      "grad_norm": 0.83644700050354,
+      "learning_rate": 3.613045218087235e-05,
+      "loss": 0.6132,
+      "step": 6942
+    },
+    {
+      "epoch": 8.88704,
+      "grad_norm": 0.8572961688041687,
+      "learning_rate": 3.612845138055222e-05,
+      "loss": 0.6198,
+      "step": 6943
+    },
+    {
+      "epoch": 8.88832,
+      "grad_norm": 0.8267669081687927,
+      "learning_rate": 3.6126450580232094e-05,
+      "loss": 0.5856,
+      "step": 6944
+    },
+    {
+      "epoch": 8.8896,
+      "grad_norm": 0.7987276315689087,
+      "learning_rate": 3.6124449779911966e-05,
+      "loss": 0.5852,
+      "step": 6945
+    },
+    {
+      "epoch": 8.89088,
+      "grad_norm": 0.8039761781692505,
+      "learning_rate": 3.612244897959184e-05,
+      "loss": 0.5562,
+      "step": 6946
+    },
+    {
+      "epoch": 8.89216,
+      "grad_norm": 0.8594339489936829,
+      "learning_rate": 3.612044817927171e-05,
+      "loss": 0.5928,
+      "step": 6947
+    },
+    {
+      "epoch": 8.89344,
+      "grad_norm": 0.7759758234024048,
+      "learning_rate": 3.611844737895158e-05,
+      "loss": 0.5363,
+      "step": 6948
+    },
+    {
+      "epoch": 8.89472,
+      "grad_norm": 0.7741097211837769,
+      "learning_rate": 3.6116446578631453e-05,
+      "loss": 0.5417,
+      "step": 6949
+    },
+    {
+      "epoch": 8.896,
+      "grad_norm": 0.800906777381897,
+      "learning_rate": 3.611444577831133e-05,
+      "loss": 0.5737,
+      "step": 6950
+    },
+    {
+      "epoch": 8.89728,
+      "grad_norm": 0.8281027674674988,
+      "learning_rate": 3.61124449779912e-05,
+      "loss": 0.5661,
+      "step": 6951
+    },
+    {
+      "epoch": 8.89856,
+      "grad_norm": 0.7662954926490784,
+      "learning_rate": 3.611044417767107e-05,
+      "loss": 0.559,
+      "step": 6952
+    },
+    {
+      "epoch": 8.89984,
+      "grad_norm": 0.8090856671333313,
+      "learning_rate": 3.610844337735094e-05,
+      "loss": 0.5599,
+      "step": 6953
+    },
+    {
+      "epoch": 8.90112,
+      "grad_norm": 0.8325416445732117,
+      "learning_rate": 3.610644257703081e-05,
+      "loss": 0.5671,
+      "step": 6954
+    },
+    {
+      "epoch": 8.9024,
+      "grad_norm": 0.8077730536460876,
+      "learning_rate": 3.6104441776710685e-05,
+      "loss": 0.6039,
+      "step": 6955
+    },
+    {
+      "epoch": 8.90368,
+      "grad_norm": 0.8278130888938904,
+      "learning_rate": 3.6102440976390556e-05,
+      "loss": 0.5771,
+      "step": 6956
+    },
+    {
+      "epoch": 8.904959999999999,
+      "grad_norm": 0.7988932132720947,
+      "learning_rate": 3.6100440176070435e-05,
+      "loss": 0.6022,
+      "step": 6957
+    },
+    {
+      "epoch": 8.90624,
+      "grad_norm": 0.8389017581939697,
+      "learning_rate": 3.609843937575031e-05,
+      "loss": 0.6001,
+      "step": 6958
+    },
+    {
+      "epoch": 8.90752,
+      "grad_norm": 0.896360456943512,
+      "learning_rate": 3.609643857543017e-05,
+      "loss": 0.6543,
+      "step": 6959
+    },
+    {
+      "epoch": 8.9088,
+      "grad_norm": 0.8242748379707336,
+      "learning_rate": 3.6094437775110044e-05,
+      "loss": 0.5962,
+      "step": 6960
+    },
+    {
+      "epoch": 8.91008,
+      "grad_norm": 0.8101686835289001,
+      "learning_rate": 3.6092436974789916e-05,
+      "loss": 0.6084,
+      "step": 6961
+    },
+    {
+      "epoch": 8.91136,
+      "grad_norm": 0.8416367769241333,
+      "learning_rate": 3.609043617446979e-05,
+      "loss": 0.5865,
+      "step": 6962
+    },
+    {
+      "epoch": 8.91264,
+      "grad_norm": 0.8185626864433289,
+      "learning_rate": 3.608843537414966e-05,
+      "loss": 0.6199,
+      "step": 6963
+    },
+    {
+      "epoch": 8.91392,
+      "grad_norm": 0.8309803009033203,
+      "learning_rate": 3.608643457382954e-05,
+      "loss": 0.6395,
+      "step": 6964
+    },
+    {
+      "epoch": 8.9152,
+      "grad_norm": 0.7879054546356201,
+      "learning_rate": 3.608443377350941e-05,
+      "loss": 0.5667,
+      "step": 6965
+    },
+    {
+      "epoch": 8.91648,
+      "grad_norm": 0.8321192860603333,
+      "learning_rate": 3.608243297318928e-05,
+      "loss": 0.6035,
+      "step": 6966
+    },
+    {
+      "epoch": 8.91776,
+      "grad_norm": 0.8073901534080505,
+      "learning_rate": 3.608043217286915e-05,
+      "loss": 0.5635,
+      "step": 6967
+    },
+    {
+      "epoch": 8.91904,
+      "grad_norm": 0.805634081363678,
+      "learning_rate": 3.607843137254902e-05,
+      "loss": 0.5788,
+      "step": 6968
+    },
+    {
+      "epoch": 8.92032,
+      "grad_norm": 0.7749924063682556,
+      "learning_rate": 3.607643057222889e-05,
+      "loss": 0.5697,
+      "step": 6969
+    },
+    {
+      "epoch": 8.9216,
+      "grad_norm": 0.7843794226646423,
+      "learning_rate": 3.607442977190876e-05,
+      "loss": 0.5763,
+      "step": 6970
+    },
+    {
+      "epoch": 8.92288,
+      "grad_norm": 0.8274489641189575,
+      "learning_rate": 3.607242897158864e-05,
+      "loss": 0.5843,
+      "step": 6971
+    },
+    {
+      "epoch": 8.92416,
+      "grad_norm": 0.834439754486084,
+      "learning_rate": 3.607042817126851e-05,
+      "loss": 0.5806,
+      "step": 6972
+    },
+    {
+      "epoch": 8.92544,
+      "grad_norm": 0.8098054528236389,
+      "learning_rate": 3.6068427370948385e-05,
+      "loss": 0.5754,
+      "step": 6973
+    },
+    {
+      "epoch": 8.92672,
+      "grad_norm": 0.8164894580841064,
+      "learning_rate": 3.606642657062826e-05,
+      "loss": 0.6191,
+      "step": 6974
+    },
+    {
+      "epoch": 8.928,
+      "grad_norm": 0.7646953463554382,
+      "learning_rate": 3.606442577030812e-05,
+      "loss": 0.5733,
+      "step": 6975
+    },
+    {
+      "epoch": 8.92928,
+      "grad_norm": 0.7695797085762024,
+      "learning_rate": 3.6062424969987994e-05,
+      "loss": 0.5586,
+      "step": 6976
+    },
+    {
+      "epoch": 8.93056,
+      "grad_norm": 0.8515230417251587,
+      "learning_rate": 3.6060424169667865e-05,
+      "loss": 0.6997,
+      "step": 6977
+    },
+    {
+      "epoch": 8.93184,
+      "grad_norm": 0.7551551461219788,
+      "learning_rate": 3.6058423369347744e-05,
+      "loss": 0.5588,
+      "step": 6978
+    },
+    {
+      "epoch": 8.93312,
+      "grad_norm": 0.8263706564903259,
+      "learning_rate": 3.6056422569027616e-05,
+      "loss": 0.6269,
+      "step": 6979
+    },
+    {
+      "epoch": 8.9344,
+      "grad_norm": 0.819164514541626,
+      "learning_rate": 3.605442176870749e-05,
+      "loss": 0.5782,
+      "step": 6980
+    },
+    {
+      "epoch": 8.93568,
+      "grad_norm": 0.875464916229248,
+      "learning_rate": 3.605242096838736e-05,
+      "loss": 0.6204,
+      "step": 6981
+    },
+    {
+      "epoch": 8.93696,
+      "grad_norm": 0.807060718536377,
+      "learning_rate": 3.605042016806723e-05,
+      "loss": 0.5681,
+      "step": 6982
+    },
+    {
+      "epoch": 8.93824,
+      "grad_norm": 0.824044942855835,
+      "learning_rate": 3.6048419367747097e-05,
+      "loss": 0.6358,
+      "step": 6983
+    },
+    {
+      "epoch": 8.93952,
+      "grad_norm": 0.7487238645553589,
+      "learning_rate": 3.604641856742697e-05,
+      "loss": 0.5667,
+      "step": 6984
+    },
+    {
+      "epoch": 8.9408,
+      "grad_norm": 0.8306474685668945,
+      "learning_rate": 3.604441776710684e-05,
+      "loss": 0.5704,
+      "step": 6985
+    },
+    {
+      "epoch": 8.94208,
+      "grad_norm": 0.7934936285018921,
+      "learning_rate": 3.604241696678672e-05,
+      "loss": 0.5734,
+      "step": 6986
+    },
+    {
+      "epoch": 8.94336,
+      "grad_norm": 0.8018743991851807,
+      "learning_rate": 3.604041616646659e-05,
+      "loss": 0.5555,
+      "step": 6987
+    },
+    {
+      "epoch": 8.94464,
+      "grad_norm": 0.8124516606330872,
+      "learning_rate": 3.603841536614646e-05,
+      "loss": 0.5972,
+      "step": 6988
+    },
+    {
+      "epoch": 8.94592,
+      "grad_norm": 0.8345234394073486,
+      "learning_rate": 3.6036414565826334e-05,
+      "loss": 0.6231,
+      "step": 6989
+    },
+    {
+      "epoch": 8.9472,
+      "grad_norm": 0.8359727263450623,
+      "learning_rate": 3.6034413765506206e-05,
+      "loss": 0.6152,
+      "step": 6990
+    },
+    {
+      "epoch": 8.94848,
+      "grad_norm": 0.7563595771789551,
+      "learning_rate": 3.603241296518607e-05,
+      "loss": 0.5544,
+      "step": 6991
+    },
+    {
+      "epoch": 8.94976,
+      "grad_norm": 0.8067274689674377,
+      "learning_rate": 3.603041216486594e-05,
+      "loss": 0.5755,
+      "step": 6992
+    },
+    {
+      "epoch": 8.95104,
+      "grad_norm": 0.8166893720626831,
+      "learning_rate": 3.602841136454582e-05,
+      "loss": 0.6486,
+      "step": 6993
+    },
+    {
+      "epoch": 8.95232,
+      "grad_norm": 0.7941920161247253,
+      "learning_rate": 3.6026410564225694e-05,
+      "loss": 0.6155,
+      "step": 6994
+    },
+    {
+      "epoch": 8.9536,
+      "grad_norm": 0.7989978790283203,
+      "learning_rate": 3.6024409763905566e-05,
+      "loss": 0.5686,
+      "step": 6995
+    },
+    {
+      "epoch": 8.95488,
+      "grad_norm": 0.8766310214996338,
+      "learning_rate": 3.602240896358544e-05,
+      "loss": 0.6189,
+      "step": 6996
+    },
+    {
+      "epoch": 8.95616,
+      "grad_norm": 0.8384754061698914,
+      "learning_rate": 3.602040816326531e-05,
+      "loss": 0.6313,
+      "step": 6997
+    },
+    {
+      "epoch": 8.95744,
+      "grad_norm": 0.8599010705947876,
+      "learning_rate": 3.601840736294518e-05,
+      "loss": 0.6274,
+      "step": 6998
+    },
+    {
+      "epoch": 8.95872,
+      "grad_norm": 0.8240278363227844,
+      "learning_rate": 3.6016406562625046e-05,
+      "loss": 0.6005,
+      "step": 6999
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.8268716335296631,
+      "learning_rate": 3.6014405762304925e-05,
+      "loss": 0.6159,
+      "step": 7000
+    },
+    {
+      "epoch": 8.96128,
+      "grad_norm": 0.8283771872520447,
+      "learning_rate": 3.60124049619848e-05,
+      "loss": 0.6447,
+      "step": 7001
+    },
+    {
+      "epoch": 8.96256,
+      "grad_norm": 0.8202353119850159,
+      "learning_rate": 3.601040416166467e-05,
+      "loss": 0.5554,
+      "step": 7002
+    },
+    {
+      "epoch": 8.96384,
+      "grad_norm": 0.8381731510162354,
+      "learning_rate": 3.600840336134454e-05,
+      "loss": 0.6646,
+      "step": 7003
+    },
+    {
+      "epoch": 8.96512,
+      "grad_norm": 0.8112981915473938,
+      "learning_rate": 3.600640256102441e-05,
+      "loss": 0.5944,
+      "step": 7004
+    },
+    {
+      "epoch": 8.9664,
+      "grad_norm": 0.8326748013496399,
+      "learning_rate": 3.6004401760704284e-05,
+      "loss": 0.6004,
+      "step": 7005
+    },
+    {
+      "epoch": 8.96768,
+      "grad_norm": 0.8164408206939697,
+      "learning_rate": 3.6002400960384156e-05,
+      "loss": 0.5891,
+      "step": 7006
+    },
+    {
+      "epoch": 8.96896,
+      "grad_norm": 0.7985873818397522,
+      "learning_rate": 3.600040016006403e-05,
+      "loss": 0.5516,
+      "step": 7007
+    },
+    {
+      "epoch": 8.97024,
+      "grad_norm": 0.8269280195236206,
+      "learning_rate": 3.59983993597439e-05,
+      "loss": 0.5673,
+      "step": 7008
+    },
+    {
+      "epoch": 8.97152,
+      "grad_norm": 0.7575619220733643,
+      "learning_rate": 3.599639855942377e-05,
+      "loss": 0.5324,
+      "step": 7009
+    },
+    {
+      "epoch": 8.9728,
+      "grad_norm": 0.8653445243835449,
+      "learning_rate": 3.5994397759103643e-05,
+      "loss": 0.6027,
+      "step": 7010
+    },
+    {
+      "epoch": 8.97408,
+      "grad_norm": 0.8530913591384888,
+      "learning_rate": 3.5992396958783515e-05,
+      "loss": 0.6428,
+      "step": 7011
+    },
+    {
+      "epoch": 8.97536,
+      "grad_norm": 0.8212682604789734,
+      "learning_rate": 3.599039615846339e-05,
+      "loss": 0.5799,
+      "step": 7012
+    },
+    {
+      "epoch": 8.97664,
+      "grad_norm": 0.8439937233924866,
+      "learning_rate": 3.598839535814326e-05,
+      "loss": 0.6195,
+      "step": 7013
+    },
+    {
+      "epoch": 8.97792,
+      "grad_norm": 0.8312199711799622,
+      "learning_rate": 3.598639455782313e-05,
+      "loss": 0.5808,
+      "step": 7014
+    },
+    {
+      "epoch": 8.9792,
+      "grad_norm": 0.8415351510047913,
+      "learning_rate": 3.5984393757503e-05,
+      "loss": 0.6143,
+      "step": 7015
+    },
+    {
+      "epoch": 8.98048,
+      "grad_norm": 0.8534009456634521,
+      "learning_rate": 3.5982392957182875e-05,
+      "loss": 0.6133,
+      "step": 7016
+    },
+    {
+      "epoch": 8.98176,
+      "grad_norm": 0.819942057132721,
+      "learning_rate": 3.5980392156862746e-05,
+      "loss": 0.62,
+      "step": 7017
+    },
+    {
+      "epoch": 8.98304,
+      "grad_norm": 0.8175826668739319,
+      "learning_rate": 3.597839135654262e-05,
+      "loss": 0.5802,
+      "step": 7018
+    },
+    {
+      "epoch": 8.98432,
+      "grad_norm": 0.8342677354812622,
+      "learning_rate": 3.597639055622249e-05,
+      "loss": 0.5873,
+      "step": 7019
+    },
+    {
+      "epoch": 8.9856,
+      "grad_norm": 0.7667478919029236,
+      "learning_rate": 3.597438975590236e-05,
+      "loss": 0.557,
+      "step": 7020
+    },
+    {
+      "epoch": 8.98688,
+      "grad_norm": 0.7987193465232849,
+      "learning_rate": 3.5972388955582234e-05,
+      "loss": 0.565,
+      "step": 7021
+    },
+    {
+      "epoch": 8.98816,
+      "grad_norm": 0.8105794787406921,
+      "learning_rate": 3.5970388155262106e-05,
+      "loss": 0.6025,
+      "step": 7022
+    },
+    {
+      "epoch": 8.98944,
+      "grad_norm": 0.8362847566604614,
+      "learning_rate": 3.596838735494198e-05,
+      "loss": 0.5946,
+      "step": 7023
+    },
+    {
+      "epoch": 8.99072,
+      "grad_norm": 0.7905120253562927,
+      "learning_rate": 3.596638655462185e-05,
+      "loss": 0.5964,
+      "step": 7024
+    },
+    {
+      "epoch": 8.992,
+      "grad_norm": 0.8283672332763672,
+      "learning_rate": 3.596438575430172e-05,
+      "loss": 0.596,
+      "step": 7025
+    },
+    {
+      "epoch": 8.99328,
+      "grad_norm": 0.8208228945732117,
+      "learning_rate": 3.596238495398159e-05,
+      "loss": 0.6088,
+      "step": 7026
+    },
+    {
+      "epoch": 8.99456,
+      "grad_norm": 0.7455883026123047,
+      "learning_rate": 3.5960384153661465e-05,
+      "loss": 0.5508,
+      "step": 7027
+    },
+    {
+      "epoch": 8.99584,
+      "grad_norm": 0.8352407217025757,
+      "learning_rate": 3.5958383353341344e-05,
+      "loss": 0.628,
+      "step": 7028
+    },
+    {
+      "epoch": 8.99712,
+      "grad_norm": 0.8079842925071716,
+      "learning_rate": 3.595638255302121e-05,
+      "loss": 0.596,
+      "step": 7029
+    },
+    {
+      "epoch": 8.9984,
+      "grad_norm": 0.8588855862617493,
+      "learning_rate": 3.595438175270108e-05,
+      "loss": 0.6203,
+      "step": 7030
+    },
+    {
+      "epoch": 8.99968,
+      "grad_norm": 0.8400841355323792,
+      "learning_rate": 3.595238095238095e-05,
+      "loss": 0.5663,
+      "step": 7031
+    },
+    {
+      "epoch": 9.00096,
+      "grad_norm": 1.7468935251235962,
+      "learning_rate": 3.5950380152060824e-05,
+      "loss": 1.0437,
+      "step": 7032
+    },
+    {
+      "epoch": 9.00224,
+      "grad_norm": 0.7875801920890808,
+      "learning_rate": 3.5948379351740696e-05,
+      "loss": 0.556,
+      "step": 7033
+    },
+    {
+      "epoch": 9.00352,
+      "grad_norm": 0.8308430910110474,
+      "learning_rate": 3.594637855142057e-05,
+      "loss": 0.5821,
+      "step": 7034
+    },
+    {
+      "epoch": 9.0048,
+      "grad_norm": 0.7866562008857727,
+      "learning_rate": 3.594437775110045e-05,
+      "loss": 0.529,
+      "step": 7035
+    },
+    {
+      "epoch": 9.00608,
+      "grad_norm": 0.8387976884841919,
+      "learning_rate": 3.594237695078032e-05,
+      "loss": 0.5738,
+      "step": 7036
+    },
+    {
+      "epoch": 9.00736,
+      "grad_norm": 0.8457949757575989,
+      "learning_rate": 3.5940376150460184e-05,
+      "loss": 0.5637,
+      "step": 7037
+    },
+    {
+      "epoch": 9.00864,
+      "grad_norm": 0.8400401473045349,
+      "learning_rate": 3.5938375350140055e-05,
+      "loss": 0.6053,
+      "step": 7038
+    },
+    {
+      "epoch": 9.00992,
+      "grad_norm": 0.8000286221504211,
+      "learning_rate": 3.593637454981993e-05,
+      "loss": 0.5456,
+      "step": 7039
+    },
+    {
+      "epoch": 9.0112,
+      "grad_norm": 0.8358906507492065,
+      "learning_rate": 3.59343737494998e-05,
+      "loss": 0.5632,
+      "step": 7040
+    },
+    {
+      "epoch": 9.01248,
+      "grad_norm": 0.8368620872497559,
+      "learning_rate": 3.593237294917967e-05,
+      "loss": 0.5464,
+      "step": 7041
+    },
+    {
+      "epoch": 9.01376,
+      "grad_norm": 0.8233303427696228,
+      "learning_rate": 3.593037214885955e-05,
+      "loss": 0.5641,
+      "step": 7042
+    },
+    {
+      "epoch": 9.01504,
+      "grad_norm": 0.8248091340065002,
+      "learning_rate": 3.592837134853942e-05,
+      "loss": 0.5642,
+      "step": 7043
+    },
+    {
+      "epoch": 9.01632,
+      "grad_norm": 0.7816410660743713,
+      "learning_rate": 3.592637054821929e-05,
+      "loss": 0.557,
+      "step": 7044
+    },
+    {
+      "epoch": 9.0176,
+      "grad_norm": 0.8148441314697266,
+      "learning_rate": 3.592436974789916e-05,
+      "loss": 0.5754,
+      "step": 7045
+    },
+    {
+      "epoch": 9.01888,
+      "grad_norm": 0.846493661403656,
+      "learning_rate": 3.592236894757903e-05,
+      "loss": 0.5328,
+      "step": 7046
+    },
+    {
+      "epoch": 9.02016,
+      "grad_norm": 0.7609879374504089,
+      "learning_rate": 3.59203681472589e-05,
+      "loss": 0.5445,
+      "step": 7047
+    },
+    {
+      "epoch": 9.02144,
+      "grad_norm": 0.7646995186805725,
+      "learning_rate": 3.5918367346938774e-05,
+      "loss": 0.5248,
+      "step": 7048
+    },
+    {
+      "epoch": 9.02272,
+      "grad_norm": 0.9177650809288025,
+      "learning_rate": 3.591636654661865e-05,
+      "loss": 0.6154,
+      "step": 7049
+    },
+    {
+      "epoch": 9.024,
+      "grad_norm": 0.8574452996253967,
+      "learning_rate": 3.5914365746298525e-05,
+      "loss": 0.5763,
+      "step": 7050
+    },
+    {
+      "epoch": 9.02528,
+      "grad_norm": 0.845577597618103,
+      "learning_rate": 3.5912364945978396e-05,
+      "loss": 0.5505,
+      "step": 7051
+    },
+    {
+      "epoch": 9.02656,
+      "grad_norm": 0.8447266817092896,
+      "learning_rate": 3.591036414565827e-05,
+      "loss": 0.5625,
+      "step": 7052
+    },
+    {
+      "epoch": 9.02784,
+      "grad_norm": 0.8364947438240051,
+      "learning_rate": 3.590836334533813e-05,
+      "loss": 0.6222,
+      "step": 7053
+    },
+    {
+      "epoch": 9.02912,
+      "grad_norm": 0.7808526158332825,
+      "learning_rate": 3.5906362545018005e-05,
+      "loss": 0.5635,
+      "step": 7054
+    },
+    {
+      "epoch": 9.0304,
+      "grad_norm": 0.8134317994117737,
+      "learning_rate": 3.590436174469788e-05,
+      "loss": 0.5632,
+      "step": 7055
+    },
+    {
+      "epoch": 9.03168,
+      "grad_norm": 0.833970844745636,
+      "learning_rate": 3.5902360944377756e-05,
+      "loss": 0.5108,
+      "step": 7056
+    },
+    {
+      "epoch": 9.03296,
+      "grad_norm": 0.8089005947113037,
+      "learning_rate": 3.590036014405763e-05,
+      "loss": 0.548,
+      "step": 7057
+    },
+    {
+      "epoch": 9.03424,
+      "grad_norm": 0.8224313855171204,
+      "learning_rate": 3.58983593437375e-05,
+      "loss": 0.5738,
+      "step": 7058
+    },
+    {
+      "epoch": 9.03552,
+      "grad_norm": 0.8225283026695251,
+      "learning_rate": 3.589635854341737e-05,
+      "loss": 0.5791,
+      "step": 7059
+    },
+    {
+      "epoch": 9.0368,
+      "grad_norm": 0.8561593890190125,
+      "learning_rate": 3.589435774309724e-05,
+      "loss": 0.6185,
+      "step": 7060
+    },
+    {
+      "epoch": 9.03808,
+      "grad_norm": 0.803023636341095,
+      "learning_rate": 3.589235694277711e-05,
+      "loss": 0.5655,
+      "step": 7061
+    },
+    {
+      "epoch": 9.03936,
+      "grad_norm": 0.8006875514984131,
+      "learning_rate": 3.589035614245698e-05,
+      "loss": 0.5293,
+      "step": 7062
+    },
+    {
+      "epoch": 9.04064,
+      "grad_norm": 0.7813231348991394,
+      "learning_rate": 3.588835534213686e-05,
+      "loss": 0.5739,
+      "step": 7063
+    },
+    {
+      "epoch": 9.04192,
+      "grad_norm": 0.7644641995429993,
+      "learning_rate": 3.588635454181673e-05,
+      "loss": 0.524,
+      "step": 7064
+    },
+    {
+      "epoch": 9.0432,
+      "grad_norm": 0.8366073369979858,
+      "learning_rate": 3.58843537414966e-05,
+      "loss": 0.5816,
+      "step": 7065
+    },
+    {
+      "epoch": 9.04448,
+      "grad_norm": 0.8413146138191223,
+      "learning_rate": 3.5882352941176474e-05,
+      "loss": 0.575,
+      "step": 7066
+    },
+    {
+      "epoch": 9.04576,
+      "grad_norm": 0.868990957736969,
+      "learning_rate": 3.5880352140856346e-05,
+      "loss": 0.5821,
+      "step": 7067
+    },
+    {
+      "epoch": 9.04704,
+      "grad_norm": 0.8294727206230164,
+      "learning_rate": 3.587835134053622e-05,
+      "loss": 0.5557,
+      "step": 7068
+    },
+    {
+      "epoch": 9.04832,
+      "grad_norm": 0.8696639537811279,
+      "learning_rate": 3.587635054021608e-05,
+      "loss": 0.6255,
+      "step": 7069
+    },
+    {
+      "epoch": 9.0496,
+      "grad_norm": 0.8515434861183167,
+      "learning_rate": 3.587434973989596e-05,
+      "loss": 0.625,
+      "step": 7070
+    },
+    {
+      "epoch": 9.05088,
+      "grad_norm": 0.8410640358924866,
+      "learning_rate": 3.5872348939575834e-05,
+      "loss": 0.6161,
+      "step": 7071
+    },
+    {
+      "epoch": 9.05216,
+      "grad_norm": 0.8619793653488159,
+      "learning_rate": 3.5870348139255705e-05,
+      "loss": 0.5977,
+      "step": 7072
+    },
+    {
+      "epoch": 9.05344,
+      "grad_norm": 0.8171403408050537,
+      "learning_rate": 3.586834733893558e-05,
+      "loss": 0.5569,
+      "step": 7073
+    },
+    {
+      "epoch": 9.05472,
+      "grad_norm": 0.823085367679596,
+      "learning_rate": 3.586634653861545e-05,
+      "loss": 0.5687,
+      "step": 7074
+    },
+    {
+      "epoch": 9.056,
+      "grad_norm": 0.8016168475151062,
+      "learning_rate": 3.586434573829532e-05,
+      "loss": 0.5622,
+      "step": 7075
+    },
+    {
+      "epoch": 9.05728,
+      "grad_norm": 0.8238281607627869,
+      "learning_rate": 3.586234493797519e-05,
+      "loss": 0.5507,
+      "step": 7076
+    },
+    {
+      "epoch": 9.05856,
+      "grad_norm": 0.859733521938324,
+      "learning_rate": 3.5860344137655065e-05,
+      "loss": 0.581,
+      "step": 7077
+    },
+    {
+      "epoch": 9.05984,
+      "grad_norm": 0.8221707940101624,
+      "learning_rate": 3.5858343337334937e-05,
+      "loss": 0.5824,
+      "step": 7078
+    },
+    {
+      "epoch": 9.06112,
+      "grad_norm": 0.818545401096344,
+      "learning_rate": 3.585634253701481e-05,
+      "loss": 0.5601,
+      "step": 7079
+    },
+    {
+      "epoch": 9.0624,
+      "grad_norm": 0.8852583765983582,
+      "learning_rate": 3.585434173669468e-05,
+      "loss": 0.584,
+      "step": 7080
+    },
+    {
+      "epoch": 9.06368,
+      "grad_norm": 0.8168684244155884,
+      "learning_rate": 3.585234093637455e-05,
+      "loss": 0.5746,
+      "step": 7081
+    },
+    {
+      "epoch": 9.06496,
+      "grad_norm": 0.77977454662323,
+      "learning_rate": 3.5850340136054424e-05,
+      "loss": 0.5343,
+      "step": 7082
+    },
+    {
+      "epoch": 9.06624,
+      "grad_norm": 0.8777265548706055,
+      "learning_rate": 3.5848339335734296e-05,
+      "loss": 0.6393,
+      "step": 7083
+    },
+    {
+      "epoch": 9.06752,
+      "grad_norm": 0.8020537495613098,
+      "learning_rate": 3.584633853541417e-05,
+      "loss": 0.5444,
+      "step": 7084
+    },
+    {
+      "epoch": 9.0688,
+      "grad_norm": 0.7844178080558777,
+      "learning_rate": 3.584433773509404e-05,
+      "loss": 0.5409,
+      "step": 7085
+    },
+    {
+      "epoch": 9.07008,
+      "grad_norm": 0.8328168392181396,
+      "learning_rate": 3.584233693477391e-05,
+      "loss": 0.5712,
+      "step": 7086
+    },
+    {
+      "epoch": 9.07136,
+      "grad_norm": 0.8295521140098572,
+      "learning_rate": 3.584033613445378e-05,
+      "loss": 0.5724,
+      "step": 7087
+    },
+    {
+      "epoch": 9.07264,
+      "grad_norm": 0.8165435791015625,
+      "learning_rate": 3.5838335334133655e-05,
+      "loss": 0.5741,
+      "step": 7088
+    },
+    {
+      "epoch": 9.07392,
+      "grad_norm": 0.7974668145179749,
+      "learning_rate": 3.583633453381353e-05,
+      "loss": 0.5751,
+      "step": 7089
+    },
+    {
+      "epoch": 9.0752,
+      "grad_norm": 0.8819810152053833,
+      "learning_rate": 3.58343337334934e-05,
+      "loss": 0.5696,
+      "step": 7090
+    },
+    {
+      "epoch": 9.07648,
+      "grad_norm": 0.8275978565216064,
+      "learning_rate": 3.583233293317327e-05,
+      "loss": 0.5667,
+      "step": 7091
+    },
+    {
+      "epoch": 9.07776,
+      "grad_norm": 0.8336565494537354,
+      "learning_rate": 3.583033213285314e-05,
+      "loss": 0.5487,
+      "step": 7092
+    },
+    {
+      "epoch": 9.079039999999999,
+      "grad_norm": 0.8257220983505249,
+      "learning_rate": 3.5828331332533014e-05,
+      "loss": 0.5878,
+      "step": 7093
+    },
+    {
+      "epoch": 9.08032,
+      "grad_norm": 0.8719106316566467,
+      "learning_rate": 3.5826330532212886e-05,
+      "loss": 0.5884,
+      "step": 7094
+    },
+    {
+      "epoch": 9.0816,
+      "grad_norm": 0.822557270526886,
+      "learning_rate": 3.582432973189276e-05,
+      "loss": 0.5495,
+      "step": 7095
+    },
+    {
+      "epoch": 9.08288,
+      "grad_norm": 0.8632369637489319,
+      "learning_rate": 3.582232893157263e-05,
+      "loss": 0.631,
+      "step": 7096
+    },
+    {
+      "epoch": 9.08416,
+      "grad_norm": 0.7848061919212341,
+      "learning_rate": 3.58203281312525e-05,
+      "loss": 0.5497,
+      "step": 7097
+    },
+    {
+      "epoch": 9.08544,
+      "grad_norm": 0.8202498555183411,
+      "learning_rate": 3.5818327330932374e-05,
+      "loss": 0.5686,
+      "step": 7098
+    },
+    {
+      "epoch": 9.08672,
+      "grad_norm": 0.8682575225830078,
+      "learning_rate": 3.5816326530612245e-05,
+      "loss": 0.5808,
+      "step": 7099
+    },
+    {
+      "epoch": 9.088,
+      "grad_norm": 0.8932950496673584,
+      "learning_rate": 3.581432573029212e-05,
+      "loss": 0.5941,
+      "step": 7100
+    },
+    {
+      "epoch": 9.08928,
+      "grad_norm": 0.8102815747261047,
+      "learning_rate": 3.581232492997199e-05,
+      "loss": 0.5704,
+      "step": 7101
+    },
+    {
+      "epoch": 9.09056,
+      "grad_norm": 0.848307192325592,
+      "learning_rate": 3.581032412965186e-05,
+      "loss": 0.5954,
+      "step": 7102
+    },
+    {
+      "epoch": 9.09184,
+      "grad_norm": 0.8251248598098755,
+      "learning_rate": 3.580832332933173e-05,
+      "loss": 0.599,
+      "step": 7103
+    },
+    {
+      "epoch": 9.09312,
+      "grad_norm": 0.8306258916854858,
+      "learning_rate": 3.5806322529011605e-05,
+      "loss": 0.5746,
+      "step": 7104
+    },
+    {
+      "epoch": 9.0944,
+      "grad_norm": 0.8358504176139832,
+      "learning_rate": 3.580432172869148e-05,
+      "loss": 0.5811,
+      "step": 7105
+    },
+    {
+      "epoch": 9.09568,
+      "grad_norm": 0.8394206762313843,
+      "learning_rate": 3.5802320928371355e-05,
+      "loss": 0.5633,
+      "step": 7106
+    },
+    {
+      "epoch": 9.09696,
+      "grad_norm": 0.7698180079460144,
+      "learning_rate": 3.580032012805122e-05,
+      "loss": 0.5287,
+      "step": 7107
+    },
+    {
+      "epoch": 9.09824,
+      "grad_norm": 0.8235006928443909,
+      "learning_rate": 3.579831932773109e-05,
+      "loss": 0.575,
+      "step": 7108
+    },
+    {
+      "epoch": 9.09952,
+      "grad_norm": 0.8552541136741638,
+      "learning_rate": 3.5796318527410964e-05,
+      "loss": 0.5888,
+      "step": 7109
+    },
+    {
+      "epoch": 9.1008,
+      "grad_norm": 0.7659354209899902,
+      "learning_rate": 3.5794317727090836e-05,
+      "loss": 0.5523,
+      "step": 7110
+    },
+    {
+      "epoch": 9.10208,
+      "grad_norm": 0.8290457129478455,
+      "learning_rate": 3.579231692677071e-05,
+      "loss": 0.5812,
+      "step": 7111
+    },
+    {
+      "epoch": 9.10336,
+      "grad_norm": 0.8734808564186096,
+      "learning_rate": 3.579031612645058e-05,
+      "loss": 0.5864,
+      "step": 7112
+    },
+    {
+      "epoch": 9.10464,
+      "grad_norm": 0.8360257148742676,
+      "learning_rate": 3.578831532613046e-05,
+      "loss": 0.6079,
+      "step": 7113
+    },
+    {
+      "epoch": 9.10592,
+      "grad_norm": 0.8263709545135498,
+      "learning_rate": 3.578631452581033e-05,
+      "loss": 0.6076,
+      "step": 7114
+    },
+    {
+      "epoch": 9.1072,
+      "grad_norm": 0.8358266949653625,
+      "learning_rate": 3.5784313725490195e-05,
+      "loss": 0.5644,
+      "step": 7115
+    },
+    {
+      "epoch": 9.10848,
+      "grad_norm": 0.9020827412605286,
+      "learning_rate": 3.578231292517007e-05,
+      "loss": 0.6079,
+      "step": 7116
+    },
+    {
+      "epoch": 9.10976,
+      "grad_norm": 0.8776836395263672,
+      "learning_rate": 3.578031212484994e-05,
+      "loss": 0.5905,
+      "step": 7117
+    },
+    {
+      "epoch": 9.11104,
+      "grad_norm": 0.8753827214241028,
+      "learning_rate": 3.577831132452981e-05,
+      "loss": 0.5962,
+      "step": 7118
+    },
+    {
+      "epoch": 9.11232,
+      "grad_norm": 0.8712731599807739,
+      "learning_rate": 3.577631052420968e-05,
+      "loss": 0.5731,
+      "step": 7119
+    },
+    {
+      "epoch": 9.1136,
+      "grad_norm": 0.8524981141090393,
+      "learning_rate": 3.577430972388956e-05,
+      "loss": 0.5583,
+      "step": 7120
+    },
+    {
+      "epoch": 9.11488,
+      "grad_norm": 0.8630044460296631,
+      "learning_rate": 3.577230892356943e-05,
+      "loss": 0.5547,
+      "step": 7121
+    },
+    {
+      "epoch": 9.11616,
+      "grad_norm": 0.7815329432487488,
+      "learning_rate": 3.5770308123249305e-05,
+      "loss": 0.5294,
+      "step": 7122
+    },
+    {
+      "epoch": 9.11744,
+      "grad_norm": 0.8376869559288025,
+      "learning_rate": 3.576830732292917e-05,
+      "loss": 0.599,
+      "step": 7123
+    },
+    {
+      "epoch": 9.11872,
+      "grad_norm": 0.8631175756454468,
+      "learning_rate": 3.576630652260904e-05,
+      "loss": 0.5887,
+      "step": 7124
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 0.8864824175834656,
+      "learning_rate": 3.5764305722288914e-05,
+      "loss": 0.571,
+      "step": 7125
+    },
+    {
+      "epoch": 9.12128,
+      "grad_norm": 0.8389583230018616,
+      "learning_rate": 3.5762304921968786e-05,
+      "loss": 0.5336,
+      "step": 7126
+    },
+    {
+      "epoch": 9.12256,
+      "grad_norm": 0.838886022567749,
+      "learning_rate": 3.5760304121648664e-05,
+      "loss": 0.5679,
+      "step": 7127
+    },
+    {
+      "epoch": 9.12384,
+      "grad_norm": 0.850913941860199,
+      "learning_rate": 3.5758303321328536e-05,
+      "loss": 0.5632,
+      "step": 7128
+    },
+    {
+      "epoch": 9.12512,
+      "grad_norm": 0.8992368578910828,
+      "learning_rate": 3.575630252100841e-05,
+      "loss": 0.598,
+      "step": 7129
+    },
+    {
+      "epoch": 9.1264,
+      "grad_norm": 0.8634032011032104,
+      "learning_rate": 3.575430172068828e-05,
+      "loss": 0.6198,
+      "step": 7130
+    },
+    {
+      "epoch": 9.12768,
+      "grad_norm": 0.8508159518241882,
+      "learning_rate": 3.5752300920368145e-05,
+      "loss": 0.5757,
+      "step": 7131
+    },
+    {
+      "epoch": 9.12896,
+      "grad_norm": 0.8230028748512268,
+      "learning_rate": 3.575030012004802e-05,
+      "loss": 0.5853,
+      "step": 7132
+    },
+    {
+      "epoch": 9.13024,
+      "grad_norm": 0.8228382468223572,
+      "learning_rate": 3.574829931972789e-05,
+      "loss": 0.5304,
+      "step": 7133
+    },
+    {
+      "epoch": 9.13152,
+      "grad_norm": 0.8506357669830322,
+      "learning_rate": 3.574629851940777e-05,
+      "loss": 0.5977,
+      "step": 7134
+    },
+    {
+      "epoch": 9.1328,
+      "grad_norm": 0.8166884779930115,
+      "learning_rate": 3.574429771908764e-05,
+      "loss": 0.5696,
+      "step": 7135
+    },
+    {
+      "epoch": 9.13408,
+      "grad_norm": 0.7951632142066956,
+      "learning_rate": 3.574229691876751e-05,
+      "loss": 0.5402,
+      "step": 7136
+    },
+    {
+      "epoch": 9.13536,
+      "grad_norm": 0.8789847493171692,
+      "learning_rate": 3.574029611844738e-05,
+      "loss": 0.5845,
+      "step": 7137
+    },
+    {
+      "epoch": 9.13664,
+      "grad_norm": 0.8529048562049866,
+      "learning_rate": 3.5738295318127255e-05,
+      "loss": 0.584,
+      "step": 7138
+    },
+    {
+      "epoch": 9.13792,
+      "grad_norm": 0.8400517702102661,
+      "learning_rate": 3.573629451780712e-05,
+      "loss": 0.585,
+      "step": 7139
+    },
+    {
+      "epoch": 9.1392,
+      "grad_norm": 0.8045387268066406,
+      "learning_rate": 3.573429371748699e-05,
+      "loss": 0.576,
+      "step": 7140
+    },
+    {
+      "epoch": 9.14048,
+      "grad_norm": 0.8386081457138062,
+      "learning_rate": 3.573229291716687e-05,
+      "loss": 0.6016,
+      "step": 7141
+    },
+    {
+      "epoch": 9.14176,
+      "grad_norm": 0.8390840291976929,
+      "learning_rate": 3.573029211684674e-05,
+      "loss": 0.5581,
+      "step": 7142
+    },
+    {
+      "epoch": 9.14304,
+      "grad_norm": 0.8972001671791077,
+      "learning_rate": 3.5728291316526614e-05,
+      "loss": 0.593,
+      "step": 7143
+    },
+    {
+      "epoch": 9.14432,
+      "grad_norm": 0.8241517543792725,
+      "learning_rate": 3.5726290516206486e-05,
+      "loss": 0.5142,
+      "step": 7144
+    },
+    {
+      "epoch": 9.1456,
+      "grad_norm": 0.8427436947822571,
+      "learning_rate": 3.572428971588636e-05,
+      "loss": 0.6082,
+      "step": 7145
+    },
+    {
+      "epoch": 9.14688,
+      "grad_norm": 0.8868518471717834,
+      "learning_rate": 3.572228891556623e-05,
+      "loss": 0.5601,
+      "step": 7146
+    },
+    {
+      "epoch": 9.14816,
+      "grad_norm": 0.8616020679473877,
+      "learning_rate": 3.5720288115246095e-05,
+      "loss": 0.589,
+      "step": 7147
+    },
+    {
+      "epoch": 9.14944,
+      "grad_norm": 0.8238210082054138,
+      "learning_rate": 3.571828731492597e-05,
+      "loss": 0.5611,
+      "step": 7148
+    },
+    {
+      "epoch": 9.15072,
+      "grad_norm": 0.7928246259689331,
+      "learning_rate": 3.5716286514605845e-05,
+      "loss": 0.5399,
+      "step": 7149
+    },
+    {
+      "epoch": 9.152,
+      "grad_norm": 0.8270377516746521,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.5782,
+      "step": 7150
+    },
+    {
+      "epoch": 9.15328,
+      "grad_norm": 0.8852272033691406,
+      "learning_rate": 3.571228491396559e-05,
+      "loss": 0.5984,
+      "step": 7151
+    },
+    {
+      "epoch": 9.15456,
+      "grad_norm": 0.8285141587257385,
+      "learning_rate": 3.571028411364546e-05,
+      "loss": 0.5448,
+      "step": 7152
+    },
+    {
+      "epoch": 9.15584,
+      "grad_norm": 0.8800240159034729,
+      "learning_rate": 3.570828331332533e-05,
+      "loss": 0.6208,
+      "step": 7153
+    },
+    {
+      "epoch": 9.15712,
+      "grad_norm": 0.8323574662208557,
+      "learning_rate": 3.5706282513005204e-05,
+      "loss": 0.5727,
+      "step": 7154
+    },
+    {
+      "epoch": 9.1584,
+      "grad_norm": 0.798857569694519,
+      "learning_rate": 3.5704281712685076e-05,
+      "loss": 0.5719,
+      "step": 7155
+    },
+    {
+      "epoch": 9.15968,
+      "grad_norm": 0.8088046312332153,
+      "learning_rate": 3.570228091236495e-05,
+      "loss": 0.545,
+      "step": 7156
+    },
+    {
+      "epoch": 9.16096,
+      "grad_norm": 0.8132175207138062,
+      "learning_rate": 3.570028011204482e-05,
+      "loss": 0.5349,
+      "step": 7157
+    },
+    {
+      "epoch": 9.16224,
+      "grad_norm": 0.8076733946800232,
+      "learning_rate": 3.569827931172469e-05,
+      "loss": 0.549,
+      "step": 7158
+    },
+    {
+      "epoch": 9.16352,
+      "grad_norm": 0.8983336687088013,
+      "learning_rate": 3.5696278511404564e-05,
+      "loss": 0.5902,
+      "step": 7159
+    },
+    {
+      "epoch": 9.1648,
+      "grad_norm": 0.8289273381233215,
+      "learning_rate": 3.5694277711084436e-05,
+      "loss": 0.5484,
+      "step": 7160
+    },
+    {
+      "epoch": 9.166080000000001,
+      "grad_norm": 0.8771588206291199,
+      "learning_rate": 3.569227691076431e-05,
+      "loss": 0.5911,
+      "step": 7161
+    },
+    {
+      "epoch": 9.16736,
+      "grad_norm": 0.8836865425109863,
+      "learning_rate": 3.569027611044418e-05,
+      "loss": 0.5916,
+      "step": 7162
+    },
+    {
+      "epoch": 9.16864,
+      "grad_norm": 0.8692944049835205,
+      "learning_rate": 3.568827531012405e-05,
+      "loss": 0.6283,
+      "step": 7163
+    },
+    {
+      "epoch": 9.16992,
+      "grad_norm": 0.8750563263893127,
+      "learning_rate": 3.568627450980392e-05,
+      "loss": 0.5961,
+      "step": 7164
+    },
+    {
+      "epoch": 9.1712,
+      "grad_norm": 0.9316475987434387,
+      "learning_rate": 3.5684273709483795e-05,
+      "loss": 0.6031,
+      "step": 7165
+    },
+    {
+      "epoch": 9.17248,
+      "grad_norm": 0.9066860675811768,
+      "learning_rate": 3.568227290916367e-05,
+      "loss": 0.5978,
+      "step": 7166
+    },
+    {
+      "epoch": 9.17376,
+      "grad_norm": 0.8408805727958679,
+      "learning_rate": 3.568027210884354e-05,
+      "loss": 0.5614,
+      "step": 7167
+    },
+    {
+      "epoch": 9.17504,
+      "grad_norm": 0.9087552428245544,
+      "learning_rate": 3.567827130852341e-05,
+      "loss": 0.6356,
+      "step": 7168
+    },
+    {
+      "epoch": 9.17632,
+      "grad_norm": 0.8625233769416809,
+      "learning_rate": 3.567627050820328e-05,
+      "loss": 0.5493,
+      "step": 7169
+    },
+    {
+      "epoch": 9.1776,
+      "grad_norm": 0.9364339709281921,
+      "learning_rate": 3.5674269707883154e-05,
+      "loss": 0.5999,
+      "step": 7170
+    },
+    {
+      "epoch": 9.17888,
+      "grad_norm": 0.895353376865387,
+      "learning_rate": 3.5672268907563026e-05,
+      "loss": 0.5865,
+      "step": 7171
+    },
+    {
+      "epoch": 9.18016,
+      "grad_norm": 0.8568089008331299,
+      "learning_rate": 3.56702681072429e-05,
+      "loss": 0.5691,
+      "step": 7172
+    },
+    {
+      "epoch": 9.18144,
+      "grad_norm": 0.9034045934677124,
+      "learning_rate": 3.566826730692277e-05,
+      "loss": 0.6174,
+      "step": 7173
+    },
+    {
+      "epoch": 9.18272,
+      "grad_norm": 0.8543916344642639,
+      "learning_rate": 3.566626650660264e-05,
+      "loss": 0.5903,
+      "step": 7174
+    },
+    {
+      "epoch": 9.184,
+      "grad_norm": 0.7896568775177002,
+      "learning_rate": 3.566426570628251e-05,
+      "loss": 0.5389,
+      "step": 7175
+    },
+    {
+      "epoch": 9.18528,
+      "grad_norm": 0.8801631927490234,
+      "learning_rate": 3.566226490596239e-05,
+      "loss": 0.5925,
+      "step": 7176
+    },
+    {
+      "epoch": 9.18656,
+      "grad_norm": 0.8208610415458679,
+      "learning_rate": 3.566026410564226e-05,
+      "loss": 0.5339,
+      "step": 7177
+    },
+    {
+      "epoch": 9.18784,
+      "grad_norm": 0.8739574551582336,
+      "learning_rate": 3.565826330532213e-05,
+      "loss": 0.5946,
+      "step": 7178
+    },
+    {
+      "epoch": 9.18912,
+      "grad_norm": 0.8522581458091736,
+      "learning_rate": 3.5656262505002e-05,
+      "loss": 0.604,
+      "step": 7179
+    },
+    {
+      "epoch": 9.1904,
+      "grad_norm": 0.8527406454086304,
+      "learning_rate": 3.565426170468187e-05,
+      "loss": 0.5349,
+      "step": 7180
+    },
+    {
+      "epoch": 9.19168,
+      "grad_norm": 0.796159029006958,
+      "learning_rate": 3.5652260904361745e-05,
+      "loss": 0.5852,
+      "step": 7181
+    },
+    {
+      "epoch": 9.19296,
+      "grad_norm": 0.8322233557701111,
+      "learning_rate": 3.5650260104041616e-05,
+      "loss": 0.5491,
+      "step": 7182
+    },
+    {
+      "epoch": 9.19424,
+      "grad_norm": 0.8693201541900635,
+      "learning_rate": 3.5648259303721495e-05,
+      "loss": 0.6308,
+      "step": 7183
+    },
+    {
+      "epoch": 9.19552,
+      "grad_norm": 0.8879520297050476,
+      "learning_rate": 3.564625850340137e-05,
+      "loss": 0.5455,
+      "step": 7184
+    },
+    {
+      "epoch": 9.1968,
+      "grad_norm": 0.8827615976333618,
+      "learning_rate": 3.564425770308123e-05,
+      "loss": 0.5712,
+      "step": 7185
+    },
+    {
+      "epoch": 9.19808,
+      "grad_norm": 0.8380709886550903,
+      "learning_rate": 3.5642256902761104e-05,
+      "loss": 0.5884,
+      "step": 7186
+    },
+    {
+      "epoch": 9.19936,
+      "grad_norm": 0.8361532092094421,
+      "learning_rate": 3.5640256102440976e-05,
+      "loss": 0.5282,
+      "step": 7187
+    },
+    {
+      "epoch": 9.20064,
+      "grad_norm": 0.8010978102684021,
+      "learning_rate": 3.563825530212085e-05,
+      "loss": 0.5372,
+      "step": 7188
+    },
+    {
+      "epoch": 9.20192,
+      "grad_norm": 0.8207326531410217,
+      "learning_rate": 3.563625450180072e-05,
+      "loss": 0.5496,
+      "step": 7189
+    },
+    {
+      "epoch": 9.2032,
+      "grad_norm": 0.9393238425254822,
+      "learning_rate": 3.56342537014806e-05,
+      "loss": 0.6691,
+      "step": 7190
+    },
+    {
+      "epoch": 9.20448,
+      "grad_norm": 0.9167293310165405,
+      "learning_rate": 3.563225290116047e-05,
+      "loss": 0.577,
+      "step": 7191
+    },
+    {
+      "epoch": 9.20576,
+      "grad_norm": 0.8757715821266174,
+      "learning_rate": 3.563025210084034e-05,
+      "loss": 0.56,
+      "step": 7192
+    },
+    {
+      "epoch": 9.20704,
+      "grad_norm": 0.8311654329299927,
+      "learning_rate": 3.562825130052021e-05,
+      "loss": 0.5641,
+      "step": 7193
+    },
+    {
+      "epoch": 9.20832,
+      "grad_norm": 0.8445587754249573,
+      "learning_rate": 3.562625050020008e-05,
+      "loss": 0.5876,
+      "step": 7194
+    },
+    {
+      "epoch": 9.2096,
+      "grad_norm": 0.805767834186554,
+      "learning_rate": 3.562424969987995e-05,
+      "loss": 0.5486,
+      "step": 7195
+    },
+    {
+      "epoch": 9.21088,
+      "grad_norm": 0.8612878918647766,
+      "learning_rate": 3.562224889955982e-05,
+      "loss": 0.5913,
+      "step": 7196
+    },
+    {
+      "epoch": 9.21216,
+      "grad_norm": 0.7950903177261353,
+      "learning_rate": 3.56202480992397e-05,
+      "loss": 0.5245,
+      "step": 7197
+    },
+    {
+      "epoch": 9.21344,
+      "grad_norm": 0.8570892214775085,
+      "learning_rate": 3.561824729891957e-05,
+      "loss": 0.5713,
+      "step": 7198
+    },
+    {
+      "epoch": 9.21472,
+      "grad_norm": 0.8819652795791626,
+      "learning_rate": 3.5616246498599445e-05,
+      "loss": 0.6082,
+      "step": 7199
+    },
+    {
+      "epoch": 9.216,
+      "grad_norm": 0.8352935910224915,
+      "learning_rate": 3.5614245698279317e-05,
+      "loss": 0.5274,
+      "step": 7200
+    },
+    {
+      "epoch": 9.21728,
+      "grad_norm": 0.8423967957496643,
+      "learning_rate": 3.561224489795918e-05,
+      "loss": 0.5569,
+      "step": 7201
+    },
+    {
+      "epoch": 9.21856,
+      "grad_norm": 0.8680324554443359,
+      "learning_rate": 3.5610244097639054e-05,
+      "loss": 0.5971,
+      "step": 7202
+    },
+    {
+      "epoch": 9.21984,
+      "grad_norm": 0.8493099808692932,
+      "learning_rate": 3.5608243297318925e-05,
+      "loss": 0.5817,
+      "step": 7203
+    },
+    {
+      "epoch": 9.22112,
+      "grad_norm": 0.8210243582725525,
+      "learning_rate": 3.5606242496998804e-05,
+      "loss": 0.5623,
+      "step": 7204
+    },
+    {
+      "epoch": 9.2224,
+      "grad_norm": 0.8306945562362671,
+      "learning_rate": 3.5604241696678676e-05,
+      "loss": 0.5609,
+      "step": 7205
+    },
+    {
+      "epoch": 9.22368,
+      "grad_norm": 0.8076730370521545,
+      "learning_rate": 3.560224089635855e-05,
+      "loss": 0.5814,
+      "step": 7206
+    },
+    {
+      "epoch": 9.22496,
+      "grad_norm": 0.8856160640716553,
+      "learning_rate": 3.560024009603842e-05,
+      "loss": 0.614,
+      "step": 7207
+    },
+    {
+      "epoch": 9.22624,
+      "grad_norm": 0.8176262974739075,
+      "learning_rate": 3.559823929571829e-05,
+      "loss": 0.5156,
+      "step": 7208
+    },
+    {
+      "epoch": 9.22752,
+      "grad_norm": 0.8237152099609375,
+      "learning_rate": 3.5596238495398157e-05,
+      "loss": 0.6039,
+      "step": 7209
+    },
+    {
+      "epoch": 9.2288,
+      "grad_norm": 0.8996145129203796,
+      "learning_rate": 3.559423769507803e-05,
+      "loss": 0.5799,
+      "step": 7210
+    },
+    {
+      "epoch": 9.23008,
+      "grad_norm": 0.8621578216552734,
+      "learning_rate": 3.55922368947579e-05,
+      "loss": 0.5983,
+      "step": 7211
+    },
+    {
+      "epoch": 9.23136,
+      "grad_norm": 0.834199845790863,
+      "learning_rate": 3.559023609443778e-05,
+      "loss": 0.584,
+      "step": 7212
+    },
+    {
+      "epoch": 9.23264,
+      "grad_norm": 0.8422051072120667,
+      "learning_rate": 3.558823529411765e-05,
+      "loss": 0.6015,
+      "step": 7213
+    },
+    {
+      "epoch": 9.23392,
+      "grad_norm": 0.8835949301719666,
+      "learning_rate": 3.558623449379752e-05,
+      "loss": 0.6181,
+      "step": 7214
+    },
+    {
+      "epoch": 9.2352,
+      "grad_norm": 0.8236482739448547,
+      "learning_rate": 3.5584233693477394e-05,
+      "loss": 0.5114,
+      "step": 7215
+    },
+    {
+      "epoch": 9.23648,
+      "grad_norm": 0.8392600417137146,
+      "learning_rate": 3.5582232893157266e-05,
+      "loss": 0.5606,
+      "step": 7216
+    },
+    {
+      "epoch": 9.23776,
+      "grad_norm": 0.8492880463600159,
+      "learning_rate": 3.558023209283713e-05,
+      "loss": 0.5595,
+      "step": 7217
+    },
+    {
+      "epoch": 9.23904,
+      "grad_norm": 0.8469056487083435,
+      "learning_rate": 3.5578231292517e-05,
+      "loss": 0.5619,
+      "step": 7218
+    },
+    {
+      "epoch": 9.24032,
+      "grad_norm": 0.8622716069221497,
+      "learning_rate": 3.557623049219688e-05,
+      "loss": 0.5764,
+      "step": 7219
+    },
+    {
+      "epoch": 9.2416,
+      "grad_norm": 0.8697147369384766,
+      "learning_rate": 3.5574229691876754e-05,
+      "loss": 0.6487,
+      "step": 7220
+    },
+    {
+      "epoch": 9.24288,
+      "grad_norm": 0.8444170355796814,
+      "learning_rate": 3.5572228891556626e-05,
+      "loss": 0.5976,
+      "step": 7221
+    },
+    {
+      "epoch": 9.24416,
+      "grad_norm": 0.9171527028083801,
+      "learning_rate": 3.55702280912365e-05,
+      "loss": 0.5975,
+      "step": 7222
+    },
+    {
+      "epoch": 9.24544,
+      "grad_norm": 0.8756519556045532,
+      "learning_rate": 3.556822729091637e-05,
+      "loss": 0.6635,
+      "step": 7223
+    },
+    {
+      "epoch": 9.24672,
+      "grad_norm": 0.8335146903991699,
+      "learning_rate": 3.556622649059624e-05,
+      "loss": 0.5707,
+      "step": 7224
+    },
+    {
+      "epoch": 9.248,
+      "grad_norm": 0.8439629673957825,
+      "learning_rate": 3.5564225690276106e-05,
+      "loss": 0.5443,
+      "step": 7225
+    },
+    {
+      "epoch": 9.24928,
+      "grad_norm": 0.8886339664459229,
+      "learning_rate": 3.5562224889955985e-05,
+      "loss": 0.6433,
+      "step": 7226
+    },
+    {
+      "epoch": 9.25056,
+      "grad_norm": 0.8498263359069824,
+      "learning_rate": 3.556022408963586e-05,
+      "loss": 0.5432,
+      "step": 7227
+    },
+    {
+      "epoch": 9.25184,
+      "grad_norm": 0.8150326609611511,
+      "learning_rate": 3.555822328931573e-05,
+      "loss": 0.5489,
+      "step": 7228
+    },
+    {
+      "epoch": 9.25312,
+      "grad_norm": 0.8487164974212646,
+      "learning_rate": 3.55562224889956e-05,
+      "loss": 0.6123,
+      "step": 7229
+    },
+    {
+      "epoch": 9.2544,
+      "grad_norm": 0.8371261358261108,
+      "learning_rate": 3.555422168867547e-05,
+      "loss": 0.586,
+      "step": 7230
+    },
+    {
+      "epoch": 9.25568,
+      "grad_norm": 0.7906522154808044,
+      "learning_rate": 3.5552220888355344e-05,
+      "loss": 0.5475,
+      "step": 7231
+    },
+    {
+      "epoch": 9.25696,
+      "grad_norm": 0.8361690640449524,
+      "learning_rate": 3.5550220088035216e-05,
+      "loss": 0.5587,
+      "step": 7232
+    },
+    {
+      "epoch": 9.25824,
+      "grad_norm": 0.8770825266838074,
+      "learning_rate": 3.554821928771509e-05,
+      "loss": 0.5292,
+      "step": 7233
+    },
+    {
+      "epoch": 9.25952,
+      "grad_norm": 0.8710092306137085,
+      "learning_rate": 3.554621848739496e-05,
+      "loss": 0.5879,
+      "step": 7234
+    },
+    {
+      "epoch": 9.2608,
+      "grad_norm": 0.9107195734977722,
+      "learning_rate": 3.554421768707483e-05,
+      "loss": 0.6258,
+      "step": 7235
+    },
+    {
+      "epoch": 9.26208,
+      "grad_norm": 0.8514677286148071,
+      "learning_rate": 3.5542216886754703e-05,
+      "loss": 0.5595,
+      "step": 7236
+    },
+    {
+      "epoch": 9.26336,
+      "grad_norm": 0.806038498878479,
+      "learning_rate": 3.5540216086434575e-05,
+      "loss": 0.5441,
+      "step": 7237
+    },
+    {
+      "epoch": 9.26464,
+      "grad_norm": 0.8175530433654785,
+      "learning_rate": 3.553821528611445e-05,
+      "loss": 0.5597,
+      "step": 7238
+    },
+    {
+      "epoch": 9.26592,
+      "grad_norm": 0.8429561257362366,
+      "learning_rate": 3.553621448579432e-05,
+      "loss": 0.6029,
+      "step": 7239
+    },
+    {
+      "epoch": 9.2672,
+      "grad_norm": 0.7659432888031006,
+      "learning_rate": 3.553421368547419e-05,
+      "loss": 0.4918,
+      "step": 7240
+    },
+    {
+      "epoch": 9.26848,
+      "grad_norm": 0.8414883613586426,
+      "learning_rate": 3.553221288515406e-05,
+      "loss": 0.5624,
+      "step": 7241
+    },
+    {
+      "epoch": 9.26976,
+      "grad_norm": 0.8417079448699951,
+      "learning_rate": 3.5530212084833935e-05,
+      "loss": 0.5585,
+      "step": 7242
+    },
+    {
+      "epoch": 9.27104,
+      "grad_norm": 0.8089073300361633,
+      "learning_rate": 3.5528211284513806e-05,
+      "loss": 0.5623,
+      "step": 7243
+    },
+    {
+      "epoch": 9.27232,
+      "grad_norm": 0.8453049063682556,
+      "learning_rate": 3.552621048419368e-05,
+      "loss": 0.5725,
+      "step": 7244
+    },
+    {
+      "epoch": 9.2736,
+      "grad_norm": 0.8788660168647766,
+      "learning_rate": 3.552420968387355e-05,
+      "loss": 0.6243,
+      "step": 7245
+    },
+    {
+      "epoch": 9.27488,
+      "grad_norm": 0.8520131707191467,
+      "learning_rate": 3.552220888355342e-05,
+      "loss": 0.5707,
+      "step": 7246
+    },
+    {
+      "epoch": 9.27616,
+      "grad_norm": 0.8138136863708496,
+      "learning_rate": 3.55202080832333e-05,
+      "loss": 0.5788,
+      "step": 7247
+    },
+    {
+      "epoch": 9.27744,
+      "grad_norm": 0.8127238750457764,
+      "learning_rate": 3.5518207282913166e-05,
+      "loss": 0.5925,
+      "step": 7248
+    },
+    {
+      "epoch": 9.27872,
+      "grad_norm": 0.7981357574462891,
+      "learning_rate": 3.551620648259304e-05,
+      "loss": 0.5755,
+      "step": 7249
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 0.8368809223175049,
+      "learning_rate": 3.551420568227291e-05,
+      "loss": 0.5962,
+      "step": 7250
+    },
+    {
+      "epoch": 9.28128,
+      "grad_norm": 0.832737386226654,
+      "learning_rate": 3.551220488195278e-05,
+      "loss": 0.5504,
+      "step": 7251
+    },
+    {
+      "epoch": 9.28256,
+      "grad_norm": 0.9122929573059082,
+      "learning_rate": 3.551020408163265e-05,
+      "loss": 0.6593,
+      "step": 7252
+    },
+    {
+      "epoch": 9.28384,
+      "grad_norm": 0.8156841397285461,
+      "learning_rate": 3.5508203281312525e-05,
+      "loss": 0.5478,
+      "step": 7253
+    },
+    {
+      "epoch": 9.28512,
+      "grad_norm": 0.8393061757087708,
+      "learning_rate": 3.5506202480992404e-05,
+      "loss": 0.5744,
+      "step": 7254
+    },
+    {
+      "epoch": 9.2864,
+      "grad_norm": 0.8752142786979675,
+      "learning_rate": 3.5504201680672275e-05,
+      "loss": 0.6099,
+      "step": 7255
+    },
+    {
+      "epoch": 9.28768,
+      "grad_norm": 0.8940780162811279,
+      "learning_rate": 3.550220088035214e-05,
+      "loss": 0.6174,
+      "step": 7256
+    },
+    {
+      "epoch": 9.28896,
+      "grad_norm": 0.8913589119911194,
+      "learning_rate": 3.550020008003201e-05,
+      "loss": 0.5515,
+      "step": 7257
+    },
+    {
+      "epoch": 9.29024,
+      "grad_norm": 0.8833445310592651,
+      "learning_rate": 3.5498199279711884e-05,
+      "loss": 0.6179,
+      "step": 7258
+    },
+    {
+      "epoch": 9.29152,
+      "grad_norm": 0.8273428678512573,
+      "learning_rate": 3.5496198479391756e-05,
+      "loss": 0.5577,
+      "step": 7259
+    },
+    {
+      "epoch": 9.2928,
+      "grad_norm": 0.8162719011306763,
+      "learning_rate": 3.549419767907163e-05,
+      "loss": 0.5771,
+      "step": 7260
+    },
+    {
+      "epoch": 9.29408,
+      "grad_norm": 0.846421480178833,
+      "learning_rate": 3.549219687875151e-05,
+      "loss": 0.5642,
+      "step": 7261
+    },
+    {
+      "epoch": 9.29536,
+      "grad_norm": 0.8472880125045776,
+      "learning_rate": 3.549019607843138e-05,
+      "loss": 0.6061,
+      "step": 7262
+    },
+    {
+      "epoch": 9.29664,
+      "grad_norm": 0.8349382877349854,
+      "learning_rate": 3.548819527811125e-05,
+      "loss": 0.6095,
+      "step": 7263
+    },
+    {
+      "epoch": 9.29792,
+      "grad_norm": 0.8894979953765869,
+      "learning_rate": 3.5486194477791115e-05,
+      "loss": 0.6172,
+      "step": 7264
+    },
+    {
+      "epoch": 9.2992,
+      "grad_norm": 0.8571387529373169,
+      "learning_rate": 3.548419367747099e-05,
+      "loss": 0.5449,
+      "step": 7265
+    },
+    {
+      "epoch": 9.30048,
+      "grad_norm": 0.898956835269928,
+      "learning_rate": 3.548219287715086e-05,
+      "loss": 0.64,
+      "step": 7266
+    },
+    {
+      "epoch": 9.30176,
+      "grad_norm": 0.8103107213973999,
+      "learning_rate": 3.548019207683073e-05,
+      "loss": 0.5698,
+      "step": 7267
+    },
+    {
+      "epoch": 9.30304,
+      "grad_norm": 0.8312423825263977,
+      "learning_rate": 3.547819127651061e-05,
+      "loss": 0.612,
+      "step": 7268
+    },
+    {
+      "epoch": 9.30432,
+      "grad_norm": 0.7823015451431274,
+      "learning_rate": 3.547619047619048e-05,
+      "loss": 0.5439,
+      "step": 7269
+    },
+    {
+      "epoch": 9.3056,
+      "grad_norm": 0.8299819231033325,
+      "learning_rate": 3.547418967587035e-05,
+      "loss": 0.5612,
+      "step": 7270
+    },
+    {
+      "epoch": 9.30688,
+      "grad_norm": 0.8400328159332275,
+      "learning_rate": 3.5472188875550225e-05,
+      "loss": 0.5443,
+      "step": 7271
+    },
+    {
+      "epoch": 9.30816,
+      "grad_norm": 0.7800350189208984,
+      "learning_rate": 3.547018807523009e-05,
+      "loss": 0.5261,
+      "step": 7272
+    },
+    {
+      "epoch": 9.30944,
+      "grad_norm": 0.847671389579773,
+      "learning_rate": 3.546818727490996e-05,
+      "loss": 0.5819,
+      "step": 7273
+    },
+    {
+      "epoch": 9.31072,
+      "grad_norm": 0.8879785537719727,
+      "learning_rate": 3.5466186474589834e-05,
+      "loss": 0.5905,
+      "step": 7274
+    },
+    {
+      "epoch": 9.312,
+      "grad_norm": 0.8334875702857971,
+      "learning_rate": 3.546418567426971e-05,
+      "loss": 0.5622,
+      "step": 7275
+    },
+    {
+      "epoch": 9.31328,
+      "grad_norm": 0.8510679006576538,
+      "learning_rate": 3.5462184873949584e-05,
+      "loss": 0.5825,
+      "step": 7276
+    },
+    {
+      "epoch": 9.31456,
+      "grad_norm": 0.8725741505622864,
+      "learning_rate": 3.5460184073629456e-05,
+      "loss": 0.6079,
+      "step": 7277
+    },
+    {
+      "epoch": 9.31584,
+      "grad_norm": 0.8539466261863708,
+      "learning_rate": 3.545818327330933e-05,
+      "loss": 0.5574,
+      "step": 7278
+    },
+    {
+      "epoch": 9.31712,
+      "grad_norm": 0.8324747681617737,
+      "learning_rate": 3.54561824729892e-05,
+      "loss": 0.5647,
+      "step": 7279
+    },
+    {
+      "epoch": 9.3184,
+      "grad_norm": 0.8533186912536621,
+      "learning_rate": 3.5454181672669065e-05,
+      "loss": 0.527,
+      "step": 7280
+    },
+    {
+      "epoch": 9.31968,
+      "grad_norm": 0.8378028273582458,
+      "learning_rate": 3.545218087234894e-05,
+      "loss": 0.5808,
+      "step": 7281
+    },
+    {
+      "epoch": 9.32096,
+      "grad_norm": 0.8582174777984619,
+      "learning_rate": 3.5450180072028816e-05,
+      "loss": 0.6096,
+      "step": 7282
+    },
+    {
+      "epoch": 9.32224,
+      "grad_norm": 0.8167418837547302,
+      "learning_rate": 3.544817927170869e-05,
+      "loss": 0.5209,
+      "step": 7283
+    },
+    {
+      "epoch": 9.32352,
+      "grad_norm": 0.8674497604370117,
+      "learning_rate": 3.544617847138856e-05,
+      "loss": 0.5823,
+      "step": 7284
+    },
+    {
+      "epoch": 9.3248,
+      "grad_norm": 0.8791519999504089,
+      "learning_rate": 3.544417767106843e-05,
+      "loss": 0.5943,
+      "step": 7285
+    },
+    {
+      "epoch": 9.32608,
+      "grad_norm": 0.8264080882072449,
+      "learning_rate": 3.54421768707483e-05,
+      "loss": 0.5731,
+      "step": 7286
+    },
+    {
+      "epoch": 9.32736,
+      "grad_norm": 0.8929938673973083,
+      "learning_rate": 3.5440176070428175e-05,
+      "loss": 0.606,
+      "step": 7287
+    },
+    {
+      "epoch": 9.32864,
+      "grad_norm": 0.8631495237350464,
+      "learning_rate": 3.543817527010804e-05,
+      "loss": 0.5336,
+      "step": 7288
+    },
+    {
+      "epoch": 9.32992,
+      "grad_norm": 0.8779585957527161,
+      "learning_rate": 3.543617446978792e-05,
+      "loss": 0.5883,
+      "step": 7289
+    },
+    {
+      "epoch": 9.3312,
+      "grad_norm": 0.8582602739334106,
+      "learning_rate": 3.543417366946779e-05,
+      "loss": 0.5969,
+      "step": 7290
+    },
+    {
+      "epoch": 9.33248,
+      "grad_norm": 0.9014308452606201,
+      "learning_rate": 3.543217286914766e-05,
+      "loss": 0.5873,
+      "step": 7291
+    },
+    {
+      "epoch": 9.33376,
+      "grad_norm": 0.8432881832122803,
+      "learning_rate": 3.5430172068827534e-05,
+      "loss": 0.5814,
+      "step": 7292
+    },
+    {
+      "epoch": 9.33504,
+      "grad_norm": 0.8336278200149536,
+      "learning_rate": 3.5428171268507406e-05,
+      "loss": 0.528,
+      "step": 7293
+    },
+    {
+      "epoch": 9.33632,
+      "grad_norm": 0.8923518657684326,
+      "learning_rate": 3.542617046818728e-05,
+      "loss": 0.6088,
+      "step": 7294
+    },
+    {
+      "epoch": 9.3376,
+      "grad_norm": 0.843570351600647,
+      "learning_rate": 3.542416966786715e-05,
+      "loss": 0.5867,
+      "step": 7295
+    },
+    {
+      "epoch": 9.33888,
+      "grad_norm": 0.8679114580154419,
+      "learning_rate": 3.542216886754702e-05,
+      "loss": 0.5836,
+      "step": 7296
+    },
+    {
+      "epoch": 9.340160000000001,
+      "grad_norm": 0.8755765557289124,
+      "learning_rate": 3.5420168067226893e-05,
+      "loss": 0.6218,
+      "step": 7297
+    },
+    {
+      "epoch": 9.34144,
+      "grad_norm": 0.8723339438438416,
+      "learning_rate": 3.5418167266906765e-05,
+      "loss": 0.5534,
+      "step": 7298
+    },
+    {
+      "epoch": 9.34272,
+      "grad_norm": 0.8771824240684509,
+      "learning_rate": 3.541616646658664e-05,
+      "loss": 0.6106,
+      "step": 7299
+    },
+    {
+      "epoch": 9.344,
+      "grad_norm": 0.7724134922027588,
+      "learning_rate": 3.541416566626651e-05,
+      "loss": 0.5511,
+      "step": 7300
+    },
+    {
+      "epoch": 9.34528,
+      "grad_norm": 0.8475801348686218,
+      "learning_rate": 3.541216486594638e-05,
+      "loss": 0.5752,
+      "step": 7301
+    },
+    {
+      "epoch": 9.34656,
+      "grad_norm": 0.8831236958503723,
+      "learning_rate": 3.541016406562625e-05,
+      "loss": 0.6072,
+      "step": 7302
+    },
+    {
+      "epoch": 9.34784,
+      "grad_norm": 0.8372252583503723,
+      "learning_rate": 3.5408163265306125e-05,
+      "loss": 0.5337,
+      "step": 7303
+    },
+    {
+      "epoch": 9.34912,
+      "grad_norm": 0.810483992099762,
+      "learning_rate": 3.5406162464985996e-05,
+      "loss": 0.6066,
+      "step": 7304
+    },
+    {
+      "epoch": 9.3504,
+      "grad_norm": 0.8280951380729675,
+      "learning_rate": 3.540416166466587e-05,
+      "loss": 0.5439,
+      "step": 7305
+    },
+    {
+      "epoch": 9.35168,
+      "grad_norm": 0.8466097712516785,
+      "learning_rate": 3.540216086434574e-05,
+      "loss": 0.6007,
+      "step": 7306
+    },
+    {
+      "epoch": 9.35296,
+      "grad_norm": 0.8468445539474487,
+      "learning_rate": 3.540016006402561e-05,
+      "loss": 0.5211,
+      "step": 7307
+    },
+    {
+      "epoch": 9.35424,
+      "grad_norm": 0.8270046710968018,
+      "learning_rate": 3.5398159263705484e-05,
+      "loss": 0.6036,
+      "step": 7308
+    },
+    {
+      "epoch": 9.35552,
+      "grad_norm": 0.8338344693183899,
+      "learning_rate": 3.5396158463385356e-05,
+      "loss": 0.5664,
+      "step": 7309
+    },
+    {
+      "epoch": 9.3568,
+      "grad_norm": 0.8763896822929382,
+      "learning_rate": 3.539415766306523e-05,
+      "loss": 0.5601,
+      "step": 7310
+    },
+    {
+      "epoch": 9.35808,
+      "grad_norm": 0.8396816849708557,
+      "learning_rate": 3.53921568627451e-05,
+      "loss": 0.5643,
+      "step": 7311
+    },
+    {
+      "epoch": 9.35936,
+      "grad_norm": 0.8239136338233948,
+      "learning_rate": 3.539015606242497e-05,
+      "loss": 0.5644,
+      "step": 7312
+    },
+    {
+      "epoch": 9.36064,
+      "grad_norm": 0.8570470809936523,
+      "learning_rate": 3.538815526210484e-05,
+      "loss": 0.5655,
+      "step": 7313
+    },
+    {
+      "epoch": 9.36192,
+      "grad_norm": 0.8580249547958374,
+      "learning_rate": 3.5386154461784715e-05,
+      "loss": 0.6384,
+      "step": 7314
+    },
+    {
+      "epoch": 9.3632,
+      "grad_norm": 0.8450345993041992,
+      "learning_rate": 3.538415366146459e-05,
+      "loss": 0.5922,
+      "step": 7315
+    },
+    {
+      "epoch": 9.36448,
+      "grad_norm": 0.8073413968086243,
+      "learning_rate": 3.538215286114446e-05,
+      "loss": 0.5208,
+      "step": 7316
+    },
+    {
+      "epoch": 9.36576,
+      "grad_norm": 0.8269434571266174,
+      "learning_rate": 3.538015206082434e-05,
+      "loss": 0.5674,
+      "step": 7317
+    },
+    {
+      "epoch": 9.36704,
+      "grad_norm": 0.831274688243866,
+      "learning_rate": 3.53781512605042e-05,
+      "loss": 0.5477,
+      "step": 7318
+    },
+    {
+      "epoch": 9.36832,
+      "grad_norm": 0.8366057872772217,
+      "learning_rate": 3.5376150460184074e-05,
+      "loss": 0.561,
+      "step": 7319
+    },
+    {
+      "epoch": 9.3696,
+      "grad_norm": 0.8297872543334961,
+      "learning_rate": 3.5374149659863946e-05,
+      "loss": 0.5819,
+      "step": 7320
+    },
+    {
+      "epoch": 9.37088,
+      "grad_norm": 0.8322229981422424,
+      "learning_rate": 3.537214885954382e-05,
+      "loss": 0.5315,
+      "step": 7321
+    },
+    {
+      "epoch": 9.37216,
+      "grad_norm": 0.8755327463150024,
+      "learning_rate": 3.537014805922369e-05,
+      "loss": 0.5705,
+      "step": 7322
+    },
+    {
+      "epoch": 9.37344,
+      "grad_norm": 0.8848744630813599,
+      "learning_rate": 3.536814725890356e-05,
+      "loss": 0.6089,
+      "step": 7323
+    },
+    {
+      "epoch": 9.37472,
+      "grad_norm": 0.7823454737663269,
+      "learning_rate": 3.5366146458583434e-05,
+      "loss": 0.5527,
+      "step": 7324
+    },
+    {
+      "epoch": 9.376,
+      "grad_norm": 0.8572680950164795,
+      "learning_rate": 3.536414565826331e-05,
+      "loss": 0.6162,
+      "step": 7325
+    },
+    {
+      "epoch": 9.37728,
+      "grad_norm": 0.8322939872741699,
+      "learning_rate": 3.536214485794318e-05,
+      "loss": 0.5732,
+      "step": 7326
+    },
+    {
+      "epoch": 9.37856,
+      "grad_norm": 0.87436842918396,
+      "learning_rate": 3.536014405762305e-05,
+      "loss": 0.6265,
+      "step": 7327
+    },
+    {
+      "epoch": 9.37984,
+      "grad_norm": 0.8274515867233276,
+      "learning_rate": 3.535814325730292e-05,
+      "loss": 0.5288,
+      "step": 7328
+    },
+    {
+      "epoch": 9.38112,
+      "grad_norm": 0.9130957126617432,
+      "learning_rate": 3.535614245698279e-05,
+      "loss": 0.6091,
+      "step": 7329
+    },
+    {
+      "epoch": 9.3824,
+      "grad_norm": 0.8239238858222961,
+      "learning_rate": 3.5354141656662665e-05,
+      "loss": 0.5281,
+      "step": 7330
+    },
+    {
+      "epoch": 9.38368,
+      "grad_norm": 0.8236826658248901,
+      "learning_rate": 3.5352140856342537e-05,
+      "loss": 0.5264,
+      "step": 7331
+    },
+    {
+      "epoch": 9.38496,
+      "grad_norm": 0.8105927109718323,
+      "learning_rate": 3.5350140056022415e-05,
+      "loss": 0.5647,
+      "step": 7332
+    },
+    {
+      "epoch": 9.38624,
+      "grad_norm": 0.8803057670593262,
+      "learning_rate": 3.534813925570229e-05,
+      "loss": 0.5641,
+      "step": 7333
+    },
+    {
+      "epoch": 9.38752,
+      "grad_norm": 0.8564363121986389,
+      "learning_rate": 3.534613845538215e-05,
+      "loss": 0.5985,
+      "step": 7334
+    },
+    {
+      "epoch": 9.3888,
+      "grad_norm": 0.8307512998580933,
+      "learning_rate": 3.5344137655062024e-05,
+      "loss": 0.5664,
+      "step": 7335
+    },
+    {
+      "epoch": 9.39008,
+      "grad_norm": 0.9065113067626953,
+      "learning_rate": 3.5342136854741896e-05,
+      "loss": 0.633,
+      "step": 7336
+    },
+    {
+      "epoch": 9.39136,
+      "grad_norm": 0.8214498162269592,
+      "learning_rate": 3.534013605442177e-05,
+      "loss": 0.5621,
+      "step": 7337
+    },
+    {
+      "epoch": 9.39264,
+      "grad_norm": 0.8665810823440552,
+      "learning_rate": 3.533813525410164e-05,
+      "loss": 0.6223,
+      "step": 7338
+    },
+    {
+      "epoch": 9.39392,
+      "grad_norm": 0.9425195455551147,
+      "learning_rate": 3.533613445378152e-05,
+      "loss": 0.6163,
+      "step": 7339
+    },
+    {
+      "epoch": 9.395199999999999,
+      "grad_norm": 0.8307021260261536,
+      "learning_rate": 3.533413365346139e-05,
+      "loss": 0.5604,
+      "step": 7340
+    },
+    {
+      "epoch": 9.39648,
+      "grad_norm": 0.8683647513389587,
+      "learning_rate": 3.533213285314126e-05,
+      "loss": 0.5811,
+      "step": 7341
+    },
+    {
+      "epoch": 9.39776,
+      "grad_norm": 0.8393630385398865,
+      "learning_rate": 3.533013205282113e-05,
+      "loss": 0.5444,
+      "step": 7342
+    },
+    {
+      "epoch": 9.39904,
+      "grad_norm": 0.809565544128418,
+      "learning_rate": 3.5328131252501e-05,
+      "loss": 0.5328,
+      "step": 7343
+    },
+    {
+      "epoch": 9.40032,
+      "grad_norm": 0.8294006586074829,
+      "learning_rate": 3.532613045218087e-05,
+      "loss": 0.5563,
+      "step": 7344
+    },
+    {
+      "epoch": 9.4016,
+      "grad_norm": 0.8536190390586853,
+      "learning_rate": 3.532412965186074e-05,
+      "loss": 0.5752,
+      "step": 7345
+    },
+    {
+      "epoch": 9.40288,
+      "grad_norm": 0.8586856722831726,
+      "learning_rate": 3.532212885154062e-05,
+      "loss": 0.6095,
+      "step": 7346
+    },
+    {
+      "epoch": 9.40416,
+      "grad_norm": 0.9051749110221863,
+      "learning_rate": 3.532012805122049e-05,
+      "loss": 0.6253,
+      "step": 7347
+    },
+    {
+      "epoch": 9.40544,
+      "grad_norm": 0.8567177057266235,
+      "learning_rate": 3.5318127250900365e-05,
+      "loss": 0.6243,
+      "step": 7348
+    },
+    {
+      "epoch": 9.40672,
+      "grad_norm": 0.8297619819641113,
+      "learning_rate": 3.531612645058024e-05,
+      "loss": 0.6051,
+      "step": 7349
+    },
+    {
+      "epoch": 9.408,
+      "grad_norm": 0.8204346895217896,
+      "learning_rate": 3.53141256502601e-05,
+      "loss": 0.5123,
+      "step": 7350
+    },
+    {
+      "epoch": 9.40928,
+      "grad_norm": 0.8355170488357544,
+      "learning_rate": 3.5312124849939974e-05,
+      "loss": 0.5724,
+      "step": 7351
+    },
+    {
+      "epoch": 9.41056,
+      "grad_norm": 0.7993974089622498,
+      "learning_rate": 3.5310124049619846e-05,
+      "loss": 0.5733,
+      "step": 7352
+    },
+    {
+      "epoch": 9.41184,
+      "grad_norm": 0.846606969833374,
+      "learning_rate": 3.5308123249299724e-05,
+      "loss": 0.5533,
+      "step": 7353
+    },
+    {
+      "epoch": 9.41312,
+      "grad_norm": 0.8249667882919312,
+      "learning_rate": 3.5306122448979596e-05,
+      "loss": 0.5561,
+      "step": 7354
+    },
+    {
+      "epoch": 9.4144,
+      "grad_norm": 0.8175135850906372,
+      "learning_rate": 3.530412164865947e-05,
+      "loss": 0.557,
+      "step": 7355
+    },
+    {
+      "epoch": 9.41568,
+      "grad_norm": 0.855964183807373,
+      "learning_rate": 3.530212084833934e-05,
+      "loss": 0.5509,
+      "step": 7356
+    },
+    {
+      "epoch": 9.41696,
+      "grad_norm": 0.8171284198760986,
+      "learning_rate": 3.530012004801921e-05,
+      "loss": 0.5227,
+      "step": 7357
+    },
+    {
+      "epoch": 9.41824,
+      "grad_norm": 0.8054320812225342,
+      "learning_rate": 3.529811924769908e-05,
+      "loss": 0.6215,
+      "step": 7358
+    },
+    {
+      "epoch": 9.41952,
+      "grad_norm": 0.8339996933937073,
+      "learning_rate": 3.529611844737895e-05,
+      "loss": 0.5524,
+      "step": 7359
+    },
+    {
+      "epoch": 9.4208,
+      "grad_norm": 0.8672724366188049,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 0.5946,
+      "step": 7360
+    },
+    {
+      "epoch": 9.42208,
+      "grad_norm": 0.799502968788147,
+      "learning_rate": 3.52921168467387e-05,
+      "loss": 0.5152,
+      "step": 7361
+    },
+    {
+      "epoch": 9.42336,
+      "grad_norm": 0.8360862731933594,
+      "learning_rate": 3.529011604641857e-05,
+      "loss": 0.5564,
+      "step": 7362
+    },
+    {
+      "epoch": 9.42464,
+      "grad_norm": 0.8385033011436462,
+      "learning_rate": 3.528811524609844e-05,
+      "loss": 0.5551,
+      "step": 7363
+    },
+    {
+      "epoch": 9.42592,
+      "grad_norm": 0.8228257894515991,
+      "learning_rate": 3.5286114445778315e-05,
+      "loss": 0.5351,
+      "step": 7364
+    },
+    {
+      "epoch": 9.4272,
+      "grad_norm": 0.8925015330314636,
+      "learning_rate": 3.5284113645458186e-05,
+      "loss": 0.6315,
+      "step": 7365
+    },
+    {
+      "epoch": 9.42848,
+      "grad_norm": 0.9042237997055054,
+      "learning_rate": 3.528211284513805e-05,
+      "loss": 0.6171,
+      "step": 7366
+    },
+    {
+      "epoch": 9.42976,
+      "grad_norm": 0.8735190033912659,
+      "learning_rate": 3.528011204481793e-05,
+      "loss": 0.581,
+      "step": 7367
+    },
+    {
+      "epoch": 9.43104,
+      "grad_norm": 0.8328735828399658,
+      "learning_rate": 3.52781112444978e-05,
+      "loss": 0.5541,
+      "step": 7368
+    },
+    {
+      "epoch": 9.43232,
+      "grad_norm": 0.8481590151786804,
+      "learning_rate": 3.5276110444177674e-05,
+      "loss": 0.5868,
+      "step": 7369
+    },
+    {
+      "epoch": 9.4336,
+      "grad_norm": 0.7884271144866943,
+      "learning_rate": 3.5274109643857546e-05,
+      "loss": 0.5346,
+      "step": 7370
+    },
+    {
+      "epoch": 9.43488,
+      "grad_norm": 0.8754736185073853,
+      "learning_rate": 3.527210884353742e-05,
+      "loss": 0.6171,
+      "step": 7371
+    },
+    {
+      "epoch": 9.43616,
+      "grad_norm": 0.8703823685646057,
+      "learning_rate": 3.527010804321729e-05,
+      "loss": 0.5973,
+      "step": 7372
+    },
+    {
+      "epoch": 9.43744,
+      "grad_norm": 0.869755208492279,
+      "learning_rate": 3.526810724289716e-05,
+      "loss": 0.5851,
+      "step": 7373
+    },
+    {
+      "epoch": 9.43872,
+      "grad_norm": 0.8053287267684937,
+      "learning_rate": 3.526610644257703e-05,
+      "loss": 0.5132,
+      "step": 7374
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 0.853288471698761,
+      "learning_rate": 3.5264105642256905e-05,
+      "loss": 0.5325,
+      "step": 7375
+    },
+    {
+      "epoch": 9.44128,
+      "grad_norm": 0.8669641613960266,
+      "learning_rate": 3.526210484193678e-05,
+      "loss": 0.5699,
+      "step": 7376
+    },
+    {
+      "epoch": 9.44256,
+      "grad_norm": 0.8685311079025269,
+      "learning_rate": 3.526010404161665e-05,
+      "loss": 0.6092,
+      "step": 7377
+    },
+    {
+      "epoch": 9.44384,
+      "grad_norm": 0.8137316703796387,
+      "learning_rate": 3.525810324129652e-05,
+      "loss": 0.5549,
+      "step": 7378
+    },
+    {
+      "epoch": 9.44512,
+      "grad_norm": 0.8705141544342041,
+      "learning_rate": 3.525610244097639e-05,
+      "loss": 0.5925,
+      "step": 7379
+    },
+    {
+      "epoch": 9.4464,
+      "grad_norm": 0.7994008660316467,
+      "learning_rate": 3.5254101640656264e-05,
+      "loss": 0.5114,
+      "step": 7380
+    },
+    {
+      "epoch": 9.44768,
+      "grad_norm": 0.8135007619857788,
+      "learning_rate": 3.5252100840336136e-05,
+      "loss": 0.5546,
+      "step": 7381
+    },
+    {
+      "epoch": 9.44896,
+      "grad_norm": 0.8146666288375854,
+      "learning_rate": 3.525010004001601e-05,
+      "loss": 0.5699,
+      "step": 7382
+    },
+    {
+      "epoch": 9.45024,
+      "grad_norm": 0.8331019878387451,
+      "learning_rate": 3.524809923969588e-05,
+      "loss": 0.6037,
+      "step": 7383
+    },
+    {
+      "epoch": 9.45152,
+      "grad_norm": 0.8365830183029175,
+      "learning_rate": 3.524609843937575e-05,
+      "loss": 0.5718,
+      "step": 7384
+    },
+    {
+      "epoch": 9.4528,
+      "grad_norm": 0.8438559770584106,
+      "learning_rate": 3.5244097639055624e-05,
+      "loss": 0.5881,
+      "step": 7385
+    },
+    {
+      "epoch": 9.45408,
+      "grad_norm": 0.8334947228431702,
+      "learning_rate": 3.5242096838735495e-05,
+      "loss": 0.5361,
+      "step": 7386
+    },
+    {
+      "epoch": 9.45536,
+      "grad_norm": 0.8178591132164001,
+      "learning_rate": 3.524009603841537e-05,
+      "loss": 0.5467,
+      "step": 7387
+    },
+    {
+      "epoch": 9.45664,
+      "grad_norm": 0.8228579759597778,
+      "learning_rate": 3.523809523809524e-05,
+      "loss": 0.549,
+      "step": 7388
+    },
+    {
+      "epoch": 9.45792,
+      "grad_norm": 0.8397416472434998,
+      "learning_rate": 3.523609443777511e-05,
+      "loss": 0.574,
+      "step": 7389
+    },
+    {
+      "epoch": 9.4592,
+      "grad_norm": 0.8239879012107849,
+      "learning_rate": 3.523409363745498e-05,
+      "loss": 0.5415,
+      "step": 7390
+    },
+    {
+      "epoch": 9.46048,
+      "grad_norm": 0.8690317273139954,
+      "learning_rate": 3.5232092837134855e-05,
+      "loss": 0.6039,
+      "step": 7391
+    },
+    {
+      "epoch": 9.46176,
+      "grad_norm": 0.7893775105476379,
+      "learning_rate": 3.523009203681473e-05,
+      "loss": 0.555,
+      "step": 7392
+    },
+    {
+      "epoch": 9.46304,
+      "grad_norm": 0.8786347508430481,
+      "learning_rate": 3.52280912364946e-05,
+      "loss": 0.5882,
+      "step": 7393
+    },
+    {
+      "epoch": 9.46432,
+      "grad_norm": 0.8203712105751038,
+      "learning_rate": 3.522609043617447e-05,
+      "loss": 0.5633,
+      "step": 7394
+    },
+    {
+      "epoch": 9.4656,
+      "grad_norm": 0.8264157176017761,
+      "learning_rate": 3.522408963585435e-05,
+      "loss": 0.548,
+      "step": 7395
+    },
+    {
+      "epoch": 9.46688,
+      "grad_norm": 0.7980100512504578,
+      "learning_rate": 3.5222088835534214e-05,
+      "loss": 0.5698,
+      "step": 7396
+    },
+    {
+      "epoch": 9.46816,
+      "grad_norm": 0.8063759803771973,
+      "learning_rate": 3.5220088035214086e-05,
+      "loss": 0.5541,
+      "step": 7397
+    },
+    {
+      "epoch": 9.46944,
+      "grad_norm": 0.8606681227684021,
+      "learning_rate": 3.521808723489396e-05,
+      "loss": 0.5892,
+      "step": 7398
+    },
+    {
+      "epoch": 9.47072,
+      "grad_norm": 0.8731750249862671,
+      "learning_rate": 3.521608643457383e-05,
+      "loss": 0.6104,
+      "step": 7399
+    },
+    {
+      "epoch": 9.472,
+      "grad_norm": 0.8435729742050171,
+      "learning_rate": 3.52140856342537e-05,
+      "loss": 0.5474,
+      "step": 7400
+    },
+    {
+      "epoch": 9.47328,
+      "grad_norm": 0.8424503207206726,
+      "learning_rate": 3.521208483393357e-05,
+      "loss": 0.6079,
+      "step": 7401
+    },
+    {
+      "epoch": 9.47456,
+      "grad_norm": 0.83389812707901,
+      "learning_rate": 3.521008403361345e-05,
+      "loss": 0.5787,
+      "step": 7402
+    },
+    {
+      "epoch": 9.47584,
+      "grad_norm": 0.8542918562889099,
+      "learning_rate": 3.5208083233293324e-05,
+      "loss": 0.5622,
+      "step": 7403
+    },
+    {
+      "epoch": 9.47712,
+      "grad_norm": 0.8500902652740479,
+      "learning_rate": 3.520608243297319e-05,
+      "loss": 0.5805,
+      "step": 7404
+    },
+    {
+      "epoch": 9.4784,
+      "grad_norm": 0.8857991099357605,
+      "learning_rate": 3.520408163265306e-05,
+      "loss": 0.5924,
+      "step": 7405
+    },
+    {
+      "epoch": 9.47968,
+      "grad_norm": 0.8328712582588196,
+      "learning_rate": 3.520208083233293e-05,
+      "loss": 0.5427,
+      "step": 7406
+    },
+    {
+      "epoch": 9.48096,
+      "grad_norm": 0.8790886998176575,
+      "learning_rate": 3.5200080032012804e-05,
+      "loss": 0.6283,
+      "step": 7407
+    },
+    {
+      "epoch": 9.482240000000001,
+      "grad_norm": 0.8296170234680176,
+      "learning_rate": 3.5198079231692676e-05,
+      "loss": 0.5678,
+      "step": 7408
+    },
+    {
+      "epoch": 9.48352,
+      "grad_norm": 0.8877446055412292,
+      "learning_rate": 3.5196078431372555e-05,
+      "loss": 0.5913,
+      "step": 7409
+    },
+    {
+      "epoch": 9.4848,
+      "grad_norm": 0.8364286422729492,
+      "learning_rate": 3.519407763105243e-05,
+      "loss": 0.5662,
+      "step": 7410
+    },
+    {
+      "epoch": 9.48608,
+      "grad_norm": 0.8386290669441223,
+      "learning_rate": 3.51920768307323e-05,
+      "loss": 0.5586,
+      "step": 7411
+    },
+    {
+      "epoch": 9.48736,
+      "grad_norm": 0.8200079798698425,
+      "learning_rate": 3.5190076030412164e-05,
+      "loss": 0.5446,
+      "step": 7412
+    },
+    {
+      "epoch": 9.48864,
+      "grad_norm": 0.8406325578689575,
+      "learning_rate": 3.5188075230092036e-05,
+      "loss": 0.5378,
+      "step": 7413
+    },
+    {
+      "epoch": 9.48992,
+      "grad_norm": 0.8740529417991638,
+      "learning_rate": 3.518607442977191e-05,
+      "loss": 0.557,
+      "step": 7414
+    },
+    {
+      "epoch": 9.4912,
+      "grad_norm": 0.8621671199798584,
+      "learning_rate": 3.518407362945178e-05,
+      "loss": 0.5946,
+      "step": 7415
+    },
+    {
+      "epoch": 9.49248,
+      "grad_norm": 0.812969446182251,
+      "learning_rate": 3.518207282913166e-05,
+      "loss": 0.5282,
+      "step": 7416
+    },
+    {
+      "epoch": 9.49376,
+      "grad_norm": 0.8162618279457092,
+      "learning_rate": 3.518007202881153e-05,
+      "loss": 0.5539,
+      "step": 7417
+    },
+    {
+      "epoch": 9.49504,
+      "grad_norm": 0.879939079284668,
+      "learning_rate": 3.51780712284914e-05,
+      "loss": 0.5933,
+      "step": 7418
+    },
+    {
+      "epoch": 9.49632,
+      "grad_norm": 0.8177329897880554,
+      "learning_rate": 3.5176070428171274e-05,
+      "loss": 0.5697,
+      "step": 7419
+    },
+    {
+      "epoch": 9.4976,
+      "grad_norm": 0.7988497018814087,
+      "learning_rate": 3.517406962785114e-05,
+      "loss": 0.4975,
+      "step": 7420
+    },
+    {
+      "epoch": 9.49888,
+      "grad_norm": 0.8101351261138916,
+      "learning_rate": 3.517206882753101e-05,
+      "loss": 0.5755,
+      "step": 7421
+    },
+    {
+      "epoch": 9.50016,
+      "grad_norm": 0.9145906567573547,
+      "learning_rate": 3.517006802721088e-05,
+      "loss": 0.6038,
+      "step": 7422
+    },
+    {
+      "epoch": 9.50144,
+      "grad_norm": 0.8204926252365112,
+      "learning_rate": 3.516806722689076e-05,
+      "loss": 0.536,
+      "step": 7423
+    },
+    {
+      "epoch": 9.50272,
+      "grad_norm": 0.8429908156394958,
+      "learning_rate": 3.516606642657063e-05,
+      "loss": 0.5637,
+      "step": 7424
+    },
+    {
+      "epoch": 9.504,
+      "grad_norm": 0.8640322685241699,
+      "learning_rate": 3.5164065626250505e-05,
+      "loss": 0.6141,
+      "step": 7425
+    },
+    {
+      "epoch": 9.505279999999999,
+      "grad_norm": 0.9030240774154663,
+      "learning_rate": 3.5162064825930377e-05,
+      "loss": 0.5929,
+      "step": 7426
+    },
+    {
+      "epoch": 9.50656,
+      "grad_norm": 0.8568662405014038,
+      "learning_rate": 3.516006402561025e-05,
+      "loss": 0.5649,
+      "step": 7427
+    },
+    {
+      "epoch": 9.50784,
+      "grad_norm": 0.8583548069000244,
+      "learning_rate": 3.5158063225290113e-05,
+      "loss": 0.5824,
+      "step": 7428
+    },
+    {
+      "epoch": 9.50912,
+      "grad_norm": 0.8739598989486694,
+      "learning_rate": 3.5156062424969985e-05,
+      "loss": 0.5895,
+      "step": 7429
+    },
+    {
+      "epoch": 9.5104,
+      "grad_norm": 0.9026753306388855,
+      "learning_rate": 3.515406162464986e-05,
+      "loss": 0.6056,
+      "step": 7430
+    },
+    {
+      "epoch": 9.51168,
+      "grad_norm": 0.8631276488304138,
+      "learning_rate": 3.5152060824329736e-05,
+      "loss": 0.5616,
+      "step": 7431
+    },
+    {
+      "epoch": 9.51296,
+      "grad_norm": 0.833394467830658,
+      "learning_rate": 3.515006002400961e-05,
+      "loss": 0.5742,
+      "step": 7432
+    },
+    {
+      "epoch": 9.514240000000001,
+      "grad_norm": 0.8503471612930298,
+      "learning_rate": 3.514805922368948e-05,
+      "loss": 0.5484,
+      "step": 7433
+    },
+    {
+      "epoch": 9.51552,
+      "grad_norm": 0.8320133686065674,
+      "learning_rate": 3.514605842336935e-05,
+      "loss": 0.5725,
+      "step": 7434
+    },
+    {
+      "epoch": 9.5168,
+      "grad_norm": 0.8564555048942566,
+      "learning_rate": 3.514405762304922e-05,
+      "loss": 0.5475,
+      "step": 7435
+    },
+    {
+      "epoch": 9.51808,
+      "grad_norm": 0.8562284708023071,
+      "learning_rate": 3.514205682272909e-05,
+      "loss": 0.6121,
+      "step": 7436
+    },
+    {
+      "epoch": 9.51936,
+      "grad_norm": 0.8364126682281494,
+      "learning_rate": 3.514005602240896e-05,
+      "loss": 0.57,
+      "step": 7437
+    },
+    {
+      "epoch": 9.52064,
+      "grad_norm": 0.8533650636672974,
+      "learning_rate": 3.513805522208884e-05,
+      "loss": 0.6145,
+      "step": 7438
+    },
+    {
+      "epoch": 9.52192,
+      "grad_norm": 0.8603209257125854,
+      "learning_rate": 3.513605442176871e-05,
+      "loss": 0.5859,
+      "step": 7439
+    },
+    {
+      "epoch": 9.5232,
+      "grad_norm": 0.8335322141647339,
+      "learning_rate": 3.513405362144858e-05,
+      "loss": 0.5862,
+      "step": 7440
+    },
+    {
+      "epoch": 9.52448,
+      "grad_norm": 0.848399817943573,
+      "learning_rate": 3.5132052821128454e-05,
+      "loss": 0.572,
+      "step": 7441
+    },
+    {
+      "epoch": 9.52576,
+      "grad_norm": 0.7791329026222229,
+      "learning_rate": 3.5130052020808326e-05,
+      "loss": 0.5761,
+      "step": 7442
+    },
+    {
+      "epoch": 9.52704,
+      "grad_norm": 0.8725833892822266,
+      "learning_rate": 3.51280512204882e-05,
+      "loss": 0.6316,
+      "step": 7443
+    },
+    {
+      "epoch": 9.52832,
+      "grad_norm": 0.8655521869659424,
+      "learning_rate": 3.512605042016806e-05,
+      "loss": 0.5992,
+      "step": 7444
+    },
+    {
+      "epoch": 9.5296,
+      "grad_norm": 0.862616777420044,
+      "learning_rate": 3.512404961984794e-05,
+      "loss": 0.5501,
+      "step": 7445
+    },
+    {
+      "epoch": 9.53088,
+      "grad_norm": 0.8571741580963135,
+      "learning_rate": 3.5122048819527814e-05,
+      "loss": 0.5958,
+      "step": 7446
+    },
+    {
+      "epoch": 9.53216,
+      "grad_norm": 0.8498180508613586,
+      "learning_rate": 3.5120048019207686e-05,
+      "loss": 0.5891,
+      "step": 7447
+    },
+    {
+      "epoch": 9.53344,
+      "grad_norm": 0.829552412033081,
+      "learning_rate": 3.511804721888756e-05,
+      "loss": 0.5575,
+      "step": 7448
+    },
+    {
+      "epoch": 9.53472,
+      "grad_norm": 0.8693800568580627,
+      "learning_rate": 3.511604641856743e-05,
+      "loss": 0.6149,
+      "step": 7449
+    },
+    {
+      "epoch": 9.536,
+      "grad_norm": 0.7874047160148621,
+      "learning_rate": 3.51140456182473e-05,
+      "loss": 0.5286,
+      "step": 7450
+    },
+    {
+      "epoch": 9.537279999999999,
+      "grad_norm": 0.8254282474517822,
+      "learning_rate": 3.511204481792717e-05,
+      "loss": 0.581,
+      "step": 7451
+    },
+    {
+      "epoch": 9.53856,
+      "grad_norm": 0.8394695520401001,
+      "learning_rate": 3.5110044017607045e-05,
+      "loss": 0.5244,
+      "step": 7452
+    },
+    {
+      "epoch": 9.53984,
+      "grad_norm": 0.8637860417366028,
+      "learning_rate": 3.510804321728692e-05,
+      "loss": 0.6142,
+      "step": 7453
+    },
+    {
+      "epoch": 9.54112,
+      "grad_norm": 0.8084328770637512,
+      "learning_rate": 3.510604241696679e-05,
+      "loss": 0.529,
+      "step": 7454
+    },
+    {
+      "epoch": 9.5424,
+      "grad_norm": 0.8361181616783142,
+      "learning_rate": 3.510404161664666e-05,
+      "loss": 0.5863,
+      "step": 7455
+    },
+    {
+      "epoch": 9.54368,
+      "grad_norm": 0.8431722521781921,
+      "learning_rate": 3.510204081632653e-05,
+      "loss": 0.5477,
+      "step": 7456
+    },
+    {
+      "epoch": 9.54496,
+      "grad_norm": 0.868948221206665,
+      "learning_rate": 3.5100040016006404e-05,
+      "loss": 0.6217,
+      "step": 7457
+    },
+    {
+      "epoch": 9.54624,
+      "grad_norm": 0.8233850002288818,
+      "learning_rate": 3.5098039215686276e-05,
+      "loss": 0.5369,
+      "step": 7458
+    },
+    {
+      "epoch": 9.54752,
+      "grad_norm": 0.8040074110031128,
+      "learning_rate": 3.509603841536615e-05,
+      "loss": 0.5127,
+      "step": 7459
+    },
+    {
+      "epoch": 9.5488,
+      "grad_norm": 0.8325415849685669,
+      "learning_rate": 3.509403761504602e-05,
+      "loss": 0.5572,
+      "step": 7460
+    },
+    {
+      "epoch": 9.55008,
+      "grad_norm": 0.8861798644065857,
+      "learning_rate": 3.509203681472589e-05,
+      "loss": 0.6159,
+      "step": 7461
+    },
+    {
+      "epoch": 9.55136,
+      "grad_norm": 0.8803839087486267,
+      "learning_rate": 3.509003601440576e-05,
+      "loss": 0.586,
+      "step": 7462
+    },
+    {
+      "epoch": 9.55264,
+      "grad_norm": 0.8698328733444214,
+      "learning_rate": 3.5088035214085635e-05,
+      "loss": 0.57,
+      "step": 7463
+    },
+    {
+      "epoch": 9.55392,
+      "grad_norm": 0.8476325273513794,
+      "learning_rate": 3.508603441376551e-05,
+      "loss": 0.5661,
+      "step": 7464
+    },
+    {
+      "epoch": 9.5552,
+      "grad_norm": 0.789214015007019,
+      "learning_rate": 3.508403361344538e-05,
+      "loss": 0.5412,
+      "step": 7465
+    },
+    {
+      "epoch": 9.55648,
+      "grad_norm": 0.851634681224823,
+      "learning_rate": 3.508203281312525e-05,
+      "loss": 0.5722,
+      "step": 7466
+    },
+    {
+      "epoch": 9.55776,
+      "grad_norm": 0.8242769241333008,
+      "learning_rate": 3.508003201280512e-05,
+      "loss": 0.5649,
+      "step": 7467
+    },
+    {
+      "epoch": 9.55904,
+      "grad_norm": 0.8163190484046936,
+      "learning_rate": 3.5078031212484995e-05,
+      "loss": 0.5694,
+      "step": 7468
+    },
+    {
+      "epoch": 9.56032,
+      "grad_norm": 0.802607536315918,
+      "learning_rate": 3.5076030412164866e-05,
+      "loss": 0.5583,
+      "step": 7469
+    },
+    {
+      "epoch": 9.5616,
+      "grad_norm": 0.8320533633232117,
+      "learning_rate": 3.507402961184474e-05,
+      "loss": 0.5443,
+      "step": 7470
+    },
+    {
+      "epoch": 9.56288,
+      "grad_norm": 0.8149722218513489,
+      "learning_rate": 3.507202881152461e-05,
+      "loss": 0.5603,
+      "step": 7471
+    },
+    {
+      "epoch": 9.56416,
+      "grad_norm": 0.8735612630844116,
+      "learning_rate": 3.507002801120448e-05,
+      "loss": 0.5986,
+      "step": 7472
+    },
+    {
+      "epoch": 9.56544,
+      "grad_norm": 0.8785853385925293,
+      "learning_rate": 3.506802721088436e-05,
+      "loss": 0.6058,
+      "step": 7473
+    },
+    {
+      "epoch": 9.56672,
+      "grad_norm": 0.8740175366401672,
+      "learning_rate": 3.5066026410564226e-05,
+      "loss": 0.5921,
+      "step": 7474
+    },
+    {
+      "epoch": 9.568,
+      "grad_norm": 0.8479311466217041,
+      "learning_rate": 3.50640256102441e-05,
+      "loss": 0.5683,
+      "step": 7475
+    },
+    {
+      "epoch": 9.56928,
+      "grad_norm": 0.786037266254425,
+      "learning_rate": 3.506202480992397e-05,
+      "loss": 0.5639,
+      "step": 7476
+    },
+    {
+      "epoch": 9.57056,
+      "grad_norm": 0.8478996753692627,
+      "learning_rate": 3.506002400960384e-05,
+      "loss": 0.5613,
+      "step": 7477
+    },
+    {
+      "epoch": 9.57184,
+      "grad_norm": 0.8028008341789246,
+      "learning_rate": 3.505802320928371e-05,
+      "loss": 0.4897,
+      "step": 7478
+    },
+    {
+      "epoch": 9.57312,
+      "grad_norm": 0.844845175743103,
+      "learning_rate": 3.5056022408963585e-05,
+      "loss": 0.5955,
+      "step": 7479
+    },
+    {
+      "epoch": 9.5744,
+      "grad_norm": 0.8595391511917114,
+      "learning_rate": 3.5054021608643464e-05,
+      "loss": 0.5998,
+      "step": 7480
+    },
+    {
+      "epoch": 9.57568,
+      "grad_norm": 0.860565185546875,
+      "learning_rate": 3.5052020808323335e-05,
+      "loss": 0.5901,
+      "step": 7481
+    },
+    {
+      "epoch": 9.57696,
+      "grad_norm": 0.8473759293556213,
+      "learning_rate": 3.50500200080032e-05,
+      "loss": 0.5797,
+      "step": 7482
+    },
+    {
+      "epoch": 9.57824,
+      "grad_norm": 0.8636260032653809,
+      "learning_rate": 3.504801920768307e-05,
+      "loss": 0.5983,
+      "step": 7483
+    },
+    {
+      "epoch": 9.57952,
+      "grad_norm": 0.8650203943252563,
+      "learning_rate": 3.5046018407362944e-05,
+      "loss": 0.5942,
+      "step": 7484
+    },
+    {
+      "epoch": 9.5808,
+      "grad_norm": 0.9003608226776123,
+      "learning_rate": 3.5044017607042816e-05,
+      "loss": 0.5908,
+      "step": 7485
+    },
+    {
+      "epoch": 9.58208,
+      "grad_norm": 0.8681960105895996,
+      "learning_rate": 3.504201680672269e-05,
+      "loss": 0.6542,
+      "step": 7486
+    },
+    {
+      "epoch": 9.58336,
+      "grad_norm": 0.8483841419219971,
+      "learning_rate": 3.5040016006402567e-05,
+      "loss": 0.5657,
+      "step": 7487
+    },
+    {
+      "epoch": 9.58464,
+      "grad_norm": 0.894481897354126,
+      "learning_rate": 3.503801520608244e-05,
+      "loss": 0.6173,
+      "step": 7488
+    },
+    {
+      "epoch": 9.58592,
+      "grad_norm": 0.7843658328056335,
+      "learning_rate": 3.503601440576231e-05,
+      "loss": 0.527,
+      "step": 7489
+    },
+    {
+      "epoch": 9.5872,
+      "grad_norm": 0.8325011134147644,
+      "learning_rate": 3.5034013605442175e-05,
+      "loss": 0.5687,
+      "step": 7490
+    },
+    {
+      "epoch": 9.58848,
+      "grad_norm": 0.8706384301185608,
+      "learning_rate": 3.503201280512205e-05,
+      "loss": 0.6229,
+      "step": 7491
+    },
+    {
+      "epoch": 9.58976,
+      "grad_norm": 0.8439401388168335,
+      "learning_rate": 3.503001200480192e-05,
+      "loss": 0.565,
+      "step": 7492
+    },
+    {
+      "epoch": 9.59104,
+      "grad_norm": 0.8194774985313416,
+      "learning_rate": 3.502801120448179e-05,
+      "loss": 0.5327,
+      "step": 7493
+    },
+    {
+      "epoch": 9.59232,
+      "grad_norm": 0.8847520351409912,
+      "learning_rate": 3.502601040416167e-05,
+      "loss": 0.6109,
+      "step": 7494
+    },
+    {
+      "epoch": 9.5936,
+      "grad_norm": 0.8577123284339905,
+      "learning_rate": 3.502400960384154e-05,
+      "loss": 0.6078,
+      "step": 7495
+    },
+    {
+      "epoch": 9.59488,
+      "grad_norm": 0.8352939486503601,
+      "learning_rate": 3.502200880352141e-05,
+      "loss": 0.5655,
+      "step": 7496
+    },
+    {
+      "epoch": 9.59616,
+      "grad_norm": 0.8935215473175049,
+      "learning_rate": 3.5020008003201285e-05,
+      "loss": 0.6405,
+      "step": 7497
+    },
+    {
+      "epoch": 9.59744,
+      "grad_norm": 0.8329689502716064,
+      "learning_rate": 3.501800720288115e-05,
+      "loss": 0.5859,
+      "step": 7498
+    },
+    {
+      "epoch": 9.59872,
+      "grad_norm": 0.8347263932228088,
+      "learning_rate": 3.501600640256102e-05,
+      "loss": 0.5564,
+      "step": 7499
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.8732603192329407,
+      "learning_rate": 3.5014005602240894e-05,
+      "loss": 0.6078,
+      "step": 7500
+    },
+    {
+      "epoch": 9.60128,
+      "grad_norm": 0.8412352800369263,
+      "learning_rate": 3.501200480192077e-05,
+      "loss": 0.5728,
+      "step": 7501
+    },
+    {
+      "epoch": 9.60256,
+      "grad_norm": 0.8667293787002563,
+      "learning_rate": 3.5010004001600644e-05,
+      "loss": 0.5968,
+      "step": 7502
+    },
+    {
+      "epoch": 9.60384,
+      "grad_norm": 0.8245180249214172,
+      "learning_rate": 3.5008003201280516e-05,
+      "loss": 0.5847,
+      "step": 7503
+    },
+    {
+      "epoch": 9.60512,
+      "grad_norm": 0.8028945326805115,
+      "learning_rate": 3.500600240096039e-05,
+      "loss": 0.5471,
+      "step": 7504
+    },
+    {
+      "epoch": 9.6064,
+      "grad_norm": 0.844236433506012,
+      "learning_rate": 3.500400160064026e-05,
+      "loss": 0.5414,
+      "step": 7505
+    },
+    {
+      "epoch": 9.60768,
+      "grad_norm": 0.8215577006340027,
+      "learning_rate": 3.5002000800320125e-05,
+      "loss": 0.5612,
+      "step": 7506
+    },
+    {
+      "epoch": 9.60896,
+      "grad_norm": 0.8602308034896851,
+      "learning_rate": 3.5e-05,
+      "loss": 0.6125,
+      "step": 7507
+    },
+    {
+      "epoch": 9.61024,
+      "grad_norm": 0.8574222922325134,
+      "learning_rate": 3.4997999199679876e-05,
+      "loss": 0.6139,
+      "step": 7508
+    },
+    {
+      "epoch": 9.61152,
+      "grad_norm": 0.8892394304275513,
+      "learning_rate": 3.499599839935975e-05,
+      "loss": 0.6059,
+      "step": 7509
+    },
+    {
+      "epoch": 9.6128,
+      "grad_norm": 0.8644453287124634,
+      "learning_rate": 3.499399759903962e-05,
+      "loss": 0.6204,
+      "step": 7510
+    },
+    {
+      "epoch": 9.61408,
+      "grad_norm": 0.8173637390136719,
+      "learning_rate": 3.499199679871949e-05,
+      "loss": 0.5308,
+      "step": 7511
+    },
+    {
+      "epoch": 9.61536,
+      "grad_norm": 0.8753309845924377,
+      "learning_rate": 3.498999599839936e-05,
+      "loss": 0.6066,
+      "step": 7512
+    },
+    {
+      "epoch": 9.61664,
+      "grad_norm": 0.8645505309104919,
+      "learning_rate": 3.4987995198079235e-05,
+      "loss": 0.5772,
+      "step": 7513
+    },
+    {
+      "epoch": 9.61792,
+      "grad_norm": 0.8549172878265381,
+      "learning_rate": 3.49859943977591e-05,
+      "loss": 0.5876,
+      "step": 7514
+    },
+    {
+      "epoch": 9.6192,
+      "grad_norm": 0.8866432309150696,
+      "learning_rate": 3.498399359743898e-05,
+      "loss": 0.5682,
+      "step": 7515
+    },
+    {
+      "epoch": 9.62048,
+      "grad_norm": 0.8328559398651123,
+      "learning_rate": 3.498199279711885e-05,
+      "loss": 0.568,
+      "step": 7516
+    },
+    {
+      "epoch": 9.62176,
+      "grad_norm": 0.8284956216812134,
+      "learning_rate": 3.497999199679872e-05,
+      "loss": 0.5592,
+      "step": 7517
+    },
+    {
+      "epoch": 9.62304,
+      "grad_norm": 0.8239821195602417,
+      "learning_rate": 3.4977991196478594e-05,
+      "loss": 0.6051,
+      "step": 7518
+    },
+    {
+      "epoch": 9.62432,
+      "grad_norm": 0.8453178405761719,
+      "learning_rate": 3.4975990396158466e-05,
+      "loss": 0.5632,
+      "step": 7519
+    },
+    {
+      "epoch": 9.6256,
+      "grad_norm": 0.8655393123626709,
+      "learning_rate": 3.497398959583834e-05,
+      "loss": 0.6542,
+      "step": 7520
+    },
+    {
+      "epoch": 9.62688,
+      "grad_norm": 0.8494473099708557,
+      "learning_rate": 3.497198879551821e-05,
+      "loss": 0.584,
+      "step": 7521
+    },
+    {
+      "epoch": 9.62816,
+      "grad_norm": 0.8517120480537415,
+      "learning_rate": 3.496998799519808e-05,
+      "loss": 0.5515,
+      "step": 7522
+    },
+    {
+      "epoch": 9.62944,
+      "grad_norm": 0.8242863416671753,
+      "learning_rate": 3.4967987194877953e-05,
+      "loss": 0.5267,
+      "step": 7523
+    },
+    {
+      "epoch": 9.63072,
+      "grad_norm": 0.8850992321968079,
+      "learning_rate": 3.4965986394557825e-05,
+      "loss": 0.6003,
+      "step": 7524
+    },
+    {
+      "epoch": 9.632,
+      "grad_norm": 0.8448317050933838,
+      "learning_rate": 3.49639855942377e-05,
+      "loss": 0.6187,
+      "step": 7525
+    },
+    {
+      "epoch": 9.63328,
+      "grad_norm": 0.8723029494285583,
+      "learning_rate": 3.496198479391757e-05,
+      "loss": 0.6129,
+      "step": 7526
+    },
+    {
+      "epoch": 9.63456,
+      "grad_norm": 0.8665981292724609,
+      "learning_rate": 3.495998399359744e-05,
+      "loss": 0.5949,
+      "step": 7527
+    },
+    {
+      "epoch": 9.63584,
+      "grad_norm": 0.8512018322944641,
+      "learning_rate": 3.495798319327731e-05,
+      "loss": 0.5528,
+      "step": 7528
+    },
+    {
+      "epoch": 9.63712,
+      "grad_norm": 0.8611388802528381,
+      "learning_rate": 3.4955982392957185e-05,
+      "loss": 0.5936,
+      "step": 7529
+    },
+    {
+      "epoch": 9.6384,
+      "grad_norm": 0.865940272808075,
+      "learning_rate": 3.4953981592637056e-05,
+      "loss": 0.6168,
+      "step": 7530
+    },
+    {
+      "epoch": 9.63968,
+      "grad_norm": 0.8968199491500854,
+      "learning_rate": 3.495198079231693e-05,
+      "loss": 0.6163,
+      "step": 7531
+    },
+    {
+      "epoch": 9.64096,
+      "grad_norm": 0.8725362420082092,
+      "learning_rate": 3.49499799919968e-05,
+      "loss": 0.6209,
+      "step": 7532
+    },
+    {
+      "epoch": 9.64224,
+      "grad_norm": 0.8303427696228027,
+      "learning_rate": 3.494797919167667e-05,
+      "loss": 0.5585,
+      "step": 7533
+    },
+    {
+      "epoch": 9.64352,
+      "grad_norm": 0.8281154036521912,
+      "learning_rate": 3.4945978391356544e-05,
+      "loss": 0.5885,
+      "step": 7534
+    },
+    {
+      "epoch": 9.6448,
+      "grad_norm": 0.8495305776596069,
+      "learning_rate": 3.4943977591036416e-05,
+      "loss": 0.6193,
+      "step": 7535
+    },
+    {
+      "epoch": 9.64608,
+      "grad_norm": 0.8367983102798462,
+      "learning_rate": 3.494197679071629e-05,
+      "loss": 0.5942,
+      "step": 7536
+    },
+    {
+      "epoch": 9.64736,
+      "grad_norm": 0.8438671231269836,
+      "learning_rate": 3.493997599039616e-05,
+      "loss": 0.5856,
+      "step": 7537
+    },
+    {
+      "epoch": 9.64864,
+      "grad_norm": 0.8848779797554016,
+      "learning_rate": 3.493797519007603e-05,
+      "loss": 0.6449,
+      "step": 7538
+    },
+    {
+      "epoch": 9.64992,
+      "grad_norm": 0.7978121042251587,
+      "learning_rate": 3.49359743897559e-05,
+      "loss": 0.573,
+      "step": 7539
+    },
+    {
+      "epoch": 9.6512,
+      "grad_norm": 0.807237446308136,
+      "learning_rate": 3.4933973589435775e-05,
+      "loss": 0.5753,
+      "step": 7540
+    },
+    {
+      "epoch": 9.65248,
+      "grad_norm": 0.7662572264671326,
+      "learning_rate": 3.493197278911565e-05,
+      "loss": 0.5361,
+      "step": 7541
+    },
+    {
+      "epoch": 9.65376,
+      "grad_norm": 0.8328047394752502,
+      "learning_rate": 3.492997198879552e-05,
+      "loss": 0.5792,
+      "step": 7542
+    },
+    {
+      "epoch": 9.65504,
+      "grad_norm": 0.7806389331817627,
+      "learning_rate": 3.492797118847539e-05,
+      "loss": 0.5287,
+      "step": 7543
+    },
+    {
+      "epoch": 9.656320000000001,
+      "grad_norm": 0.8597953915596008,
+      "learning_rate": 3.492597038815526e-05,
+      "loss": 0.5512,
+      "step": 7544
+    },
+    {
+      "epoch": 9.6576,
+      "grad_norm": 0.8092467784881592,
+      "learning_rate": 3.4923969587835134e-05,
+      "loss": 0.557,
+      "step": 7545
+    },
+    {
+      "epoch": 9.65888,
+      "grad_norm": 0.8465959429740906,
+      "learning_rate": 3.4921968787515006e-05,
+      "loss": 0.5901,
+      "step": 7546
+    },
+    {
+      "epoch": 9.66016,
+      "grad_norm": 0.8790938258171082,
+      "learning_rate": 3.491996798719488e-05,
+      "loss": 0.5985,
+      "step": 7547
+    },
+    {
+      "epoch": 9.66144,
+      "grad_norm": 0.8606868982315063,
+      "learning_rate": 3.491796718687475e-05,
+      "loss": 0.6082,
+      "step": 7548
+    },
+    {
+      "epoch": 9.66272,
+      "grad_norm": 0.8462215065956116,
+      "learning_rate": 3.491596638655462e-05,
+      "loss": 0.6326,
+      "step": 7549
+    },
+    {
+      "epoch": 9.664,
+      "grad_norm": 0.8208281993865967,
+      "learning_rate": 3.4913965586234494e-05,
+      "loss": 0.5717,
+      "step": 7550
+    },
+    {
+      "epoch": 9.66528,
+      "grad_norm": 0.8237935900688171,
+      "learning_rate": 3.491196478591437e-05,
+      "loss": 0.5861,
+      "step": 7551
+    },
+    {
+      "epoch": 9.66656,
+      "grad_norm": 0.8763028979301453,
+      "learning_rate": 3.490996398559424e-05,
+      "loss": 0.5965,
+      "step": 7552
+    },
+    {
+      "epoch": 9.66784,
+      "grad_norm": 0.7981544733047485,
+      "learning_rate": 3.490796318527411e-05,
+      "loss": 0.51,
+      "step": 7553
+    },
+    {
+      "epoch": 9.66912,
+      "grad_norm": 0.8098160624504089,
+      "learning_rate": 3.490596238495398e-05,
+      "loss": 0.5858,
+      "step": 7554
+    },
+    {
+      "epoch": 9.6704,
+      "grad_norm": 0.8530492782592773,
+      "learning_rate": 3.490396158463385e-05,
+      "loss": 0.5727,
+      "step": 7555
+    },
+    {
+      "epoch": 9.67168,
+      "grad_norm": 0.8604266047477722,
+      "learning_rate": 3.4901960784313725e-05,
+      "loss": 0.6063,
+      "step": 7556
+    },
+    {
+      "epoch": 9.67296,
+      "grad_norm": 0.8214526772499084,
+      "learning_rate": 3.4899959983993597e-05,
+      "loss": 0.549,
+      "step": 7557
+    },
+    {
+      "epoch": 9.67424,
+      "grad_norm": 0.7836440205574036,
+      "learning_rate": 3.4897959183673475e-05,
+      "loss": 0.5446,
+      "step": 7558
+    },
+    {
+      "epoch": 9.67552,
+      "grad_norm": 0.8219870924949646,
+      "learning_rate": 3.489595838335335e-05,
+      "loss": 0.5766,
+      "step": 7559
+    },
+    {
+      "epoch": 9.6768,
+      "grad_norm": 0.8548493385314941,
+      "learning_rate": 3.489395758303321e-05,
+      "loss": 0.5894,
+      "step": 7560
+    },
+    {
+      "epoch": 9.67808,
+      "grad_norm": 0.8581690788269043,
+      "learning_rate": 3.4891956782713084e-05,
+      "loss": 0.549,
+      "step": 7561
+    },
+    {
+      "epoch": 9.679359999999999,
+      "grad_norm": 0.858836829662323,
+      "learning_rate": 3.4889955982392956e-05,
+      "loss": 0.5783,
+      "step": 7562
+    },
+    {
+      "epoch": 9.68064,
+      "grad_norm": 0.840040385723114,
+      "learning_rate": 3.488795518207283e-05,
+      "loss": 0.6081,
+      "step": 7563
+    },
+    {
+      "epoch": 9.68192,
+      "grad_norm": 0.8569283485412598,
+      "learning_rate": 3.48859543817527e-05,
+      "loss": 0.61,
+      "step": 7564
+    },
+    {
+      "epoch": 9.6832,
+      "grad_norm": 0.843917191028595,
+      "learning_rate": 3.488395358143258e-05,
+      "loss": 0.5821,
+      "step": 7565
+    },
+    {
+      "epoch": 9.68448,
+      "grad_norm": 0.8357688784599304,
+      "learning_rate": 3.488195278111245e-05,
+      "loss": 0.5624,
+      "step": 7566
+    },
+    {
+      "epoch": 9.68576,
+      "grad_norm": 0.8503354787826538,
+      "learning_rate": 3.487995198079232e-05,
+      "loss": 0.6566,
+      "step": 7567
+    },
+    {
+      "epoch": 9.68704,
+      "grad_norm": 0.8254701495170593,
+      "learning_rate": 3.487795118047219e-05,
+      "loss": 0.5381,
+      "step": 7568
+    },
+    {
+      "epoch": 9.688320000000001,
+      "grad_norm": 0.899823784828186,
+      "learning_rate": 3.487595038015206e-05,
+      "loss": 0.578,
+      "step": 7569
+    },
+    {
+      "epoch": 9.6896,
+      "grad_norm": 0.8426713943481445,
+      "learning_rate": 3.487394957983193e-05,
+      "loss": 0.5818,
+      "step": 7570
+    },
+    {
+      "epoch": 9.69088,
+      "grad_norm": 0.846393346786499,
+      "learning_rate": 3.48719487795118e-05,
+      "loss": 0.5821,
+      "step": 7571
+    },
+    {
+      "epoch": 9.69216,
+      "grad_norm": 0.8248345255851746,
+      "learning_rate": 3.486994797919168e-05,
+      "loss": 0.5701,
+      "step": 7572
+    },
+    {
+      "epoch": 9.69344,
+      "grad_norm": 0.8395203948020935,
+      "learning_rate": 3.486794717887155e-05,
+      "loss": 0.5862,
+      "step": 7573
+    },
+    {
+      "epoch": 9.69472,
+      "grad_norm": 0.8293047547340393,
+      "learning_rate": 3.4865946378551425e-05,
+      "loss": 0.5585,
+      "step": 7574
+    },
+    {
+      "epoch": 9.696,
+      "grad_norm": 0.848100483417511,
+      "learning_rate": 3.48639455782313e-05,
+      "loss": 0.5687,
+      "step": 7575
+    },
+    {
+      "epoch": 9.69728,
+      "grad_norm": 0.8797847032546997,
+      "learning_rate": 3.486194477791116e-05,
+      "loss": 0.6331,
+      "step": 7576
+    },
+    {
+      "epoch": 9.69856,
+      "grad_norm": 0.7995453476905823,
+      "learning_rate": 3.4859943977591034e-05,
+      "loss": 0.5394,
+      "step": 7577
+    },
+    {
+      "epoch": 9.69984,
+      "grad_norm": 0.8336317539215088,
+      "learning_rate": 3.4857943177270906e-05,
+      "loss": 0.6146,
+      "step": 7578
+    },
+    {
+      "epoch": 9.70112,
+      "grad_norm": 0.8199133276939392,
+      "learning_rate": 3.4855942376950784e-05,
+      "loss": 0.596,
+      "step": 7579
+    },
+    {
+      "epoch": 9.7024,
+      "grad_norm": 0.8550775647163391,
+      "learning_rate": 3.4853941576630656e-05,
+      "loss": 0.5881,
+      "step": 7580
+    },
+    {
+      "epoch": 9.70368,
+      "grad_norm": 0.8737037777900696,
+      "learning_rate": 3.485194077631053e-05,
+      "loss": 0.6125,
+      "step": 7581
+    },
+    {
+      "epoch": 9.70496,
+      "grad_norm": 0.8511414527893066,
+      "learning_rate": 3.48499399759904e-05,
+      "loss": 0.5995,
+      "step": 7582
+    },
+    {
+      "epoch": 9.70624,
+      "grad_norm": 0.8362718224525452,
+      "learning_rate": 3.484793917567027e-05,
+      "loss": 0.5418,
+      "step": 7583
+    },
+    {
+      "epoch": 9.70752,
+      "grad_norm": 0.8965070247650146,
+      "learning_rate": 3.484593837535014e-05,
+      "loss": 0.6493,
+      "step": 7584
+    },
+    {
+      "epoch": 9.7088,
+      "grad_norm": 0.8419719338417053,
+      "learning_rate": 3.484393757503001e-05,
+      "loss": 0.5813,
+      "step": 7585
+    },
+    {
+      "epoch": 9.71008,
+      "grad_norm": 0.903337299823761,
+      "learning_rate": 3.484193677470989e-05,
+      "loss": 0.6155,
+      "step": 7586
+    },
+    {
+      "epoch": 9.711359999999999,
+      "grad_norm": 0.8451673984527588,
+      "learning_rate": 3.483993597438976e-05,
+      "loss": 0.5438,
+      "step": 7587
+    },
+    {
+      "epoch": 9.71264,
+      "grad_norm": 0.8772710561752319,
+      "learning_rate": 3.483793517406963e-05,
+      "loss": 0.5555,
+      "step": 7588
+    },
+    {
+      "epoch": 9.71392,
+      "grad_norm": 0.8326647877693176,
+      "learning_rate": 3.48359343737495e-05,
+      "loss": 0.6242,
+      "step": 7589
+    },
+    {
+      "epoch": 9.7152,
+      "grad_norm": 0.8710974454879761,
+      "learning_rate": 3.4833933573429375e-05,
+      "loss": 0.6274,
+      "step": 7590
+    },
+    {
+      "epoch": 9.71648,
+      "grad_norm": 0.8127773404121399,
+      "learning_rate": 3.4831932773109246e-05,
+      "loss": 0.6033,
+      "step": 7591
+    },
+    {
+      "epoch": 9.71776,
+      "grad_norm": 0.7887704372406006,
+      "learning_rate": 3.482993197278911e-05,
+      "loss": 0.5284,
+      "step": 7592
+    },
+    {
+      "epoch": 9.71904,
+      "grad_norm": 0.8618784546852112,
+      "learning_rate": 3.482793117246899e-05,
+      "loss": 0.6029,
+      "step": 7593
+    },
+    {
+      "epoch": 9.72032,
+      "grad_norm": 0.8521010875701904,
+      "learning_rate": 3.482593037214886e-05,
+      "loss": 0.5912,
+      "step": 7594
+    },
+    {
+      "epoch": 9.7216,
+      "grad_norm": 0.8535512685775757,
+      "learning_rate": 3.4823929571828734e-05,
+      "loss": 0.5922,
+      "step": 7595
+    },
+    {
+      "epoch": 9.72288,
+      "grad_norm": 0.8280578255653381,
+      "learning_rate": 3.4821928771508606e-05,
+      "loss": 0.5867,
+      "step": 7596
+    },
+    {
+      "epoch": 9.72416,
+      "grad_norm": 0.7792149782180786,
+      "learning_rate": 3.481992797118848e-05,
+      "loss": 0.5363,
+      "step": 7597
+    },
+    {
+      "epoch": 9.72544,
+      "grad_norm": 0.7704604864120483,
+      "learning_rate": 3.481792717086835e-05,
+      "loss": 0.5437,
+      "step": 7598
+    },
+    {
+      "epoch": 9.72672,
+      "grad_norm": 0.819395899772644,
+      "learning_rate": 3.481592637054822e-05,
+      "loss": 0.5802,
+      "step": 7599
+    },
+    {
+      "epoch": 9.728,
+      "grad_norm": 0.83621746301651,
+      "learning_rate": 3.481392557022809e-05,
+      "loss": 0.579,
+      "step": 7600
+    },
+    {
+      "epoch": 9.72928,
+      "grad_norm": 0.8944879770278931,
+      "learning_rate": 3.4811924769907965e-05,
+      "loss": 0.6284,
+      "step": 7601
+    },
+    {
+      "epoch": 9.73056,
+      "grad_norm": 0.8720524311065674,
+      "learning_rate": 3.480992396958784e-05,
+      "loss": 0.5917,
+      "step": 7602
+    },
+    {
+      "epoch": 9.73184,
+      "grad_norm": 0.877223789691925,
+      "learning_rate": 3.480792316926771e-05,
+      "loss": 0.591,
+      "step": 7603
+    },
+    {
+      "epoch": 9.73312,
+      "grad_norm": 0.9013481140136719,
+      "learning_rate": 3.480592236894758e-05,
+      "loss": 0.6077,
+      "step": 7604
+    },
+    {
+      "epoch": 9.7344,
+      "grad_norm": 0.8513638973236084,
+      "learning_rate": 3.480392156862745e-05,
+      "loss": 0.5542,
+      "step": 7605
+    },
+    {
+      "epoch": 9.73568,
+      "grad_norm": 0.862791121006012,
+      "learning_rate": 3.4801920768307324e-05,
+      "loss": 0.5792,
+      "step": 7606
+    },
+    {
+      "epoch": 9.73696,
+      "grad_norm": 0.8830035328865051,
+      "learning_rate": 3.4799919967987196e-05,
+      "loss": 0.5655,
+      "step": 7607
+    },
+    {
+      "epoch": 9.73824,
+      "grad_norm": 0.8427029252052307,
+      "learning_rate": 3.479791916766707e-05,
+      "loss": 0.5665,
+      "step": 7608
+    },
+    {
+      "epoch": 9.73952,
+      "grad_norm": 0.8503885269165039,
+      "learning_rate": 3.479591836734694e-05,
+      "loss": 0.5891,
+      "step": 7609
+    },
+    {
+      "epoch": 9.7408,
+      "grad_norm": 0.8855993151664734,
+      "learning_rate": 3.479391756702681e-05,
+      "loss": 0.603,
+      "step": 7610
+    },
+    {
+      "epoch": 9.74208,
+      "grad_norm": 0.8236071467399597,
+      "learning_rate": 3.4791916766706684e-05,
+      "loss": 0.5416,
+      "step": 7611
+    },
+    {
+      "epoch": 9.74336,
+      "grad_norm": 0.8632338643074036,
+      "learning_rate": 3.4789915966386555e-05,
+      "loss": 0.572,
+      "step": 7612
+    },
+    {
+      "epoch": 9.74464,
+      "grad_norm": 0.8438730239868164,
+      "learning_rate": 3.478791516606643e-05,
+      "loss": 0.5588,
+      "step": 7613
+    },
+    {
+      "epoch": 9.74592,
+      "grad_norm": 0.8554427027702332,
+      "learning_rate": 3.4785914365746306e-05,
+      "loss": 0.596,
+      "step": 7614
+    },
+    {
+      "epoch": 9.7472,
+      "grad_norm": 0.8517599701881409,
+      "learning_rate": 3.478391356542617e-05,
+      "loss": 0.5597,
+      "step": 7615
+    },
+    {
+      "epoch": 9.74848,
+      "grad_norm": 0.8286646604537964,
+      "learning_rate": 3.478191276510604e-05,
+      "loss": 0.5932,
+      "step": 7616
+    },
+    {
+      "epoch": 9.74976,
+      "grad_norm": 0.7939943075180054,
+      "learning_rate": 3.4779911964785915e-05,
+      "loss": 0.5925,
+      "step": 7617
+    },
+    {
+      "epoch": 9.75104,
+      "grad_norm": 0.8272203803062439,
+      "learning_rate": 3.4777911164465787e-05,
+      "loss": 0.6005,
+      "step": 7618
+    },
+    {
+      "epoch": 9.75232,
+      "grad_norm": 0.8649276494979858,
+      "learning_rate": 3.477591036414566e-05,
+      "loss": 0.5847,
+      "step": 7619
+    },
+    {
+      "epoch": 9.7536,
+      "grad_norm": 0.7952982783317566,
+      "learning_rate": 3.477390956382553e-05,
+      "loss": 0.5359,
+      "step": 7620
+    },
+    {
+      "epoch": 9.75488,
+      "grad_norm": 0.8271358609199524,
+      "learning_rate": 3.477190876350541e-05,
+      "loss": 0.5964,
+      "step": 7621
+    },
+    {
+      "epoch": 9.75616,
+      "grad_norm": 0.8045120239257812,
+      "learning_rate": 3.476990796318528e-05,
+      "loss": 0.5676,
+      "step": 7622
+    },
+    {
+      "epoch": 9.75744,
+      "grad_norm": 0.7775219082832336,
+      "learning_rate": 3.4767907162865146e-05,
+      "loss": 0.5519,
+      "step": 7623
+    },
+    {
+      "epoch": 9.75872,
+      "grad_norm": 0.8189329504966736,
+      "learning_rate": 3.476590636254502e-05,
+      "loss": 0.5834,
+      "step": 7624
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 0.8617966771125793,
+      "learning_rate": 3.476390556222489e-05,
+      "loss": 0.5863,
+      "step": 7625
+    },
+    {
+      "epoch": 9.76128,
+      "grad_norm": 0.8166400790214539,
+      "learning_rate": 3.476190476190476e-05,
+      "loss": 0.5754,
+      "step": 7626
+    },
+    {
+      "epoch": 9.76256,
+      "grad_norm": 0.8374921679496765,
+      "learning_rate": 3.475990396158463e-05,
+      "loss": 0.557,
+      "step": 7627
+    },
+    {
+      "epoch": 9.76384,
+      "grad_norm": 0.7682600617408752,
+      "learning_rate": 3.475790316126451e-05,
+      "loss": 0.5077,
+      "step": 7628
+    },
+    {
+      "epoch": 9.76512,
+      "grad_norm": 0.8810880780220032,
+      "learning_rate": 3.4755902360944384e-05,
+      "loss": 0.6657,
+      "step": 7629
+    },
+    {
+      "epoch": 9.7664,
+      "grad_norm": 0.8713139295578003,
+      "learning_rate": 3.4753901560624256e-05,
+      "loss": 0.6198,
+      "step": 7630
+    },
+    {
+      "epoch": 9.76768,
+      "grad_norm": 0.858737051486969,
+      "learning_rate": 3.475190076030412e-05,
+      "loss": 0.5409,
+      "step": 7631
+    },
+    {
+      "epoch": 9.76896,
+      "grad_norm": 0.7983132004737854,
+      "learning_rate": 3.474989995998399e-05,
+      "loss": 0.5621,
+      "step": 7632
+    },
+    {
+      "epoch": 9.77024,
+      "grad_norm": 0.8178649544715881,
+      "learning_rate": 3.4747899159663864e-05,
+      "loss": 0.5632,
+      "step": 7633
+    },
+    {
+      "epoch": 9.77152,
+      "grad_norm": 0.8813498020172119,
+      "learning_rate": 3.4745898359343736e-05,
+      "loss": 0.6399,
+      "step": 7634
+    },
+    {
+      "epoch": 9.7728,
+      "grad_norm": 0.8513997197151184,
+      "learning_rate": 3.4743897559023615e-05,
+      "loss": 0.5805,
+      "step": 7635
+    },
+    {
+      "epoch": 9.77408,
+      "grad_norm": 0.8140328526496887,
+      "learning_rate": 3.474189675870349e-05,
+      "loss": 0.534,
+      "step": 7636
+    },
+    {
+      "epoch": 9.77536,
+      "grad_norm": 0.8354536890983582,
+      "learning_rate": 3.473989595838336e-05,
+      "loss": 0.5771,
+      "step": 7637
+    },
+    {
+      "epoch": 9.77664,
+      "grad_norm": 0.8029643297195435,
+      "learning_rate": 3.473789515806323e-05,
+      "loss": 0.527,
+      "step": 7638
+    },
+    {
+      "epoch": 9.77792,
+      "grad_norm": 0.8642273545265198,
+      "learning_rate": 3.4735894357743096e-05,
+      "loss": 0.591,
+      "step": 7639
+    },
+    {
+      "epoch": 9.7792,
+      "grad_norm": 0.8148753643035889,
+      "learning_rate": 3.473389355742297e-05,
+      "loss": 0.5682,
+      "step": 7640
+    },
+    {
+      "epoch": 9.78048,
+      "grad_norm": 0.8324840664863586,
+      "learning_rate": 3.473189275710284e-05,
+      "loss": 0.5479,
+      "step": 7641
+    },
+    {
+      "epoch": 9.78176,
+      "grad_norm": 0.911037027835846,
+      "learning_rate": 3.472989195678272e-05,
+      "loss": 0.6622,
+      "step": 7642
+    },
+    {
+      "epoch": 9.78304,
+      "grad_norm": 0.847952663898468,
+      "learning_rate": 3.472789115646259e-05,
+      "loss": 0.5716,
+      "step": 7643
+    },
+    {
+      "epoch": 9.78432,
+      "grad_norm": 0.8314871191978455,
+      "learning_rate": 3.472589035614246e-05,
+      "loss": 0.626,
+      "step": 7644
+    },
+    {
+      "epoch": 9.7856,
+      "grad_norm": 0.8641082048416138,
+      "learning_rate": 3.4723889555822333e-05,
+      "loss": 0.5465,
+      "step": 7645
+    },
+    {
+      "epoch": 9.78688,
+      "grad_norm": 0.834967851638794,
+      "learning_rate": 3.4721888755502205e-05,
+      "loss": 0.5653,
+      "step": 7646
+    },
+    {
+      "epoch": 9.78816,
+      "grad_norm": 0.8525474667549133,
+      "learning_rate": 3.471988795518207e-05,
+      "loss": 0.5715,
+      "step": 7647
+    },
+    {
+      "epoch": 9.78944,
+      "grad_norm": 0.8522567749023438,
+      "learning_rate": 3.471788715486194e-05,
+      "loss": 0.5699,
+      "step": 7648
+    },
+    {
+      "epoch": 9.79072,
+      "grad_norm": 0.8890289068222046,
+      "learning_rate": 3.471588635454182e-05,
+      "loss": 0.6542,
+      "step": 7649
+    },
+    {
+      "epoch": 9.792,
+      "grad_norm": 0.8257919549942017,
+      "learning_rate": 3.471388555422169e-05,
+      "loss": 0.5972,
+      "step": 7650
+    },
+    {
+      "epoch": 9.79328,
+      "grad_norm": 0.8120242357254028,
+      "learning_rate": 3.4711884753901565e-05,
+      "loss": 0.5841,
+      "step": 7651
+    },
+    {
+      "epoch": 9.79456,
+      "grad_norm": 0.8151296973228455,
+      "learning_rate": 3.4709883953581436e-05,
+      "loss": 0.5583,
+      "step": 7652
+    },
+    {
+      "epoch": 9.79584,
+      "grad_norm": 0.8686169981956482,
+      "learning_rate": 3.470788315326131e-05,
+      "loss": 0.5816,
+      "step": 7653
+    },
+    {
+      "epoch": 9.79712,
+      "grad_norm": 0.871562123298645,
+      "learning_rate": 3.470588235294118e-05,
+      "loss": 0.5792,
+      "step": 7654
+    },
+    {
+      "epoch": 9.7984,
+      "grad_norm": 0.8428971767425537,
+      "learning_rate": 3.4703881552621045e-05,
+      "loss": 0.517,
+      "step": 7655
+    },
+    {
+      "epoch": 9.79968,
+      "grad_norm": 0.8845090866088867,
+      "learning_rate": 3.470188075230092e-05,
+      "loss": 0.6303,
+      "step": 7656
+    },
+    {
+      "epoch": 9.80096,
+      "grad_norm": 0.8464072942733765,
+      "learning_rate": 3.4699879951980796e-05,
+      "loss": 0.5476,
+      "step": 7657
+    },
+    {
+      "epoch": 9.80224,
+      "grad_norm": 0.871966540813446,
+      "learning_rate": 3.469787915166067e-05,
+      "loss": 0.5654,
+      "step": 7658
+    },
+    {
+      "epoch": 9.80352,
+      "grad_norm": 0.811346173286438,
+      "learning_rate": 3.469587835134054e-05,
+      "loss": 0.5738,
+      "step": 7659
+    },
+    {
+      "epoch": 9.8048,
+      "grad_norm": 0.8650268316268921,
+      "learning_rate": 3.469387755102041e-05,
+      "loss": 0.6266,
+      "step": 7660
+    },
+    {
+      "epoch": 9.80608,
+      "grad_norm": 0.7988438010215759,
+      "learning_rate": 3.469187675070028e-05,
+      "loss": 0.5255,
+      "step": 7661
+    },
+    {
+      "epoch": 9.80736,
+      "grad_norm": 0.8700106739997864,
+      "learning_rate": 3.4689875950380155e-05,
+      "loss": 0.6087,
+      "step": 7662
+    },
+    {
+      "epoch": 9.80864,
+      "grad_norm": 0.8239459991455078,
+      "learning_rate": 3.468787515006002e-05,
+      "loss": 0.5602,
+      "step": 7663
+    },
+    {
+      "epoch": 9.80992,
+      "grad_norm": 0.9134540557861328,
+      "learning_rate": 3.46858743497399e-05,
+      "loss": 0.5968,
+      "step": 7664
+    },
+    {
+      "epoch": 9.8112,
+      "grad_norm": 0.9253342747688293,
+      "learning_rate": 3.468387354941977e-05,
+      "loss": 0.6128,
+      "step": 7665
+    },
+    {
+      "epoch": 9.81248,
+      "grad_norm": 0.7687677145004272,
+      "learning_rate": 3.468187274909964e-05,
+      "loss": 0.5191,
+      "step": 7666
+    },
+    {
+      "epoch": 9.81376,
+      "grad_norm": 0.8132002949714661,
+      "learning_rate": 3.4679871948779514e-05,
+      "loss": 0.5736,
+      "step": 7667
+    },
+    {
+      "epoch": 9.81504,
+      "grad_norm": 0.8032669425010681,
+      "learning_rate": 3.4677871148459386e-05,
+      "loss": 0.5311,
+      "step": 7668
+    },
+    {
+      "epoch": 9.81632,
+      "grad_norm": 0.8530933260917664,
+      "learning_rate": 3.467587034813926e-05,
+      "loss": 0.5699,
+      "step": 7669
+    },
+    {
+      "epoch": 9.8176,
+      "grad_norm": 0.8100473880767822,
+      "learning_rate": 3.467386954781913e-05,
+      "loss": 0.5431,
+      "step": 7670
+    },
+    {
+      "epoch": 9.81888,
+      "grad_norm": 0.8088340759277344,
+      "learning_rate": 3.4671868747499e-05,
+      "loss": 0.5454,
+      "step": 7671
+    },
+    {
+      "epoch": 9.82016,
+      "grad_norm": 0.8359904289245605,
+      "learning_rate": 3.4669867947178874e-05,
+      "loss": 0.567,
+      "step": 7672
+    },
+    {
+      "epoch": 9.821439999999999,
+      "grad_norm": 0.8769454956054688,
+      "learning_rate": 3.4667867146858745e-05,
+      "loss": 0.5966,
+      "step": 7673
+    },
+    {
+      "epoch": 9.82272,
+      "grad_norm": 0.8174176812171936,
+      "learning_rate": 3.466586634653862e-05,
+      "loss": 0.5452,
+      "step": 7674
+    },
+    {
+      "epoch": 9.824,
+      "grad_norm": 0.8501536846160889,
+      "learning_rate": 3.466386554621849e-05,
+      "loss": 0.5929,
+      "step": 7675
+    },
+    {
+      "epoch": 9.82528,
+      "grad_norm": 0.8151215314865112,
+      "learning_rate": 3.466186474589836e-05,
+      "loss": 0.552,
+      "step": 7676
+    },
+    {
+      "epoch": 9.82656,
+      "grad_norm": 0.8890595436096191,
+      "learning_rate": 3.465986394557823e-05,
+      "loss": 0.6142,
+      "step": 7677
+    },
+    {
+      "epoch": 9.82784,
+      "grad_norm": 0.8509846329689026,
+      "learning_rate": 3.4657863145258105e-05,
+      "loss": 0.6358,
+      "step": 7678
+    },
+    {
+      "epoch": 9.82912,
+      "grad_norm": 0.8409616947174072,
+      "learning_rate": 3.465586234493798e-05,
+      "loss": 0.5594,
+      "step": 7679
+    },
+    {
+      "epoch": 9.830400000000001,
+      "grad_norm": 0.8827511072158813,
+      "learning_rate": 3.465386154461785e-05,
+      "loss": 0.6054,
+      "step": 7680
+    },
+    {
+      "epoch": 9.83168,
+      "grad_norm": 0.8123102784156799,
+      "learning_rate": 3.465186074429772e-05,
+      "loss": 0.5833,
+      "step": 7681
+    },
+    {
+      "epoch": 9.83296,
+      "grad_norm": 0.8210862874984741,
+      "learning_rate": 3.464985994397759e-05,
+      "loss": 0.6082,
+      "step": 7682
+    },
+    {
+      "epoch": 9.83424,
+      "grad_norm": 0.8209624886512756,
+      "learning_rate": 3.4647859143657464e-05,
+      "loss": 0.615,
+      "step": 7683
+    },
+    {
+      "epoch": 9.83552,
+      "grad_norm": 0.8852274417877197,
+      "learning_rate": 3.4645858343337336e-05,
+      "loss": 0.6337,
+      "step": 7684
+    },
+    {
+      "epoch": 9.8368,
+      "grad_norm": 0.8162413239479065,
+      "learning_rate": 3.464385754301721e-05,
+      "loss": 0.5424,
+      "step": 7685
+    },
+    {
+      "epoch": 9.83808,
+      "grad_norm": 0.8327953219413757,
+      "learning_rate": 3.464185674269708e-05,
+      "loss": 0.5759,
+      "step": 7686
+    },
+    {
+      "epoch": 9.83936,
+      "grad_norm": 0.8537192940711975,
+      "learning_rate": 3.463985594237695e-05,
+      "loss": 0.5714,
+      "step": 7687
+    },
+    {
+      "epoch": 9.84064,
+      "grad_norm": 0.8469899892807007,
+      "learning_rate": 3.463785514205682e-05,
+      "loss": 0.6088,
+      "step": 7688
+    },
+    {
+      "epoch": 9.84192,
+      "grad_norm": 0.8121939301490784,
+      "learning_rate": 3.4635854341736695e-05,
+      "loss": 0.5807,
+      "step": 7689
+    },
+    {
+      "epoch": 9.8432,
+      "grad_norm": 0.8225154280662537,
+      "learning_rate": 3.463385354141657e-05,
+      "loss": 0.5833,
+      "step": 7690
+    },
+    {
+      "epoch": 9.84448,
+      "grad_norm": 0.8836968541145325,
+      "learning_rate": 3.463185274109644e-05,
+      "loss": 0.5574,
+      "step": 7691
+    },
+    {
+      "epoch": 9.84576,
+      "grad_norm": 0.8945218920707703,
+      "learning_rate": 3.462985194077632e-05,
+      "loss": 0.6273,
+      "step": 7692
+    },
+    {
+      "epoch": 9.84704,
+      "grad_norm": 0.8279548287391663,
+      "learning_rate": 3.462785114045618e-05,
+      "loss": 0.5263,
+      "step": 7693
+    },
+    {
+      "epoch": 9.84832,
+      "grad_norm": 0.8554354906082153,
+      "learning_rate": 3.4625850340136054e-05,
+      "loss": 0.5709,
+      "step": 7694
+    },
+    {
+      "epoch": 9.8496,
+      "grad_norm": 0.8173097372055054,
+      "learning_rate": 3.4623849539815926e-05,
+      "loss": 0.5876,
+      "step": 7695
+    },
+    {
+      "epoch": 9.85088,
+      "grad_norm": 0.8793294429779053,
+      "learning_rate": 3.46218487394958e-05,
+      "loss": 0.627,
+      "step": 7696
+    },
+    {
+      "epoch": 9.85216,
+      "grad_norm": 0.7973728775978088,
+      "learning_rate": 3.461984793917567e-05,
+      "loss": 0.5243,
+      "step": 7697
+    },
+    {
+      "epoch": 9.853439999999999,
+      "grad_norm": 0.8718621730804443,
+      "learning_rate": 3.461784713885554e-05,
+      "loss": 0.5843,
+      "step": 7698
+    },
+    {
+      "epoch": 9.85472,
+      "grad_norm": 0.8487060070037842,
+      "learning_rate": 3.461584633853542e-05,
+      "loss": 0.5407,
+      "step": 7699
+    },
+    {
+      "epoch": 9.856,
+      "grad_norm": 0.8882485032081604,
+      "learning_rate": 3.461384553821529e-05,
+      "loss": 0.6204,
+      "step": 7700
+    },
+    {
+      "epoch": 9.85728,
+      "grad_norm": 0.8720253705978394,
+      "learning_rate": 3.461184473789516e-05,
+      "loss": 0.5962,
+      "step": 7701
+    },
+    {
+      "epoch": 9.85856,
+      "grad_norm": 0.8400206565856934,
+      "learning_rate": 3.460984393757503e-05,
+      "loss": 0.5695,
+      "step": 7702
+    },
+    {
+      "epoch": 9.85984,
+      "grad_norm": 0.9012857675552368,
+      "learning_rate": 3.46078431372549e-05,
+      "loss": 0.5909,
+      "step": 7703
+    },
+    {
+      "epoch": 9.86112,
+      "grad_norm": 0.8660620450973511,
+      "learning_rate": 3.460584233693477e-05,
+      "loss": 0.5646,
+      "step": 7704
+    },
+    {
+      "epoch": 9.862400000000001,
+      "grad_norm": 0.8514848947525024,
+      "learning_rate": 3.4603841536614645e-05,
+      "loss": 0.5869,
+      "step": 7705
+    },
+    {
+      "epoch": 9.86368,
+      "grad_norm": 0.8337460160255432,
+      "learning_rate": 3.4601840736294524e-05,
+      "loss": 0.5438,
+      "step": 7706
+    },
+    {
+      "epoch": 9.86496,
+      "grad_norm": 0.8334992527961731,
+      "learning_rate": 3.4599839935974395e-05,
+      "loss": 0.5825,
+      "step": 7707
+    },
+    {
+      "epoch": 9.86624,
+      "grad_norm": 0.8814105987548828,
+      "learning_rate": 3.459783913565427e-05,
+      "loss": 0.6023,
+      "step": 7708
+    },
+    {
+      "epoch": 9.86752,
+      "grad_norm": 0.9168762564659119,
+      "learning_rate": 3.459583833533413e-05,
+      "loss": 0.6323,
+      "step": 7709
+    },
+    {
+      "epoch": 9.8688,
+      "grad_norm": 0.8501449227333069,
+      "learning_rate": 3.4593837535014004e-05,
+      "loss": 0.5978,
+      "step": 7710
+    },
+    {
+      "epoch": 9.87008,
+      "grad_norm": 0.8778765201568604,
+      "learning_rate": 3.4591836734693876e-05,
+      "loss": 0.5654,
+      "step": 7711
+    },
+    {
+      "epoch": 9.87136,
+      "grad_norm": 0.8818456530570984,
+      "learning_rate": 3.458983593437375e-05,
+      "loss": 0.6219,
+      "step": 7712
+    },
+    {
+      "epoch": 9.87264,
+      "grad_norm": 0.8822683095932007,
+      "learning_rate": 3.4587835134053627e-05,
+      "loss": 0.6086,
+      "step": 7713
+    },
+    {
+      "epoch": 9.87392,
+      "grad_norm": 0.8153172731399536,
+      "learning_rate": 3.45858343337335e-05,
+      "loss": 0.568,
+      "step": 7714
+    },
+    {
+      "epoch": 9.8752,
+      "grad_norm": 0.8170477151870728,
+      "learning_rate": 3.458383353341337e-05,
+      "loss": 0.5734,
+      "step": 7715
+    },
+    {
+      "epoch": 9.87648,
+      "grad_norm": 0.8518624901771545,
+      "learning_rate": 3.458183273309324e-05,
+      "loss": 0.5697,
+      "step": 7716
+    },
+    {
+      "epoch": 9.87776,
+      "grad_norm": 0.8774121999740601,
+      "learning_rate": 3.457983193277311e-05,
+      "loss": 0.5983,
+      "step": 7717
+    },
+    {
+      "epoch": 9.87904,
+      "grad_norm": 0.851138710975647,
+      "learning_rate": 3.457783113245298e-05,
+      "loss": 0.6161,
+      "step": 7718
+    },
+    {
+      "epoch": 9.88032,
+      "grad_norm": 0.8621139526367188,
+      "learning_rate": 3.457583033213285e-05,
+      "loss": 0.6147,
+      "step": 7719
+    },
+    {
+      "epoch": 9.8816,
+      "grad_norm": 0.8349597454071045,
+      "learning_rate": 3.457382953181273e-05,
+      "loss": 0.5587,
+      "step": 7720
+    },
+    {
+      "epoch": 9.88288,
+      "grad_norm": 0.8482023477554321,
+      "learning_rate": 3.45718287314926e-05,
+      "loss": 0.5783,
+      "step": 7721
+    },
+    {
+      "epoch": 9.88416,
+      "grad_norm": 0.8426772952079773,
+      "learning_rate": 3.456982793117247e-05,
+      "loss": 0.5994,
+      "step": 7722
+    },
+    {
+      "epoch": 9.88544,
+      "grad_norm": 0.8138001561164856,
+      "learning_rate": 3.4567827130852345e-05,
+      "loss": 0.5165,
+      "step": 7723
+    },
+    {
+      "epoch": 9.88672,
+      "grad_norm": 0.838672399520874,
+      "learning_rate": 3.456582633053222e-05,
+      "loss": 0.5827,
+      "step": 7724
+    },
+    {
+      "epoch": 9.888,
+      "grad_norm": 0.8754329085350037,
+      "learning_rate": 3.456382553021208e-05,
+      "loss": 0.5673,
+      "step": 7725
+    },
+    {
+      "epoch": 9.88928,
+      "grad_norm": 0.856153130531311,
+      "learning_rate": 3.4561824729891954e-05,
+      "loss": 0.5857,
+      "step": 7726
+    },
+    {
+      "epoch": 9.89056,
+      "grad_norm": 0.8428956866264343,
+      "learning_rate": 3.455982392957183e-05,
+      "loss": 0.5538,
+      "step": 7727
+    },
+    {
+      "epoch": 9.89184,
+      "grad_norm": 0.8644785284996033,
+      "learning_rate": 3.4557823129251704e-05,
+      "loss": 0.5754,
+      "step": 7728
+    },
+    {
+      "epoch": 9.89312,
+      "grad_norm": 0.8557632565498352,
+      "learning_rate": 3.4555822328931576e-05,
+      "loss": 0.5598,
+      "step": 7729
+    },
+    {
+      "epoch": 9.8944,
+      "grad_norm": 0.84462970495224,
+      "learning_rate": 3.455382152861145e-05,
+      "loss": 0.59,
+      "step": 7730
+    },
+    {
+      "epoch": 9.89568,
+      "grad_norm": 0.8369764089584351,
+      "learning_rate": 3.455182072829132e-05,
+      "loss": 0.6021,
+      "step": 7731
+    },
+    {
+      "epoch": 9.89696,
+      "grad_norm": 0.8819150328636169,
+      "learning_rate": 3.454981992797119e-05,
+      "loss": 0.6122,
+      "step": 7732
+    },
+    {
+      "epoch": 9.89824,
+      "grad_norm": 0.8815730214118958,
+      "learning_rate": 3.454781912765106e-05,
+      "loss": 0.6513,
+      "step": 7733
+    },
+    {
+      "epoch": 9.89952,
+      "grad_norm": 0.8620604276657104,
+      "learning_rate": 3.4545818327330936e-05,
+      "loss": 0.5683,
+      "step": 7734
+    },
+    {
+      "epoch": 9.9008,
+      "grad_norm": 0.8745002746582031,
+      "learning_rate": 3.454381752701081e-05,
+      "loss": 0.5996,
+      "step": 7735
+    },
+    {
+      "epoch": 9.90208,
+      "grad_norm": 0.8810462355613708,
+      "learning_rate": 3.454181672669068e-05,
+      "loss": 0.6145,
+      "step": 7736
+    },
+    {
+      "epoch": 9.90336,
+      "grad_norm": 0.8663910031318665,
+      "learning_rate": 3.453981592637055e-05,
+      "loss": 0.5592,
+      "step": 7737
+    },
+    {
+      "epoch": 9.90464,
+      "grad_norm": 0.8494516611099243,
+      "learning_rate": 3.453781512605042e-05,
+      "loss": 0.5281,
+      "step": 7738
+    },
+    {
+      "epoch": 9.90592,
+      "grad_norm": 0.8416398763656616,
+      "learning_rate": 3.4535814325730295e-05,
+      "loss": 0.5736,
+      "step": 7739
+    },
+    {
+      "epoch": 9.9072,
+      "grad_norm": 0.7921009659767151,
+      "learning_rate": 3.453381352541017e-05,
+      "loss": 0.5327,
+      "step": 7740
+    },
+    {
+      "epoch": 9.90848,
+      "grad_norm": 0.8950327038764954,
+      "learning_rate": 3.453181272509004e-05,
+      "loss": 0.6223,
+      "step": 7741
+    },
+    {
+      "epoch": 9.90976,
+      "grad_norm": 0.8658446669578552,
+      "learning_rate": 3.452981192476991e-05,
+      "loss": 0.6088,
+      "step": 7742
+    },
+    {
+      "epoch": 9.91104,
+      "grad_norm": 0.8409397006034851,
+      "learning_rate": 3.452781112444978e-05,
+      "loss": 0.6228,
+      "step": 7743
+    },
+    {
+      "epoch": 9.91232,
+      "grad_norm": 0.873226523399353,
+      "learning_rate": 3.4525810324129654e-05,
+      "loss": 0.6051,
+      "step": 7744
+    },
+    {
+      "epoch": 9.9136,
+      "grad_norm": 0.8407812714576721,
+      "learning_rate": 3.4523809523809526e-05,
+      "loss": 0.5404,
+      "step": 7745
+    },
+    {
+      "epoch": 9.91488,
+      "grad_norm": 0.8529120087623596,
+      "learning_rate": 3.45218087234894e-05,
+      "loss": 0.6056,
+      "step": 7746
+    },
+    {
+      "epoch": 9.91616,
+      "grad_norm": 0.88858562707901,
+      "learning_rate": 3.451980792316927e-05,
+      "loss": 0.5821,
+      "step": 7747
+    },
+    {
+      "epoch": 9.91744,
+      "grad_norm": 0.8706713318824768,
+      "learning_rate": 3.451780712284914e-05,
+      "loss": 0.6058,
+      "step": 7748
+    },
+    {
+      "epoch": 9.91872,
+      "grad_norm": 0.8645205497741699,
+      "learning_rate": 3.451580632252901e-05,
+      "loss": 0.634,
+      "step": 7749
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 0.9036076068878174,
+      "learning_rate": 3.4513805522208885e-05,
+      "loss": 0.6029,
+      "step": 7750
+    },
+    {
+      "epoch": 9.92128,
+      "grad_norm": 0.8720131516456604,
+      "learning_rate": 3.451180472188876e-05,
+      "loss": 0.587,
+      "step": 7751
+    },
+    {
+      "epoch": 9.92256,
+      "grad_norm": 0.8317989110946655,
+      "learning_rate": 3.450980392156863e-05,
+      "loss": 0.5854,
+      "step": 7752
+    },
+    {
+      "epoch": 9.92384,
+      "grad_norm": 0.8542910814285278,
+      "learning_rate": 3.45078031212485e-05,
+      "loss": 0.6074,
+      "step": 7753
+    },
+    {
+      "epoch": 9.92512,
+      "grad_norm": 0.858959972858429,
+      "learning_rate": 3.450580232092837e-05,
+      "loss": 0.5892,
+      "step": 7754
+    },
+    {
+      "epoch": 9.9264,
+      "grad_norm": 0.8159505128860474,
+      "learning_rate": 3.4503801520608245e-05,
+      "loss": 0.5781,
+      "step": 7755
+    },
+    {
+      "epoch": 9.92768,
+      "grad_norm": 0.810766339302063,
+      "learning_rate": 3.4501800720288116e-05,
+      "loss": 0.5439,
+      "step": 7756
+    },
+    {
+      "epoch": 9.92896,
+      "grad_norm": 0.8826687335968018,
+      "learning_rate": 3.449979991996799e-05,
+      "loss": 0.6347,
+      "step": 7757
+    },
+    {
+      "epoch": 9.93024,
+      "grad_norm": 0.8329678773880005,
+      "learning_rate": 3.449779911964786e-05,
+      "loss": 0.5471,
+      "step": 7758
+    },
+    {
+      "epoch": 9.93152,
+      "grad_norm": 0.8527499437332153,
+      "learning_rate": 3.449579831932773e-05,
+      "loss": 0.5741,
+      "step": 7759
+    },
+    {
+      "epoch": 9.9328,
+      "grad_norm": 0.8892127871513367,
+      "learning_rate": 3.4493797519007604e-05,
+      "loss": 0.6161,
+      "step": 7760
+    },
+    {
+      "epoch": 9.93408,
+      "grad_norm": 0.8425252437591553,
+      "learning_rate": 3.4491796718687476e-05,
+      "loss": 0.6091,
+      "step": 7761
+    },
+    {
+      "epoch": 9.93536,
+      "grad_norm": 0.8560076355934143,
+      "learning_rate": 3.4489795918367354e-05,
+      "loss": 0.6142,
+      "step": 7762
+    },
+    {
+      "epoch": 9.93664,
+      "grad_norm": 0.8586561679840088,
+      "learning_rate": 3.448779511804722e-05,
+      "loss": 0.6231,
+      "step": 7763
+    },
+    {
+      "epoch": 9.93792,
+      "grad_norm": 0.8342392444610596,
+      "learning_rate": 3.448579431772709e-05,
+      "loss": 0.546,
+      "step": 7764
+    },
+    {
+      "epoch": 9.9392,
+      "grad_norm": 0.788269579410553,
+      "learning_rate": 3.448379351740696e-05,
+      "loss": 0.5268,
+      "step": 7765
+    },
+    {
+      "epoch": 9.94048,
+      "grad_norm": 0.8121354579925537,
+      "learning_rate": 3.4481792717086835e-05,
+      "loss": 0.5221,
+      "step": 7766
+    },
+    {
+      "epoch": 9.94176,
+      "grad_norm": 0.8493075966835022,
+      "learning_rate": 3.447979191676671e-05,
+      "loss": 0.6281,
+      "step": 7767
+    },
+    {
+      "epoch": 9.94304,
+      "grad_norm": 0.8820042014122009,
+      "learning_rate": 3.447779111644658e-05,
+      "loss": 0.5822,
+      "step": 7768
+    },
+    {
+      "epoch": 9.94432,
+      "grad_norm": 0.8810234665870667,
+      "learning_rate": 3.447579031612645e-05,
+      "loss": 0.5933,
+      "step": 7769
+    },
+    {
+      "epoch": 9.9456,
+      "grad_norm": 0.8822959065437317,
+      "learning_rate": 3.447378951580633e-05,
+      "loss": 0.6068,
+      "step": 7770
+    },
+    {
+      "epoch": 9.94688,
+      "grad_norm": 0.8374966979026794,
+      "learning_rate": 3.4471788715486194e-05,
+      "loss": 0.5737,
+      "step": 7771
+    },
+    {
+      "epoch": 9.94816,
+      "grad_norm": 0.8867483139038086,
+      "learning_rate": 3.4469787915166066e-05,
+      "loss": 0.5899,
+      "step": 7772
+    },
+    {
+      "epoch": 9.94944,
+      "grad_norm": 0.8893982172012329,
+      "learning_rate": 3.446778711484594e-05,
+      "loss": 0.5771,
+      "step": 7773
+    },
+    {
+      "epoch": 9.95072,
+      "grad_norm": 0.8742989897727966,
+      "learning_rate": 3.446578631452581e-05,
+      "loss": 0.5844,
+      "step": 7774
+    },
+    {
+      "epoch": 9.952,
+      "grad_norm": 0.8614414930343628,
+      "learning_rate": 3.446378551420568e-05,
+      "loss": 0.5596,
+      "step": 7775
+    },
+    {
+      "epoch": 9.95328,
+      "grad_norm": 0.8370686173439026,
+      "learning_rate": 3.4461784713885554e-05,
+      "loss": 0.5797,
+      "step": 7776
+    },
+    {
+      "epoch": 9.95456,
+      "grad_norm": 0.8885643482208252,
+      "learning_rate": 3.445978391356543e-05,
+      "loss": 0.5866,
+      "step": 7777
+    },
+    {
+      "epoch": 9.95584,
+      "grad_norm": 0.8820523619651794,
+      "learning_rate": 3.4457783113245304e-05,
+      "loss": 0.5999,
+      "step": 7778
+    },
+    {
+      "epoch": 9.95712,
+      "grad_norm": 0.8967301249504089,
+      "learning_rate": 3.445578231292517e-05,
+      "loss": 0.6284,
+      "step": 7779
+    },
+    {
+      "epoch": 9.9584,
+      "grad_norm": 0.8450223207473755,
+      "learning_rate": 3.445378151260504e-05,
+      "loss": 0.5917,
+      "step": 7780
+    },
+    {
+      "epoch": 9.95968,
+      "grad_norm": 0.8768996596336365,
+      "learning_rate": 3.445178071228491e-05,
+      "loss": 0.5879,
+      "step": 7781
+    },
+    {
+      "epoch": 9.96096,
+      "grad_norm": 0.876092255115509,
+      "learning_rate": 3.4449779911964785e-05,
+      "loss": 0.609,
+      "step": 7782
+    },
+    {
+      "epoch": 9.96224,
+      "grad_norm": 0.799235463142395,
+      "learning_rate": 3.4447779111644657e-05,
+      "loss": 0.5266,
+      "step": 7783
+    },
+    {
+      "epoch": 9.96352,
+      "grad_norm": 0.8335480690002441,
+      "learning_rate": 3.4445778311324535e-05,
+      "loss": 0.579,
+      "step": 7784
+    },
+    {
+      "epoch": 9.9648,
+      "grad_norm": 0.8067023158073425,
+      "learning_rate": 3.444377751100441e-05,
+      "loss": 0.5569,
+      "step": 7785
+    },
+    {
+      "epoch": 9.96608,
+      "grad_norm": 0.8609577417373657,
+      "learning_rate": 3.444177671068428e-05,
+      "loss": 0.614,
+      "step": 7786
+    },
+    {
+      "epoch": 9.96736,
+      "grad_norm": 0.8286531567573547,
+      "learning_rate": 3.4439775910364144e-05,
+      "loss": 0.582,
+      "step": 7787
+    },
+    {
+      "epoch": 9.96864,
+      "grad_norm": 0.8231421113014221,
+      "learning_rate": 3.4437775110044016e-05,
+      "loss": 0.5541,
+      "step": 7788
+    },
+    {
+      "epoch": 9.96992,
+      "grad_norm": 0.8715162873268127,
+      "learning_rate": 3.443577430972389e-05,
+      "loss": 0.6195,
+      "step": 7789
+    },
+    {
+      "epoch": 9.9712,
+      "grad_norm": 0.9087401032447815,
+      "learning_rate": 3.443377350940376e-05,
+      "loss": 0.593,
+      "step": 7790
+    },
+    {
+      "epoch": 9.972480000000001,
+      "grad_norm": 0.8596914410591125,
+      "learning_rate": 3.443177270908364e-05,
+      "loss": 0.5851,
+      "step": 7791
+    },
+    {
+      "epoch": 9.97376,
+      "grad_norm": 0.8268222808837891,
+      "learning_rate": 3.442977190876351e-05,
+      "loss": 0.5725,
+      "step": 7792
+    },
+    {
+      "epoch": 9.97504,
+      "grad_norm": 0.8293636441230774,
+      "learning_rate": 3.442777110844338e-05,
+      "loss": 0.5897,
+      "step": 7793
+    },
+    {
+      "epoch": 9.97632,
+      "grad_norm": 0.8125051259994507,
+      "learning_rate": 3.4425770308123254e-05,
+      "loss": 0.5753,
+      "step": 7794
+    },
+    {
+      "epoch": 9.9776,
+      "grad_norm": 0.8295629024505615,
+      "learning_rate": 3.442376950780312e-05,
+      "loss": 0.5983,
+      "step": 7795
+    },
+    {
+      "epoch": 9.97888,
+      "grad_norm": 0.828607439994812,
+      "learning_rate": 3.442176870748299e-05,
+      "loss": 0.6585,
+      "step": 7796
+    },
+    {
+      "epoch": 9.98016,
+      "grad_norm": 0.8404040932655334,
+      "learning_rate": 3.441976790716286e-05,
+      "loss": 0.5979,
+      "step": 7797
+    },
+    {
+      "epoch": 9.98144,
+      "grad_norm": 0.8014954924583435,
+      "learning_rate": 3.441776710684274e-05,
+      "loss": 0.5512,
+      "step": 7798
+    },
+    {
+      "epoch": 9.98272,
+      "grad_norm": 0.8161661028862,
+      "learning_rate": 3.441576630652261e-05,
+      "loss": 0.5654,
+      "step": 7799
+    },
+    {
+      "epoch": 9.984,
+      "grad_norm": 0.8457411527633667,
+      "learning_rate": 3.4413765506202485e-05,
+      "loss": 0.6289,
+      "step": 7800
+    },
+    {
+      "epoch": 9.98528,
+      "grad_norm": 0.8675429224967957,
+      "learning_rate": 3.441176470588236e-05,
+      "loss": 0.5657,
+      "step": 7801
+    },
+    {
+      "epoch": 9.98656,
+      "grad_norm": 0.8058546185493469,
+      "learning_rate": 3.440976390556223e-05,
+      "loss": 0.5645,
+      "step": 7802
+    },
+    {
+      "epoch": 9.98784,
+      "grad_norm": 0.7946271896362305,
+      "learning_rate": 3.4407763105242094e-05,
+      "loss": 0.5822,
+      "step": 7803
+    },
+    {
+      "epoch": 9.98912,
+      "grad_norm": 0.7849445939064026,
+      "learning_rate": 3.4405762304921965e-05,
+      "loss": 0.5638,
+      "step": 7804
+    },
+    {
+      "epoch": 9.9904,
+      "grad_norm": 0.8515406847000122,
+      "learning_rate": 3.4403761504601844e-05,
+      "loss": 0.5734,
+      "step": 7805
+    },
+    {
+      "epoch": 9.99168,
+      "grad_norm": 0.8528926968574524,
+      "learning_rate": 3.4401760704281716e-05,
+      "loss": 0.5722,
+      "step": 7806
+    },
+    {
+      "epoch": 9.99296,
+      "grad_norm": 0.851125180721283,
+      "learning_rate": 3.439975990396159e-05,
+      "loss": 0.5603,
+      "step": 7807
+    },
+    {
+      "epoch": 9.99424,
+      "grad_norm": 0.8529646396636963,
+      "learning_rate": 3.439775910364146e-05,
+      "loss": 0.6152,
+      "step": 7808
+    },
+    {
+      "epoch": 9.995519999999999,
+      "grad_norm": 0.8701699376106262,
+      "learning_rate": 3.439575830332133e-05,
+      "loss": 0.6082,
+      "step": 7809
+    },
+    {
+      "epoch": 9.9968,
+      "grad_norm": 0.875981330871582,
+      "learning_rate": 3.4393757503001203e-05,
+      "loss": 0.5978,
+      "step": 7810
+    },
+    {
+      "epoch": 9.99808,
+      "grad_norm": 0.8291494250297546,
+      "learning_rate": 3.439175670268107e-05,
+      "loss": 0.5422,
+      "step": 7811
+    },
+    {
+      "epoch": 9.99936,
+      "grad_norm": 0.9472531080245972,
+      "learning_rate": 3.438975590236095e-05,
+      "loss": 0.6558,
+      "step": 7812
+    },
+    {
+      "epoch": 10.00064,
+      "grad_norm": 1.5498021841049194,
+      "learning_rate": 3.438775510204082e-05,
+      "loss": 0.8758,
+      "step": 7813
+    },
+    {
+      "epoch": 10.00192,
+      "grad_norm": 0.8499518632888794,
+      "learning_rate": 3.438575430172069e-05,
+      "loss": 0.5858,
+      "step": 7814
+    },
+    {
+      "epoch": 10.0032,
+      "grad_norm": 0.7815194129943848,
+      "learning_rate": 3.438375350140056e-05,
+      "loss": 0.5487,
+      "step": 7815
+    },
+    {
+      "epoch": 10.00448,
+      "grad_norm": 0.8824237585067749,
+      "learning_rate": 3.4381752701080435e-05,
+      "loss": 0.65,
+      "step": 7816
+    },
+    {
+      "epoch": 10.00576,
+      "grad_norm": 0.7906972169876099,
+      "learning_rate": 3.4379751900760306e-05,
+      "loss": 0.5486,
+      "step": 7817
+    },
+    {
+      "epoch": 10.00704,
+      "grad_norm": 0.8036819100379944,
+      "learning_rate": 3.437775110044018e-05,
+      "loss": 0.5068,
+      "step": 7818
+    },
+    {
+      "epoch": 10.00832,
+      "grad_norm": 0.7495970129966736,
+      "learning_rate": 3.437575030012005e-05,
+      "loss": 0.5221,
+      "step": 7819
+    },
+    {
+      "epoch": 10.0096,
+      "grad_norm": 0.8430529832839966,
+      "learning_rate": 3.437374949979992e-05,
+      "loss": 0.5377,
+      "step": 7820
+    },
+    {
+      "epoch": 10.01088,
+      "grad_norm": 0.8672014474868774,
+      "learning_rate": 3.4371748699479794e-05,
+      "loss": 0.5683,
+      "step": 7821
+    },
+    {
+      "epoch": 10.01216,
+      "grad_norm": 0.904216468334198,
+      "learning_rate": 3.4369747899159666e-05,
+      "loss": 0.6345,
+      "step": 7822
+    },
+    {
+      "epoch": 10.01344,
+      "grad_norm": 0.9090774655342102,
+      "learning_rate": 3.436774709883954e-05,
+      "loss": 0.6341,
+      "step": 7823
+    },
+    {
+      "epoch": 10.01472,
+      "grad_norm": 0.857877790927887,
+      "learning_rate": 3.436574629851941e-05,
+      "loss": 0.5991,
+      "step": 7824
+    },
+    {
+      "epoch": 10.016,
+      "grad_norm": 0.8711366057395935,
+      "learning_rate": 3.436374549819928e-05,
+      "loss": 0.5882,
+      "step": 7825
+    },
+    {
+      "epoch": 10.01728,
+      "grad_norm": 0.8916945457458496,
+      "learning_rate": 3.436174469787915e-05,
+      "loss": 0.5846,
+      "step": 7826
+    },
+    {
+      "epoch": 10.01856,
+      "grad_norm": 0.8220417499542236,
+      "learning_rate": 3.4359743897559025e-05,
+      "loss": 0.5259,
+      "step": 7827
+    },
+    {
+      "epoch": 10.01984,
+      "grad_norm": 0.8267784118652344,
+      "learning_rate": 3.43577430972389e-05,
+      "loss": 0.5125,
+      "step": 7828
+    },
+    {
+      "epoch": 10.02112,
+      "grad_norm": 0.8478044271469116,
+      "learning_rate": 3.435574229691877e-05,
+      "loss": 0.564,
+      "step": 7829
+    },
+    {
+      "epoch": 10.0224,
+      "grad_norm": 0.8476915955543518,
+      "learning_rate": 3.435374149659864e-05,
+      "loss": 0.514,
+      "step": 7830
+    },
+    {
+      "epoch": 10.02368,
+      "grad_norm": 0.8940284252166748,
+      "learning_rate": 3.435174069627851e-05,
+      "loss": 0.6042,
+      "step": 7831
+    },
+    {
+      "epoch": 10.02496,
+      "grad_norm": 0.8573498129844666,
+      "learning_rate": 3.4349739895958384e-05,
+      "loss": 0.5649,
+      "step": 7832
+    },
+    {
+      "epoch": 10.02624,
+      "grad_norm": 0.8762027621269226,
+      "learning_rate": 3.4347739095638256e-05,
+      "loss": 0.5969,
+      "step": 7833
+    },
+    {
+      "epoch": 10.02752,
+      "grad_norm": 0.9091726541519165,
+      "learning_rate": 3.434573829531813e-05,
+      "loss": 0.5614,
+      "step": 7834
+    },
+    {
+      "epoch": 10.0288,
+      "grad_norm": 0.8687475919723511,
+      "learning_rate": 3.4343737494998e-05,
+      "loss": 0.5659,
+      "step": 7835
+    },
+    {
+      "epoch": 10.03008,
+      "grad_norm": 0.8800902366638184,
+      "learning_rate": 3.434173669467787e-05,
+      "loss": 0.5798,
+      "step": 7836
+    },
+    {
+      "epoch": 10.03136,
+      "grad_norm": 0.8568121194839478,
+      "learning_rate": 3.4339735894357744e-05,
+      "loss": 0.5399,
+      "step": 7837
+    },
+    {
+      "epoch": 10.03264,
+      "grad_norm": 0.8370882272720337,
+      "learning_rate": 3.4337735094037615e-05,
+      "loss": 0.5406,
+      "step": 7838
+    },
+    {
+      "epoch": 10.03392,
+      "grad_norm": 0.8754110932350159,
+      "learning_rate": 3.433573429371749e-05,
+      "loss": 0.5842,
+      "step": 7839
+    },
+    {
+      "epoch": 10.0352,
+      "grad_norm": 0.8230734467506409,
+      "learning_rate": 3.4333733493397366e-05,
+      "loss": 0.529,
+      "step": 7840
+    },
+    {
+      "epoch": 10.03648,
+      "grad_norm": 0.8830897212028503,
+      "learning_rate": 3.433173269307723e-05,
+      "loss": 0.5796,
+      "step": 7841
+    },
+    {
+      "epoch": 10.03776,
+      "grad_norm": 0.8082362413406372,
+      "learning_rate": 3.43297318927571e-05,
+      "loss": 0.5195,
+      "step": 7842
+    },
+    {
+      "epoch": 10.03904,
+      "grad_norm": 0.8439376354217529,
+      "learning_rate": 3.4327731092436975e-05,
+      "loss": 0.5555,
+      "step": 7843
+    },
+    {
+      "epoch": 10.04032,
+      "grad_norm": 0.8426663279533386,
+      "learning_rate": 3.4325730292116847e-05,
+      "loss": 0.5987,
+      "step": 7844
+    },
+    {
+      "epoch": 10.0416,
+      "grad_norm": 0.8852523565292358,
+      "learning_rate": 3.432372949179672e-05,
+      "loss": 0.5595,
+      "step": 7845
+    },
+    {
+      "epoch": 10.04288,
+      "grad_norm": 0.83525151014328,
+      "learning_rate": 3.432172869147659e-05,
+      "loss": 0.5076,
+      "step": 7846
+    },
+    {
+      "epoch": 10.04416,
+      "grad_norm": 0.879931628704071,
+      "learning_rate": 3.431972789115647e-05,
+      "loss": 0.56,
+      "step": 7847
+    },
+    {
+      "epoch": 10.04544,
+      "grad_norm": 0.8788473606109619,
+      "learning_rate": 3.431772709083634e-05,
+      "loss": 0.6005,
+      "step": 7848
+    },
+    {
+      "epoch": 10.04672,
+      "grad_norm": 0.8380979895591736,
+      "learning_rate": 3.4315726290516206e-05,
+      "loss": 0.5227,
+      "step": 7849
+    },
+    {
+      "epoch": 10.048,
+      "grad_norm": 0.8248594999313354,
+      "learning_rate": 3.431372549019608e-05,
+      "loss": 0.5403,
+      "step": 7850
+    },
+    {
+      "epoch": 10.04928,
+      "grad_norm": 0.8815642595291138,
+      "learning_rate": 3.431172468987595e-05,
+      "loss": 0.5672,
+      "step": 7851
+    },
+    {
+      "epoch": 10.05056,
+      "grad_norm": 0.8581164479255676,
+      "learning_rate": 3.430972388955582e-05,
+      "loss": 0.5726,
+      "step": 7852
+    },
+    {
+      "epoch": 10.05184,
+      "grad_norm": 0.9075201749801636,
+      "learning_rate": 3.430772308923569e-05,
+      "loss": 0.6214,
+      "step": 7853
+    },
+    {
+      "epoch": 10.05312,
+      "grad_norm": 0.918493390083313,
+      "learning_rate": 3.430572228891557e-05,
+      "loss": 0.591,
+      "step": 7854
+    },
+    {
+      "epoch": 10.0544,
+      "grad_norm": 0.9461652040481567,
+      "learning_rate": 3.4303721488595444e-05,
+      "loss": 0.641,
+      "step": 7855
+    },
+    {
+      "epoch": 10.05568,
+      "grad_norm": 0.9013059735298157,
+      "learning_rate": 3.4301720688275316e-05,
+      "loss": 0.548,
+      "step": 7856
+    },
+    {
+      "epoch": 10.05696,
+      "grad_norm": 0.8635231256484985,
+      "learning_rate": 3.429971988795518e-05,
+      "loss": 0.5381,
+      "step": 7857
+    },
+    {
+      "epoch": 10.05824,
+      "grad_norm": 0.8652742505073547,
+      "learning_rate": 3.429771908763505e-05,
+      "loss": 0.526,
+      "step": 7858
+    },
+    {
+      "epoch": 10.05952,
+      "grad_norm": 0.8748012781143188,
+      "learning_rate": 3.4295718287314924e-05,
+      "loss": 0.5851,
+      "step": 7859
+    },
+    {
+      "epoch": 10.0608,
+      "grad_norm": 0.8530634641647339,
+      "learning_rate": 3.4293717486994796e-05,
+      "loss": 0.5469,
+      "step": 7860
+    },
+    {
+      "epoch": 10.06208,
+      "grad_norm": 0.8816656470298767,
+      "learning_rate": 3.4291716686674675e-05,
+      "loss": 0.5817,
+      "step": 7861
+    },
+    {
+      "epoch": 10.06336,
+      "grad_norm": 0.8042013049125671,
+      "learning_rate": 3.428971588635455e-05,
+      "loss": 0.5568,
+      "step": 7862
+    },
+    {
+      "epoch": 10.06464,
+      "grad_norm": 0.8870890736579895,
+      "learning_rate": 3.428771508603442e-05,
+      "loss": 0.5667,
+      "step": 7863
+    },
+    {
+      "epoch": 10.06592,
+      "grad_norm": 0.9021799564361572,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": 0.5946,
+      "step": 7864
+    },
+    {
+      "epoch": 10.0672,
+      "grad_norm": 0.8616315126419067,
+      "learning_rate": 3.4283713485394156e-05,
+      "loss": 0.5375,
+      "step": 7865
+    },
+    {
+      "epoch": 10.06848,
+      "grad_norm": 0.8459718823432922,
+      "learning_rate": 3.428171268507403e-05,
+      "loss": 0.5569,
+      "step": 7866
+    },
+    {
+      "epoch": 10.06976,
+      "grad_norm": 0.8223601579666138,
+      "learning_rate": 3.42797118847539e-05,
+      "loss": 0.5202,
+      "step": 7867
+    },
+    {
+      "epoch": 10.07104,
+      "grad_norm": 0.8322831392288208,
+      "learning_rate": 3.427771108443378e-05,
+      "loss": 0.5596,
+      "step": 7868
+    },
+    {
+      "epoch": 10.07232,
+      "grad_norm": 0.8532194495201111,
+      "learning_rate": 3.427571028411365e-05,
+      "loss": 0.562,
+      "step": 7869
+    },
+    {
+      "epoch": 10.0736,
+      "grad_norm": 0.8365527391433716,
+      "learning_rate": 3.427370948379352e-05,
+      "loss": 0.5869,
+      "step": 7870
+    },
+    {
+      "epoch": 10.07488,
+      "grad_norm": 0.8544728755950928,
+      "learning_rate": 3.4271708683473393e-05,
+      "loss": 0.562,
+      "step": 7871
+    },
+    {
+      "epoch": 10.07616,
+      "grad_norm": 0.8916680216789246,
+      "learning_rate": 3.4269707883153265e-05,
+      "loss": 0.6106,
+      "step": 7872
+    },
+    {
+      "epoch": 10.07744,
+      "grad_norm": 0.904686450958252,
+      "learning_rate": 3.426770708283313e-05,
+      "loss": 0.5728,
+      "step": 7873
+    },
+    {
+      "epoch": 10.07872,
+      "grad_norm": 0.9149865508079529,
+      "learning_rate": 3.4265706282513e-05,
+      "loss": 0.5738,
+      "step": 7874
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 0.905720591545105,
+      "learning_rate": 3.426370548219288e-05,
+      "loss": 0.5414,
+      "step": 7875
+    },
+    {
+      "epoch": 10.08128,
+      "grad_norm": 0.9203237295150757,
+      "learning_rate": 3.426170468187275e-05,
+      "loss": 0.5853,
+      "step": 7876
+    },
+    {
+      "epoch": 10.08256,
+      "grad_norm": 0.8600616455078125,
+      "learning_rate": 3.4259703881552625e-05,
+      "loss": 0.5506,
+      "step": 7877
+    },
+    {
+      "epoch": 10.08384,
+      "grad_norm": 0.8730953335762024,
+      "learning_rate": 3.4257703081232496e-05,
+      "loss": 0.5553,
+      "step": 7878
+    },
+    {
+      "epoch": 10.08512,
+      "grad_norm": 0.8567367792129517,
+      "learning_rate": 3.425570228091237e-05,
+      "loss": 0.5875,
+      "step": 7879
+    },
+    {
+      "epoch": 10.0864,
+      "grad_norm": 0.8998702168464661,
+      "learning_rate": 3.425370148059224e-05,
+      "loss": 0.6049,
+      "step": 7880
+    },
+    {
+      "epoch": 10.08768,
+      "grad_norm": 0.88094162940979,
+      "learning_rate": 3.4251700680272105e-05,
+      "loss": 0.5715,
+      "step": 7881
+    },
+    {
+      "epoch": 10.08896,
+      "grad_norm": 0.8907934427261353,
+      "learning_rate": 3.424969987995198e-05,
+      "loss": 0.5553,
+      "step": 7882
+    },
+    {
+      "epoch": 10.09024,
+      "grad_norm": 0.809238076210022,
+      "learning_rate": 3.4247699079631856e-05,
+      "loss": 0.5389,
+      "step": 7883
+    },
+    {
+      "epoch": 10.09152,
+      "grad_norm": 0.8654890060424805,
+      "learning_rate": 3.424569827931173e-05,
+      "loss": 0.5753,
+      "step": 7884
+    },
+    {
+      "epoch": 10.0928,
+      "grad_norm": 0.8068164587020874,
+      "learning_rate": 3.42436974789916e-05,
+      "loss": 0.5082,
+      "step": 7885
+    },
+    {
+      "epoch": 10.09408,
+      "grad_norm": 0.8308097124099731,
+      "learning_rate": 3.424169667867147e-05,
+      "loss": 0.5334,
+      "step": 7886
+    },
+    {
+      "epoch": 10.09536,
+      "grad_norm": 0.889793336391449,
+      "learning_rate": 3.423969587835134e-05,
+      "loss": 0.5774,
+      "step": 7887
+    },
+    {
+      "epoch": 10.09664,
+      "grad_norm": 0.9080507159233093,
+      "learning_rate": 3.4237695078031215e-05,
+      "loss": 0.5919,
+      "step": 7888
+    },
+    {
+      "epoch": 10.09792,
+      "grad_norm": 0.8671941757202148,
+      "learning_rate": 3.423569427771108e-05,
+      "loss": 0.5619,
+      "step": 7889
+    },
+    {
+      "epoch": 10.0992,
+      "grad_norm": 0.8323997259140015,
+      "learning_rate": 3.423369347739096e-05,
+      "loss": 0.4927,
+      "step": 7890
+    },
+    {
+      "epoch": 10.10048,
+      "grad_norm": 0.8521079421043396,
+      "learning_rate": 3.423169267707083e-05,
+      "loss": 0.5689,
+      "step": 7891
+    },
+    {
+      "epoch": 10.10176,
+      "grad_norm": 0.8631630539894104,
+      "learning_rate": 3.42296918767507e-05,
+      "loss": 0.5655,
+      "step": 7892
+    },
+    {
+      "epoch": 10.10304,
+      "grad_norm": 0.8886538147926331,
+      "learning_rate": 3.4227691076430574e-05,
+      "loss": 0.5851,
+      "step": 7893
+    },
+    {
+      "epoch": 10.10432,
+      "grad_norm": 0.8930172920227051,
+      "learning_rate": 3.4225690276110446e-05,
+      "loss": 0.6251,
+      "step": 7894
+    },
+    {
+      "epoch": 10.1056,
+      "grad_norm": 0.8463165760040283,
+      "learning_rate": 3.422368947579032e-05,
+      "loss": 0.4898,
+      "step": 7895
+    },
+    {
+      "epoch": 10.10688,
+      "grad_norm": 0.893409252166748,
+      "learning_rate": 3.422168867547019e-05,
+      "loss": 0.5761,
+      "step": 7896
+    },
+    {
+      "epoch": 10.10816,
+      "grad_norm": 0.8559973239898682,
+      "learning_rate": 3.421968787515006e-05,
+      "loss": 0.5351,
+      "step": 7897
+    },
+    {
+      "epoch": 10.10944,
+      "grad_norm": 0.8258970379829407,
+      "learning_rate": 3.4217687074829934e-05,
+      "loss": 0.5773,
+      "step": 7898
+    },
+    {
+      "epoch": 10.11072,
+      "grad_norm": 0.8773029446601868,
+      "learning_rate": 3.4215686274509805e-05,
+      "loss": 0.547,
+      "step": 7899
+    },
+    {
+      "epoch": 10.112,
+      "grad_norm": 0.8930992484092712,
+      "learning_rate": 3.421368547418968e-05,
+      "loss": 0.6053,
+      "step": 7900
+    },
+    {
+      "epoch": 10.11328,
+      "grad_norm": 0.872535228729248,
+      "learning_rate": 3.421168467386955e-05,
+      "loss": 0.5813,
+      "step": 7901
+    },
+    {
+      "epoch": 10.11456,
+      "grad_norm": 0.8744470477104187,
+      "learning_rate": 3.420968387354942e-05,
+      "loss": 0.5333,
+      "step": 7902
+    },
+    {
+      "epoch": 10.11584,
+      "grad_norm": 0.8636131882667542,
+      "learning_rate": 3.420768307322929e-05,
+      "loss": 0.5746,
+      "step": 7903
+    },
+    {
+      "epoch": 10.11712,
+      "grad_norm": 0.865611732006073,
+      "learning_rate": 3.4205682272909165e-05,
+      "loss": 0.561,
+      "step": 7904
+    },
+    {
+      "epoch": 10.1184,
+      "grad_norm": 0.8732993006706238,
+      "learning_rate": 3.4203681472589037e-05,
+      "loss": 0.6067,
+      "step": 7905
+    },
+    {
+      "epoch": 10.11968,
+      "grad_norm": 0.8363077044487,
+      "learning_rate": 3.420168067226891e-05,
+      "loss": 0.5354,
+      "step": 7906
+    },
+    {
+      "epoch": 10.12096,
+      "grad_norm": 0.9122660756111145,
+      "learning_rate": 3.419967987194878e-05,
+      "loss": 0.6047,
+      "step": 7907
+    },
+    {
+      "epoch": 10.12224,
+      "grad_norm": 0.8874605298042297,
+      "learning_rate": 3.419767907162865e-05,
+      "loss": 0.6021,
+      "step": 7908
+    },
+    {
+      "epoch": 10.12352,
+      "grad_norm": 0.9380055665969849,
+      "learning_rate": 3.4195678271308524e-05,
+      "loss": 0.6033,
+      "step": 7909
+    },
+    {
+      "epoch": 10.1248,
+      "grad_norm": 0.8816724419593811,
+      "learning_rate": 3.4193677470988396e-05,
+      "loss": 0.5775,
+      "step": 7910
+    },
+    {
+      "epoch": 10.12608,
+      "grad_norm": 0.8694825172424316,
+      "learning_rate": 3.4191676670668274e-05,
+      "loss": 0.5716,
+      "step": 7911
+    },
+    {
+      "epoch": 10.12736,
+      "grad_norm": 0.9232937693595886,
+      "learning_rate": 3.418967587034814e-05,
+      "loss": 0.6145,
+      "step": 7912
+    },
+    {
+      "epoch": 10.12864,
+      "grad_norm": 0.8252090215682983,
+      "learning_rate": 3.418767507002801e-05,
+      "loss": 0.5711,
+      "step": 7913
+    },
+    {
+      "epoch": 10.12992,
+      "grad_norm": 0.8287693858146667,
+      "learning_rate": 3.418567426970788e-05,
+      "loss": 0.5349,
+      "step": 7914
+    },
+    {
+      "epoch": 10.1312,
+      "grad_norm": 0.8445566892623901,
+      "learning_rate": 3.4183673469387755e-05,
+      "loss": 0.5614,
+      "step": 7915
+    },
+    {
+      "epoch": 10.13248,
+      "grad_norm": 0.886237382888794,
+      "learning_rate": 3.418167266906763e-05,
+      "loss": 0.5865,
+      "step": 7916
+    },
+    {
+      "epoch": 10.13376,
+      "grad_norm": 0.8767721652984619,
+      "learning_rate": 3.41796718687475e-05,
+      "loss": 0.5316,
+      "step": 7917
+    },
+    {
+      "epoch": 10.13504,
+      "grad_norm": 0.907900333404541,
+      "learning_rate": 3.417767106842738e-05,
+      "loss": 0.5782,
+      "step": 7918
+    },
+    {
+      "epoch": 10.13632,
+      "grad_norm": 0.8434520363807678,
+      "learning_rate": 3.417567026810725e-05,
+      "loss": 0.5359,
+      "step": 7919
+    },
+    {
+      "epoch": 10.1376,
+      "grad_norm": 0.8087551593780518,
+      "learning_rate": 3.4173669467787114e-05,
+      "loss": 0.5508,
+      "step": 7920
+    },
+    {
+      "epoch": 10.13888,
+      "grad_norm": 0.9005576968193054,
+      "learning_rate": 3.4171668667466986e-05,
+      "loss": 0.5558,
+      "step": 7921
+    },
+    {
+      "epoch": 10.14016,
+      "grad_norm": 0.8877005577087402,
+      "learning_rate": 3.416966786714686e-05,
+      "loss": 0.615,
+      "step": 7922
+    },
+    {
+      "epoch": 10.14144,
+      "grad_norm": 0.8587065935134888,
+      "learning_rate": 3.416766706682673e-05,
+      "loss": 0.5714,
+      "step": 7923
+    },
+    {
+      "epoch": 10.14272,
+      "grad_norm": 0.8844937086105347,
+      "learning_rate": 3.41656662665066e-05,
+      "loss": 0.6407,
+      "step": 7924
+    },
+    {
+      "epoch": 10.144,
+      "grad_norm": 0.8781236410140991,
+      "learning_rate": 3.416366546618648e-05,
+      "loss": 0.5428,
+      "step": 7925
+    },
+    {
+      "epoch": 10.14528,
+      "grad_norm": 0.8121967315673828,
+      "learning_rate": 3.416166466586635e-05,
+      "loss": 0.5061,
+      "step": 7926
+    },
+    {
+      "epoch": 10.14656,
+      "grad_norm": 0.8259449005126953,
+      "learning_rate": 3.4159663865546224e-05,
+      "loss": 0.5297,
+      "step": 7927
+    },
+    {
+      "epoch": 10.14784,
+      "grad_norm": 0.9252973794937134,
+      "learning_rate": 3.415766306522609e-05,
+      "loss": 0.6159,
+      "step": 7928
+    },
+    {
+      "epoch": 10.14912,
+      "grad_norm": 0.8705994486808777,
+      "learning_rate": 3.415566226490596e-05,
+      "loss": 0.5131,
+      "step": 7929
+    },
+    {
+      "epoch": 10.1504,
+      "grad_norm": 0.892468273639679,
+      "learning_rate": 3.415366146458583e-05,
+      "loss": 0.551,
+      "step": 7930
+    },
+    {
+      "epoch": 10.15168,
+      "grad_norm": 0.8754770755767822,
+      "learning_rate": 3.4151660664265705e-05,
+      "loss": 0.5457,
+      "step": 7931
+    },
+    {
+      "epoch": 10.15296,
+      "grad_norm": 0.8621395826339722,
+      "learning_rate": 3.4149659863945583e-05,
+      "loss": 0.5292,
+      "step": 7932
+    },
+    {
+      "epoch": 10.15424,
+      "grad_norm": 0.9692301750183105,
+      "learning_rate": 3.4147659063625455e-05,
+      "loss": 0.5985,
+      "step": 7933
+    },
+    {
+      "epoch": 10.15552,
+      "grad_norm": 0.9365867376327515,
+      "learning_rate": 3.414565826330533e-05,
+      "loss": 0.5769,
+      "step": 7934
+    },
+    {
+      "epoch": 10.1568,
+      "grad_norm": 0.9413188695907593,
+      "learning_rate": 3.41436574629852e-05,
+      "loss": 0.6406,
+      "step": 7935
+    },
+    {
+      "epoch": 10.15808,
+      "grad_norm": 0.8697437047958374,
+      "learning_rate": 3.4141656662665064e-05,
+      "loss": 0.5793,
+      "step": 7936
+    },
+    {
+      "epoch": 10.15936,
+      "grad_norm": 0.8574464917182922,
+      "learning_rate": 3.4139655862344936e-05,
+      "loss": 0.5234,
+      "step": 7937
+    },
+    {
+      "epoch": 10.16064,
+      "grad_norm": 0.8575087785720825,
+      "learning_rate": 3.413765506202481e-05,
+      "loss": 0.5734,
+      "step": 7938
+    },
+    {
+      "epoch": 10.16192,
+      "grad_norm": 0.8977473974227905,
+      "learning_rate": 3.4135654261704686e-05,
+      "loss": 0.5894,
+      "step": 7939
+    },
+    {
+      "epoch": 10.1632,
+      "grad_norm": 0.8591358661651611,
+      "learning_rate": 3.413365346138456e-05,
+      "loss": 0.527,
+      "step": 7940
+    },
+    {
+      "epoch": 10.16448,
+      "grad_norm": 0.8251833915710449,
+      "learning_rate": 3.413165266106443e-05,
+      "loss": 0.5393,
+      "step": 7941
+    },
+    {
+      "epoch": 10.16576,
+      "grad_norm": 0.8660216331481934,
+      "learning_rate": 3.41296518607443e-05,
+      "loss": 0.5797,
+      "step": 7942
+    },
+    {
+      "epoch": 10.16704,
+      "grad_norm": 0.9057691693305969,
+      "learning_rate": 3.4127651060424174e-05,
+      "loss": 0.5394,
+      "step": 7943
+    },
+    {
+      "epoch": 10.16832,
+      "grad_norm": 0.914899468421936,
+      "learning_rate": 3.412565026010404e-05,
+      "loss": 0.5747,
+      "step": 7944
+    },
+    {
+      "epoch": 10.1696,
+      "grad_norm": 0.838821291923523,
+      "learning_rate": 3.412364945978391e-05,
+      "loss": 0.5453,
+      "step": 7945
+    },
+    {
+      "epoch": 10.17088,
+      "grad_norm": 0.8535134792327881,
+      "learning_rate": 3.412164865946379e-05,
+      "loss": 0.5618,
+      "step": 7946
+    },
+    {
+      "epoch": 10.17216,
+      "grad_norm": 0.8193386793136597,
+      "learning_rate": 3.411964785914366e-05,
+      "loss": 0.5303,
+      "step": 7947
+    },
+    {
+      "epoch": 10.17344,
+      "grad_norm": 0.8792878985404968,
+      "learning_rate": 3.411764705882353e-05,
+      "loss": 0.6145,
+      "step": 7948
+    },
+    {
+      "epoch": 10.17472,
+      "grad_norm": 0.9004140496253967,
+      "learning_rate": 3.4115646258503405e-05,
+      "loss": 0.6126,
+      "step": 7949
+    },
+    {
+      "epoch": 10.176,
+      "grad_norm": 0.883179247379303,
+      "learning_rate": 3.411364545818328e-05,
+      "loss": 0.5655,
+      "step": 7950
+    },
+    {
+      "epoch": 10.17728,
+      "grad_norm": 0.8625328540802002,
+      "learning_rate": 3.411164465786315e-05,
+      "loss": 0.5993,
+      "step": 7951
+    },
+    {
+      "epoch": 10.17856,
+      "grad_norm": 0.8304584622383118,
+      "learning_rate": 3.4109643857543014e-05,
+      "loss": 0.5177,
+      "step": 7952
+    },
+    {
+      "epoch": 10.17984,
+      "grad_norm": 0.8241145610809326,
+      "learning_rate": 3.410764305722289e-05,
+      "loss": 0.5112,
+      "step": 7953
+    },
+    {
+      "epoch": 10.18112,
+      "grad_norm": 0.8981744647026062,
+      "learning_rate": 3.4105642256902764e-05,
+      "loss": 0.5897,
+      "step": 7954
+    },
+    {
+      "epoch": 10.1824,
+      "grad_norm": 0.8619903326034546,
+      "learning_rate": 3.4103641456582636e-05,
+      "loss": 0.5543,
+      "step": 7955
+    },
+    {
+      "epoch": 10.18368,
+      "grad_norm": 0.8140372037887573,
+      "learning_rate": 3.410164065626251e-05,
+      "loss": 0.5214,
+      "step": 7956
+    },
+    {
+      "epoch": 10.18496,
+      "grad_norm": 0.8518925309181213,
+      "learning_rate": 3.409963985594238e-05,
+      "loss": 0.5465,
+      "step": 7957
+    },
+    {
+      "epoch": 10.18624,
+      "grad_norm": 0.8702125549316406,
+      "learning_rate": 3.409763905562225e-05,
+      "loss": 0.5621,
+      "step": 7958
+    },
+    {
+      "epoch": 10.18752,
+      "grad_norm": 0.8971076011657715,
+      "learning_rate": 3.4095638255302124e-05,
+      "loss": 0.5631,
+      "step": 7959
+    },
+    {
+      "epoch": 10.1888,
+      "grad_norm": 0.853391706943512,
+      "learning_rate": 3.4093637454981995e-05,
+      "loss": 0.5682,
+      "step": 7960
+    },
+    {
+      "epoch": 10.19008,
+      "grad_norm": 0.8766993284225464,
+      "learning_rate": 3.409163665466187e-05,
+      "loss": 0.6276,
+      "step": 7961
+    },
+    {
+      "epoch": 10.19136,
+      "grad_norm": 0.8604820370674133,
+      "learning_rate": 3.408963585434174e-05,
+      "loss": 0.5688,
+      "step": 7962
+    },
+    {
+      "epoch": 10.19264,
+      "grad_norm": 0.8074341416358948,
+      "learning_rate": 3.408763505402161e-05,
+      "loss": 0.5106,
+      "step": 7963
+    },
+    {
+      "epoch": 10.19392,
+      "grad_norm": 0.871636152267456,
+      "learning_rate": 3.408563425370148e-05,
+      "loss": 0.6026,
+      "step": 7964
+    },
+    {
+      "epoch": 10.1952,
+      "grad_norm": 0.8580559492111206,
+      "learning_rate": 3.4083633453381355e-05,
+      "loss": 0.5621,
+      "step": 7965
+    },
+    {
+      "epoch": 10.19648,
+      "grad_norm": 0.8855502605438232,
+      "learning_rate": 3.408163265306123e-05,
+      "loss": 0.5359,
+      "step": 7966
+    },
+    {
+      "epoch": 10.19776,
+      "grad_norm": 0.8627590537071228,
+      "learning_rate": 3.40796318527411e-05,
+      "loss": 0.521,
+      "step": 7967
+    },
+    {
+      "epoch": 10.19904,
+      "grad_norm": 0.7877764105796814,
+      "learning_rate": 3.407763105242097e-05,
+      "loss": 0.5293,
+      "step": 7968
+    },
+    {
+      "epoch": 10.20032,
+      "grad_norm": 0.8240257501602173,
+      "learning_rate": 3.407563025210084e-05,
+      "loss": 0.5453,
+      "step": 7969
+    },
+    {
+      "epoch": 10.2016,
+      "grad_norm": 0.8509385585784912,
+      "learning_rate": 3.4073629451780714e-05,
+      "loss": 0.5745,
+      "step": 7970
+    },
+    {
+      "epoch": 10.20288,
+      "grad_norm": 0.8994958400726318,
+      "learning_rate": 3.4071628651460586e-05,
+      "loss": 0.5561,
+      "step": 7971
+    },
+    {
+      "epoch": 10.20416,
+      "grad_norm": 0.8885303139686584,
+      "learning_rate": 3.406962785114046e-05,
+      "loss": 0.5537,
+      "step": 7972
+    },
+    {
+      "epoch": 10.20544,
+      "grad_norm": 0.9128597378730774,
+      "learning_rate": 3.406762705082033e-05,
+      "loss": 0.5795,
+      "step": 7973
+    },
+    {
+      "epoch": 10.20672,
+      "grad_norm": 0.83772212266922,
+      "learning_rate": 3.40656262505002e-05,
+      "loss": 0.5156,
+      "step": 7974
+    },
+    {
+      "epoch": 10.208,
+      "grad_norm": 0.822971522808075,
+      "learning_rate": 3.406362545018007e-05,
+      "loss": 0.5237,
+      "step": 7975
+    },
+    {
+      "epoch": 10.20928,
+      "grad_norm": 0.898842453956604,
+      "learning_rate": 3.4061624649859945e-05,
+      "loss": 0.5979,
+      "step": 7976
+    },
+    {
+      "epoch": 10.21056,
+      "grad_norm": 0.9250447154045105,
+      "learning_rate": 3.405962384953982e-05,
+      "loss": 0.6017,
+      "step": 7977
+    },
+    {
+      "epoch": 10.21184,
+      "grad_norm": 0.8869206309318542,
+      "learning_rate": 3.405762304921969e-05,
+      "loss": 0.5366,
+      "step": 7978
+    },
+    {
+      "epoch": 10.21312,
+      "grad_norm": 0.863950788974762,
+      "learning_rate": 3.405562224889956e-05,
+      "loss": 0.5374,
+      "step": 7979
+    },
+    {
+      "epoch": 10.2144,
+      "grad_norm": 0.81676185131073,
+      "learning_rate": 3.405362144857943e-05,
+      "loss": 0.5335,
+      "step": 7980
+    },
+    {
+      "epoch": 10.21568,
+      "grad_norm": 0.8401233553886414,
+      "learning_rate": 3.405162064825931e-05,
+      "loss": 0.5484,
+      "step": 7981
+    },
+    {
+      "epoch": 10.21696,
+      "grad_norm": 0.8678064346313477,
+      "learning_rate": 3.4049619847939176e-05,
+      "loss": 0.5231,
+      "step": 7982
+    },
+    {
+      "epoch": 10.21824,
+      "grad_norm": 0.8489272594451904,
+      "learning_rate": 3.404761904761905e-05,
+      "loss": 0.4638,
+      "step": 7983
+    },
+    {
+      "epoch": 10.21952,
+      "grad_norm": 0.8029497265815735,
+      "learning_rate": 3.404561824729892e-05,
+      "loss": 0.5081,
+      "step": 7984
+    },
+    {
+      "epoch": 10.2208,
+      "grad_norm": 0.8994140625,
+      "learning_rate": 3.404361744697879e-05,
+      "loss": 0.5386,
+      "step": 7985
+    },
+    {
+      "epoch": 10.22208,
+      "grad_norm": 0.8933800458908081,
+      "learning_rate": 3.4041616646658664e-05,
+      "loss": 0.5312,
+      "step": 7986
+    },
+    {
+      "epoch": 10.22336,
+      "grad_norm": 0.8817566633224487,
+      "learning_rate": 3.4039615846338536e-05,
+      "loss": 0.5578,
+      "step": 7987
+    },
+    {
+      "epoch": 10.22464,
+      "grad_norm": 0.8852986693382263,
+      "learning_rate": 3.403761504601841e-05,
+      "loss": 0.5669,
+      "step": 7988
+    },
+    {
+      "epoch": 10.22592,
+      "grad_norm": 0.9889851212501526,
+      "learning_rate": 3.4035614245698286e-05,
+      "loss": 0.5101,
+      "step": 7989
+    },
+    {
+      "epoch": 10.2272,
+      "grad_norm": 0.8555750846862793,
+      "learning_rate": 3.403361344537815e-05,
+      "loss": 0.551,
+      "step": 7990
+    },
+    {
+      "epoch": 10.22848,
+      "grad_norm": 0.838453471660614,
+      "learning_rate": 3.403161264505802e-05,
+      "loss": 0.565,
+      "step": 7991
+    },
+    {
+      "epoch": 10.22976,
+      "grad_norm": 0.8447315096855164,
+      "learning_rate": 3.4029611844737895e-05,
+      "loss": 0.5671,
+      "step": 7992
+    },
+    {
+      "epoch": 10.23104,
+      "grad_norm": 0.8619872331619263,
+      "learning_rate": 3.402761104441777e-05,
+      "loss": 0.5877,
+      "step": 7993
+    },
+    {
+      "epoch": 10.23232,
+      "grad_norm": 0.8690614104270935,
+      "learning_rate": 3.402561024409764e-05,
+      "loss": 0.5583,
+      "step": 7994
+    },
+    {
+      "epoch": 10.2336,
+      "grad_norm": 0.9469643831253052,
+      "learning_rate": 3.402360944377751e-05,
+      "loss": 0.5224,
+      "step": 7995
+    },
+    {
+      "epoch": 10.23488,
+      "grad_norm": 0.8962279558181763,
+      "learning_rate": 3.402160864345739e-05,
+      "loss": 0.5714,
+      "step": 7996
+    },
+    {
+      "epoch": 10.23616,
+      "grad_norm": 0.9197521805763245,
+      "learning_rate": 3.401960784313726e-05,
+      "loss": 0.5755,
+      "step": 7997
+    },
+    {
+      "epoch": 10.23744,
+      "grad_norm": 0.8115956783294678,
+      "learning_rate": 3.4017607042817126e-05,
+      "loss": 0.5475,
+      "step": 7998
+    },
+    {
+      "epoch": 10.23872,
+      "grad_norm": 0.7984319925308228,
+      "learning_rate": 3.4015606242497e-05,
+      "loss": 0.5479,
+      "step": 7999
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 0.8857104778289795,
+      "learning_rate": 3.401360544217687e-05,
+      "loss": 0.5641,
+      "step": 8000
+    },
+    {
+      "epoch": 10.24128,
+      "grad_norm": 0.9119358062744141,
+      "learning_rate": 3.401160464185674e-05,
+      "loss": 0.5903,
+      "step": 8001
+    },
+    {
+      "epoch": 10.24256,
+      "grad_norm": 0.8577554821968079,
+      "learning_rate": 3.4009603841536613e-05,
+      "loss": 0.5551,
+      "step": 8002
+    },
+    {
+      "epoch": 10.24384,
+      "grad_norm": 0.9047631025314331,
+      "learning_rate": 3.400760304121649e-05,
+      "loss": 0.551,
+      "step": 8003
+    },
+    {
+      "epoch": 10.24512,
+      "grad_norm": 0.8967153429985046,
+      "learning_rate": 3.4005602240896364e-05,
+      "loss": 0.5531,
+      "step": 8004
+    },
+    {
+      "epoch": 10.2464,
+      "grad_norm": 0.8773406744003296,
+      "learning_rate": 3.4003601440576236e-05,
+      "loss": 0.5648,
+      "step": 8005
+    },
+    {
+      "epoch": 10.24768,
+      "grad_norm": 0.903235137462616,
+      "learning_rate": 3.40016006402561e-05,
+      "loss": 0.5552,
+      "step": 8006
+    },
+    {
+      "epoch": 10.24896,
+      "grad_norm": 0.8304867148399353,
+      "learning_rate": 3.399959983993597e-05,
+      "loss": 0.5472,
+      "step": 8007
+    },
+    {
+      "epoch": 10.25024,
+      "grad_norm": 0.844079852104187,
+      "learning_rate": 3.3997599039615845e-05,
+      "loss": 0.5607,
+      "step": 8008
+    },
+    {
+      "epoch": 10.25152,
+      "grad_norm": 0.8768244981765747,
+      "learning_rate": 3.3995598239295716e-05,
+      "loss": 0.5559,
+      "step": 8009
+    },
+    {
+      "epoch": 10.2528,
+      "grad_norm": 0.8504285216331482,
+      "learning_rate": 3.3993597438975595e-05,
+      "loss": 0.5555,
+      "step": 8010
+    },
+    {
+      "epoch": 10.25408,
+      "grad_norm": 0.8665524125099182,
+      "learning_rate": 3.399159663865547e-05,
+      "loss": 0.5473,
+      "step": 8011
+    },
+    {
+      "epoch": 10.25536,
+      "grad_norm": 0.8851114511489868,
+      "learning_rate": 3.398959583833534e-05,
+      "loss": 0.6117,
+      "step": 8012
+    },
+    {
+      "epoch": 10.25664,
+      "grad_norm": 0.8923035860061646,
+      "learning_rate": 3.398759503801521e-05,
+      "loss": 0.5575,
+      "step": 8013
+    },
+    {
+      "epoch": 10.25792,
+      "grad_norm": 0.8378768563270569,
+      "learning_rate": 3.3985594237695076e-05,
+      "loss": 0.5484,
+      "step": 8014
+    },
+    {
+      "epoch": 10.2592,
+      "grad_norm": 0.8756462931632996,
+      "learning_rate": 3.398359343737495e-05,
+      "loss": 0.562,
+      "step": 8015
+    },
+    {
+      "epoch": 10.26048,
+      "grad_norm": 0.8708872199058533,
+      "learning_rate": 3.398159263705482e-05,
+      "loss": 0.5702,
+      "step": 8016
+    },
+    {
+      "epoch": 10.26176,
+      "grad_norm": 0.865534245967865,
+      "learning_rate": 3.39795918367347e-05,
+      "loss": 0.5521,
+      "step": 8017
+    },
+    {
+      "epoch": 10.26304,
+      "grad_norm": 0.8553348779678345,
+      "learning_rate": 3.397759103641457e-05,
+      "loss": 0.5661,
+      "step": 8018
+    },
+    {
+      "epoch": 10.26432,
+      "grad_norm": 0.9413052797317505,
+      "learning_rate": 3.397559023609444e-05,
+      "loss": 0.5919,
+      "step": 8019
+    },
+    {
+      "epoch": 10.2656,
+      "grad_norm": 0.8924047350883484,
+      "learning_rate": 3.3973589435774314e-05,
+      "loss": 0.5855,
+      "step": 8020
+    },
+    {
+      "epoch": 10.26688,
+      "grad_norm": 0.8545451164245605,
+      "learning_rate": 3.3971588635454186e-05,
+      "loss": 0.5515,
+      "step": 8021
+    },
+    {
+      "epoch": 10.26816,
+      "grad_norm": 0.8800926208496094,
+      "learning_rate": 3.396958783513405e-05,
+      "loss": 0.537,
+      "step": 8022
+    },
+    {
+      "epoch": 10.26944,
+      "grad_norm": 0.9025660753250122,
+      "learning_rate": 3.396758703481392e-05,
+      "loss": 0.6178,
+      "step": 8023
+    },
+    {
+      "epoch": 10.27072,
+      "grad_norm": 0.8830463290214539,
+      "learning_rate": 3.39655862344938e-05,
+      "loss": 0.555,
+      "step": 8024
+    },
+    {
+      "epoch": 10.272,
+      "grad_norm": 0.9167059659957886,
+      "learning_rate": 3.396358543417367e-05,
+      "loss": 0.5678,
+      "step": 8025
+    },
+    {
+      "epoch": 10.27328,
+      "grad_norm": 0.8567641377449036,
+      "learning_rate": 3.3961584633853545e-05,
+      "loss": 0.5473,
+      "step": 8026
+    },
+    {
+      "epoch": 10.27456,
+      "grad_norm": 0.8600786328315735,
+      "learning_rate": 3.395958383353342e-05,
+      "loss": 0.5562,
+      "step": 8027
+    },
+    {
+      "epoch": 10.27584,
+      "grad_norm": 0.8937509655952454,
+      "learning_rate": 3.395758303321329e-05,
+      "loss": 0.5721,
+      "step": 8028
+    },
+    {
+      "epoch": 10.27712,
+      "grad_norm": 0.8415324091911316,
+      "learning_rate": 3.395558223289316e-05,
+      "loss": 0.5219,
+      "step": 8029
+    },
+    {
+      "epoch": 10.2784,
+      "grad_norm": 0.888176679611206,
+      "learning_rate": 3.3953581432573025e-05,
+      "loss": 0.5618,
+      "step": 8030
+    },
+    {
+      "epoch": 10.27968,
+      "grad_norm": 0.8729084730148315,
+      "learning_rate": 3.3951580632252904e-05,
+      "loss": 0.581,
+      "step": 8031
+    },
+    {
+      "epoch": 10.28096,
+      "grad_norm": 0.8753156661987305,
+      "learning_rate": 3.3949579831932776e-05,
+      "loss": 0.5667,
+      "step": 8032
+    },
+    {
+      "epoch": 10.28224,
+      "grad_norm": 0.8683957457542419,
+      "learning_rate": 3.394757903161265e-05,
+      "loss": 0.5274,
+      "step": 8033
+    },
+    {
+      "epoch": 10.28352,
+      "grad_norm": 0.8913008570671082,
+      "learning_rate": 3.394557823129252e-05,
+      "loss": 0.5799,
+      "step": 8034
+    },
+    {
+      "epoch": 10.2848,
+      "grad_norm": 0.8697010278701782,
+      "learning_rate": 3.394357743097239e-05,
+      "loss": 0.56,
+      "step": 8035
+    },
+    {
+      "epoch": 10.28608,
+      "grad_norm": 0.8997796177864075,
+      "learning_rate": 3.394157663065226e-05,
+      "loss": 0.6066,
+      "step": 8036
+    },
+    {
+      "epoch": 10.28736,
+      "grad_norm": 0.8587310314178467,
+      "learning_rate": 3.3939575830332135e-05,
+      "loss": 0.5758,
+      "step": 8037
+    },
+    {
+      "epoch": 10.288640000000001,
+      "grad_norm": 0.8550379276275635,
+      "learning_rate": 3.393757503001201e-05,
+      "loss": 0.5858,
+      "step": 8038
+    },
+    {
+      "epoch": 10.28992,
+      "grad_norm": 0.8495590686798096,
+      "learning_rate": 3.393557422969188e-05,
+      "loss": 0.5774,
+      "step": 8039
+    },
+    {
+      "epoch": 10.2912,
+      "grad_norm": 0.8640943765640259,
+      "learning_rate": 3.393357342937175e-05,
+      "loss": 0.5839,
+      "step": 8040
+    },
+    {
+      "epoch": 10.29248,
+      "grad_norm": 0.8572250604629517,
+      "learning_rate": 3.393157262905162e-05,
+      "loss": 0.5521,
+      "step": 8041
+    },
+    {
+      "epoch": 10.29376,
+      "grad_norm": 0.8335790634155273,
+      "learning_rate": 3.3929571828731495e-05,
+      "loss": 0.5296,
+      "step": 8042
+    },
+    {
+      "epoch": 10.29504,
+      "grad_norm": 0.861184298992157,
+      "learning_rate": 3.3927571028411366e-05,
+      "loss": 0.5687,
+      "step": 8043
+    },
+    {
+      "epoch": 10.29632,
+      "grad_norm": 0.8524270057678223,
+      "learning_rate": 3.392557022809124e-05,
+      "loss": 0.5577,
+      "step": 8044
+    },
+    {
+      "epoch": 10.2976,
+      "grad_norm": 0.9047152996063232,
+      "learning_rate": 3.392356942777111e-05,
+      "loss": 0.5752,
+      "step": 8045
+    },
+    {
+      "epoch": 10.29888,
+      "grad_norm": 0.8878473043441772,
+      "learning_rate": 3.392156862745098e-05,
+      "loss": 0.5627,
+      "step": 8046
+    },
+    {
+      "epoch": 10.30016,
+      "grad_norm": 0.8869286775588989,
+      "learning_rate": 3.3919567827130854e-05,
+      "loss": 0.587,
+      "step": 8047
+    },
+    {
+      "epoch": 10.30144,
+      "grad_norm": 0.8654095530509949,
+      "learning_rate": 3.3917567026810726e-05,
+      "loss": 0.5427,
+      "step": 8048
+    },
+    {
+      "epoch": 10.30272,
+      "grad_norm": 0.9054012298583984,
+      "learning_rate": 3.39155662264906e-05,
+      "loss": 0.5907,
+      "step": 8049
+    },
+    {
+      "epoch": 10.304,
+      "grad_norm": 0.852177083492279,
+      "learning_rate": 3.391356542617047e-05,
+      "loss": 0.5445,
+      "step": 8050
+    },
+    {
+      "epoch": 10.30528,
+      "grad_norm": 0.9118196964263916,
+      "learning_rate": 3.391156462585034e-05,
+      "loss": 0.5557,
+      "step": 8051
+    },
+    {
+      "epoch": 10.30656,
+      "grad_norm": 0.9239182472229004,
+      "learning_rate": 3.390956382553021e-05,
+      "loss": 0.5647,
+      "step": 8052
+    },
+    {
+      "epoch": 10.30784,
+      "grad_norm": 0.875339925289154,
+      "learning_rate": 3.3907563025210085e-05,
+      "loss": 0.5532,
+      "step": 8053
+    },
+    {
+      "epoch": 10.30912,
+      "grad_norm": 0.9014837145805359,
+      "learning_rate": 3.390556222488996e-05,
+      "loss": 0.5569,
+      "step": 8054
+    },
+    {
+      "epoch": 10.3104,
+      "grad_norm": 0.9332119226455688,
+      "learning_rate": 3.390356142456983e-05,
+      "loss": 0.5645,
+      "step": 8055
+    },
+    {
+      "epoch": 10.31168,
+      "grad_norm": 0.8813993334770203,
+      "learning_rate": 3.39015606242497e-05,
+      "loss": 0.5751,
+      "step": 8056
+    },
+    {
+      "epoch": 10.31296,
+      "grad_norm": 0.8453924655914307,
+      "learning_rate": 3.389955982392957e-05,
+      "loss": 0.5304,
+      "step": 8057
+    },
+    {
+      "epoch": 10.31424,
+      "grad_norm": 0.8781421780586243,
+      "learning_rate": 3.3897559023609444e-05,
+      "loss": 0.5555,
+      "step": 8058
+    },
+    {
+      "epoch": 10.31552,
+      "grad_norm": 0.819387674331665,
+      "learning_rate": 3.389555822328932e-05,
+      "loss": 0.5362,
+      "step": 8059
+    },
+    {
+      "epoch": 10.3168,
+      "grad_norm": 0.838315486907959,
+      "learning_rate": 3.389355742296919e-05,
+      "loss": 0.5609,
+      "step": 8060
+    },
+    {
+      "epoch": 10.31808,
+      "grad_norm": 0.8653327226638794,
+      "learning_rate": 3.389155662264906e-05,
+      "loss": 0.5313,
+      "step": 8061
+    },
+    {
+      "epoch": 10.31936,
+      "grad_norm": 0.8249607682228088,
+      "learning_rate": 3.388955582232893e-05,
+      "loss": 0.5308,
+      "step": 8062
+    },
+    {
+      "epoch": 10.32064,
+      "grad_norm": 0.8277181386947632,
+      "learning_rate": 3.3887555022008804e-05,
+      "loss": 0.5428,
+      "step": 8063
+    },
+    {
+      "epoch": 10.32192,
+      "grad_norm": 0.8770848512649536,
+      "learning_rate": 3.3885554221688675e-05,
+      "loss": 0.536,
+      "step": 8064
+    },
+    {
+      "epoch": 10.3232,
+      "grad_norm": 0.8789455890655518,
+      "learning_rate": 3.388355342136855e-05,
+      "loss": 0.5509,
+      "step": 8065
+    },
+    {
+      "epoch": 10.32448,
+      "grad_norm": 0.8777965307235718,
+      "learning_rate": 3.3881552621048426e-05,
+      "loss": 0.5601,
+      "step": 8066
+    },
+    {
+      "epoch": 10.32576,
+      "grad_norm": 0.846284031867981,
+      "learning_rate": 3.38795518207283e-05,
+      "loss": 0.5322,
+      "step": 8067
+    },
+    {
+      "epoch": 10.32704,
+      "grad_norm": 0.820878267288208,
+      "learning_rate": 3.387755102040816e-05,
+      "loss": 0.4972,
+      "step": 8068
+    },
+    {
+      "epoch": 10.32832,
+      "grad_norm": 0.8530818819999695,
+      "learning_rate": 3.3875550220088035e-05,
+      "loss": 0.5465,
+      "step": 8069
+    },
+    {
+      "epoch": 10.3296,
+      "grad_norm": 0.8610104322433472,
+      "learning_rate": 3.3873549419767907e-05,
+      "loss": 0.5503,
+      "step": 8070
+    },
+    {
+      "epoch": 10.33088,
+      "grad_norm": 0.883931040763855,
+      "learning_rate": 3.387154861944778e-05,
+      "loss": 0.5977,
+      "step": 8071
+    },
+    {
+      "epoch": 10.33216,
+      "grad_norm": 0.8754240870475769,
+      "learning_rate": 3.386954781912765e-05,
+      "loss": 0.5968,
+      "step": 8072
+    },
+    {
+      "epoch": 10.33344,
+      "grad_norm": 0.8416239023208618,
+      "learning_rate": 3.386754701880753e-05,
+      "loss": 0.5346,
+      "step": 8073
+    },
+    {
+      "epoch": 10.33472,
+      "grad_norm": 0.8367804884910583,
+      "learning_rate": 3.38655462184874e-05,
+      "loss": 0.5026,
+      "step": 8074
+    },
+    {
+      "epoch": 10.336,
+      "grad_norm": 0.8462991118431091,
+      "learning_rate": 3.386354541816727e-05,
+      "loss": 0.563,
+      "step": 8075
+    },
+    {
+      "epoch": 10.33728,
+      "grad_norm": 0.8666642308235168,
+      "learning_rate": 3.386154461784714e-05,
+      "loss": 0.5549,
+      "step": 8076
+    },
+    {
+      "epoch": 10.33856,
+      "grad_norm": 0.8881992101669312,
+      "learning_rate": 3.385954381752701e-05,
+      "loss": 0.5675,
+      "step": 8077
+    },
+    {
+      "epoch": 10.33984,
+      "grad_norm": 0.8590701222419739,
+      "learning_rate": 3.385754301720688e-05,
+      "loss": 0.5806,
+      "step": 8078
+    },
+    {
+      "epoch": 10.34112,
+      "grad_norm": 0.8979669809341431,
+      "learning_rate": 3.385554221688675e-05,
+      "loss": 0.5928,
+      "step": 8079
+    },
+    {
+      "epoch": 10.3424,
+      "grad_norm": 0.8830470442771912,
+      "learning_rate": 3.385354141656663e-05,
+      "loss": 0.5631,
+      "step": 8080
+    },
+    {
+      "epoch": 10.343679999999999,
+      "grad_norm": 0.8694333434104919,
+      "learning_rate": 3.3851540616246504e-05,
+      "loss": 0.5683,
+      "step": 8081
+    },
+    {
+      "epoch": 10.34496,
+      "grad_norm": 0.8495916128158569,
+      "learning_rate": 3.3849539815926376e-05,
+      "loss": 0.5719,
+      "step": 8082
+    },
+    {
+      "epoch": 10.34624,
+      "grad_norm": 0.8509037494659424,
+      "learning_rate": 3.384753901560625e-05,
+      "loss": 0.5438,
+      "step": 8083
+    },
+    {
+      "epoch": 10.34752,
+      "grad_norm": 0.8636916875839233,
+      "learning_rate": 3.384553821528611e-05,
+      "loss": 0.5433,
+      "step": 8084
+    },
+    {
+      "epoch": 10.3488,
+      "grad_norm": 0.8579983711242676,
+      "learning_rate": 3.3843537414965984e-05,
+      "loss": 0.5244,
+      "step": 8085
+    },
+    {
+      "epoch": 10.35008,
+      "grad_norm": 0.8326935172080994,
+      "learning_rate": 3.3841536614645856e-05,
+      "loss": 0.4873,
+      "step": 8086
+    },
+    {
+      "epoch": 10.35136,
+      "grad_norm": 0.8900673389434814,
+      "learning_rate": 3.3839535814325735e-05,
+      "loss": 0.6081,
+      "step": 8087
+    },
+    {
+      "epoch": 10.35264,
+      "grad_norm": 0.8790429830551147,
+      "learning_rate": 3.383753501400561e-05,
+      "loss": 0.5785,
+      "step": 8088
+    },
+    {
+      "epoch": 10.35392,
+      "grad_norm": 0.8508909344673157,
+      "learning_rate": 3.383553421368548e-05,
+      "loss": 0.5545,
+      "step": 8089
+    },
+    {
+      "epoch": 10.3552,
+      "grad_norm": 0.9315157532691956,
+      "learning_rate": 3.383353341336535e-05,
+      "loss": 0.5911,
+      "step": 8090
+    },
+    {
+      "epoch": 10.35648,
+      "grad_norm": 0.8326758146286011,
+      "learning_rate": 3.383153261304522e-05,
+      "loss": 0.5483,
+      "step": 8091
+    },
+    {
+      "epoch": 10.35776,
+      "grad_norm": 0.8776389956474304,
+      "learning_rate": 3.382953181272509e-05,
+      "loss": 0.5853,
+      "step": 8092
+    },
+    {
+      "epoch": 10.35904,
+      "grad_norm": 0.8722195625305176,
+      "learning_rate": 3.382753101240496e-05,
+      "loss": 0.5785,
+      "step": 8093
+    },
+    {
+      "epoch": 10.36032,
+      "grad_norm": 0.9000146389007568,
+      "learning_rate": 3.382553021208484e-05,
+      "loss": 0.5833,
+      "step": 8094
+    },
+    {
+      "epoch": 10.3616,
+      "grad_norm": 0.8690416812896729,
+      "learning_rate": 3.382352941176471e-05,
+      "loss": 0.5583,
+      "step": 8095
+    },
+    {
+      "epoch": 10.36288,
+      "grad_norm": 0.8674702048301697,
+      "learning_rate": 3.382152861144458e-05,
+      "loss": 0.5792,
+      "step": 8096
+    },
+    {
+      "epoch": 10.36416,
+      "grad_norm": 0.9020072221755981,
+      "learning_rate": 3.3819527811124453e-05,
+      "loss": 0.5737,
+      "step": 8097
+    },
+    {
+      "epoch": 10.36544,
+      "grad_norm": 0.9401588439941406,
+      "learning_rate": 3.3817527010804325e-05,
+      "loss": 0.6055,
+      "step": 8098
+    },
+    {
+      "epoch": 10.36672,
+      "grad_norm": 0.8760440349578857,
+      "learning_rate": 3.38155262104842e-05,
+      "loss": 0.5584,
+      "step": 8099
+    },
+    {
+      "epoch": 10.368,
+      "grad_norm": 0.8204819560050964,
+      "learning_rate": 3.381352541016406e-05,
+      "loss": 0.5278,
+      "step": 8100
+    },
+    {
+      "epoch": 10.36928,
+      "grad_norm": 0.9087020754814148,
+      "learning_rate": 3.3811524609843934e-05,
+      "loss": 0.5957,
+      "step": 8101
+    },
+    {
+      "epoch": 10.37056,
+      "grad_norm": 0.9149252772331238,
+      "learning_rate": 3.380952380952381e-05,
+      "loss": 0.5722,
+      "step": 8102
+    },
+    {
+      "epoch": 10.37184,
+      "grad_norm": 0.8594633936882019,
+      "learning_rate": 3.3807523009203685e-05,
+      "loss": 0.5866,
+      "step": 8103
+    },
+    {
+      "epoch": 10.37312,
+      "grad_norm": 0.8941963315010071,
+      "learning_rate": 3.3805522208883556e-05,
+      "loss": 0.566,
+      "step": 8104
+    },
+    {
+      "epoch": 10.3744,
+      "grad_norm": 0.9061945676803589,
+      "learning_rate": 3.380352140856343e-05,
+      "loss": 0.5989,
+      "step": 8105
+    },
+    {
+      "epoch": 10.37568,
+      "grad_norm": 0.9072801470756531,
+      "learning_rate": 3.38015206082433e-05,
+      "loss": 0.5341,
+      "step": 8106
+    },
+    {
+      "epoch": 10.37696,
+      "grad_norm": 0.8797597289085388,
+      "learning_rate": 3.379951980792317e-05,
+      "loss": 0.5658,
+      "step": 8107
+    },
+    {
+      "epoch": 10.37824,
+      "grad_norm": 0.8986260294914246,
+      "learning_rate": 3.379751900760304e-05,
+      "loss": 0.5663,
+      "step": 8108
+    },
+    {
+      "epoch": 10.37952,
+      "grad_norm": 0.870966374874115,
+      "learning_rate": 3.3795518207282916e-05,
+      "loss": 0.5759,
+      "step": 8109
+    },
+    {
+      "epoch": 10.3808,
+      "grad_norm": 0.8527510166168213,
+      "learning_rate": 3.379351740696279e-05,
+      "loss": 0.5566,
+      "step": 8110
+    },
+    {
+      "epoch": 10.38208,
+      "grad_norm": 0.8812639117240906,
+      "learning_rate": 3.379151660664266e-05,
+      "loss": 0.5787,
+      "step": 8111
+    },
+    {
+      "epoch": 10.38336,
+      "grad_norm": 0.9007368683815002,
+      "learning_rate": 3.378951580632253e-05,
+      "loss": 0.6065,
+      "step": 8112
+    },
+    {
+      "epoch": 10.38464,
+      "grad_norm": 0.908555269241333,
+      "learning_rate": 3.37875150060024e-05,
+      "loss": 0.5707,
+      "step": 8113
+    },
+    {
+      "epoch": 10.38592,
+      "grad_norm": 0.8893713355064392,
+      "learning_rate": 3.3785514205682275e-05,
+      "loss": 0.6067,
+      "step": 8114
+    },
+    {
+      "epoch": 10.3872,
+      "grad_norm": 0.8574310541152954,
+      "learning_rate": 3.378351340536215e-05,
+      "loss": 0.5643,
+      "step": 8115
+    },
+    {
+      "epoch": 10.38848,
+      "grad_norm": 0.823165774345398,
+      "learning_rate": 3.378151260504202e-05,
+      "loss": 0.5808,
+      "step": 8116
+    },
+    {
+      "epoch": 10.38976,
+      "grad_norm": 0.818540096282959,
+      "learning_rate": 3.377951180472189e-05,
+      "loss": 0.5709,
+      "step": 8117
+    },
+    {
+      "epoch": 10.39104,
+      "grad_norm": 0.8618648052215576,
+      "learning_rate": 3.377751100440176e-05,
+      "loss": 0.5565,
+      "step": 8118
+    },
+    {
+      "epoch": 10.39232,
+      "grad_norm": 0.869215726852417,
+      "learning_rate": 3.3775510204081634e-05,
+      "loss": 0.5499,
+      "step": 8119
+    },
+    {
+      "epoch": 10.3936,
+      "grad_norm": 0.9559123516082764,
+      "learning_rate": 3.3773509403761506e-05,
+      "loss": 0.6168,
+      "step": 8120
+    },
+    {
+      "epoch": 10.39488,
+      "grad_norm": 0.924443244934082,
+      "learning_rate": 3.377150860344138e-05,
+      "loss": 0.5954,
+      "step": 8121
+    },
+    {
+      "epoch": 10.39616,
+      "grad_norm": 0.8861650824546814,
+      "learning_rate": 3.376950780312125e-05,
+      "loss": 0.5793,
+      "step": 8122
+    },
+    {
+      "epoch": 10.39744,
+      "grad_norm": 0.9248014092445374,
+      "learning_rate": 3.376750700280112e-05,
+      "loss": 0.5581,
+      "step": 8123
+    },
+    {
+      "epoch": 10.39872,
+      "grad_norm": 0.8819634914398193,
+      "learning_rate": 3.3765506202480994e-05,
+      "loss": 0.5263,
+      "step": 8124
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 0.8656063079833984,
+      "learning_rate": 3.3763505402160865e-05,
+      "loss": 0.5241,
+      "step": 8125
+    },
+    {
+      "epoch": 10.40128,
+      "grad_norm": 0.8893786668777466,
+      "learning_rate": 3.376150460184074e-05,
+      "loss": 0.5451,
+      "step": 8126
+    },
+    {
+      "epoch": 10.40256,
+      "grad_norm": 0.9260717034339905,
+      "learning_rate": 3.375950380152061e-05,
+      "loss": 0.5505,
+      "step": 8127
+    },
+    {
+      "epoch": 10.40384,
+      "grad_norm": 0.912977397441864,
+      "learning_rate": 3.375750300120048e-05,
+      "loss": 0.6363,
+      "step": 8128
+    },
+    {
+      "epoch": 10.40512,
+      "grad_norm": 0.8732292652130127,
+      "learning_rate": 3.375550220088035e-05,
+      "loss": 0.5778,
+      "step": 8129
+    },
+    {
+      "epoch": 10.4064,
+      "grad_norm": 0.8404425978660583,
+      "learning_rate": 3.3753501400560225e-05,
+      "loss": 0.5693,
+      "step": 8130
+    },
+    {
+      "epoch": 10.40768,
+      "grad_norm": 0.9023398160934448,
+      "learning_rate": 3.3751500600240097e-05,
+      "loss": 0.6112,
+      "step": 8131
+    },
+    {
+      "epoch": 10.40896,
+      "grad_norm": 0.9250540137290955,
+      "learning_rate": 3.374949979991997e-05,
+      "loss": 0.6067,
+      "step": 8132
+    },
+    {
+      "epoch": 10.41024,
+      "grad_norm": 0.9005985856056213,
+      "learning_rate": 3.374749899959984e-05,
+      "loss": 0.5746,
+      "step": 8133
+    },
+    {
+      "epoch": 10.41152,
+      "grad_norm": 0.9189587235450745,
+      "learning_rate": 3.374549819927971e-05,
+      "loss": 0.5847,
+      "step": 8134
+    },
+    {
+      "epoch": 10.4128,
+      "grad_norm": 0.8882416486740112,
+      "learning_rate": 3.3743497398959584e-05,
+      "loss": 0.6156,
+      "step": 8135
+    },
+    {
+      "epoch": 10.41408,
+      "grad_norm": 0.845780074596405,
+      "learning_rate": 3.3741496598639456e-05,
+      "loss": 0.549,
+      "step": 8136
+    },
+    {
+      "epoch": 10.41536,
+      "grad_norm": 0.9068146347999573,
+      "learning_rate": 3.3739495798319334e-05,
+      "loss": 0.5735,
+      "step": 8137
+    },
+    {
+      "epoch": 10.41664,
+      "grad_norm": 0.8547993898391724,
+      "learning_rate": 3.37374949979992e-05,
+      "loss": 0.5472,
+      "step": 8138
+    },
+    {
+      "epoch": 10.41792,
+      "grad_norm": 0.8667715191841125,
+      "learning_rate": 3.373549419767907e-05,
+      "loss": 0.5533,
+      "step": 8139
+    },
+    {
+      "epoch": 10.4192,
+      "grad_norm": 0.8902264833450317,
+      "learning_rate": 3.373349339735894e-05,
+      "loss": 0.5751,
+      "step": 8140
+    },
+    {
+      "epoch": 10.42048,
+      "grad_norm": 0.8540170788764954,
+      "learning_rate": 3.3731492597038815e-05,
+      "loss": 0.5179,
+      "step": 8141
+    },
+    {
+      "epoch": 10.42176,
+      "grad_norm": 0.8607949614524841,
+      "learning_rate": 3.372949179671869e-05,
+      "loss": 0.5825,
+      "step": 8142
+    },
+    {
+      "epoch": 10.42304,
+      "grad_norm": 0.8747577667236328,
+      "learning_rate": 3.372749099639856e-05,
+      "loss": 0.6228,
+      "step": 8143
+    },
+    {
+      "epoch": 10.42432,
+      "grad_norm": 0.8700942993164062,
+      "learning_rate": 3.372549019607844e-05,
+      "loss": 0.5888,
+      "step": 8144
+    },
+    {
+      "epoch": 10.4256,
+      "grad_norm": 0.9046246409416199,
+      "learning_rate": 3.372348939575831e-05,
+      "loss": 0.5769,
+      "step": 8145
+    },
+    {
+      "epoch": 10.42688,
+      "grad_norm": 0.8095414042472839,
+      "learning_rate": 3.3721488595438174e-05,
+      "loss": 0.5288,
+      "step": 8146
+    },
+    {
+      "epoch": 10.42816,
+      "grad_norm": 0.8345249891281128,
+      "learning_rate": 3.3719487795118046e-05,
+      "loss": 0.5943,
+      "step": 8147
+    },
+    {
+      "epoch": 10.42944,
+      "grad_norm": 0.9272936582565308,
+      "learning_rate": 3.371748699479792e-05,
+      "loss": 0.5317,
+      "step": 8148
+    },
+    {
+      "epoch": 10.43072,
+      "grad_norm": 0.9539486169815063,
+      "learning_rate": 3.371548619447779e-05,
+      "loss": 0.6475,
+      "step": 8149
+    },
+    {
+      "epoch": 10.432,
+      "grad_norm": Infinity,
+      "learning_rate": 3.371548619447779e-05,
+      "loss": 0.6468,
+      "step": 8150
+    },
+    {
+      "epoch": 10.43328,
+      "grad_norm": 0.8678048253059387,
+      "learning_rate": 3.371348539415766e-05,
+      "loss": 0.5463,
+      "step": 8151
+    },
+    {
+      "epoch": 10.43456,
+      "grad_norm": 0.8489457964897156,
+      "learning_rate": 3.371148459383754e-05,
+      "loss": 0.5703,
+      "step": 8152
+    },
+    {
+      "epoch": 10.43584,
+      "grad_norm": 0.8788950443267822,
+      "learning_rate": 3.370948379351741e-05,
+      "loss": 0.5346,
+      "step": 8153
+    },
+    {
+      "epoch": 10.43712,
+      "grad_norm": 0.853437066078186,
+      "learning_rate": 3.3707482993197284e-05,
+      "loss": 0.5659,
+      "step": 8154
+    },
+    {
+      "epoch": 10.4384,
+      "grad_norm": 0.8777585625648499,
+      "learning_rate": 3.370548219287715e-05,
+      "loss": 0.5578,
+      "step": 8155
+    },
+    {
+      "epoch": 10.43968,
+      "grad_norm": 0.8524881601333618,
+      "learning_rate": 3.370348139255702e-05,
+      "loss": 0.5656,
+      "step": 8156
+    },
+    {
+      "epoch": 10.44096,
+      "grad_norm": 0.8162695169448853,
+      "learning_rate": 3.370148059223689e-05,
+      "loss": 0.5092,
+      "step": 8157
+    },
+    {
+      "epoch": 10.44224,
+      "grad_norm": 0.8364541530609131,
+      "learning_rate": 3.3699479791916765e-05,
+      "loss": 0.5346,
+      "step": 8158
+    },
+    {
+      "epoch": 10.44352,
+      "grad_norm": 0.9183416962623596,
+      "learning_rate": 3.3697478991596643e-05,
+      "loss": 0.6277,
+      "step": 8159
+    },
+    {
+      "epoch": 10.4448,
+      "grad_norm": 0.8532844185829163,
+      "learning_rate": 3.3695478191276515e-05,
+      "loss": 0.5318,
+      "step": 8160
+    },
+    {
+      "epoch": 10.44608,
+      "grad_norm": 0.8698902726173401,
+      "learning_rate": 3.369347739095639e-05,
+      "loss": 0.5872,
+      "step": 8161
+    },
+    {
+      "epoch": 10.44736,
+      "grad_norm": 0.8348177671432495,
+      "learning_rate": 3.369147659063626e-05,
+      "loss": 0.5381,
+      "step": 8162
+    },
+    {
+      "epoch": 10.44864,
+      "grad_norm": 0.8752634525299072,
+      "learning_rate": 3.3689475790316124e-05,
+      "loss": 0.5587,
+      "step": 8163
+    },
+    {
+      "epoch": 10.44992,
+      "grad_norm": 0.8772228360176086,
+      "learning_rate": 3.3687474989995996e-05,
+      "loss": 0.5763,
+      "step": 8164
+    },
+    {
+      "epoch": 10.4512,
+      "grad_norm": 0.8985121250152588,
+      "learning_rate": 3.368547418967587e-05,
+      "loss": 0.5443,
+      "step": 8165
+    },
+    {
+      "epoch": 10.45248,
+      "grad_norm": 0.8824835419654846,
+      "learning_rate": 3.3683473389355746e-05,
+      "loss": 0.5583,
+      "step": 8166
+    },
+    {
+      "epoch": 10.45376,
+      "grad_norm": 0.8776131272315979,
+      "learning_rate": 3.368147258903562e-05,
+      "loss": 0.5331,
+      "step": 8167
+    },
+    {
+      "epoch": 10.45504,
+      "grad_norm": 0.9394381046295166,
+      "learning_rate": 3.367947178871549e-05,
+      "loss": 0.6115,
+      "step": 8168
+    },
+    {
+      "epoch": 10.45632,
+      "grad_norm": 0.8873891234397888,
+      "learning_rate": 3.367747098839536e-05,
+      "loss": 0.5574,
+      "step": 8169
+    },
+    {
+      "epoch": 10.4576,
+      "grad_norm": 0.8328064680099487,
+      "learning_rate": 3.3675470188075234e-05,
+      "loss": 0.5499,
+      "step": 8170
+    },
+    {
+      "epoch": 10.45888,
+      "grad_norm": 0.8790708184242249,
+      "learning_rate": 3.36734693877551e-05,
+      "loss": 0.5231,
+      "step": 8171
+    },
+    {
+      "epoch": 10.46016,
+      "grad_norm": 0.8684185743331909,
+      "learning_rate": 3.367146858743497e-05,
+      "loss": 0.5494,
+      "step": 8172
+    },
+    {
+      "epoch": 10.46144,
+      "grad_norm": 0.9176812767982483,
+      "learning_rate": 3.366946778711485e-05,
+      "loss": 0.5929,
+      "step": 8173
+    },
+    {
+      "epoch": 10.462720000000001,
+      "grad_norm": 0.9031935930252075,
+      "learning_rate": 3.366746698679472e-05,
+      "loss": 0.5999,
+      "step": 8174
+    },
+    {
+      "epoch": 10.464,
+      "grad_norm": 0.8639822006225586,
+      "learning_rate": 3.366546618647459e-05,
+      "loss": 0.6266,
+      "step": 8175
+    },
+    {
+      "epoch": 10.46528,
+      "grad_norm": 0.8309516310691833,
+      "learning_rate": 3.3663465386154465e-05,
+      "loss": 0.5368,
+      "step": 8176
+    },
+    {
+      "epoch": 10.46656,
+      "grad_norm": 0.8735124468803406,
+      "learning_rate": 3.366146458583434e-05,
+      "loss": 0.5759,
+      "step": 8177
+    },
+    {
+      "epoch": 10.46784,
+      "grad_norm": 0.850908100605011,
+      "learning_rate": 3.365946378551421e-05,
+      "loss": 0.5377,
+      "step": 8178
+    },
+    {
+      "epoch": 10.46912,
+      "grad_norm": 0.8890109062194824,
+      "learning_rate": 3.3657462985194074e-05,
+      "loss": 0.5821,
+      "step": 8179
+    },
+    {
+      "epoch": 10.4704,
+      "grad_norm": 0.8895400762557983,
+      "learning_rate": 3.365546218487395e-05,
+      "loss": 0.5535,
+      "step": 8180
+    },
+    {
+      "epoch": 10.47168,
+      "grad_norm": 0.8601114153862,
+      "learning_rate": 3.3653461384553824e-05,
+      "loss": 0.5324,
+      "step": 8181
+    },
+    {
+      "epoch": 10.47296,
+      "grad_norm": 0.8928337693214417,
+      "learning_rate": 3.3651460584233696e-05,
+      "loss": 0.5868,
+      "step": 8182
+    },
+    {
+      "epoch": 10.47424,
+      "grad_norm": 0.8805640935897827,
+      "learning_rate": 3.364945978391357e-05,
+      "loss": 0.5735,
+      "step": 8183
+    },
+    {
+      "epoch": 10.47552,
+      "grad_norm": 0.8599642515182495,
+      "learning_rate": 3.364745898359344e-05,
+      "loss": 0.5431,
+      "step": 8184
+    },
+    {
+      "epoch": 10.4768,
+      "grad_norm": 0.9143441319465637,
+      "learning_rate": 3.364545818327331e-05,
+      "loss": 0.5903,
+      "step": 8185
+    },
+    {
+      "epoch": 10.47808,
+      "grad_norm": 0.8206641674041748,
+      "learning_rate": 3.3643457382953184e-05,
+      "loss": 0.5216,
+      "step": 8186
+    },
+    {
+      "epoch": 10.47936,
+      "grad_norm": 0.7796454429626465,
+      "learning_rate": 3.3641456582633055e-05,
+      "loss": 0.4986,
+      "step": 8187
+    },
+    {
+      "epoch": 10.48064,
+      "grad_norm": 0.8821196556091309,
+      "learning_rate": 3.363945578231293e-05,
+      "loss": 0.5665,
+      "step": 8188
+    },
+    {
+      "epoch": 10.48192,
+      "grad_norm": 0.86763596534729,
+      "learning_rate": 3.36374549819928e-05,
+      "loss": 0.5393,
+      "step": 8189
+    },
+    {
+      "epoch": 10.4832,
+      "grad_norm": 0.8729016184806824,
+      "learning_rate": 3.363545418167267e-05,
+      "loss": 0.5357,
+      "step": 8190
+    },
+    {
+      "epoch": 10.48448,
+      "grad_norm": 0.9105653166770935,
+      "learning_rate": 3.363345338135254e-05,
+      "loss": 0.6309,
+      "step": 8191
+    },
+    {
+      "epoch": 10.48576,
+      "grad_norm": 0.8702036142349243,
+      "learning_rate": 3.3631452581032415e-05,
+      "loss": 0.5913,
+      "step": 8192
+    },
+    {
+      "epoch": 10.48704,
+      "grad_norm": 0.8215315937995911,
+      "learning_rate": 3.3629451780712287e-05,
+      "loss": 0.585,
+      "step": 8193
+    },
+    {
+      "epoch": 10.48832,
+      "grad_norm": 0.8497202396392822,
+      "learning_rate": 3.362745098039216e-05,
+      "loss": 0.5519,
+      "step": 8194
+    },
+    {
+      "epoch": 10.4896,
+      "grad_norm": 0.8402096629142761,
+      "learning_rate": 3.362545018007203e-05,
+      "loss": 0.5624,
+      "step": 8195
+    },
+    {
+      "epoch": 10.49088,
+      "grad_norm": 0.8736252784729004,
+      "learning_rate": 3.36234493797519e-05,
+      "loss": 0.5404,
+      "step": 8196
+    },
+    {
+      "epoch": 10.49216,
+      "grad_norm": 0.8728154301643372,
+      "learning_rate": 3.3621448579431774e-05,
+      "loss": 0.5412,
+      "step": 8197
+    },
+    {
+      "epoch": 10.49344,
+      "grad_norm": 0.8873513340950012,
+      "learning_rate": 3.3619447779111646e-05,
+      "loss": 0.5671,
+      "step": 8198
+    },
+    {
+      "epoch": 10.49472,
+      "grad_norm": 0.873191237449646,
+      "learning_rate": 3.361744697879152e-05,
+      "loss": 0.5645,
+      "step": 8199
+    },
+    {
+      "epoch": 10.496,
+      "grad_norm": 0.8937346339225769,
+      "learning_rate": 3.361544617847139e-05,
+      "loss": 0.604,
+      "step": 8200
+    },
+    {
+      "epoch": 10.49728,
+      "grad_norm": 0.8788394927978516,
+      "learning_rate": 3.361344537815127e-05,
+      "loss": 0.6088,
+      "step": 8201
+    },
+    {
+      "epoch": 10.49856,
+      "grad_norm": 0.8717678189277649,
+      "learning_rate": 3.361144457783113e-05,
+      "loss": 0.5475,
+      "step": 8202
+    },
+    {
+      "epoch": 10.49984,
+      "grad_norm": 0.8855169415473938,
+      "learning_rate": 3.3609443777511005e-05,
+      "loss": 0.6181,
+      "step": 8203
+    },
+    {
+      "epoch": 10.50112,
+      "grad_norm": 0.9005426168441772,
+      "learning_rate": 3.360744297719088e-05,
+      "loss": 0.5924,
+      "step": 8204
+    },
+    {
+      "epoch": 10.5024,
+      "grad_norm": 0.9078197479248047,
+      "learning_rate": 3.360544217687075e-05,
+      "loss": 0.6005,
+      "step": 8205
+    },
+    {
+      "epoch": 10.50368,
+      "grad_norm": 0.9592050909996033,
+      "learning_rate": 3.360344137655062e-05,
+      "loss": 0.6359,
+      "step": 8206
+    },
+    {
+      "epoch": 10.50496,
+      "grad_norm": 0.8957341313362122,
+      "learning_rate": 3.360144057623049e-05,
+      "loss": 0.5806,
+      "step": 8207
+    },
+    {
+      "epoch": 10.50624,
+      "grad_norm": 0.8797503709793091,
+      "learning_rate": 3.359943977591037e-05,
+      "loss": 0.5798,
+      "step": 8208
+    },
+    {
+      "epoch": 10.50752,
+      "grad_norm": 0.8420235514640808,
+      "learning_rate": 3.359743897559024e-05,
+      "loss": 0.5894,
+      "step": 8209
+    },
+    {
+      "epoch": 10.5088,
+      "grad_norm": 0.8644899725914001,
+      "learning_rate": 3.359543817527011e-05,
+      "loss": 0.553,
+      "step": 8210
+    },
+    {
+      "epoch": 10.51008,
+      "grad_norm": 0.8995040059089661,
+      "learning_rate": 3.359343737494998e-05,
+      "loss": 0.6117,
+      "step": 8211
+    },
+    {
+      "epoch": 10.51136,
+      "grad_norm": 0.8395155668258667,
+      "learning_rate": 3.359143657462985e-05,
+      "loss": 0.5678,
+      "step": 8212
+    },
+    {
+      "epoch": 10.51264,
+      "grad_norm": 0.8515924215316772,
+      "learning_rate": 3.3589435774309724e-05,
+      "loss": 0.5409,
+      "step": 8213
+    },
+    {
+      "epoch": 10.51392,
+      "grad_norm": 0.9288252592086792,
+      "learning_rate": 3.3587434973989596e-05,
+      "loss": 0.6328,
+      "step": 8214
+    },
+    {
+      "epoch": 10.5152,
+      "grad_norm": 0.9206863045692444,
+      "learning_rate": 3.358543417366947e-05,
+      "loss": 0.6071,
+      "step": 8215
+    },
+    {
+      "epoch": 10.51648,
+      "grad_norm": 0.875196099281311,
+      "learning_rate": 3.3583433373349346e-05,
+      "loss": 0.5762,
+      "step": 8216
+    },
+    {
+      "epoch": 10.517759999999999,
+      "grad_norm": 0.8598700165748596,
+      "learning_rate": 3.358143257302921e-05,
+      "loss": 0.5428,
+      "step": 8217
+    },
+    {
+      "epoch": 10.51904,
+      "grad_norm": 0.8348386287689209,
+      "learning_rate": 3.357943177270908e-05,
+      "loss": 0.5135,
+      "step": 8218
+    },
+    {
+      "epoch": 10.52032,
+      "grad_norm": 0.8748103380203247,
+      "learning_rate": 3.3577430972388955e-05,
+      "loss": 0.5582,
+      "step": 8219
+    },
+    {
+      "epoch": 10.5216,
+      "grad_norm": 0.8388493657112122,
+      "learning_rate": 3.357543017206883e-05,
+      "loss": 0.5594,
+      "step": 8220
+    },
+    {
+      "epoch": 10.52288,
+      "grad_norm": 0.8430709838867188,
+      "learning_rate": 3.35734293717487e-05,
+      "loss": 0.5304,
+      "step": 8221
+    },
+    {
+      "epoch": 10.52416,
+      "grad_norm": 0.8487758040428162,
+      "learning_rate": 3.357142857142857e-05,
+      "loss": 0.5249,
+      "step": 8222
+    },
+    {
+      "epoch": 10.52544,
+      "grad_norm": 0.8581976890563965,
+      "learning_rate": 3.356942777110845e-05,
+      "loss": 0.5209,
+      "step": 8223
+    },
+    {
+      "epoch": 10.52672,
+      "grad_norm": 0.9251878261566162,
+      "learning_rate": 3.356742697078832e-05,
+      "loss": 0.5878,
+      "step": 8224
+    },
+    {
+      "epoch": 10.528,
+      "grad_norm": 0.8627904057502747,
+      "learning_rate": 3.3565426170468186e-05,
+      "loss": 0.5766,
+      "step": 8225
+    },
+    {
+      "epoch": 10.52928,
+      "grad_norm": 0.8664613962173462,
+      "learning_rate": 3.356342537014806e-05,
+      "loss": 0.562,
+      "step": 8226
+    },
+    {
+      "epoch": 10.53056,
+      "grad_norm": 0.8714502453804016,
+      "learning_rate": 3.356142456982793e-05,
+      "loss": 0.578,
+      "step": 8227
+    },
+    {
+      "epoch": 10.53184,
+      "grad_norm": 0.8915185332298279,
+      "learning_rate": 3.35594237695078e-05,
+      "loss": 0.5542,
+      "step": 8228
+    },
+    {
+      "epoch": 10.53312,
+      "grad_norm": 0.9236972332000732,
+      "learning_rate": 3.3557422969187673e-05,
+      "loss": 0.5776,
+      "step": 8229
+    },
+    {
+      "epoch": 10.5344,
+      "grad_norm": 0.8826823234558105,
+      "learning_rate": 3.355542216886755e-05,
+      "loss": 0.5879,
+      "step": 8230
+    },
+    {
+      "epoch": 10.53568,
+      "grad_norm": 0.8796812295913696,
+      "learning_rate": 3.3553421368547424e-05,
+      "loss": 0.5992,
+      "step": 8231
+    },
+    {
+      "epoch": 10.53696,
+      "grad_norm": 0.8763803243637085,
+      "learning_rate": 3.3551420568227296e-05,
+      "loss": 0.5546,
+      "step": 8232
+    },
+    {
+      "epoch": 10.53824,
+      "grad_norm": 0.8898423910140991,
+      "learning_rate": 3.354941976790716e-05,
+      "loss": 0.5755,
+      "step": 8233
+    },
+    {
+      "epoch": 10.53952,
+      "grad_norm": 0.8547391295433044,
+      "learning_rate": 3.354741896758703e-05,
+      "loss": 0.5844,
+      "step": 8234
+    },
+    {
+      "epoch": 10.5408,
+      "grad_norm": 0.8550713658332825,
+      "learning_rate": 3.3545418167266905e-05,
+      "loss": 0.5609,
+      "step": 8235
+    },
+    {
+      "epoch": 10.54208,
+      "grad_norm": 0.8670467138290405,
+      "learning_rate": 3.3543417366946776e-05,
+      "loss": 0.5439,
+      "step": 8236
+    },
+    {
+      "epoch": 10.54336,
+      "grad_norm": 0.8298435807228088,
+      "learning_rate": 3.3541416566626655e-05,
+      "loss": 0.5424,
+      "step": 8237
+    },
+    {
+      "epoch": 10.54464,
+      "grad_norm": 0.8586260676383972,
+      "learning_rate": 3.353941576630653e-05,
+      "loss": 0.5316,
+      "step": 8238
+    },
+    {
+      "epoch": 10.54592,
+      "grad_norm": 0.8656241297721863,
+      "learning_rate": 3.35374149659864e-05,
+      "loss": 0.5515,
+      "step": 8239
+    },
+    {
+      "epoch": 10.5472,
+      "grad_norm": 0.8878941535949707,
+      "learning_rate": 3.353541416566627e-05,
+      "loss": 0.5874,
+      "step": 8240
+    },
+    {
+      "epoch": 10.54848,
+      "grad_norm": 0.900540828704834,
+      "learning_rate": 3.3533413365346136e-05,
+      "loss": 0.6261,
+      "step": 8241
+    },
+    {
+      "epoch": 10.54976,
+      "grad_norm": 0.8868071436882019,
+      "learning_rate": 3.353141256502601e-05,
+      "loss": 0.5749,
+      "step": 8242
+    },
+    {
+      "epoch": 10.55104,
+      "grad_norm": 0.9406156539916992,
+      "learning_rate": 3.352941176470588e-05,
+      "loss": 0.6071,
+      "step": 8243
+    },
+    {
+      "epoch": 10.55232,
+      "grad_norm": 0.9124932289123535,
+      "learning_rate": 3.352741096438576e-05,
+      "loss": 0.587,
+      "step": 8244
+    },
+    {
+      "epoch": 10.5536,
+      "grad_norm": 0.8528494834899902,
+      "learning_rate": 3.352541016406563e-05,
+      "loss": 0.5266,
+      "step": 8245
+    },
+    {
+      "epoch": 10.55488,
+      "grad_norm": 0.8507541418075562,
+      "learning_rate": 3.35234093637455e-05,
+      "loss": 0.5906,
+      "step": 8246
+    },
+    {
+      "epoch": 10.55616,
+      "grad_norm": 0.8238681554794312,
+      "learning_rate": 3.3521408563425374e-05,
+      "loss": 0.5421,
+      "step": 8247
+    },
+    {
+      "epoch": 10.55744,
+      "grad_norm": 0.8529794812202454,
+      "learning_rate": 3.3519407763105245e-05,
+      "loss": 0.5699,
+      "step": 8248
+    },
+    {
+      "epoch": 10.55872,
+      "grad_norm": 0.8948842883110046,
+      "learning_rate": 3.351740696278511e-05,
+      "loss": 0.5987,
+      "step": 8249
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 0.8481034636497498,
+      "learning_rate": 3.351540616246498e-05,
+      "loss": 0.5355,
+      "step": 8250
+    },
+    {
+      "epoch": 10.56128,
+      "grad_norm": 0.900606632232666,
+      "learning_rate": 3.351340536214486e-05,
+      "loss": 0.6198,
+      "step": 8251
+    },
+    {
+      "epoch": 10.56256,
+      "grad_norm": 0.8651003241539001,
+      "learning_rate": 3.351140456182473e-05,
+      "loss": 0.5792,
+      "step": 8252
+    },
+    {
+      "epoch": 10.56384,
+      "grad_norm": 0.8270814418792725,
+      "learning_rate": 3.3509403761504605e-05,
+      "loss": 0.5138,
+      "step": 8253
+    },
+    {
+      "epoch": 10.56512,
+      "grad_norm": 0.8760477900505066,
+      "learning_rate": 3.350740296118448e-05,
+      "loss": 0.5928,
+      "step": 8254
+    },
+    {
+      "epoch": 10.5664,
+      "grad_norm": 0.889542818069458,
+      "learning_rate": 3.350540216086435e-05,
+      "loss": 0.6012,
+      "step": 8255
+    },
+    {
+      "epoch": 10.56768,
+      "grad_norm": 0.8778030872344971,
+      "learning_rate": 3.350340136054422e-05,
+      "loss": 0.5747,
+      "step": 8256
+    },
+    {
+      "epoch": 10.56896,
+      "grad_norm": 0.889215350151062,
+      "learning_rate": 3.3501400560224085e-05,
+      "loss": 0.5577,
+      "step": 8257
+    },
+    {
+      "epoch": 10.57024,
+      "grad_norm": 0.8863917589187622,
+      "learning_rate": 3.3499399759903964e-05,
+      "loss": 0.5355,
+      "step": 8258
+    },
+    {
+      "epoch": 10.57152,
+      "grad_norm": 0.8980902433395386,
+      "learning_rate": 3.3497398959583836e-05,
+      "loss": 0.5563,
+      "step": 8259
+    },
+    {
+      "epoch": 10.5728,
+      "grad_norm": 0.8849428296089172,
+      "learning_rate": 3.349539815926371e-05,
+      "loss": 0.578,
+      "step": 8260
+    },
+    {
+      "epoch": 10.57408,
+      "grad_norm": 0.8660392761230469,
+      "learning_rate": 3.349339735894358e-05,
+      "loss": 0.5187,
+      "step": 8261
+    },
+    {
+      "epoch": 10.57536,
+      "grad_norm": 0.891211986541748,
+      "learning_rate": 3.349139655862345e-05,
+      "loss": 0.5778,
+      "step": 8262
+    },
+    {
+      "epoch": 10.57664,
+      "grad_norm": 0.8924466967582703,
+      "learning_rate": 3.348939575830332e-05,
+      "loss": 0.5561,
+      "step": 8263
+    },
+    {
+      "epoch": 10.57792,
+      "grad_norm": 0.9162275791168213,
+      "learning_rate": 3.3487394957983195e-05,
+      "loss": 0.6179,
+      "step": 8264
+    },
+    {
+      "epoch": 10.5792,
+      "grad_norm": 0.9216055870056152,
+      "learning_rate": 3.348539415766307e-05,
+      "loss": 0.5975,
+      "step": 8265
+    },
+    {
+      "epoch": 10.58048,
+      "grad_norm": 0.8706624507904053,
+      "learning_rate": 3.348339335734294e-05,
+      "loss": 0.5412,
+      "step": 8266
+    },
+    {
+      "epoch": 10.58176,
+      "grad_norm": 0.8571588397026062,
+      "learning_rate": 3.348139255702281e-05,
+      "loss": 0.5807,
+      "step": 8267
+    },
+    {
+      "epoch": 10.58304,
+      "grad_norm": 0.7969219088554382,
+      "learning_rate": 3.347939175670268e-05,
+      "loss": 0.5632,
+      "step": 8268
+    },
+    {
+      "epoch": 10.58432,
+      "grad_norm": 0.8372544050216675,
+      "learning_rate": 3.3477390956382554e-05,
+      "loss": 0.5497,
+      "step": 8269
+    },
+    {
+      "epoch": 10.5856,
+      "grad_norm": 0.8220162987709045,
+      "learning_rate": 3.3475390156062426e-05,
+      "loss": 0.5106,
+      "step": 8270
+    },
+    {
+      "epoch": 10.58688,
+      "grad_norm": 0.8827033042907715,
+      "learning_rate": 3.34733893557423e-05,
+      "loss": 0.5885,
+      "step": 8271
+    },
+    {
+      "epoch": 10.58816,
+      "grad_norm": 0.8495452404022217,
+      "learning_rate": 3.347138855542217e-05,
+      "loss": 0.565,
+      "step": 8272
+    },
+    {
+      "epoch": 10.58944,
+      "grad_norm": 0.8994190692901611,
+      "learning_rate": 3.346938775510204e-05,
+      "loss": 0.5612,
+      "step": 8273
+    },
+    {
+      "epoch": 10.59072,
+      "grad_norm": 0.8603847622871399,
+      "learning_rate": 3.3467386954781914e-05,
+      "loss": 0.5775,
+      "step": 8274
+    },
+    {
+      "epoch": 10.592,
+      "grad_norm": 0.884907066822052,
+      "learning_rate": 3.3465386154461786e-05,
+      "loss": 0.5231,
+      "step": 8275
+    },
+    {
+      "epoch": 10.59328,
+      "grad_norm": 0.9084495306015015,
+      "learning_rate": 3.346338535414166e-05,
+      "loss": 0.5845,
+      "step": 8276
+    },
+    {
+      "epoch": 10.59456,
+      "grad_norm": 0.8667944669723511,
+      "learning_rate": 3.346138455382153e-05,
+      "loss": 0.5664,
+      "step": 8277
+    },
+    {
+      "epoch": 10.59584,
+      "grad_norm": 0.8516685366630554,
+      "learning_rate": 3.34593837535014e-05,
+      "loss": 0.5483,
+      "step": 8278
+    },
+    {
+      "epoch": 10.59712,
+      "grad_norm": 0.8369115591049194,
+      "learning_rate": 3.345738295318128e-05,
+      "loss": 0.505,
+      "step": 8279
+    },
+    {
+      "epoch": 10.5984,
+      "grad_norm": 0.8858165740966797,
+      "learning_rate": 3.3455382152861145e-05,
+      "loss": 0.5714,
+      "step": 8280
+    },
+    {
+      "epoch": 10.59968,
+      "grad_norm": 0.858709990978241,
+      "learning_rate": 3.345338135254102e-05,
+      "loss": 0.5921,
+      "step": 8281
+    },
+    {
+      "epoch": 10.60096,
+      "grad_norm": 0.8412848114967346,
+      "learning_rate": 3.345138055222089e-05,
+      "loss": 0.5848,
+      "step": 8282
+    },
+    {
+      "epoch": 10.60224,
+      "grad_norm": 0.8578768968582153,
+      "learning_rate": 3.344937975190076e-05,
+      "loss": 0.564,
+      "step": 8283
+    },
+    {
+      "epoch": 10.60352,
+      "grad_norm": 0.8731163144111633,
+      "learning_rate": 3.344737895158063e-05,
+      "loss": 0.5947,
+      "step": 8284
+    },
+    {
+      "epoch": 10.604800000000001,
+      "grad_norm": 0.8939322829246521,
+      "learning_rate": 3.3445378151260504e-05,
+      "loss": 0.6458,
+      "step": 8285
+    },
+    {
+      "epoch": 10.60608,
+      "grad_norm": 0.8193719983100891,
+      "learning_rate": 3.344337735094038e-05,
+      "loss": 0.5601,
+      "step": 8286
+    },
+    {
+      "epoch": 10.60736,
+      "grad_norm": 0.9018811583518982,
+      "learning_rate": 3.3441376550620255e-05,
+      "loss": 0.5914,
+      "step": 8287
+    },
+    {
+      "epoch": 10.60864,
+      "grad_norm": 0.8852116465568542,
+      "learning_rate": 3.343937575030012e-05,
+      "loss": 0.5545,
+      "step": 8288
+    },
+    {
+      "epoch": 10.60992,
+      "grad_norm": 0.880398690700531,
+      "learning_rate": 3.343737494997999e-05,
+      "loss": 0.5642,
+      "step": 8289
+    },
+    {
+      "epoch": 10.6112,
+      "grad_norm": 0.8978055715560913,
+      "learning_rate": 3.3435374149659863e-05,
+      "loss": 0.5649,
+      "step": 8290
+    },
+    {
+      "epoch": 10.61248,
+      "grad_norm": 0.8889503479003906,
+      "learning_rate": 3.3433373349339735e-05,
+      "loss": 0.6023,
+      "step": 8291
+    },
+    {
+      "epoch": 10.61376,
+      "grad_norm": 0.8328986167907715,
+      "learning_rate": 3.343137254901961e-05,
+      "loss": 0.5532,
+      "step": 8292
+    },
+    {
+      "epoch": 10.61504,
+      "grad_norm": 0.8667716979980469,
+      "learning_rate": 3.3429371748699486e-05,
+      "loss": 0.5735,
+      "step": 8293
+    },
+    {
+      "epoch": 10.61632,
+      "grad_norm": 0.8680749535560608,
+      "learning_rate": 3.342737094837936e-05,
+      "loss": 0.5629,
+      "step": 8294
+    },
+    {
+      "epoch": 10.6176,
+      "grad_norm": 0.8346037268638611,
+      "learning_rate": 3.342537014805923e-05,
+      "loss": 0.5411,
+      "step": 8295
+    },
+    {
+      "epoch": 10.61888,
+      "grad_norm": 0.8703474402427673,
+      "learning_rate": 3.3423369347739095e-05,
+      "loss": 0.5365,
+      "step": 8296
+    },
+    {
+      "epoch": 10.62016,
+      "grad_norm": 0.8694249391555786,
+      "learning_rate": 3.3421368547418966e-05,
+      "loss": 0.5453,
+      "step": 8297
+    },
+    {
+      "epoch": 10.62144,
+      "grad_norm": 0.8796177506446838,
+      "learning_rate": 3.341936774709884e-05,
+      "loss": 0.5339,
+      "step": 8298
+    },
+    {
+      "epoch": 10.62272,
+      "grad_norm": 0.8828015923500061,
+      "learning_rate": 3.341736694677871e-05,
+      "loss": 0.5838,
+      "step": 8299
+    },
+    {
+      "epoch": 10.624,
+      "grad_norm": 0.9480679035186768,
+      "learning_rate": 3.341536614645859e-05,
+      "loss": 0.5491,
+      "step": 8300
+    },
+    {
+      "epoch": 10.62528,
+      "grad_norm": 0.8716679811477661,
+      "learning_rate": 3.341336534613846e-05,
+      "loss": 0.5377,
+      "step": 8301
+    },
+    {
+      "epoch": 10.62656,
+      "grad_norm": 0.886888861656189,
+      "learning_rate": 3.341136454581833e-05,
+      "loss": 0.5493,
+      "step": 8302
+    },
+    {
+      "epoch": 10.627839999999999,
+      "grad_norm": 0.8986707925796509,
+      "learning_rate": 3.3409363745498204e-05,
+      "loss": 0.5805,
+      "step": 8303
+    },
+    {
+      "epoch": 10.62912,
+      "grad_norm": 0.8717333078384399,
+      "learning_rate": 3.340736294517807e-05,
+      "loss": 0.5042,
+      "step": 8304
+    },
+    {
+      "epoch": 10.6304,
+      "grad_norm": 0.8657734394073486,
+      "learning_rate": 3.340536214485794e-05,
+      "loss": 0.5976,
+      "step": 8305
+    },
+    {
+      "epoch": 10.63168,
+      "grad_norm": 0.8697702288627625,
+      "learning_rate": 3.340336134453781e-05,
+      "loss": 0.5653,
+      "step": 8306
+    },
+    {
+      "epoch": 10.63296,
+      "grad_norm": 0.9345014691352844,
+      "learning_rate": 3.340136054421769e-05,
+      "loss": 0.6546,
+      "step": 8307
+    },
+    {
+      "epoch": 10.63424,
+      "grad_norm": 0.8431648015975952,
+      "learning_rate": 3.3399359743897564e-05,
+      "loss": 0.5652,
+      "step": 8308
+    },
+    {
+      "epoch": 10.63552,
+      "grad_norm": 0.8228232264518738,
+      "learning_rate": 3.3397358943577436e-05,
+      "loss": 0.5742,
+      "step": 8309
+    },
+    {
+      "epoch": 10.636800000000001,
+      "grad_norm": 0.8282391428947449,
+      "learning_rate": 3.339535814325731e-05,
+      "loss": 0.5692,
+      "step": 8310
+    },
+    {
+      "epoch": 10.63808,
+      "grad_norm": 0.8612817525863647,
+      "learning_rate": 3.339335734293718e-05,
+      "loss": 0.5897,
+      "step": 8311
+    },
+    {
+      "epoch": 10.63936,
+      "grad_norm": 0.8568882942199707,
+      "learning_rate": 3.3391356542617044e-05,
+      "loss": 0.5675,
+      "step": 8312
+    },
+    {
+      "epoch": 10.64064,
+      "grad_norm": 0.8818540573120117,
+      "learning_rate": 3.3389355742296916e-05,
+      "loss": 0.5862,
+      "step": 8313
+    },
+    {
+      "epoch": 10.64192,
+      "grad_norm": 0.859569787979126,
+      "learning_rate": 3.3387354941976795e-05,
+      "loss": 0.5456,
+      "step": 8314
+    },
+    {
+      "epoch": 10.6432,
+      "grad_norm": 0.9021491408348083,
+      "learning_rate": 3.338535414165667e-05,
+      "loss": 0.6043,
+      "step": 8315
+    },
+    {
+      "epoch": 10.64448,
+      "grad_norm": 0.8722798228263855,
+      "learning_rate": 3.338335334133654e-05,
+      "loss": 0.5441,
+      "step": 8316
+    },
+    {
+      "epoch": 10.64576,
+      "grad_norm": 0.8792587518692017,
+      "learning_rate": 3.338135254101641e-05,
+      "loss": 0.5534,
+      "step": 8317
+    },
+    {
+      "epoch": 10.64704,
+      "grad_norm": 0.8752608895301819,
+      "learning_rate": 3.337935174069628e-05,
+      "loss": 0.537,
+      "step": 8318
+    },
+    {
+      "epoch": 10.64832,
+      "grad_norm": 0.8951905369758606,
+      "learning_rate": 3.3377350940376154e-05,
+      "loss": 0.6054,
+      "step": 8319
+    },
+    {
+      "epoch": 10.6496,
+      "grad_norm": 0.8960387110710144,
+      "learning_rate": 3.337535014005602e-05,
+      "loss": 0.6164,
+      "step": 8320
+    },
+    {
+      "epoch": 10.65088,
+      "grad_norm": 0.9029424786567688,
+      "learning_rate": 3.33733493397359e-05,
+      "loss": 0.5642,
+      "step": 8321
+    },
+    {
+      "epoch": 10.65216,
+      "grad_norm": 0.8857460618019104,
+      "learning_rate": 3.337134853941577e-05,
+      "loss": 0.5833,
+      "step": 8322
+    },
+    {
+      "epoch": 10.65344,
+      "grad_norm": 0.8373080492019653,
+      "learning_rate": 3.336934773909564e-05,
+      "loss": 0.5349,
+      "step": 8323
+    },
+    {
+      "epoch": 10.65472,
+      "grad_norm": 0.8566232919692993,
+      "learning_rate": 3.336734693877551e-05,
+      "loss": 0.5648,
+      "step": 8324
+    },
+    {
+      "epoch": 10.656,
+      "grad_norm": 0.894591212272644,
+      "learning_rate": 3.3365346138455385e-05,
+      "loss": 0.5989,
+      "step": 8325
+    },
+    {
+      "epoch": 10.65728,
+      "grad_norm": 0.8619808554649353,
+      "learning_rate": 3.336334533813526e-05,
+      "loss": 0.535,
+      "step": 8326
+    },
+    {
+      "epoch": 10.65856,
+      "grad_norm": 0.8474307060241699,
+      "learning_rate": 3.336134453781513e-05,
+      "loss": 0.5814,
+      "step": 8327
+    },
+    {
+      "epoch": 10.659839999999999,
+      "grad_norm": 0.8520254492759705,
+      "learning_rate": 3.3359343737494994e-05,
+      "loss": 0.522,
+      "step": 8328
+    },
+    {
+      "epoch": 10.66112,
+      "grad_norm": 0.8767806887626648,
+      "learning_rate": 3.335734293717487e-05,
+      "loss": 0.5766,
+      "step": 8329
+    },
+    {
+      "epoch": 10.6624,
+      "grad_norm": 0.8880060911178589,
+      "learning_rate": 3.3355342136854745e-05,
+      "loss": 0.6238,
+      "step": 8330
+    },
+    {
+      "epoch": 10.66368,
+      "grad_norm": 0.8470461964607239,
+      "learning_rate": 3.3353341336534616e-05,
+      "loss": 0.5878,
+      "step": 8331
+    },
+    {
+      "epoch": 10.66496,
+      "grad_norm": 0.8690903782844543,
+      "learning_rate": 3.335134053621449e-05,
+      "loss": 0.6082,
+      "step": 8332
+    },
+    {
+      "epoch": 10.66624,
+      "grad_norm": 0.8474588990211487,
+      "learning_rate": 3.334933973589436e-05,
+      "loss": 0.5873,
+      "step": 8333
+    },
+    {
+      "epoch": 10.66752,
+      "grad_norm": 0.9330595135688782,
+      "learning_rate": 3.334733893557423e-05,
+      "loss": 0.6094,
+      "step": 8334
+    },
+    {
+      "epoch": 10.6688,
+      "grad_norm": 0.8823012709617615,
+      "learning_rate": 3.3345338135254104e-05,
+      "loss": 0.5635,
+      "step": 8335
+    },
+    {
+      "epoch": 10.67008,
+      "grad_norm": 0.8572724461555481,
+      "learning_rate": 3.3343337334933976e-05,
+      "loss": 0.5932,
+      "step": 8336
+    },
+    {
+      "epoch": 10.67136,
+      "grad_norm": 0.8401781916618347,
+      "learning_rate": 3.334133653461385e-05,
+      "loss": 0.595,
+      "step": 8337
+    },
+    {
+      "epoch": 10.67264,
+      "grad_norm": 0.841372549533844,
+      "learning_rate": 3.333933573429372e-05,
+      "loss": 0.5606,
+      "step": 8338
+    },
+    {
+      "epoch": 10.67392,
+      "grad_norm": 0.8536092042922974,
+      "learning_rate": 3.333733493397359e-05,
+      "loss": 0.5635,
+      "step": 8339
+    },
+    {
+      "epoch": 10.6752,
+      "grad_norm": 0.8724260330200195,
+      "learning_rate": 3.333533413365346e-05,
+      "loss": 0.5582,
+      "step": 8340
+    },
+    {
+      "epoch": 10.67648,
+      "grad_norm": 0.8848631978034973,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5671,
+      "step": 8341
+    },
+    {
+      "epoch": 10.67776,
+      "grad_norm": 0.8863515257835388,
+      "learning_rate": 3.333133253301321e-05,
+      "loss": 0.6179,
+      "step": 8342
+    },
+    {
+      "epoch": 10.67904,
+      "grad_norm": 0.8705168962478638,
+      "learning_rate": 3.332933173269308e-05,
+      "loss": 0.5777,
+      "step": 8343
+    },
+    {
+      "epoch": 10.68032,
+      "grad_norm": 0.845525324344635,
+      "learning_rate": 3.332733093237295e-05,
+      "loss": 0.5936,
+      "step": 8344
+    },
+    {
+      "epoch": 10.6816,
+      "grad_norm": 0.9055159687995911,
+      "learning_rate": 3.332533013205282e-05,
+      "loss": 0.5915,
+      "step": 8345
+    },
+    {
+      "epoch": 10.68288,
+      "grad_norm": 0.8923260569572449,
+      "learning_rate": 3.3323329331732694e-05,
+      "loss": 0.5929,
+      "step": 8346
+    },
+    {
+      "epoch": 10.68416,
+      "grad_norm": 0.8504287600517273,
+      "learning_rate": 3.3321328531412566e-05,
+      "loss": 0.5632,
+      "step": 8347
+    },
+    {
+      "epoch": 10.68544,
+      "grad_norm": 0.8821974992752075,
+      "learning_rate": 3.331932773109244e-05,
+      "loss": 0.5745,
+      "step": 8348
+    },
+    {
+      "epoch": 10.68672,
+      "grad_norm": 0.8966748118400574,
+      "learning_rate": 3.331732693077231e-05,
+      "loss": 0.5867,
+      "step": 8349
+    },
+    {
+      "epoch": 10.688,
+      "grad_norm": 0.8681886196136475,
+      "learning_rate": 3.331532613045218e-05,
+      "loss": 0.5741,
+      "step": 8350
+    },
+    {
+      "epoch": 10.68928,
+      "grad_norm": 0.8703950643539429,
+      "learning_rate": 3.3313325330132053e-05,
+      "loss": 0.5322,
+      "step": 8351
+    },
+    {
+      "epoch": 10.69056,
+      "grad_norm": 0.8685113787651062,
+      "learning_rate": 3.3311324529811925e-05,
+      "loss": 0.5438,
+      "step": 8352
+    },
+    {
+      "epoch": 10.69184,
+      "grad_norm": 0.8648349642753601,
+      "learning_rate": 3.33093237294918e-05,
+      "loss": 0.5632,
+      "step": 8353
+    },
+    {
+      "epoch": 10.69312,
+      "grad_norm": 0.906882107257843,
+      "learning_rate": 3.330732292917167e-05,
+      "loss": 0.5535,
+      "step": 8354
+    },
+    {
+      "epoch": 10.6944,
+      "grad_norm": 0.8382284045219421,
+      "learning_rate": 3.330532212885154e-05,
+      "loss": 0.542,
+      "step": 8355
+    },
+    {
+      "epoch": 10.69568,
+      "grad_norm": 0.90635746717453,
+      "learning_rate": 3.330332132853141e-05,
+      "loss": 0.588,
+      "step": 8356
+    },
+    {
+      "epoch": 10.69696,
+      "grad_norm": 0.8943015336990356,
+      "learning_rate": 3.330132052821129e-05,
+      "loss": 0.5763,
+      "step": 8357
+    },
+    {
+      "epoch": 10.69824,
+      "grad_norm": 0.8443808555603027,
+      "learning_rate": 3.3299319727891156e-05,
+      "loss": 0.4964,
+      "step": 8358
+    },
+    {
+      "epoch": 10.69952,
+      "grad_norm": 0.8682435750961304,
+      "learning_rate": 3.329731892757103e-05,
+      "loss": 0.5687,
+      "step": 8359
+    },
+    {
+      "epoch": 10.7008,
+      "grad_norm": 0.856095552444458,
+      "learning_rate": 3.32953181272509e-05,
+      "loss": 0.5713,
+      "step": 8360
+    },
+    {
+      "epoch": 10.70208,
+      "grad_norm": 0.8573371171951294,
+      "learning_rate": 3.329331732693077e-05,
+      "loss": 0.5471,
+      "step": 8361
+    },
+    {
+      "epoch": 10.70336,
+      "grad_norm": 0.8584128618240356,
+      "learning_rate": 3.3291316526610644e-05,
+      "loss": 0.545,
+      "step": 8362
+    },
+    {
+      "epoch": 10.70464,
+      "grad_norm": 0.827085554599762,
+      "learning_rate": 3.3289315726290516e-05,
+      "loss": 0.5662,
+      "step": 8363
+    },
+    {
+      "epoch": 10.70592,
+      "grad_norm": 0.852120578289032,
+      "learning_rate": 3.3287314925970394e-05,
+      "loss": 0.5583,
+      "step": 8364
+    },
+    {
+      "epoch": 10.7072,
+      "grad_norm": 0.8772450685501099,
+      "learning_rate": 3.3285314125650266e-05,
+      "loss": 0.5958,
+      "step": 8365
+    },
+    {
+      "epoch": 10.70848,
+      "grad_norm": 0.7987073063850403,
+      "learning_rate": 3.328331332533013e-05,
+      "loss": 0.5179,
+      "step": 8366
+    },
+    {
+      "epoch": 10.70976,
+      "grad_norm": 0.8500874638557434,
+      "learning_rate": 3.328131252501e-05,
+      "loss": 0.5829,
+      "step": 8367
+    },
+    {
+      "epoch": 10.71104,
+      "grad_norm": 0.8709688782691956,
+      "learning_rate": 3.3279311724689875e-05,
+      "loss": 0.5892,
+      "step": 8368
+    },
+    {
+      "epoch": 10.71232,
+      "grad_norm": 0.8327182531356812,
+      "learning_rate": 3.327731092436975e-05,
+      "loss": 0.5589,
+      "step": 8369
+    },
+    {
+      "epoch": 10.7136,
+      "grad_norm": 0.8227571845054626,
+      "learning_rate": 3.327531012404962e-05,
+      "loss": 0.5639,
+      "step": 8370
+    },
+    {
+      "epoch": 10.71488,
+      "grad_norm": 0.8822687268257141,
+      "learning_rate": 3.32733093237295e-05,
+      "loss": 0.5801,
+      "step": 8371
+    },
+    {
+      "epoch": 10.71616,
+      "grad_norm": 0.8199798464775085,
+      "learning_rate": 3.327130852340937e-05,
+      "loss": 0.5418,
+      "step": 8372
+    },
+    {
+      "epoch": 10.71744,
+      "grad_norm": 0.877302885055542,
+      "learning_rate": 3.326930772308924e-05,
+      "loss": 0.5717,
+      "step": 8373
+    },
+    {
+      "epoch": 10.71872,
+      "grad_norm": 0.874665379524231,
+      "learning_rate": 3.3267306922769106e-05,
+      "loss": 0.558,
+      "step": 8374
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 0.8584824800491333,
+      "learning_rate": 3.326530612244898e-05,
+      "loss": 0.5602,
+      "step": 8375
+    },
+    {
+      "epoch": 10.72128,
+      "grad_norm": 0.8538774251937866,
+      "learning_rate": 3.326330532212885e-05,
+      "loss": 0.5614,
+      "step": 8376
+    },
+    {
+      "epoch": 10.72256,
+      "grad_norm": 0.8546411991119385,
+      "learning_rate": 3.326130452180872e-05,
+      "loss": 0.6141,
+      "step": 8377
+    },
+    {
+      "epoch": 10.72384,
+      "grad_norm": 0.8563506007194519,
+      "learning_rate": 3.32593037214886e-05,
+      "loss": 0.5317,
+      "step": 8378
+    },
+    {
+      "epoch": 10.72512,
+      "grad_norm": 0.88761305809021,
+      "learning_rate": 3.325730292116847e-05,
+      "loss": 0.5783,
+      "step": 8379
+    },
+    {
+      "epoch": 10.7264,
+      "grad_norm": 0.8833502531051636,
+      "learning_rate": 3.3255302120848344e-05,
+      "loss": 0.5822,
+      "step": 8380
+    },
+    {
+      "epoch": 10.72768,
+      "grad_norm": 0.8859793543815613,
+      "learning_rate": 3.3253301320528216e-05,
+      "loss": 0.5652,
+      "step": 8381
+    },
+    {
+      "epoch": 10.72896,
+      "grad_norm": 0.855240523815155,
+      "learning_rate": 3.325130052020808e-05,
+      "loss": 0.5449,
+      "step": 8382
+    },
+    {
+      "epoch": 10.73024,
+      "grad_norm": 0.8389564156532288,
+      "learning_rate": 3.324929971988795e-05,
+      "loss": 0.5591,
+      "step": 8383
+    },
+    {
+      "epoch": 10.73152,
+      "grad_norm": 0.8849216103553772,
+      "learning_rate": 3.3247298919567825e-05,
+      "loss": 0.5663,
+      "step": 8384
+    },
+    {
+      "epoch": 10.7328,
+      "grad_norm": 0.8661803007125854,
+      "learning_rate": 3.3245298119247703e-05,
+      "loss": 0.59,
+      "step": 8385
+    },
+    {
+      "epoch": 10.73408,
+      "grad_norm": 0.8539361953735352,
+      "learning_rate": 3.3243297318927575e-05,
+      "loss": 0.5407,
+      "step": 8386
+    },
+    {
+      "epoch": 10.73536,
+      "grad_norm": 0.9314575791358948,
+      "learning_rate": 3.324129651860745e-05,
+      "loss": 0.6113,
+      "step": 8387
+    },
+    {
+      "epoch": 10.73664,
+      "grad_norm": 0.8329246044158936,
+      "learning_rate": 3.323929571828732e-05,
+      "loss": 0.5317,
+      "step": 8388
+    },
+    {
+      "epoch": 10.73792,
+      "grad_norm": 0.8746315240859985,
+      "learning_rate": 3.323729491796719e-05,
+      "loss": 0.6104,
+      "step": 8389
+    },
+    {
+      "epoch": 10.7392,
+      "grad_norm": 0.8568903803825378,
+      "learning_rate": 3.3235294117647056e-05,
+      "loss": 0.6033,
+      "step": 8390
+    },
+    {
+      "epoch": 10.74048,
+      "grad_norm": 0.8712742924690247,
+      "learning_rate": 3.323329331732693e-05,
+      "loss": 0.5644,
+      "step": 8391
+    },
+    {
+      "epoch": 10.74176,
+      "grad_norm": 0.8397703170776367,
+      "learning_rate": 3.3231292517006806e-05,
+      "loss": 0.5512,
+      "step": 8392
+    },
+    {
+      "epoch": 10.74304,
+      "grad_norm": 0.9338214993476868,
+      "learning_rate": 3.322929171668668e-05,
+      "loss": 0.6489,
+      "step": 8393
+    },
+    {
+      "epoch": 10.74432,
+      "grad_norm": 0.8752598762512207,
+      "learning_rate": 3.322729091636655e-05,
+      "loss": 0.5417,
+      "step": 8394
+    },
+    {
+      "epoch": 10.7456,
+      "grad_norm": 0.8597055077552795,
+      "learning_rate": 3.322529011604642e-05,
+      "loss": 0.5591,
+      "step": 8395
+    },
+    {
+      "epoch": 10.74688,
+      "grad_norm": 0.9333181381225586,
+      "learning_rate": 3.3223289315726294e-05,
+      "loss": 0.5923,
+      "step": 8396
+    },
+    {
+      "epoch": 10.74816,
+      "grad_norm": 0.8479388356208801,
+      "learning_rate": 3.3221288515406166e-05,
+      "loss": 0.5427,
+      "step": 8397
+    },
+    {
+      "epoch": 10.74944,
+      "grad_norm": 0.8617178201675415,
+      "learning_rate": 3.321928771508603e-05,
+      "loss": 0.5473,
+      "step": 8398
+    },
+    {
+      "epoch": 10.75072,
+      "grad_norm": 0.9343951344490051,
+      "learning_rate": 3.321728691476591e-05,
+      "loss": 0.6115,
+      "step": 8399
+    },
+    {
+      "epoch": 10.752,
+      "grad_norm": 0.9055424928665161,
+      "learning_rate": 3.321528611444578e-05,
+      "loss": 0.5647,
+      "step": 8400
+    },
+    {
+      "epoch": 10.75328,
+      "grad_norm": 0.9039073586463928,
+      "learning_rate": 3.321328531412565e-05,
+      "loss": 0.5382,
+      "step": 8401
+    },
+    {
+      "epoch": 10.75456,
+      "grad_norm": 0.8732048869132996,
+      "learning_rate": 3.3211284513805525e-05,
+      "loss": 0.5488,
+      "step": 8402
+    },
+    {
+      "epoch": 10.75584,
+      "grad_norm": 0.859381914138794,
+      "learning_rate": 3.32092837134854e-05,
+      "loss": 0.6105,
+      "step": 8403
+    },
+    {
+      "epoch": 10.75712,
+      "grad_norm": 0.8635976910591125,
+      "learning_rate": 3.320728291316527e-05,
+      "loss": 0.5932,
+      "step": 8404
+    },
+    {
+      "epoch": 10.7584,
+      "grad_norm": 0.8763898611068726,
+      "learning_rate": 3.320528211284514e-05,
+      "loss": 0.5703,
+      "step": 8405
+    },
+    {
+      "epoch": 10.75968,
+      "grad_norm": 0.828618049621582,
+      "learning_rate": 3.320328131252501e-05,
+      "loss": 0.5675,
+      "step": 8406
+    },
+    {
+      "epoch": 10.76096,
+      "grad_norm": 0.894950807094574,
+      "learning_rate": 3.3201280512204884e-05,
+      "loss": 0.5832,
+      "step": 8407
+    },
+    {
+      "epoch": 10.76224,
+      "grad_norm": 0.8377487659454346,
+      "learning_rate": 3.3199279711884756e-05,
+      "loss": 0.5568,
+      "step": 8408
+    },
+    {
+      "epoch": 10.76352,
+      "grad_norm": 0.8179535269737244,
+      "learning_rate": 3.319727891156463e-05,
+      "loss": 0.4754,
+      "step": 8409
+    },
+    {
+      "epoch": 10.7648,
+      "grad_norm": 0.9088795185089111,
+      "learning_rate": 3.31952781112445e-05,
+      "loss": 0.6044,
+      "step": 8410
+    },
+    {
+      "epoch": 10.76608,
+      "grad_norm": 0.9149488806724548,
+      "learning_rate": 3.319327731092437e-05,
+      "loss": 0.577,
+      "step": 8411
+    },
+    {
+      "epoch": 10.76736,
+      "grad_norm": 0.9027832746505737,
+      "learning_rate": 3.3191276510604244e-05,
+      "loss": 0.5866,
+      "step": 8412
+    },
+    {
+      "epoch": 10.76864,
+      "grad_norm": 0.86416095495224,
+      "learning_rate": 3.3189275710284115e-05,
+      "loss": 0.6003,
+      "step": 8413
+    },
+    {
+      "epoch": 10.76992,
+      "grad_norm": 0.8570460677146912,
+      "learning_rate": 3.318727490996399e-05,
+      "loss": 0.562,
+      "step": 8414
+    },
+    {
+      "epoch": 10.7712,
+      "grad_norm": 0.9128470420837402,
+      "learning_rate": 3.318527410964386e-05,
+      "loss": 0.5915,
+      "step": 8415
+    },
+    {
+      "epoch": 10.77248,
+      "grad_norm": 0.8856536746025085,
+      "learning_rate": 3.318327330932373e-05,
+      "loss": 0.5406,
+      "step": 8416
+    },
+    {
+      "epoch": 10.77376,
+      "grad_norm": 0.8812784552574158,
+      "learning_rate": 3.31812725090036e-05,
+      "loss": 0.5233,
+      "step": 8417
+    },
+    {
+      "epoch": 10.77504,
+      "grad_norm": 0.8757975697517395,
+      "learning_rate": 3.3179271708683475e-05,
+      "loss": 0.5947,
+      "step": 8418
+    },
+    {
+      "epoch": 10.77632,
+      "grad_norm": 0.8653748035430908,
+      "learning_rate": 3.3177270908363347e-05,
+      "loss": 0.5823,
+      "step": 8419
+    },
+    {
+      "epoch": 10.7776,
+      "grad_norm": 0.9042747020721436,
+      "learning_rate": 3.317527010804322e-05,
+      "loss": 0.5583,
+      "step": 8420
+    },
+    {
+      "epoch": 10.778880000000001,
+      "grad_norm": 0.8836710453033447,
+      "learning_rate": 3.317326930772309e-05,
+      "loss": 0.563,
+      "step": 8421
+    },
+    {
+      "epoch": 10.78016,
+      "grad_norm": 0.8876301050186157,
+      "learning_rate": 3.317126850740296e-05,
+      "loss": 0.5814,
+      "step": 8422
+    },
+    {
+      "epoch": 10.78144,
+      "grad_norm": 0.8692103028297424,
+      "learning_rate": 3.3169267707082834e-05,
+      "loss": 0.6207,
+      "step": 8423
+    },
+    {
+      "epoch": 10.78272,
+      "grad_norm": 0.8895595669746399,
+      "learning_rate": 3.3167266906762706e-05,
+      "loss": 0.566,
+      "step": 8424
+    },
+    {
+      "epoch": 10.784,
+      "grad_norm": 0.8768266439437866,
+      "learning_rate": 3.316526610644258e-05,
+      "loss": 0.6218,
+      "step": 8425
+    },
+    {
+      "epoch": 10.78528,
+      "grad_norm": 0.8848718404769897,
+      "learning_rate": 3.316326530612245e-05,
+      "loss": 0.5943,
+      "step": 8426
+    },
+    {
+      "epoch": 10.78656,
+      "grad_norm": 0.9084532260894775,
+      "learning_rate": 3.316126450580233e-05,
+      "loss": 0.6096,
+      "step": 8427
+    },
+    {
+      "epoch": 10.78784,
+      "grad_norm": 0.9083433747291565,
+      "learning_rate": 3.315926370548219e-05,
+      "loss": 0.5836,
+      "step": 8428
+    },
+    {
+      "epoch": 10.78912,
+      "grad_norm": 0.8622574806213379,
+      "learning_rate": 3.3157262905162065e-05,
+      "loss": 0.5541,
+      "step": 8429
+    },
+    {
+      "epoch": 10.7904,
+      "grad_norm": 0.8447574973106384,
+      "learning_rate": 3.315526210484194e-05,
+      "loss": 0.5544,
+      "step": 8430
+    },
+    {
+      "epoch": 10.79168,
+      "grad_norm": 0.8831068873405457,
+      "learning_rate": 3.315326130452181e-05,
+      "loss": 0.6255,
+      "step": 8431
+    },
+    {
+      "epoch": 10.79296,
+      "grad_norm": 0.8802228569984436,
+      "learning_rate": 3.315126050420168e-05,
+      "loss": 0.6056,
+      "step": 8432
+    },
+    {
+      "epoch": 10.79424,
+      "grad_norm": 0.8990511298179626,
+      "learning_rate": 3.314925970388155e-05,
+      "loss": 0.6257,
+      "step": 8433
+    },
+    {
+      "epoch": 10.79552,
+      "grad_norm": 0.8636667728424072,
+      "learning_rate": 3.3147258903561424e-05,
+      "loss": 0.5596,
+      "step": 8434
+    },
+    {
+      "epoch": 10.7968,
+      "grad_norm": 0.8775346279144287,
+      "learning_rate": 3.31452581032413e-05,
+      "loss": 0.5596,
+      "step": 8435
+    },
+    {
+      "epoch": 10.79808,
+      "grad_norm": 0.8895041942596436,
+      "learning_rate": 3.314325730292117e-05,
+      "loss": 0.5812,
+      "step": 8436
+    },
+    {
+      "epoch": 10.79936,
+      "grad_norm": 0.8645225167274475,
+      "learning_rate": 3.314125650260104e-05,
+      "loss": 0.566,
+      "step": 8437
+    },
+    {
+      "epoch": 10.80064,
+      "grad_norm": 0.9078739285469055,
+      "learning_rate": 3.313925570228091e-05,
+      "loss": 0.5976,
+      "step": 8438
+    },
+    {
+      "epoch": 10.801919999999999,
+      "grad_norm": 0.8740204572677612,
+      "learning_rate": 3.3137254901960784e-05,
+      "loss": 0.5386,
+      "step": 8439
+    },
+    {
+      "epoch": 10.8032,
+      "grad_norm": 0.8575959205627441,
+      "learning_rate": 3.3135254101640656e-05,
+      "loss": 0.5371,
+      "step": 8440
+    },
+    {
+      "epoch": 10.80448,
+      "grad_norm": 0.8617302179336548,
+      "learning_rate": 3.313325330132053e-05,
+      "loss": 0.5515,
+      "step": 8441
+    },
+    {
+      "epoch": 10.80576,
+      "grad_norm": 0.8753884434700012,
+      "learning_rate": 3.3131252501000406e-05,
+      "loss": 0.5821,
+      "step": 8442
+    },
+    {
+      "epoch": 10.80704,
+      "grad_norm": 0.8543848991394043,
+      "learning_rate": 3.312925170068028e-05,
+      "loss": 0.5948,
+      "step": 8443
+    },
+    {
+      "epoch": 10.80832,
+      "grad_norm": 0.8967148661613464,
+      "learning_rate": 3.312725090036014e-05,
+      "loss": 0.6083,
+      "step": 8444
+    },
+    {
+      "epoch": 10.8096,
+      "grad_norm": 0.8336367011070251,
+      "learning_rate": 3.3125250100040015e-05,
+      "loss": 0.5127,
+      "step": 8445
+    },
+    {
+      "epoch": 10.810880000000001,
+      "grad_norm": 0.8468740582466125,
+      "learning_rate": 3.312324929971989e-05,
+      "loss": 0.5522,
+      "step": 8446
+    },
+    {
+      "epoch": 10.81216,
+      "grad_norm": 0.8677701354026794,
+      "learning_rate": 3.312124849939976e-05,
+      "loss": 0.5554,
+      "step": 8447
+    },
+    {
+      "epoch": 10.81344,
+      "grad_norm": 0.8767905235290527,
+      "learning_rate": 3.311924769907963e-05,
+      "loss": 0.5572,
+      "step": 8448
+    },
+    {
+      "epoch": 10.81472,
+      "grad_norm": 0.8894631266593933,
+      "learning_rate": 3.311724689875951e-05,
+      "loss": 0.6002,
+      "step": 8449
+    },
+    {
+      "epoch": 10.816,
+      "grad_norm": 0.8528721332550049,
+      "learning_rate": 3.311524609843938e-05,
+      "loss": 0.5644,
+      "step": 8450
+    },
+    {
+      "epoch": 10.81728,
+      "grad_norm": 0.8421396017074585,
+      "learning_rate": 3.311324529811925e-05,
+      "loss": 0.5392,
+      "step": 8451
+    },
+    {
+      "epoch": 10.81856,
+      "grad_norm": 0.8507756590843201,
+      "learning_rate": 3.311124449779912e-05,
+      "loss": 0.5513,
+      "step": 8452
+    },
+    {
+      "epoch": 10.81984,
+      "grad_norm": 0.8500514626502991,
+      "learning_rate": 3.310924369747899e-05,
+      "loss": 0.6198,
+      "step": 8453
+    },
+    {
+      "epoch": 10.82112,
+      "grad_norm": 0.8615267872810364,
+      "learning_rate": 3.310724289715886e-05,
+      "loss": 0.5162,
+      "step": 8454
+    },
+    {
+      "epoch": 10.8224,
+      "grad_norm": 0.8663539290428162,
+      "learning_rate": 3.310524209683873e-05,
+      "loss": 0.5811,
+      "step": 8455
+    },
+    {
+      "epoch": 10.82368,
+      "grad_norm": 0.8774354457855225,
+      "learning_rate": 3.310324129651861e-05,
+      "loss": 0.6085,
+      "step": 8456
+    },
+    {
+      "epoch": 10.82496,
+      "grad_norm": 0.9009751081466675,
+      "learning_rate": 3.3101240496198484e-05,
+      "loss": 0.5154,
+      "step": 8457
+    },
+    {
+      "epoch": 10.82624,
+      "grad_norm": 0.9249484539031982,
+      "learning_rate": 3.3099239695878356e-05,
+      "loss": 0.6029,
+      "step": 8458
+    },
+    {
+      "epoch": 10.82752,
+      "grad_norm": 0.8451379537582397,
+      "learning_rate": 3.309723889555823e-05,
+      "loss": 0.5601,
+      "step": 8459
+    },
+    {
+      "epoch": 10.8288,
+      "grad_norm": 0.905768871307373,
+      "learning_rate": 3.309523809523809e-05,
+      "loss": 0.5652,
+      "step": 8460
+    },
+    {
+      "epoch": 10.83008,
+      "grad_norm": 0.9199037551879883,
+      "learning_rate": 3.3093237294917965e-05,
+      "loss": 0.6089,
+      "step": 8461
+    },
+    {
+      "epoch": 10.83136,
+      "grad_norm": 0.8506740927696228,
+      "learning_rate": 3.3091236494597836e-05,
+      "loss": 0.5242,
+      "step": 8462
+    },
+    {
+      "epoch": 10.83264,
+      "grad_norm": 0.9114420413970947,
+      "learning_rate": 3.3089235694277715e-05,
+      "loss": 0.6026,
+      "step": 8463
+    },
+    {
+      "epoch": 10.833919999999999,
+      "grad_norm": 0.8378769755363464,
+      "learning_rate": 3.308723489395759e-05,
+      "loss": 0.5428,
+      "step": 8464
+    },
+    {
+      "epoch": 10.8352,
+      "grad_norm": 0.863136887550354,
+      "learning_rate": 3.308523409363746e-05,
+      "loss": 0.5569,
+      "step": 8465
+    },
+    {
+      "epoch": 10.83648,
+      "grad_norm": 0.9558273553848267,
+      "learning_rate": 3.308323329331733e-05,
+      "loss": 0.6102,
+      "step": 8466
+    },
+    {
+      "epoch": 10.83776,
+      "grad_norm": 0.860105037689209,
+      "learning_rate": 3.30812324929972e-05,
+      "loss": 0.5872,
+      "step": 8467
+    },
+    {
+      "epoch": 10.83904,
+      "grad_norm": 0.8709381222724915,
+      "learning_rate": 3.307923169267707e-05,
+      "loss": 0.5548,
+      "step": 8468
+    },
+    {
+      "epoch": 10.84032,
+      "grad_norm": 0.8705414533615112,
+      "learning_rate": 3.307723089235694e-05,
+      "loss": 0.5317,
+      "step": 8469
+    },
+    {
+      "epoch": 10.8416,
+      "grad_norm": 0.8116998672485352,
+      "learning_rate": 3.307523009203682e-05,
+      "loss": 0.5055,
+      "step": 8470
+    },
+    {
+      "epoch": 10.84288,
+      "grad_norm": 0.9043382406234741,
+      "learning_rate": 3.307322929171669e-05,
+      "loss": 0.5995,
+      "step": 8471
+    },
+    {
+      "epoch": 10.84416,
+      "grad_norm": 0.9392409920692444,
+      "learning_rate": 3.307122849139656e-05,
+      "loss": 0.6222,
+      "step": 8472
+    },
+    {
+      "epoch": 10.84544,
+      "grad_norm": 0.8821365833282471,
+      "learning_rate": 3.3069227691076434e-05,
+      "loss": 0.5192,
+      "step": 8473
+    },
+    {
+      "epoch": 10.84672,
+      "grad_norm": 0.885803759098053,
+      "learning_rate": 3.3067226890756305e-05,
+      "loss": 0.6162,
+      "step": 8474
+    },
+    {
+      "epoch": 10.848,
+      "grad_norm": 0.8577411770820618,
+      "learning_rate": 3.306522609043618e-05,
+      "loss": 0.5609,
+      "step": 8475
+    },
+    {
+      "epoch": 10.84928,
+      "grad_norm": 0.8510453104972839,
+      "learning_rate": 3.306322529011604e-05,
+      "loss": 0.5378,
+      "step": 8476
+    },
+    {
+      "epoch": 10.85056,
+      "grad_norm": 0.8527485728263855,
+      "learning_rate": 3.306122448979592e-05,
+      "loss": 0.558,
+      "step": 8477
+    },
+    {
+      "epoch": 10.85184,
+      "grad_norm": 0.8647722601890564,
+      "learning_rate": 3.305922368947579e-05,
+      "loss": 0.5802,
+      "step": 8478
+    },
+    {
+      "epoch": 10.85312,
+      "grad_norm": 0.8825690746307373,
+      "learning_rate": 3.3057222889155665e-05,
+      "loss": 0.5959,
+      "step": 8479
+    },
+    {
+      "epoch": 10.8544,
+      "grad_norm": 0.8674288392066956,
+      "learning_rate": 3.3055222088835537e-05,
+      "loss": 0.5621,
+      "step": 8480
+    },
+    {
+      "epoch": 10.85568,
+      "grad_norm": 0.8390064239501953,
+      "learning_rate": 3.305322128851541e-05,
+      "loss": 0.5452,
+      "step": 8481
+    },
+    {
+      "epoch": 10.85696,
+      "grad_norm": 0.8398364186286926,
+      "learning_rate": 3.305122048819528e-05,
+      "loss": 0.5649,
+      "step": 8482
+    },
+    {
+      "epoch": 10.85824,
+      "grad_norm": 0.7934215664863586,
+      "learning_rate": 3.304921968787515e-05,
+      "loss": 0.5222,
+      "step": 8483
+    },
+    {
+      "epoch": 10.85952,
+      "grad_norm": 0.8218225240707397,
+      "learning_rate": 3.3047218887555024e-05,
+      "loss": 0.5509,
+      "step": 8484
+    },
+    {
+      "epoch": 10.8608,
+      "grad_norm": 0.8352612257003784,
+      "learning_rate": 3.3045218087234896e-05,
+      "loss": 0.564,
+      "step": 8485
+    },
+    {
+      "epoch": 10.86208,
+      "grad_norm": 0.8846644759178162,
+      "learning_rate": 3.304321728691477e-05,
+      "loss": 0.5925,
+      "step": 8486
+    },
+    {
+      "epoch": 10.86336,
+      "grad_norm": 0.8664666414260864,
+      "learning_rate": 3.304121648659464e-05,
+      "loss": 0.573,
+      "step": 8487
+    },
+    {
+      "epoch": 10.86464,
+      "grad_norm": 0.8052683472633362,
+      "learning_rate": 3.303921568627451e-05,
+      "loss": 0.5172,
+      "step": 8488
+    },
+    {
+      "epoch": 10.86592,
+      "grad_norm": 0.8650311827659607,
+      "learning_rate": 3.303721488595438e-05,
+      "loss": 0.5449,
+      "step": 8489
+    },
+    {
+      "epoch": 10.8672,
+      "grad_norm": 0.8478010296821594,
+      "learning_rate": 3.3035214085634255e-05,
+      "loss": 0.5509,
+      "step": 8490
+    },
+    {
+      "epoch": 10.86848,
+      "grad_norm": 0.8610655665397644,
+      "learning_rate": 3.303321328531413e-05,
+      "loss": 0.5669,
+      "step": 8491
+    },
+    {
+      "epoch": 10.86976,
+      "grad_norm": 0.8204941153526306,
+      "learning_rate": 3.3031212484994e-05,
+      "loss": 0.5607,
+      "step": 8492
+    },
+    {
+      "epoch": 10.87104,
+      "grad_norm": 0.8423011302947998,
+      "learning_rate": 3.302921168467387e-05,
+      "loss": 0.5431,
+      "step": 8493
+    },
+    {
+      "epoch": 10.87232,
+      "grad_norm": 0.8661743998527527,
+      "learning_rate": 3.302721088435374e-05,
+      "loss": 0.5609,
+      "step": 8494
+    },
+    {
+      "epoch": 10.8736,
+      "grad_norm": 0.8518742918968201,
+      "learning_rate": 3.3025210084033614e-05,
+      "loss": 0.5223,
+      "step": 8495
+    },
+    {
+      "epoch": 10.87488,
+      "grad_norm": 0.9045655131340027,
+      "learning_rate": 3.3023209283713486e-05,
+      "loss": 0.6088,
+      "step": 8496
+    },
+    {
+      "epoch": 10.87616,
+      "grad_norm": 0.8613651990890503,
+      "learning_rate": 3.302120848339336e-05,
+      "loss": 0.5155,
+      "step": 8497
+    },
+    {
+      "epoch": 10.87744,
+      "grad_norm": 0.8599672317504883,
+      "learning_rate": 3.301920768307323e-05,
+      "loss": 0.5743,
+      "step": 8498
+    },
+    {
+      "epoch": 10.87872,
+      "grad_norm": 0.8626627922058105,
+      "learning_rate": 3.30172068827531e-05,
+      "loss": 0.5878,
+      "step": 8499
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 0.879513144493103,
+      "learning_rate": 3.3015206082432974e-05,
+      "loss": 0.5954,
+      "step": 8500
+    },
+    {
+      "epoch": 10.88128,
+      "grad_norm": 0.8912549614906311,
+      "learning_rate": 3.3013205282112846e-05,
+      "loss": 0.585,
+      "step": 8501
+    },
+    {
+      "epoch": 10.88256,
+      "grad_norm": 0.8111445903778076,
+      "learning_rate": 3.301120448179272e-05,
+      "loss": 0.5391,
+      "step": 8502
+    },
+    {
+      "epoch": 10.88384,
+      "grad_norm": 0.8916809558868408,
+      "learning_rate": 3.300920368147259e-05,
+      "loss": 0.6075,
+      "step": 8503
+    },
+    {
+      "epoch": 10.88512,
+      "grad_norm": 0.8808960318565369,
+      "learning_rate": 3.300720288115246e-05,
+      "loss": 0.5037,
+      "step": 8504
+    },
+    {
+      "epoch": 10.8864,
+      "grad_norm": 0.8821715116500854,
+      "learning_rate": 3.300520208083234e-05,
+      "loss": 0.566,
+      "step": 8505
+    },
+    {
+      "epoch": 10.88768,
+      "grad_norm": 0.8724703192710876,
+      "learning_rate": 3.3003201280512205e-05,
+      "loss": 0.5715,
+      "step": 8506
+    },
+    {
+      "epoch": 10.88896,
+      "grad_norm": 0.8815937638282776,
+      "learning_rate": 3.300120048019208e-05,
+      "loss": 0.5345,
+      "step": 8507
+    },
+    {
+      "epoch": 10.89024,
+      "grad_norm": 0.8287563920021057,
+      "learning_rate": 3.299919967987195e-05,
+      "loss": 0.551,
+      "step": 8508
+    },
+    {
+      "epoch": 10.89152,
+      "grad_norm": 0.9234015941619873,
+      "learning_rate": 3.299719887955182e-05,
+      "loss": 0.6083,
+      "step": 8509
+    },
+    {
+      "epoch": 10.8928,
+      "grad_norm": 0.8626712560653687,
+      "learning_rate": 3.299519807923169e-05,
+      "loss": 0.6007,
+      "step": 8510
+    },
+    {
+      "epoch": 10.89408,
+      "grad_norm": 0.8290971517562866,
+      "learning_rate": 3.2993197278911564e-05,
+      "loss": 0.5234,
+      "step": 8511
+    },
+    {
+      "epoch": 10.89536,
+      "grad_norm": 0.8788360357284546,
+      "learning_rate": 3.299119647859144e-05,
+      "loss": 0.6131,
+      "step": 8512
+    },
+    {
+      "epoch": 10.89664,
+      "grad_norm": 0.8960014581680298,
+      "learning_rate": 3.2989195678271315e-05,
+      "loss": 0.6024,
+      "step": 8513
+    },
+    {
+      "epoch": 10.89792,
+      "grad_norm": 0.8566673398017883,
+      "learning_rate": 3.298719487795118e-05,
+      "loss": 0.5515,
+      "step": 8514
+    },
+    {
+      "epoch": 10.8992,
+      "grad_norm": 0.8075211048126221,
+      "learning_rate": 3.298519407763105e-05,
+      "loss": 0.516,
+      "step": 8515
+    },
+    {
+      "epoch": 10.90048,
+      "grad_norm": 0.9277984499931335,
+      "learning_rate": 3.2983193277310923e-05,
+      "loss": 0.6206,
+      "step": 8516
+    },
+    {
+      "epoch": 10.90176,
+      "grad_norm": 0.8750211596488953,
+      "learning_rate": 3.2981192476990795e-05,
+      "loss": 0.5697,
+      "step": 8517
+    },
+    {
+      "epoch": 10.90304,
+      "grad_norm": 0.8238059282302856,
+      "learning_rate": 3.297919167667067e-05,
+      "loss": 0.541,
+      "step": 8518
+    },
+    {
+      "epoch": 10.90432,
+      "grad_norm": 0.8247233629226685,
+      "learning_rate": 3.2977190876350546e-05,
+      "loss": 0.5711,
+      "step": 8519
+    },
+    {
+      "epoch": 10.9056,
+      "grad_norm": 0.880142092704773,
+      "learning_rate": 3.297519007603042e-05,
+      "loss": 0.581,
+      "step": 8520
+    },
+    {
+      "epoch": 10.90688,
+      "grad_norm": 0.9421783089637756,
+      "learning_rate": 3.297318927571029e-05,
+      "loss": 0.6042,
+      "step": 8521
+    },
+    {
+      "epoch": 10.90816,
+      "grad_norm": 0.8348096609115601,
+      "learning_rate": 3.2971188475390155e-05,
+      "loss": 0.5075,
+      "step": 8522
+    },
+    {
+      "epoch": 10.90944,
+      "grad_norm": 0.8904414772987366,
+      "learning_rate": 3.2969187675070026e-05,
+      "loss": 0.6436,
+      "step": 8523
+    },
+    {
+      "epoch": 10.91072,
+      "grad_norm": 0.927683413028717,
+      "learning_rate": 3.29671868747499e-05,
+      "loss": 0.6266,
+      "step": 8524
+    },
+    {
+      "epoch": 10.912,
+      "grad_norm": 0.8911395072937012,
+      "learning_rate": 3.296518607442977e-05,
+      "loss": 0.6085,
+      "step": 8525
+    },
+    {
+      "epoch": 10.91328,
+      "grad_norm": 0.8726085424423218,
+      "learning_rate": 3.296318527410965e-05,
+      "loss": 0.5798,
+      "step": 8526
+    },
+    {
+      "epoch": 10.91456,
+      "grad_norm": 0.8867003321647644,
+      "learning_rate": 3.296118447378952e-05,
+      "loss": 0.5685,
+      "step": 8527
+    },
+    {
+      "epoch": 10.91584,
+      "grad_norm": 0.877424418926239,
+      "learning_rate": 3.295918367346939e-05,
+      "loss": 0.5986,
+      "step": 8528
+    },
+    {
+      "epoch": 10.91712,
+      "grad_norm": 0.8679222464561462,
+      "learning_rate": 3.2957182873149264e-05,
+      "loss": 0.5735,
+      "step": 8529
+    },
+    {
+      "epoch": 10.9184,
+      "grad_norm": 0.8904789090156555,
+      "learning_rate": 3.295518207282913e-05,
+      "loss": 0.567,
+      "step": 8530
+    },
+    {
+      "epoch": 10.91968,
+      "grad_norm": 0.8913958668708801,
+      "learning_rate": 3.2953181272509e-05,
+      "loss": 0.57,
+      "step": 8531
+    },
+    {
+      "epoch": 10.920960000000001,
+      "grad_norm": 0.8957118391990662,
+      "learning_rate": 3.295118047218887e-05,
+      "loss": 0.5692,
+      "step": 8532
+    },
+    {
+      "epoch": 10.92224,
+      "grad_norm": 0.8906075358390808,
+      "learning_rate": 3.294917967186875e-05,
+      "loss": 0.5645,
+      "step": 8533
+    },
+    {
+      "epoch": 10.92352,
+      "grad_norm": 0.8821467757225037,
+      "learning_rate": 3.2947178871548624e-05,
+      "loss": 0.5799,
+      "step": 8534
+    },
+    {
+      "epoch": 10.9248,
+      "grad_norm": 0.8894703984260559,
+      "learning_rate": 3.2945178071228495e-05,
+      "loss": 0.6076,
+      "step": 8535
+    },
+    {
+      "epoch": 10.92608,
+      "grad_norm": 0.8252606987953186,
+      "learning_rate": 3.294317727090837e-05,
+      "loss": 0.5762,
+      "step": 8536
+    },
+    {
+      "epoch": 10.92736,
+      "grad_norm": 0.8460403680801392,
+      "learning_rate": 3.294117647058824e-05,
+      "loss": 0.5853,
+      "step": 8537
+    },
+    {
+      "epoch": 10.92864,
+      "grad_norm": 0.8825834393501282,
+      "learning_rate": 3.2939175670268104e-05,
+      "loss": 0.5628,
+      "step": 8538
+    },
+    {
+      "epoch": 10.92992,
+      "grad_norm": 0.8889744877815247,
+      "learning_rate": 3.2937174869947976e-05,
+      "loss": 0.549,
+      "step": 8539
+    },
+    {
+      "epoch": 10.9312,
+      "grad_norm": 0.9132182002067566,
+      "learning_rate": 3.2935174069627855e-05,
+      "loss": 0.5821,
+      "step": 8540
+    },
+    {
+      "epoch": 10.93248,
+      "grad_norm": 0.8521642684936523,
+      "learning_rate": 3.2933173269307727e-05,
+      "loss": 0.5378,
+      "step": 8541
+    },
+    {
+      "epoch": 10.93376,
+      "grad_norm": 0.8864872455596924,
+      "learning_rate": 3.29311724689876e-05,
+      "loss": 0.5589,
+      "step": 8542
+    },
+    {
+      "epoch": 10.93504,
+      "grad_norm": 0.8883858919143677,
+      "learning_rate": 3.292917166866747e-05,
+      "loss": 0.5716,
+      "step": 8543
+    },
+    {
+      "epoch": 10.93632,
+      "grad_norm": 0.862106442451477,
+      "learning_rate": 3.292717086834734e-05,
+      "loss": 0.5487,
+      "step": 8544
+    },
+    {
+      "epoch": 10.9376,
+      "grad_norm": 0.9219768643379211,
+      "learning_rate": 3.2925170068027214e-05,
+      "loss": 0.5911,
+      "step": 8545
+    },
+    {
+      "epoch": 10.93888,
+      "grad_norm": 0.9102250337600708,
+      "learning_rate": 3.292316926770708e-05,
+      "loss": 0.5712,
+      "step": 8546
+    },
+    {
+      "epoch": 10.94016,
+      "grad_norm": 0.8525840640068054,
+      "learning_rate": 3.292116846738695e-05,
+      "loss": 0.5885,
+      "step": 8547
+    },
+    {
+      "epoch": 10.94144,
+      "grad_norm": 0.907727837562561,
+      "learning_rate": 3.291916766706683e-05,
+      "loss": 0.6011,
+      "step": 8548
+    },
+    {
+      "epoch": 10.94272,
+      "grad_norm": 0.8702722191810608,
+      "learning_rate": 3.29171668667467e-05,
+      "loss": 0.564,
+      "step": 8549
+    },
+    {
+      "epoch": 10.943999999999999,
+      "grad_norm": 0.8578301668167114,
+      "learning_rate": 3.291516606642657e-05,
+      "loss": 0.5763,
+      "step": 8550
+    },
+    {
+      "epoch": 10.94528,
+      "grad_norm": 0.9135550260543823,
+      "learning_rate": 3.2913165266106445e-05,
+      "loss": 0.6035,
+      "step": 8551
+    },
+    {
+      "epoch": 10.94656,
+      "grad_norm": 0.8668347001075745,
+      "learning_rate": 3.291116446578632e-05,
+      "loss": 0.5407,
+      "step": 8552
+    },
+    {
+      "epoch": 10.94784,
+      "grad_norm": 0.8684234619140625,
+      "learning_rate": 3.290916366546619e-05,
+      "loss": 0.5913,
+      "step": 8553
+    },
+    {
+      "epoch": 10.94912,
+      "grad_norm": 0.8825991749763489,
+      "learning_rate": 3.2907162865146054e-05,
+      "loss": 0.6135,
+      "step": 8554
+    },
+    {
+      "epoch": 10.9504,
+      "grad_norm": 0.9004281163215637,
+      "learning_rate": 3.290516206482593e-05,
+      "loss": 0.6063,
+      "step": 8555
+    },
+    {
+      "epoch": 10.95168,
+      "grad_norm": 0.8196146488189697,
+      "learning_rate": 3.2903161264505804e-05,
+      "loss": 0.5095,
+      "step": 8556
+    },
+    {
+      "epoch": 10.952960000000001,
+      "grad_norm": 0.8968302011489868,
+      "learning_rate": 3.2901160464185676e-05,
+      "loss": 0.6126,
+      "step": 8557
+    },
+    {
+      "epoch": 10.95424,
+      "grad_norm": 0.8384824991226196,
+      "learning_rate": 3.289915966386555e-05,
+      "loss": 0.5662,
+      "step": 8558
+    },
+    {
+      "epoch": 10.95552,
+      "grad_norm": 0.9103649258613586,
+      "learning_rate": 3.289715886354542e-05,
+      "loss": 0.5788,
+      "step": 8559
+    },
+    {
+      "epoch": 10.9568,
+      "grad_norm": 0.8952497243881226,
+      "learning_rate": 3.289515806322529e-05,
+      "loss": 0.5809,
+      "step": 8560
+    },
+    {
+      "epoch": 10.95808,
+      "grad_norm": 0.8801275491714478,
+      "learning_rate": 3.2893157262905164e-05,
+      "loss": 0.5753,
+      "step": 8561
+    },
+    {
+      "epoch": 10.95936,
+      "grad_norm": 0.9225260615348816,
+      "learning_rate": 3.2891156462585036e-05,
+      "loss": 0.601,
+      "step": 8562
+    },
+    {
+      "epoch": 10.96064,
+      "grad_norm": 0.8567471504211426,
+      "learning_rate": 3.288915566226491e-05,
+      "loss": 0.554,
+      "step": 8563
+    },
+    {
+      "epoch": 10.96192,
+      "grad_norm": 0.8874462842941284,
+      "learning_rate": 3.288715486194478e-05,
+      "loss": 0.5497,
+      "step": 8564
+    },
+    {
+      "epoch": 10.9632,
+      "grad_norm": 0.916522741317749,
+      "learning_rate": 3.288515406162465e-05,
+      "loss": 0.6097,
+      "step": 8565
+    },
+    {
+      "epoch": 10.96448,
+      "grad_norm": 0.890692412853241,
+      "learning_rate": 3.288315326130452e-05,
+      "loss": 0.5895,
+      "step": 8566
+    },
+    {
+      "epoch": 10.96576,
+      "grad_norm": 0.8528810739517212,
+      "learning_rate": 3.2881152460984395e-05,
+      "loss": 0.5664,
+      "step": 8567
+    },
+    {
+      "epoch": 10.96704,
+      "grad_norm": 0.8634024858474731,
+      "learning_rate": 3.287915166066427e-05,
+      "loss": 0.598,
+      "step": 8568
+    },
+    {
+      "epoch": 10.96832,
+      "grad_norm": 0.8372824788093567,
+      "learning_rate": 3.287715086034414e-05,
+      "loss": 0.5313,
+      "step": 8569
+    },
+    {
+      "epoch": 10.9696,
+      "grad_norm": 0.8880471587181091,
+      "learning_rate": 3.287515006002401e-05,
+      "loss": 0.5818,
+      "step": 8570
+    },
+    {
+      "epoch": 10.97088,
+      "grad_norm": 0.8914214372634888,
+      "learning_rate": 3.287314925970388e-05,
+      "loss": 0.5708,
+      "step": 8571
+    },
+    {
+      "epoch": 10.97216,
+      "grad_norm": 0.8861146569252014,
+      "learning_rate": 3.2871148459383754e-05,
+      "loss": 0.5029,
+      "step": 8572
+    },
+    {
+      "epoch": 10.97344,
+      "grad_norm": 0.9242550730705261,
+      "learning_rate": 3.2869147659063626e-05,
+      "loss": 0.5788,
+      "step": 8573
+    },
+    {
+      "epoch": 10.97472,
+      "grad_norm": 0.8695968389511108,
+      "learning_rate": 3.28671468587435e-05,
+      "loss": 0.5828,
+      "step": 8574
+    },
+    {
+      "epoch": 10.975999999999999,
+      "grad_norm": 0.8502150177955627,
+      "learning_rate": 3.286514605842337e-05,
+      "loss": 0.5377,
+      "step": 8575
+    },
+    {
+      "epoch": 10.97728,
+      "grad_norm": 0.8788812160491943,
+      "learning_rate": 3.286314525810325e-05,
+      "loss": 0.5599,
+      "step": 8576
+    },
+    {
+      "epoch": 10.97856,
+      "grad_norm": 0.9098305106163025,
+      "learning_rate": 3.2861144457783113e-05,
+      "loss": 0.5665,
+      "step": 8577
+    },
+    {
+      "epoch": 10.97984,
+      "grad_norm": 0.9447456002235413,
+      "learning_rate": 3.2859143657462985e-05,
+      "loss": 0.5948,
+      "step": 8578
+    },
+    {
+      "epoch": 10.98112,
+      "grad_norm": 0.9016488194465637,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 0.5681,
+      "step": 8579
+    },
+    {
+      "epoch": 10.9824,
+      "grad_norm": 0.8744751811027527,
+      "learning_rate": 3.285514205682273e-05,
+      "loss": 0.5673,
+      "step": 8580
+    },
+    {
+      "epoch": 10.98368,
+      "grad_norm": 0.8944303393363953,
+      "learning_rate": 3.28531412565026e-05,
+      "loss": 0.5808,
+      "step": 8581
+    },
+    {
+      "epoch": 10.98496,
+      "grad_norm": 0.8822252750396729,
+      "learning_rate": 3.285114045618247e-05,
+      "loss": 0.5607,
+      "step": 8582
+    },
+    {
+      "epoch": 10.98624,
+      "grad_norm": 0.8522640466690063,
+      "learning_rate": 3.284913965586235e-05,
+      "loss": 0.5971,
+      "step": 8583
+    },
+    {
+      "epoch": 10.98752,
+      "grad_norm": 0.874326765537262,
+      "learning_rate": 3.284713885554222e-05,
+      "loss": 0.6094,
+      "step": 8584
+    },
+    {
+      "epoch": 10.9888,
+      "grad_norm": 0.9082483053207397,
+      "learning_rate": 3.284513805522209e-05,
+      "loss": 0.5866,
+      "step": 8585
+    },
+    {
+      "epoch": 10.99008,
+      "grad_norm": 0.8725292086601257,
+      "learning_rate": 3.284313725490196e-05,
+      "loss": 0.6204,
+      "step": 8586
+    },
+    {
+      "epoch": 10.99136,
+      "grad_norm": 0.8516504168510437,
+      "learning_rate": 3.284113645458183e-05,
+      "loss": 0.5942,
+      "step": 8587
+    },
+    {
+      "epoch": 10.99264,
+      "grad_norm": 0.8405760526657104,
+      "learning_rate": 3.2839135654261704e-05,
+      "loss": 0.6035,
+      "step": 8588
+    },
+    {
+      "epoch": 10.99392,
+      "grad_norm": 0.8437188863754272,
+      "learning_rate": 3.2837134853941576e-05,
+      "loss": 0.5071,
+      "step": 8589
+    },
+    {
+      "epoch": 10.9952,
+      "grad_norm": 0.877285897731781,
+      "learning_rate": 3.2835134053621454e-05,
+      "loss": 0.5761,
+      "step": 8590
+    },
+    {
+      "epoch": 10.99648,
+      "grad_norm": 0.8691837191581726,
+      "learning_rate": 3.2833133253301326e-05,
+      "loss": 0.6239,
+      "step": 8591
+    },
+    {
+      "epoch": 10.99776,
+      "grad_norm": 0.8213376402854919,
+      "learning_rate": 3.28311324529812e-05,
+      "loss": 0.5429,
+      "step": 8592
+    },
+    {
+      "epoch": 10.99904,
+      "grad_norm": 0.8809442520141602,
+      "learning_rate": 3.282913165266106e-05,
+      "loss": 0.5888,
+      "step": 8593
+    },
+    {
+      "epoch": 11.00032,
+      "grad_norm": 1.9187315702438354,
+      "learning_rate": 3.2827130852340935e-05,
+      "loss": 1.0473,
+      "step": 8594
+    },
+    {
+      "epoch": 11.0016,
+      "grad_norm": 0.9014028310775757,
+      "learning_rate": 3.282513005202081e-05,
+      "loss": 0.587,
+      "step": 8595
+    },
+    {
+      "epoch": 11.00288,
+      "grad_norm": 0.8248893022537231,
+      "learning_rate": 3.282312925170068e-05,
+      "loss": 0.5534,
+      "step": 8596
+    },
+    {
+      "epoch": 11.00416,
+      "grad_norm": 0.8243193626403809,
+      "learning_rate": 3.282112845138056e-05,
+      "loss": 0.5334,
+      "step": 8597
+    },
+    {
+      "epoch": 11.00544,
+      "grad_norm": 0.8348228931427002,
+      "learning_rate": 3.281912765106043e-05,
+      "loss": 0.5808,
+      "step": 8598
+    },
+    {
+      "epoch": 11.00672,
+      "grad_norm": 0.8530436754226685,
+      "learning_rate": 3.28171268507403e-05,
+      "loss": 0.5644,
+      "step": 8599
+    },
+    {
+      "epoch": 11.008,
+      "grad_norm": 0.8829576969146729,
+      "learning_rate": 3.281512605042017e-05,
+      "loss": 0.5946,
+      "step": 8600
+    },
+    {
+      "epoch": 11.00928,
+      "grad_norm": 0.8608317971229553,
+      "learning_rate": 3.281312525010004e-05,
+      "loss": 0.5479,
+      "step": 8601
+    },
+    {
+      "epoch": 11.01056,
+      "grad_norm": 0.8292250037193298,
+      "learning_rate": 3.281112444977991e-05,
+      "loss": 0.5278,
+      "step": 8602
+    },
+    {
+      "epoch": 11.01184,
+      "grad_norm": 0.8977212905883789,
+      "learning_rate": 3.280912364945978e-05,
+      "loss": 0.6198,
+      "step": 8603
+    },
+    {
+      "epoch": 11.01312,
+      "grad_norm": 0.866463840007782,
+      "learning_rate": 3.280712284913966e-05,
+      "loss": 0.5493,
+      "step": 8604
+    },
+    {
+      "epoch": 11.0144,
+      "grad_norm": 0.9034777879714966,
+      "learning_rate": 3.280512204881953e-05,
+      "loss": 0.5457,
+      "step": 8605
+    },
+    {
+      "epoch": 11.01568,
+      "grad_norm": 0.9087190628051758,
+      "learning_rate": 3.2803121248499404e-05,
+      "loss": 0.5557,
+      "step": 8606
+    },
+    {
+      "epoch": 11.01696,
+      "grad_norm": 0.8566074967384338,
+      "learning_rate": 3.2801120448179276e-05,
+      "loss": 0.5722,
+      "step": 8607
+    },
+    {
+      "epoch": 11.01824,
+      "grad_norm": 0.8810505270957947,
+      "learning_rate": 3.279911964785915e-05,
+      "loss": 0.5842,
+      "step": 8608
+    },
+    {
+      "epoch": 11.01952,
+      "grad_norm": 0.8587162494659424,
+      "learning_rate": 3.279711884753901e-05,
+      "loss": 0.4996,
+      "step": 8609
+    },
+    {
+      "epoch": 11.0208,
+      "grad_norm": 0.8678077459335327,
+      "learning_rate": 3.2795118047218885e-05,
+      "loss": 0.5706,
+      "step": 8610
+    },
+    {
+      "epoch": 11.02208,
+      "grad_norm": 0.8607105612754822,
+      "learning_rate": 3.279311724689876e-05,
+      "loss": 0.548,
+      "step": 8611
+    },
+    {
+      "epoch": 11.02336,
+      "grad_norm": 0.8053077459335327,
+      "learning_rate": 3.2791116446578635e-05,
+      "loss": 0.5258,
+      "step": 8612
+    },
+    {
+      "epoch": 11.02464,
+      "grad_norm": 0.8794430494308472,
+      "learning_rate": 3.278911564625851e-05,
+      "loss": 0.5362,
+      "step": 8613
+    },
+    {
+      "epoch": 11.02592,
+      "grad_norm": 0.8815926313400269,
+      "learning_rate": 3.278711484593838e-05,
+      "loss": 0.5306,
+      "step": 8614
+    },
+    {
+      "epoch": 11.0272,
+      "grad_norm": 0.8498107194900513,
+      "learning_rate": 3.278511404561825e-05,
+      "loss": 0.5102,
+      "step": 8615
+    },
+    {
+      "epoch": 11.02848,
+      "grad_norm": 0.907062292098999,
+      "learning_rate": 3.278311324529812e-05,
+      "loss": 0.5752,
+      "step": 8616
+    },
+    {
+      "epoch": 11.02976,
+      "grad_norm": 0.8796223998069763,
+      "learning_rate": 3.278111244497799e-05,
+      "loss": 0.5416,
+      "step": 8617
+    },
+    {
+      "epoch": 11.03104,
+      "grad_norm": 0.9161748290061951,
+      "learning_rate": 3.2779111644657866e-05,
+      "loss": 0.5676,
+      "step": 8618
+    },
+    {
+      "epoch": 11.03232,
+      "grad_norm": 0.8488770127296448,
+      "learning_rate": 3.277711084433774e-05,
+      "loss": 0.5681,
+      "step": 8619
+    },
+    {
+      "epoch": 11.0336,
+      "grad_norm": 0.9294846057891846,
+      "learning_rate": 3.277511004401761e-05,
+      "loss": 0.5358,
+      "step": 8620
+    },
+    {
+      "epoch": 11.03488,
+      "grad_norm": 0.8868327736854553,
+      "learning_rate": 3.277310924369748e-05,
+      "loss": 0.4881,
+      "step": 8621
+    },
+    {
+      "epoch": 11.03616,
+      "grad_norm": 0.8899376392364502,
+      "learning_rate": 3.2771108443377354e-05,
+      "loss": 0.533,
+      "step": 8622
+    },
+    {
+      "epoch": 11.03744,
+      "grad_norm": 0.8953622579574585,
+      "learning_rate": 3.2769107643057226e-05,
+      "loss": 0.5389,
+      "step": 8623
+    },
+    {
+      "epoch": 11.03872,
+      "grad_norm": 0.9292336106300354,
+      "learning_rate": 3.27671068427371e-05,
+      "loss": 0.606,
+      "step": 8624
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 0.9286232590675354,
+      "learning_rate": 3.276510604241697e-05,
+      "loss": 0.5897,
+      "step": 8625
+    },
+    {
+      "epoch": 11.04128,
+      "grad_norm": 0.9098997712135315,
+      "learning_rate": 3.276310524209684e-05,
+      "loss": 0.5316,
+      "step": 8626
+    },
+    {
+      "epoch": 11.04256,
+      "grad_norm": 0.9626008868217468,
+      "learning_rate": 3.276110444177671e-05,
+      "loss": 0.6035,
+      "step": 8627
+    },
+    {
+      "epoch": 11.04384,
+      "grad_norm": 0.8677993416786194,
+      "learning_rate": 3.2759103641456585e-05,
+      "loss": 0.4975,
+      "step": 8628
+    },
+    {
+      "epoch": 11.04512,
+      "grad_norm": 0.8999881148338318,
+      "learning_rate": 3.275710284113646e-05,
+      "loss": 0.5494,
+      "step": 8629
+    },
+    {
+      "epoch": 11.0464,
+      "grad_norm": 0.8779304623603821,
+      "learning_rate": 3.275510204081633e-05,
+      "loss": 0.5525,
+      "step": 8630
+    },
+    {
+      "epoch": 11.04768,
+      "grad_norm": 0.9179341793060303,
+      "learning_rate": 3.27531012404962e-05,
+      "loss": 0.5784,
+      "step": 8631
+    },
+    {
+      "epoch": 11.04896,
+      "grad_norm": 0.9072324633598328,
+      "learning_rate": 3.275110044017607e-05,
+      "loss": 0.548,
+      "step": 8632
+    },
+    {
+      "epoch": 11.05024,
+      "grad_norm": 0.9047489166259766,
+      "learning_rate": 3.2749099639855944e-05,
+      "loss": 0.5374,
+      "step": 8633
+    },
+    {
+      "epoch": 11.05152,
+      "grad_norm": 0.9042777419090271,
+      "learning_rate": 3.2747098839535816e-05,
+      "loss": 0.6008,
+      "step": 8634
+    },
+    {
+      "epoch": 11.0528,
+      "grad_norm": 0.913502037525177,
+      "learning_rate": 3.274509803921569e-05,
+      "loss": 0.6139,
+      "step": 8635
+    },
+    {
+      "epoch": 11.05408,
+      "grad_norm": 0.8988363146781921,
+      "learning_rate": 3.274309723889556e-05,
+      "loss": 0.5735,
+      "step": 8636
+    },
+    {
+      "epoch": 11.05536,
+      "grad_norm": 0.9248592257499695,
+      "learning_rate": 3.274109643857543e-05,
+      "loss": 0.6024,
+      "step": 8637
+    },
+    {
+      "epoch": 11.05664,
+      "grad_norm": 0.8595258593559265,
+      "learning_rate": 3.2739095638255303e-05,
+      "loss": 0.5686,
+      "step": 8638
+    },
+    {
+      "epoch": 11.05792,
+      "grad_norm": 0.917407751083374,
+      "learning_rate": 3.2737094837935175e-05,
+      "loss": 0.5686,
+      "step": 8639
+    },
+    {
+      "epoch": 11.0592,
+      "grad_norm": 0.8839895725250244,
+      "learning_rate": 3.273509403761505e-05,
+      "loss": 0.5334,
+      "step": 8640
+    },
+    {
+      "epoch": 11.06048,
+      "grad_norm": 0.8676347136497498,
+      "learning_rate": 3.273309323729492e-05,
+      "loss": 0.5396,
+      "step": 8641
+    },
+    {
+      "epoch": 11.06176,
+      "grad_norm": 0.880967915058136,
+      "learning_rate": 3.273109243697479e-05,
+      "loss": 0.514,
+      "step": 8642
+    },
+    {
+      "epoch": 11.06304,
+      "grad_norm": 0.9367587566375732,
+      "learning_rate": 3.272909163665466e-05,
+      "loss": 0.5731,
+      "step": 8643
+    },
+    {
+      "epoch": 11.06432,
+      "grad_norm": 0.8796616196632385,
+      "learning_rate": 3.2727090836334535e-05,
+      "loss": 0.5887,
+      "step": 8644
+    },
+    {
+      "epoch": 11.0656,
+      "grad_norm": 0.8505174517631531,
+      "learning_rate": 3.2725090036014406e-05,
+      "loss": 0.5534,
+      "step": 8645
+    },
+    {
+      "epoch": 11.06688,
+      "grad_norm": 0.8879563808441162,
+      "learning_rate": 3.2723089235694285e-05,
+      "loss": 0.6019,
+      "step": 8646
+    },
+    {
+      "epoch": 11.06816,
+      "grad_norm": 0.876809298992157,
+      "learning_rate": 3.272108843537415e-05,
+      "loss": 0.5354,
+      "step": 8647
+    },
+    {
+      "epoch": 11.06944,
+      "grad_norm": 0.8911082148551941,
+      "learning_rate": 3.271908763505402e-05,
+      "loss": 0.5154,
+      "step": 8648
+    },
+    {
+      "epoch": 11.07072,
+      "grad_norm": 0.9009726643562317,
+      "learning_rate": 3.2717086834733894e-05,
+      "loss": 0.5128,
+      "step": 8649
+    },
+    {
+      "epoch": 11.072,
+      "grad_norm": 0.8957163691520691,
+      "learning_rate": 3.2715086034413766e-05,
+      "loss": 0.5322,
+      "step": 8650
+    },
+    {
+      "epoch": 11.07328,
+      "grad_norm": 0.866563618183136,
+      "learning_rate": 3.271308523409364e-05,
+      "loss": 0.5464,
+      "step": 8651
+    },
+    {
+      "epoch": 11.07456,
+      "grad_norm": 0.9009678959846497,
+      "learning_rate": 3.271108443377351e-05,
+      "loss": 0.536,
+      "step": 8652
+    },
+    {
+      "epoch": 11.07584,
+      "grad_norm": 0.9172965288162231,
+      "learning_rate": 3.270908363345339e-05,
+      "loss": 0.5667,
+      "step": 8653
+    },
+    {
+      "epoch": 11.07712,
+      "grad_norm": 0.9287314414978027,
+      "learning_rate": 3.270708283313326e-05,
+      "loss": 0.5507,
+      "step": 8654
+    },
+    {
+      "epoch": 11.0784,
+      "grad_norm": 0.9205261468887329,
+      "learning_rate": 3.2705082032813125e-05,
+      "loss": 0.5491,
+      "step": 8655
+    },
+    {
+      "epoch": 11.07968,
+      "grad_norm": 0.8569768667221069,
+      "learning_rate": 3.2703081232493e-05,
+      "loss": 0.5158,
+      "step": 8656
+    },
+    {
+      "epoch": 11.08096,
+      "grad_norm": 0.9120553731918335,
+      "learning_rate": 3.270108043217287e-05,
+      "loss": 0.5894,
+      "step": 8657
+    },
+    {
+      "epoch": 11.08224,
+      "grad_norm": 0.9037409424781799,
+      "learning_rate": 3.269907963185274e-05,
+      "loss": 0.5456,
+      "step": 8658
+    },
+    {
+      "epoch": 11.08352,
+      "grad_norm": 0.9417985081672668,
+      "learning_rate": 3.269707883153261e-05,
+      "loss": 0.5491,
+      "step": 8659
+    },
+    {
+      "epoch": 11.0848,
+      "grad_norm": 0.8706283569335938,
+      "learning_rate": 3.2695078031212484e-05,
+      "loss": 0.5391,
+      "step": 8660
+    },
+    {
+      "epoch": 11.08608,
+      "grad_norm": 0.8442224860191345,
+      "learning_rate": 3.269307723089236e-05,
+      "loss": 0.5301,
+      "step": 8661
+    },
+    {
+      "epoch": 11.08736,
+      "grad_norm": 0.8892581462860107,
+      "learning_rate": 3.2691076430572235e-05,
+      "loss": 0.5239,
+      "step": 8662
+    },
+    {
+      "epoch": 11.08864,
+      "grad_norm": 0.8483731746673584,
+      "learning_rate": 3.26890756302521e-05,
+      "loss": 0.5079,
+      "step": 8663
+    },
+    {
+      "epoch": 11.08992,
+      "grad_norm": 0.8920766711235046,
+      "learning_rate": 3.268707482993197e-05,
+      "loss": 0.5682,
+      "step": 8664
+    },
+    {
+      "epoch": 11.0912,
+      "grad_norm": 0.8351317048072815,
+      "learning_rate": 3.2685074029611844e-05,
+      "loss": 0.5189,
+      "step": 8665
+    },
+    {
+      "epoch": 11.09248,
+      "grad_norm": 0.9149410724639893,
+      "learning_rate": 3.2683073229291715e-05,
+      "loss": 0.5959,
+      "step": 8666
+    },
+    {
+      "epoch": 11.09376,
+      "grad_norm": 0.8564448356628418,
+      "learning_rate": 3.268107242897159e-05,
+      "loss": 0.5124,
+      "step": 8667
+    },
+    {
+      "epoch": 11.09504,
+      "grad_norm": 0.8573926687240601,
+      "learning_rate": 3.2679071628651466e-05,
+      "loss": 0.5366,
+      "step": 8668
+    },
+    {
+      "epoch": 11.09632,
+      "grad_norm": 0.8379302620887756,
+      "learning_rate": 3.267707082833134e-05,
+      "loss": 0.5218,
+      "step": 8669
+    },
+    {
+      "epoch": 11.0976,
+      "grad_norm": 0.8685084581375122,
+      "learning_rate": 3.267507002801121e-05,
+      "loss": 0.5322,
+      "step": 8670
+    },
+    {
+      "epoch": 11.09888,
+      "grad_norm": 0.8932461738586426,
+      "learning_rate": 3.2673069227691075e-05,
+      "loss": 0.59,
+      "step": 8671
+    },
+    {
+      "epoch": 11.10016,
+      "grad_norm": 0.8610022664070129,
+      "learning_rate": 3.267106842737095e-05,
+      "loss": 0.5244,
+      "step": 8672
+    },
+    {
+      "epoch": 11.10144,
+      "grad_norm": 0.8831630349159241,
+      "learning_rate": 3.266906762705082e-05,
+      "loss": 0.609,
+      "step": 8673
+    },
+    {
+      "epoch": 11.10272,
+      "grad_norm": 0.899043619632721,
+      "learning_rate": 3.266706682673069e-05,
+      "loss": 0.552,
+      "step": 8674
+    },
+    {
+      "epoch": 11.104,
+      "grad_norm": 0.9248834848403931,
+      "learning_rate": 3.266506602641057e-05,
+      "loss": 0.5851,
+      "step": 8675
+    },
+    {
+      "epoch": 11.10528,
+      "grad_norm": 0.8615071177482605,
+      "learning_rate": 3.266306522609044e-05,
+      "loss": 0.5664,
+      "step": 8676
+    },
+    {
+      "epoch": 11.10656,
+      "grad_norm": 0.882163405418396,
+      "learning_rate": 3.266106442577031e-05,
+      "loss": 0.5823,
+      "step": 8677
+    },
+    {
+      "epoch": 11.10784,
+      "grad_norm": 0.8114548921585083,
+      "learning_rate": 3.2659063625450185e-05,
+      "loss": 0.5001,
+      "step": 8678
+    },
+    {
+      "epoch": 11.10912,
+      "grad_norm": 0.9176467657089233,
+      "learning_rate": 3.265706282513005e-05,
+      "loss": 0.566,
+      "step": 8679
+    },
+    {
+      "epoch": 11.1104,
+      "grad_norm": 0.867563009262085,
+      "learning_rate": 3.265506202480992e-05,
+      "loss": 0.5233,
+      "step": 8680
+    },
+    {
+      "epoch": 11.11168,
+      "grad_norm": 0.8672628402709961,
+      "learning_rate": 3.265306122448979e-05,
+      "loss": 0.5412,
+      "step": 8681
+    },
+    {
+      "epoch": 11.11296,
+      "grad_norm": 0.9063531160354614,
+      "learning_rate": 3.265106042416967e-05,
+      "loss": 0.5753,
+      "step": 8682
+    },
+    {
+      "epoch": 11.11424,
+      "grad_norm": 0.86628258228302,
+      "learning_rate": 3.2649059623849544e-05,
+      "loss": 0.5739,
+      "step": 8683
+    },
+    {
+      "epoch": 11.11552,
+      "grad_norm": 0.8533140420913696,
+      "learning_rate": 3.2647058823529416e-05,
+      "loss": 0.555,
+      "step": 8684
+    },
+    {
+      "epoch": 11.1168,
+      "grad_norm": 0.8930099606513977,
+      "learning_rate": 3.264505802320929e-05,
+      "loss": 0.5193,
+      "step": 8685
+    },
+    {
+      "epoch": 11.11808,
+      "grad_norm": 0.8562926650047302,
+      "learning_rate": 3.264305722288916e-05,
+      "loss": 0.5209,
+      "step": 8686
+    },
+    {
+      "epoch": 11.11936,
+      "grad_norm": 0.8850371241569519,
+      "learning_rate": 3.2641056422569024e-05,
+      "loss": 0.5525,
+      "step": 8687
+    },
+    {
+      "epoch": 11.12064,
+      "grad_norm": 0.932697594165802,
+      "learning_rate": 3.2639055622248896e-05,
+      "loss": 0.5613,
+      "step": 8688
+    },
+    {
+      "epoch": 11.12192,
+      "grad_norm": 0.8724458813667297,
+      "learning_rate": 3.2637054821928775e-05,
+      "loss": 0.5146,
+      "step": 8689
+    },
+    {
+      "epoch": 11.1232,
+      "grad_norm": 0.8727782368659973,
+      "learning_rate": 3.263505402160865e-05,
+      "loss": 0.5339,
+      "step": 8690
+    },
+    {
+      "epoch": 11.12448,
+      "grad_norm": 0.916933536529541,
+      "learning_rate": 3.263305322128852e-05,
+      "loss": 0.5998,
+      "step": 8691
+    },
+    {
+      "epoch": 11.12576,
+      "grad_norm": 0.875407338142395,
+      "learning_rate": 3.263105242096839e-05,
+      "loss": 0.5418,
+      "step": 8692
+    },
+    {
+      "epoch": 11.12704,
+      "grad_norm": 0.8921369910240173,
+      "learning_rate": 3.262905162064826e-05,
+      "loss": 0.5805,
+      "step": 8693
+    },
+    {
+      "epoch": 11.12832,
+      "grad_norm": 0.9058687090873718,
+      "learning_rate": 3.2627050820328134e-05,
+      "loss": 0.5751,
+      "step": 8694
+    },
+    {
+      "epoch": 11.1296,
+      "grad_norm": 0.9492460489273071,
+      "learning_rate": 3.2625050020008e-05,
+      "loss": 0.5572,
+      "step": 8695
+    },
+    {
+      "epoch": 11.13088,
+      "grad_norm": 0.9078418016433716,
+      "learning_rate": 3.262304921968788e-05,
+      "loss": 0.5816,
+      "step": 8696
+    },
+    {
+      "epoch": 11.13216,
+      "grad_norm": 0.8583430051803589,
+      "learning_rate": 3.262104841936775e-05,
+      "loss": 0.5219,
+      "step": 8697
+    },
+    {
+      "epoch": 11.13344,
+      "grad_norm": 0.9013339281082153,
+      "learning_rate": 3.261904761904762e-05,
+      "loss": 0.5521,
+      "step": 8698
+    },
+    {
+      "epoch": 11.13472,
+      "grad_norm": 0.9228495955467224,
+      "learning_rate": 3.2617046818727494e-05,
+      "loss": 0.6093,
+      "step": 8699
+    },
+    {
+      "epoch": 11.136,
+      "grad_norm": 0.895261287689209,
+      "learning_rate": 3.2615046018407365e-05,
+      "loss": 0.5856,
+      "step": 8700
+    },
+    {
+      "epoch": 11.13728,
+      "grad_norm": 0.9044747948646545,
+      "learning_rate": 3.261304521808724e-05,
+      "loss": 0.5621,
+      "step": 8701
+    },
+    {
+      "epoch": 11.13856,
+      "grad_norm": 0.9378688335418701,
+      "learning_rate": 3.261104441776711e-05,
+      "loss": 0.6047,
+      "step": 8702
+    },
+    {
+      "epoch": 11.13984,
+      "grad_norm": 0.941558301448822,
+      "learning_rate": 3.260904361744698e-05,
+      "loss": 0.5467,
+      "step": 8703
+    },
+    {
+      "epoch": 11.14112,
+      "grad_norm": 0.9449127912521362,
+      "learning_rate": 3.260704281712685e-05,
+      "loss": 0.5336,
+      "step": 8704
+    },
+    {
+      "epoch": 11.1424,
+      "grad_norm": 0.8693781495094299,
+      "learning_rate": 3.2605042016806725e-05,
+      "loss": 0.522,
+      "step": 8705
+    },
+    {
+      "epoch": 11.14368,
+      "grad_norm": 0.8828169107437134,
+      "learning_rate": 3.2603041216486597e-05,
+      "loss": 0.557,
+      "step": 8706
+    },
+    {
+      "epoch": 11.14496,
+      "grad_norm": 0.9226519465446472,
+      "learning_rate": 3.260104041616647e-05,
+      "loss": 0.5246,
+      "step": 8707
+    },
+    {
+      "epoch": 11.14624,
+      "grad_norm": 0.9169709086418152,
+      "learning_rate": 3.259903961584634e-05,
+      "loss": 0.58,
+      "step": 8708
+    },
+    {
+      "epoch": 11.14752,
+      "grad_norm": 0.92818683385849,
+      "learning_rate": 3.259703881552621e-05,
+      "loss": 0.5937,
+      "step": 8709
+    },
+    {
+      "epoch": 11.1488,
+      "grad_norm": 0.9431915879249573,
+      "learning_rate": 3.2595038015206084e-05,
+      "loss": 0.5702,
+      "step": 8710
+    },
+    {
+      "epoch": 11.150079999999999,
+      "grad_norm": 0.8843738436698914,
+      "learning_rate": 3.2593037214885956e-05,
+      "loss": 0.5305,
+      "step": 8711
+    },
+    {
+      "epoch": 11.15136,
+      "grad_norm": 0.8925315737724304,
+      "learning_rate": 3.259103641456583e-05,
+      "loss": 0.5114,
+      "step": 8712
+    },
+    {
+      "epoch": 11.15264,
+      "grad_norm": 0.8767265677452087,
+      "learning_rate": 3.25890356142457e-05,
+      "loss": 0.5199,
+      "step": 8713
+    },
+    {
+      "epoch": 11.15392,
+      "grad_norm": 0.8850144743919373,
+      "learning_rate": 3.258703481392557e-05,
+      "loss": 0.5388,
+      "step": 8714
+    },
+    {
+      "epoch": 11.1552,
+      "grad_norm": 0.8759915232658386,
+      "learning_rate": 3.258503401360544e-05,
+      "loss": 0.5628,
+      "step": 8715
+    },
+    {
+      "epoch": 11.15648,
+      "grad_norm": 0.8618879318237305,
+      "learning_rate": 3.2583033213285315e-05,
+      "loss": 0.5094,
+      "step": 8716
+    },
+    {
+      "epoch": 11.15776,
+      "grad_norm": 0.9156239628791809,
+      "learning_rate": 3.258103241296519e-05,
+      "loss": 0.562,
+      "step": 8717
+    },
+    {
+      "epoch": 11.15904,
+      "grad_norm": 0.8669613599777222,
+      "learning_rate": 3.257903161264506e-05,
+      "loss": 0.5478,
+      "step": 8718
+    },
+    {
+      "epoch": 11.16032,
+      "grad_norm": 0.9046533107757568,
+      "learning_rate": 3.257703081232493e-05,
+      "loss": 0.6049,
+      "step": 8719
+    },
+    {
+      "epoch": 11.1616,
+      "grad_norm": 0.9057307243347168,
+      "learning_rate": 3.25750300120048e-05,
+      "loss": 0.5175,
+      "step": 8720
+    },
+    {
+      "epoch": 11.16288,
+      "grad_norm": 0.8702899813652039,
+      "learning_rate": 3.2573029211684674e-05,
+      "loss": 0.5061,
+      "step": 8721
+    },
+    {
+      "epoch": 11.16416,
+      "grad_norm": 0.8357275724411011,
+      "learning_rate": 3.2571028411364546e-05,
+      "loss": 0.5112,
+      "step": 8722
+    },
+    {
+      "epoch": 11.16544,
+      "grad_norm": 0.9093285799026489,
+      "learning_rate": 3.256902761104442e-05,
+      "loss": 0.5853,
+      "step": 8723
+    },
+    {
+      "epoch": 11.16672,
+      "grad_norm": 0.8935590386390686,
+      "learning_rate": 3.25670268107243e-05,
+      "loss": 0.5726,
+      "step": 8724
+    },
+    {
+      "epoch": 11.168,
+      "grad_norm": 0.8492559194564819,
+      "learning_rate": 3.256502601040416e-05,
+      "loss": 0.522,
+      "step": 8725
+    },
+    {
+      "epoch": 11.16928,
+      "grad_norm": 0.8798608779907227,
+      "learning_rate": 3.2563025210084034e-05,
+      "loss": 0.5661,
+      "step": 8726
+    },
+    {
+      "epoch": 11.17056,
+      "grad_norm": 0.8920543193817139,
+      "learning_rate": 3.2561024409763906e-05,
+      "loss": 0.5535,
+      "step": 8727
+    },
+    {
+      "epoch": 11.17184,
+      "grad_norm": 0.8616754412651062,
+      "learning_rate": 3.255902360944378e-05,
+      "loss": 0.5294,
+      "step": 8728
+    },
+    {
+      "epoch": 11.17312,
+      "grad_norm": 0.8341376185417175,
+      "learning_rate": 3.255702280912365e-05,
+      "loss": 0.4899,
+      "step": 8729
+    },
+    {
+      "epoch": 11.1744,
+      "grad_norm": 0.8843587636947632,
+      "learning_rate": 3.255502200880352e-05,
+      "loss": 0.5537,
+      "step": 8730
+    },
+    {
+      "epoch": 11.17568,
+      "grad_norm": 0.8609508872032166,
+      "learning_rate": 3.25530212084834e-05,
+      "loss": 0.5028,
+      "step": 8731
+    },
+    {
+      "epoch": 11.17696,
+      "grad_norm": 0.8849518299102783,
+      "learning_rate": 3.255102040816327e-05,
+      "loss": 0.5173,
+      "step": 8732
+    },
+    {
+      "epoch": 11.17824,
+      "grad_norm": 0.874070942401886,
+      "learning_rate": 3.254901960784314e-05,
+      "loss": 0.5143,
+      "step": 8733
+    },
+    {
+      "epoch": 11.17952,
+      "grad_norm": 0.9157163500785828,
+      "learning_rate": 3.254701880752301e-05,
+      "loss": 0.5429,
+      "step": 8734
+    },
+    {
+      "epoch": 11.1808,
+      "grad_norm": 0.90493243932724,
+      "learning_rate": 3.254501800720288e-05,
+      "loss": 0.5794,
+      "step": 8735
+    },
+    {
+      "epoch": 11.18208,
+      "grad_norm": 0.9423990249633789,
+      "learning_rate": 3.254301720688275e-05,
+      "loss": 0.5889,
+      "step": 8736
+    },
+    {
+      "epoch": 11.18336,
+      "grad_norm": 0.950400710105896,
+      "learning_rate": 3.2541016406562624e-05,
+      "loss": 0.5909,
+      "step": 8737
+    },
+    {
+      "epoch": 11.18464,
+      "grad_norm": 0.946723997592926,
+      "learning_rate": 3.25390156062425e-05,
+      "loss": 0.5764,
+      "step": 8738
+    },
+    {
+      "epoch": 11.18592,
+      "grad_norm": 0.9119811058044434,
+      "learning_rate": 3.2537014805922375e-05,
+      "loss": 0.5859,
+      "step": 8739
+    },
+    {
+      "epoch": 11.1872,
+      "grad_norm": 0.8999208807945251,
+      "learning_rate": 3.2535014005602246e-05,
+      "loss": 0.5725,
+      "step": 8740
+    },
+    {
+      "epoch": 11.18848,
+      "grad_norm": 0.8947361707687378,
+      "learning_rate": 3.253301320528211e-05,
+      "loss": 0.5432,
+      "step": 8741
+    },
+    {
+      "epoch": 11.18976,
+      "grad_norm": 0.9105396270751953,
+      "learning_rate": 3.253101240496198e-05,
+      "loss": 0.5896,
+      "step": 8742
+    },
+    {
+      "epoch": 11.19104,
+      "grad_norm": 0.9032384753227234,
+      "learning_rate": 3.2529011604641855e-05,
+      "loss": 0.5516,
+      "step": 8743
+    },
+    {
+      "epoch": 11.19232,
+      "grad_norm": 0.966087281703949,
+      "learning_rate": 3.252701080432173e-05,
+      "loss": 0.5906,
+      "step": 8744
+    },
+    {
+      "epoch": 11.1936,
+      "grad_norm": 0.9482411742210388,
+      "learning_rate": 3.2525010004001606e-05,
+      "loss": 0.597,
+      "step": 8745
+    },
+    {
+      "epoch": 11.19488,
+      "grad_norm": 0.8811086416244507,
+      "learning_rate": 3.252300920368148e-05,
+      "loss": 0.54,
+      "step": 8746
+    },
+    {
+      "epoch": 11.19616,
+      "grad_norm": 0.8933411240577698,
+      "learning_rate": 3.252100840336135e-05,
+      "loss": 0.5465,
+      "step": 8747
+    },
+    {
+      "epoch": 11.19744,
+      "grad_norm": 0.8633106350898743,
+      "learning_rate": 3.251900760304122e-05,
+      "loss": 0.5178,
+      "step": 8748
+    },
+    {
+      "epoch": 11.19872,
+      "grad_norm": 0.8998020887374878,
+      "learning_rate": 3.2517006802721086e-05,
+      "loss": 0.5437,
+      "step": 8749
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 0.9244217276573181,
+      "learning_rate": 3.251500600240096e-05,
+      "loss": 0.6415,
+      "step": 8750
+    },
+    {
+      "epoch": 11.20128,
+      "grad_norm": 0.9325177073478699,
+      "learning_rate": 3.251300520208083e-05,
+      "loss": 0.6122,
+      "step": 8751
+    },
+    {
+      "epoch": 11.20256,
+      "grad_norm": 0.9503816962242126,
+      "learning_rate": 3.251100440176071e-05,
+      "loss": 0.5848,
+      "step": 8752
+    },
+    {
+      "epoch": 11.20384,
+      "grad_norm": 0.9142863154411316,
+      "learning_rate": 3.250900360144058e-05,
+      "loss": 0.5554,
+      "step": 8753
+    },
+    {
+      "epoch": 11.20512,
+      "grad_norm": 0.8959869742393494,
+      "learning_rate": 3.250700280112045e-05,
+      "loss": 0.6131,
+      "step": 8754
+    },
+    {
+      "epoch": 11.2064,
+      "grad_norm": 0.9044204354286194,
+      "learning_rate": 3.2505002000800324e-05,
+      "loss": 0.5885,
+      "step": 8755
+    },
+    {
+      "epoch": 11.20768,
+      "grad_norm": 0.9072441458702087,
+      "learning_rate": 3.2503001200480196e-05,
+      "loss": 0.5396,
+      "step": 8756
+    },
+    {
+      "epoch": 11.20896,
+      "grad_norm": 0.8490614295005798,
+      "learning_rate": 3.250100040016006e-05,
+      "loss": 0.5353,
+      "step": 8757
+    },
+    {
+      "epoch": 11.21024,
+      "grad_norm": 0.9115065336227417,
+      "learning_rate": 3.249899959983993e-05,
+      "loss": 0.5137,
+      "step": 8758
+    },
+    {
+      "epoch": 11.21152,
+      "grad_norm": 0.8661530017852783,
+      "learning_rate": 3.249699879951981e-05,
+      "loss": 0.4879,
+      "step": 8759
+    },
+    {
+      "epoch": 11.2128,
+      "grad_norm": 0.8535701036453247,
+      "learning_rate": 3.2494997999199684e-05,
+      "loss": 0.5335,
+      "step": 8760
+    },
+    {
+      "epoch": 11.21408,
+      "grad_norm": 0.9392260313034058,
+      "learning_rate": 3.2492997198879555e-05,
+      "loss": 0.5892,
+      "step": 8761
+    },
+    {
+      "epoch": 11.21536,
+      "grad_norm": 0.861599862575531,
+      "learning_rate": 3.249099639855943e-05,
+      "loss": 0.5172,
+      "step": 8762
+    },
+    {
+      "epoch": 11.21664,
+      "grad_norm": 0.8953681588172913,
+      "learning_rate": 3.24889955982393e-05,
+      "loss": 0.5388,
+      "step": 8763
+    },
+    {
+      "epoch": 11.21792,
+      "grad_norm": 0.9016774892807007,
+      "learning_rate": 3.248699479791917e-05,
+      "loss": 0.5661,
+      "step": 8764
+    },
+    {
+      "epoch": 11.2192,
+      "grad_norm": 0.8995947241783142,
+      "learning_rate": 3.2484993997599036e-05,
+      "loss": 0.5247,
+      "step": 8765
+    },
+    {
+      "epoch": 11.22048,
+      "grad_norm": 0.9032994508743286,
+      "learning_rate": 3.2482993197278915e-05,
+      "loss": 0.5577,
+      "step": 8766
+    },
+    {
+      "epoch": 11.22176,
+      "grad_norm": 0.934908390045166,
+      "learning_rate": 3.2480992396958787e-05,
+      "loss": 0.6034,
+      "step": 8767
+    },
+    {
+      "epoch": 11.22304,
+      "grad_norm": 0.8706194758415222,
+      "learning_rate": 3.247899159663866e-05,
+      "loss": 0.5512,
+      "step": 8768
+    },
+    {
+      "epoch": 11.22432,
+      "grad_norm": 0.8422589302062988,
+      "learning_rate": 3.247699079631853e-05,
+      "loss": 0.5348,
+      "step": 8769
+    },
+    {
+      "epoch": 11.2256,
+      "grad_norm": 0.8409394025802612,
+      "learning_rate": 3.24749899959984e-05,
+      "loss": 0.545,
+      "step": 8770
+    },
+    {
+      "epoch": 11.22688,
+      "grad_norm": 0.9218395948410034,
+      "learning_rate": 3.2472989195678274e-05,
+      "loss": 0.6042,
+      "step": 8771
+    },
+    {
+      "epoch": 11.22816,
+      "grad_norm": 0.8771111965179443,
+      "learning_rate": 3.2470988395358146e-05,
+      "loss": 0.5602,
+      "step": 8772
+    },
+    {
+      "epoch": 11.22944,
+      "grad_norm": 0.8485012054443359,
+      "learning_rate": 3.246898759503801e-05,
+      "loss": 0.5159,
+      "step": 8773
+    },
+    {
+      "epoch": 11.23072,
+      "grad_norm": 0.9108744263648987,
+      "learning_rate": 3.246698679471789e-05,
+      "loss": 0.5771,
+      "step": 8774
+    },
+    {
+      "epoch": 11.232,
+      "grad_norm": 0.8793061971664429,
+      "learning_rate": 3.246498599439776e-05,
+      "loss": 0.5697,
+      "step": 8775
+    },
+    {
+      "epoch": 11.23328,
+      "grad_norm": 0.8831557631492615,
+      "learning_rate": 3.246298519407763e-05,
+      "loss": 0.5513,
+      "step": 8776
+    },
+    {
+      "epoch": 11.23456,
+      "grad_norm": 0.9184591770172119,
+      "learning_rate": 3.2460984393757505e-05,
+      "loss": 0.5967,
+      "step": 8777
+    },
+    {
+      "epoch": 11.23584,
+      "grad_norm": 0.9362067580223083,
+      "learning_rate": 3.245898359343738e-05,
+      "loss": 0.5762,
+      "step": 8778
+    },
+    {
+      "epoch": 11.23712,
+      "grad_norm": 0.9164170026779175,
+      "learning_rate": 3.245698279311725e-05,
+      "loss": 0.5353,
+      "step": 8779
+    },
+    {
+      "epoch": 11.2384,
+      "grad_norm": 0.878761351108551,
+      "learning_rate": 3.245498199279712e-05,
+      "loss": 0.5491,
+      "step": 8780
+    },
+    {
+      "epoch": 11.23968,
+      "grad_norm": 0.828332245349884,
+      "learning_rate": 3.245298119247699e-05,
+      "loss": 0.505,
+      "step": 8781
+    },
+    {
+      "epoch": 11.24096,
+      "grad_norm": 0.9329909682273865,
+      "learning_rate": 3.2450980392156864e-05,
+      "loss": 0.5247,
+      "step": 8782
+    },
+    {
+      "epoch": 11.24224,
+      "grad_norm": 0.9171594977378845,
+      "learning_rate": 3.2448979591836736e-05,
+      "loss": 0.5647,
+      "step": 8783
+    },
+    {
+      "epoch": 11.24352,
+      "grad_norm": 0.829479455947876,
+      "learning_rate": 3.244697879151661e-05,
+      "loss": 0.4752,
+      "step": 8784
+    },
+    {
+      "epoch": 11.2448,
+      "grad_norm": 0.8708236813545227,
+      "learning_rate": 3.244497799119648e-05,
+      "loss": 0.5795,
+      "step": 8785
+    },
+    {
+      "epoch": 11.24608,
+      "grad_norm": 0.8896415829658508,
+      "learning_rate": 3.244297719087635e-05,
+      "loss": 0.5318,
+      "step": 8786
+    },
+    {
+      "epoch": 11.24736,
+      "grad_norm": 0.9375750422477722,
+      "learning_rate": 3.2440976390556224e-05,
+      "loss": 0.612,
+      "step": 8787
+    },
+    {
+      "epoch": 11.24864,
+      "grad_norm": 0.9271597862243652,
+      "learning_rate": 3.2438975590236096e-05,
+      "loss": 0.5192,
+      "step": 8788
+    },
+    {
+      "epoch": 11.24992,
+      "grad_norm": 0.894335925579071,
+      "learning_rate": 3.243697478991597e-05,
+      "loss": 0.6151,
+      "step": 8789
+    },
+    {
+      "epoch": 11.2512,
+      "grad_norm": 0.9000656008720398,
+      "learning_rate": 3.243497398959584e-05,
+      "loss": 0.5774,
+      "step": 8790
+    },
+    {
+      "epoch": 11.25248,
+      "grad_norm": 0.8798941969871521,
+      "learning_rate": 3.243297318927571e-05,
+      "loss": 0.5594,
+      "step": 8791
+    },
+    {
+      "epoch": 11.25376,
+      "grad_norm": 0.8531872034072876,
+      "learning_rate": 3.243097238895558e-05,
+      "loss": 0.5439,
+      "step": 8792
+    },
+    {
+      "epoch": 11.25504,
+      "grad_norm": 0.9348339438438416,
+      "learning_rate": 3.2428971588635455e-05,
+      "loss": 0.5349,
+      "step": 8793
+    },
+    {
+      "epoch": 11.25632,
+      "grad_norm": 0.8829602003097534,
+      "learning_rate": 3.242697078831533e-05,
+      "loss": 0.5285,
+      "step": 8794
+    },
+    {
+      "epoch": 11.2576,
+      "grad_norm": 0.9161067605018616,
+      "learning_rate": 3.24249699879952e-05,
+      "loss": 0.5868,
+      "step": 8795
+    },
+    {
+      "epoch": 11.25888,
+      "grad_norm": 0.8983628153800964,
+      "learning_rate": 3.242296918767507e-05,
+      "loss": 0.5213,
+      "step": 8796
+    },
+    {
+      "epoch": 11.26016,
+      "grad_norm": 0.9068379998207092,
+      "learning_rate": 3.242096838735494e-05,
+      "loss": 0.5583,
+      "step": 8797
+    },
+    {
+      "epoch": 11.26144,
+      "grad_norm": 0.9481262564659119,
+      "learning_rate": 3.2418967587034814e-05,
+      "loss": 0.5554,
+      "step": 8798
+    },
+    {
+      "epoch": 11.26272,
+      "grad_norm": 0.8567753434181213,
+      "learning_rate": 3.2416966786714686e-05,
+      "loss": 0.4905,
+      "step": 8799
+    },
+    {
+      "epoch": 11.264,
+      "grad_norm": 0.9208746552467346,
+      "learning_rate": 3.241496598639456e-05,
+      "loss": 0.5791,
+      "step": 8800
+    },
+    {
+      "epoch": 11.26528,
+      "grad_norm": 0.9474934935569763,
+      "learning_rate": 3.241296518607443e-05,
+      "loss": 0.5849,
+      "step": 8801
+    },
+    {
+      "epoch": 11.26656,
+      "grad_norm": 0.9702956676483154,
+      "learning_rate": 3.241096438575431e-05,
+      "loss": 0.62,
+      "step": 8802
+    },
+    {
+      "epoch": 11.26784,
+      "grad_norm": 0.8868454694747925,
+      "learning_rate": 3.2408963585434173e-05,
+      "loss": 0.5779,
+      "step": 8803
+    },
+    {
+      "epoch": 11.269120000000001,
+      "grad_norm": 0.9131936430931091,
+      "learning_rate": 3.2406962785114045e-05,
+      "loss": 0.5787,
+      "step": 8804
+    },
+    {
+      "epoch": 11.2704,
+      "grad_norm": 0.8842905163764954,
+      "learning_rate": 3.240496198479392e-05,
+      "loss": 0.5514,
+      "step": 8805
+    },
+    {
+      "epoch": 11.27168,
+      "grad_norm": 0.8582538366317749,
+      "learning_rate": 3.240296118447379e-05,
+      "loss": 0.5355,
+      "step": 8806
+    },
+    {
+      "epoch": 11.27296,
+      "grad_norm": 0.8297027945518494,
+      "learning_rate": 3.240096038415366e-05,
+      "loss": 0.5481,
+      "step": 8807
+    },
+    {
+      "epoch": 11.27424,
+      "grad_norm": 0.9267979264259338,
+      "learning_rate": 3.239895958383353e-05,
+      "loss": 0.5676,
+      "step": 8808
+    },
+    {
+      "epoch": 11.27552,
+      "grad_norm": 0.9307857155799866,
+      "learning_rate": 3.239695878351341e-05,
+      "loss": 0.5526,
+      "step": 8809
+    },
+    {
+      "epoch": 11.2768,
+      "grad_norm": 0.9614000916481018,
+      "learning_rate": 3.239495798319328e-05,
+      "loss": 0.583,
+      "step": 8810
+    },
+    {
+      "epoch": 11.27808,
+      "grad_norm": 0.8482030630111694,
+      "learning_rate": 3.239295718287315e-05,
+      "loss": 0.5473,
+      "step": 8811
+    },
+    {
+      "epoch": 11.27936,
+      "grad_norm": 0.8771170377731323,
+      "learning_rate": 3.239095638255302e-05,
+      "loss": 0.5808,
+      "step": 8812
+    },
+    {
+      "epoch": 11.28064,
+      "grad_norm": 0.8574709296226501,
+      "learning_rate": 3.238895558223289e-05,
+      "loss": 0.55,
+      "step": 8813
+    },
+    {
+      "epoch": 11.28192,
+      "grad_norm": 0.8876118659973145,
+      "learning_rate": 3.2386954781912764e-05,
+      "loss": 0.5451,
+      "step": 8814
+    },
+    {
+      "epoch": 11.2832,
+      "grad_norm": 0.874653697013855,
+      "learning_rate": 3.2384953981592636e-05,
+      "loss": 0.5076,
+      "step": 8815
+    },
+    {
+      "epoch": 11.28448,
+      "grad_norm": 0.9255512356758118,
+      "learning_rate": 3.2382953181272514e-05,
+      "loss": 0.5806,
+      "step": 8816
+    },
+    {
+      "epoch": 11.28576,
+      "grad_norm": 0.9429183602333069,
+      "learning_rate": 3.2380952380952386e-05,
+      "loss": 0.5884,
+      "step": 8817
+    },
+    {
+      "epoch": 11.28704,
+      "grad_norm": 0.9244290590286255,
+      "learning_rate": 3.237895158063226e-05,
+      "loss": 0.5938,
+      "step": 8818
+    },
+    {
+      "epoch": 11.28832,
+      "grad_norm": 0.8580917716026306,
+      "learning_rate": 3.237695078031212e-05,
+      "loss": 0.55,
+      "step": 8819
+    },
+    {
+      "epoch": 11.2896,
+      "grad_norm": 0.9350975155830383,
+      "learning_rate": 3.2374949979991995e-05,
+      "loss": 0.5704,
+      "step": 8820
+    },
+    {
+      "epoch": 11.29088,
+      "grad_norm": 0.913230836391449,
+      "learning_rate": 3.237294917967187e-05,
+      "loss": 0.5657,
+      "step": 8821
+    },
+    {
+      "epoch": 11.292159999999999,
+      "grad_norm": 0.8917736411094666,
+      "learning_rate": 3.237094837935174e-05,
+      "loss": 0.5393,
+      "step": 8822
+    },
+    {
+      "epoch": 11.29344,
+      "grad_norm": 0.8485799431800842,
+      "learning_rate": 3.236894757903162e-05,
+      "loss": 0.5264,
+      "step": 8823
+    },
+    {
+      "epoch": 11.29472,
+      "grad_norm": 0.8522817492485046,
+      "learning_rate": 3.236694677871149e-05,
+      "loss": 0.4902,
+      "step": 8824
+    },
+    {
+      "epoch": 11.296,
+      "grad_norm": 0.8695195317268372,
+      "learning_rate": 3.236494597839136e-05,
+      "loss": 0.5453,
+      "step": 8825
+    },
+    {
+      "epoch": 11.29728,
+      "grad_norm": 0.9237691164016724,
+      "learning_rate": 3.236294517807123e-05,
+      "loss": 0.5709,
+      "step": 8826
+    },
+    {
+      "epoch": 11.29856,
+      "grad_norm": 0.9328168034553528,
+      "learning_rate": 3.23609443777511e-05,
+      "loss": 0.5593,
+      "step": 8827
+    },
+    {
+      "epoch": 11.29984,
+      "grad_norm": 0.8827159404754639,
+      "learning_rate": 3.235894357743097e-05,
+      "loss": 0.528,
+      "step": 8828
+    },
+    {
+      "epoch": 11.30112,
+      "grad_norm": 0.9112002849578857,
+      "learning_rate": 3.235694277711084e-05,
+      "loss": 0.557,
+      "step": 8829
+    },
+    {
+      "epoch": 11.3024,
+      "grad_norm": 0.8474133610725403,
+      "learning_rate": 3.235494197679072e-05,
+      "loss": 0.5448,
+      "step": 8830
+    },
+    {
+      "epoch": 11.30368,
+      "grad_norm": 0.9159875512123108,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.5638,
+      "step": 8831
+    },
+    {
+      "epoch": 11.30496,
+      "grad_norm": 0.8455161452293396,
+      "learning_rate": 3.2350940376150464e-05,
+      "loss": 0.5317,
+      "step": 8832
+    },
+    {
+      "epoch": 11.30624,
+      "grad_norm": 0.8944229483604431,
+      "learning_rate": 3.2348939575830336e-05,
+      "loss": 0.5455,
+      "step": 8833
+    },
+    {
+      "epoch": 11.30752,
+      "grad_norm": 0.9140881896018982,
+      "learning_rate": 3.234693877551021e-05,
+      "loss": 0.565,
+      "step": 8834
+    },
+    {
+      "epoch": 11.3088,
+      "grad_norm": 0.9187982082366943,
+      "learning_rate": 3.234493797519007e-05,
+      "loss": 0.5593,
+      "step": 8835
+    },
+    {
+      "epoch": 11.31008,
+      "grad_norm": 0.9055234789848328,
+      "learning_rate": 3.2342937174869945e-05,
+      "loss": 0.5166,
+      "step": 8836
+    },
+    {
+      "epoch": 11.31136,
+      "grad_norm": 0.9232755303382874,
+      "learning_rate": 3.234093637454982e-05,
+      "loss": 0.5993,
+      "step": 8837
+    },
+    {
+      "epoch": 11.31264,
+      "grad_norm": 0.8765431046485901,
+      "learning_rate": 3.2338935574229695e-05,
+      "loss": 0.5717,
+      "step": 8838
+    },
+    {
+      "epoch": 11.31392,
+      "grad_norm": 0.848947286605835,
+      "learning_rate": 3.233693477390957e-05,
+      "loss": 0.528,
+      "step": 8839
+    },
+    {
+      "epoch": 11.3152,
+      "grad_norm": 0.9066827297210693,
+      "learning_rate": 3.233493397358944e-05,
+      "loss": 0.5882,
+      "step": 8840
+    },
+    {
+      "epoch": 11.31648,
+      "grad_norm": 0.9069925546646118,
+      "learning_rate": 3.233293317326931e-05,
+      "loss": 0.5502,
+      "step": 8841
+    },
+    {
+      "epoch": 11.31776,
+      "grad_norm": 0.9095439314842224,
+      "learning_rate": 3.233093237294918e-05,
+      "loss": 0.5638,
+      "step": 8842
+    },
+    {
+      "epoch": 11.31904,
+      "grad_norm": 0.9278164505958557,
+      "learning_rate": 3.232893157262905e-05,
+      "loss": 0.5834,
+      "step": 8843
+    },
+    {
+      "epoch": 11.32032,
+      "grad_norm": 0.8815664052963257,
+      "learning_rate": 3.2326930772308926e-05,
+      "loss": 0.5516,
+      "step": 8844
+    },
+    {
+      "epoch": 11.3216,
+      "grad_norm": 0.8748176097869873,
+      "learning_rate": 3.23249299719888e-05,
+      "loss": 0.5715,
+      "step": 8845
+    },
+    {
+      "epoch": 11.32288,
+      "grad_norm": 0.8980875015258789,
+      "learning_rate": 3.232292917166867e-05,
+      "loss": 0.5218,
+      "step": 8846
+    },
+    {
+      "epoch": 11.32416,
+      "grad_norm": 0.9095767736434937,
+      "learning_rate": 3.232092837134854e-05,
+      "loss": 0.5353,
+      "step": 8847
+    },
+    {
+      "epoch": 11.32544,
+      "grad_norm": 0.9545304775238037,
+      "learning_rate": 3.2318927571028414e-05,
+      "loss": 0.5816,
+      "step": 8848
+    },
+    {
+      "epoch": 11.32672,
+      "grad_norm": 0.8944869637489319,
+      "learning_rate": 3.2316926770708286e-05,
+      "loss": 0.5624,
+      "step": 8849
+    },
+    {
+      "epoch": 11.328,
+      "grad_norm": 0.891927182674408,
+      "learning_rate": 3.231492597038816e-05,
+      "loss": 0.523,
+      "step": 8850
+    },
+    {
+      "epoch": 11.32928,
+      "grad_norm": 0.9130933284759521,
+      "learning_rate": 3.231292517006803e-05,
+      "loss": 0.6285,
+      "step": 8851
+    },
+    {
+      "epoch": 11.33056,
+      "grad_norm": 0.8824868202209473,
+      "learning_rate": 3.23109243697479e-05,
+      "loss": 0.5385,
+      "step": 8852
+    },
+    {
+      "epoch": 11.33184,
+      "grad_norm": 0.8892396688461304,
+      "learning_rate": 3.230892356942777e-05,
+      "loss": 0.5261,
+      "step": 8853
+    },
+    {
+      "epoch": 11.33312,
+      "grad_norm": 0.9578300714492798,
+      "learning_rate": 3.2306922769107645e-05,
+      "loss": 0.635,
+      "step": 8854
+    },
+    {
+      "epoch": 11.3344,
+      "grad_norm": 0.9021438956260681,
+      "learning_rate": 3.230492196878752e-05,
+      "loss": 0.5346,
+      "step": 8855
+    },
+    {
+      "epoch": 11.33568,
+      "grad_norm": 0.8745192289352417,
+      "learning_rate": 3.230292116846739e-05,
+      "loss": 0.5481,
+      "step": 8856
+    },
+    {
+      "epoch": 11.33696,
+      "grad_norm": 0.8894162178039551,
+      "learning_rate": 3.230092036814726e-05,
+      "loss": 0.558,
+      "step": 8857
+    },
+    {
+      "epoch": 11.33824,
+      "grad_norm": 0.8988181352615356,
+      "learning_rate": 3.229891956782713e-05,
+      "loss": 0.5633,
+      "step": 8858
+    },
+    {
+      "epoch": 11.33952,
+      "grad_norm": 0.8719711899757385,
+      "learning_rate": 3.2296918767507004e-05,
+      "loss": 0.5066,
+      "step": 8859
+    },
+    {
+      "epoch": 11.3408,
+      "grad_norm": 0.8456899523735046,
+      "learning_rate": 3.2294917967186876e-05,
+      "loss": 0.5139,
+      "step": 8860
+    },
+    {
+      "epoch": 11.34208,
+      "grad_norm": 0.8919529914855957,
+      "learning_rate": 3.229291716686675e-05,
+      "loss": 0.5203,
+      "step": 8861
+    },
+    {
+      "epoch": 11.34336,
+      "grad_norm": 0.8657916188240051,
+      "learning_rate": 3.229091636654662e-05,
+      "loss": 0.5298,
+      "step": 8862
+    },
+    {
+      "epoch": 11.34464,
+      "grad_norm": 0.8953869938850403,
+      "learning_rate": 3.228891556622649e-05,
+      "loss": 0.5217,
+      "step": 8863
+    },
+    {
+      "epoch": 11.34592,
+      "grad_norm": 0.9137585759162903,
+      "learning_rate": 3.2286914765906363e-05,
+      "loss": 0.5738,
+      "step": 8864
+    },
+    {
+      "epoch": 11.3472,
+      "grad_norm": 0.8777027726173401,
+      "learning_rate": 3.228491396558624e-05,
+      "loss": 0.5342,
+      "step": 8865
+    },
+    {
+      "epoch": 11.34848,
+      "grad_norm": 0.9374188780784607,
+      "learning_rate": 3.228291316526611e-05,
+      "loss": 0.6148,
+      "step": 8866
+    },
+    {
+      "epoch": 11.34976,
+      "grad_norm": 0.8601318001747131,
+      "learning_rate": 3.228091236494598e-05,
+      "loss": 0.5121,
+      "step": 8867
+    },
+    {
+      "epoch": 11.35104,
+      "grad_norm": 0.904672384262085,
+      "learning_rate": 3.227891156462585e-05,
+      "loss": 0.5718,
+      "step": 8868
+    },
+    {
+      "epoch": 11.35232,
+      "grad_norm": 0.9223089218139648,
+      "learning_rate": 3.227691076430572e-05,
+      "loss": 0.5815,
+      "step": 8869
+    },
+    {
+      "epoch": 11.3536,
+      "grad_norm": 0.8737932443618774,
+      "learning_rate": 3.2274909963985595e-05,
+      "loss": 0.5285,
+      "step": 8870
+    },
+    {
+      "epoch": 11.35488,
+      "grad_norm": 0.9372346997261047,
+      "learning_rate": 3.2272909163665466e-05,
+      "loss": 0.5748,
+      "step": 8871
+    },
+    {
+      "epoch": 11.35616,
+      "grad_norm": 0.9131039977073669,
+      "learning_rate": 3.2270908363345345e-05,
+      "loss": 0.5606,
+      "step": 8872
+    },
+    {
+      "epoch": 11.35744,
+      "grad_norm": 0.9296916127204895,
+      "learning_rate": 3.226890756302522e-05,
+      "loss": 0.576,
+      "step": 8873
+    },
+    {
+      "epoch": 11.35872,
+      "grad_norm": 0.9245988130569458,
+      "learning_rate": 3.226690676270508e-05,
+      "loss": 0.5462,
+      "step": 8874
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 0.934669017791748,
+      "learning_rate": 3.2264905962384954e-05,
+      "loss": 0.555,
+      "step": 8875
+    },
+    {
+      "epoch": 11.36128,
+      "grad_norm": 0.8804442882537842,
+      "learning_rate": 3.2262905162064826e-05,
+      "loss": 0.5157,
+      "step": 8876
+    },
+    {
+      "epoch": 11.36256,
+      "grad_norm": 0.8725820779800415,
+      "learning_rate": 3.22609043617447e-05,
+      "loss": 0.5829,
+      "step": 8877
+    },
+    {
+      "epoch": 11.36384,
+      "grad_norm": 0.9546533823013306,
+      "learning_rate": 3.225890356142457e-05,
+      "loss": 0.5668,
+      "step": 8878
+    },
+    {
+      "epoch": 11.36512,
+      "grad_norm": 0.9215765595436096,
+      "learning_rate": 3.225690276110445e-05,
+      "loss": 0.546,
+      "step": 8879
+    },
+    {
+      "epoch": 11.3664,
+      "grad_norm": 0.9249255061149597,
+      "learning_rate": 3.225490196078432e-05,
+      "loss": 0.5562,
+      "step": 8880
+    },
+    {
+      "epoch": 11.36768,
+      "grad_norm": 0.9244142174720764,
+      "learning_rate": 3.225290116046419e-05,
+      "loss": 0.54,
+      "step": 8881
+    },
+    {
+      "epoch": 11.36896,
+      "grad_norm": 0.9065011143684387,
+      "learning_rate": 3.225090036014406e-05,
+      "loss": 0.5555,
+      "step": 8882
+    },
+    {
+      "epoch": 11.37024,
+      "grad_norm": 0.8759673237800598,
+      "learning_rate": 3.224889955982393e-05,
+      "loss": 0.5506,
+      "step": 8883
+    },
+    {
+      "epoch": 11.37152,
+      "grad_norm": 0.9149131774902344,
+      "learning_rate": 3.22468987595038e-05,
+      "loss": 0.5635,
+      "step": 8884
+    },
+    {
+      "epoch": 11.3728,
+      "grad_norm": 0.9005359411239624,
+      "learning_rate": 3.224489795918367e-05,
+      "loss": 0.5289,
+      "step": 8885
+    },
+    {
+      "epoch": 11.37408,
+      "grad_norm": 0.9132285714149475,
+      "learning_rate": 3.2242897158863544e-05,
+      "loss": 0.5287,
+      "step": 8886
+    },
+    {
+      "epoch": 11.37536,
+      "grad_norm": 0.9003539681434631,
+      "learning_rate": 3.224089635854342e-05,
+      "loss": 0.521,
+      "step": 8887
+    },
+    {
+      "epoch": 11.37664,
+      "grad_norm": 0.8908035755157471,
+      "learning_rate": 3.2238895558223295e-05,
+      "loss": 0.5668,
+      "step": 8888
+    },
+    {
+      "epoch": 11.37792,
+      "grad_norm": 0.9106180667877197,
+      "learning_rate": 3.223689475790317e-05,
+      "loss": 0.5502,
+      "step": 8889
+    },
+    {
+      "epoch": 11.3792,
+      "grad_norm": 0.9311418533325195,
+      "learning_rate": 3.223489395758303e-05,
+      "loss": 0.5703,
+      "step": 8890
+    },
+    {
+      "epoch": 11.38048,
+      "grad_norm": 0.8639375567436218,
+      "learning_rate": 3.2232893157262904e-05,
+      "loss": 0.5366,
+      "step": 8891
+    },
+    {
+      "epoch": 11.38176,
+      "grad_norm": 0.903183102607727,
+      "learning_rate": 3.2230892356942775e-05,
+      "loss": 0.5783,
+      "step": 8892
+    },
+    {
+      "epoch": 11.38304,
+      "grad_norm": 0.8958691954612732,
+      "learning_rate": 3.222889155662265e-05,
+      "loss": 0.5197,
+      "step": 8893
+    },
+    {
+      "epoch": 11.38432,
+      "grad_norm": 0.95194411277771,
+      "learning_rate": 3.2226890756302526e-05,
+      "loss": 0.6212,
+      "step": 8894
+    },
+    {
+      "epoch": 11.3856,
+      "grad_norm": 0.8696224689483643,
+      "learning_rate": 3.22248899559824e-05,
+      "loss": 0.5376,
+      "step": 8895
+    },
+    {
+      "epoch": 11.38688,
+      "grad_norm": 0.925094485282898,
+      "learning_rate": 3.222288915566227e-05,
+      "loss": 0.5839,
+      "step": 8896
+    },
+    {
+      "epoch": 11.38816,
+      "grad_norm": 0.8680374026298523,
+      "learning_rate": 3.222088835534214e-05,
+      "loss": 0.4927,
+      "step": 8897
+    },
+    {
+      "epoch": 11.38944,
+      "grad_norm": 0.9025076031684875,
+      "learning_rate": 3.2218887555022007e-05,
+      "loss": 0.5683,
+      "step": 8898
+    },
+    {
+      "epoch": 11.39072,
+      "grad_norm": 0.8952671885490417,
+      "learning_rate": 3.221688675470188e-05,
+      "loss": 0.5648,
+      "step": 8899
+    },
+    {
+      "epoch": 11.392,
+      "grad_norm": 0.916085422039032,
+      "learning_rate": 3.221488595438175e-05,
+      "loss": 0.5825,
+      "step": 8900
+    },
+    {
+      "epoch": 11.39328,
+      "grad_norm": 0.8768932223320007,
+      "learning_rate": 3.221288515406163e-05,
+      "loss": 0.5205,
+      "step": 8901
+    },
+    {
+      "epoch": 11.39456,
+      "grad_norm": 0.912806510925293,
+      "learning_rate": 3.22108843537415e-05,
+      "loss": 0.5078,
+      "step": 8902
+    },
+    {
+      "epoch": 11.39584,
+      "grad_norm": 0.9842191338539124,
+      "learning_rate": 3.220888355342137e-05,
+      "loss": 0.6436,
+      "step": 8903
+    },
+    {
+      "epoch": 11.39712,
+      "grad_norm": 0.9304628372192383,
+      "learning_rate": 3.2206882753101244e-05,
+      "loss": 0.5789,
+      "step": 8904
+    },
+    {
+      "epoch": 11.3984,
+      "grad_norm": 0.9298278093338013,
+      "learning_rate": 3.2204881952781116e-05,
+      "loss": 0.575,
+      "step": 8905
+    },
+    {
+      "epoch": 11.39968,
+      "grad_norm": 0.9112979173660278,
+      "learning_rate": 3.220288115246098e-05,
+      "loss": 0.5536,
+      "step": 8906
+    },
+    {
+      "epoch": 11.40096,
+      "grad_norm": 0.9044650793075562,
+      "learning_rate": 3.220088035214085e-05,
+      "loss": 0.5697,
+      "step": 8907
+    },
+    {
+      "epoch": 11.40224,
+      "grad_norm": 0.90861576795578,
+      "learning_rate": 3.219887955182073e-05,
+      "loss": 0.5748,
+      "step": 8908
+    },
+    {
+      "epoch": 11.40352,
+      "grad_norm": 0.8984891176223755,
+      "learning_rate": 3.2196878751500604e-05,
+      "loss": 0.5191,
+      "step": 8909
+    },
+    {
+      "epoch": 11.4048,
+      "grad_norm": 0.9326877593994141,
+      "learning_rate": 3.2194877951180476e-05,
+      "loss": 0.5591,
+      "step": 8910
+    },
+    {
+      "epoch": 11.40608,
+      "grad_norm": 0.9329106211662292,
+      "learning_rate": 3.219287715086035e-05,
+      "loss": 0.5585,
+      "step": 8911
+    },
+    {
+      "epoch": 11.40736,
+      "grad_norm": 0.917129397392273,
+      "learning_rate": 3.219087635054022e-05,
+      "loss": 0.5843,
+      "step": 8912
+    },
+    {
+      "epoch": 11.40864,
+      "grad_norm": 0.8364509344100952,
+      "learning_rate": 3.218887555022009e-05,
+      "loss": 0.5127,
+      "step": 8913
+    },
+    {
+      "epoch": 11.40992,
+      "grad_norm": 0.8630757927894592,
+      "learning_rate": 3.2186874749899956e-05,
+      "loss": 0.5589,
+      "step": 8914
+    },
+    {
+      "epoch": 11.411200000000001,
+      "grad_norm": 0.9007800221443176,
+      "learning_rate": 3.2184873949579835e-05,
+      "loss": 0.6069,
+      "step": 8915
+    },
+    {
+      "epoch": 11.41248,
+      "grad_norm": 0.9004875421524048,
+      "learning_rate": 3.218287314925971e-05,
+      "loss": 0.5776,
+      "step": 8916
+    },
+    {
+      "epoch": 11.41376,
+      "grad_norm": 0.8987116813659668,
+      "learning_rate": 3.218087234893958e-05,
+      "loss": 0.5216,
+      "step": 8917
+    },
+    {
+      "epoch": 11.41504,
+      "grad_norm": 0.8875414133071899,
+      "learning_rate": 3.217887154861945e-05,
+      "loss": 0.5412,
+      "step": 8918
+    },
+    {
+      "epoch": 11.41632,
+      "grad_norm": 0.8848508596420288,
+      "learning_rate": 3.217687074829932e-05,
+      "loss": 0.5271,
+      "step": 8919
+    },
+    {
+      "epoch": 11.4176,
+      "grad_norm": 0.9203560948371887,
+      "learning_rate": 3.2174869947979194e-05,
+      "loss": 0.5973,
+      "step": 8920
+    },
+    {
+      "epoch": 11.41888,
+      "grad_norm": 0.9155365228652954,
+      "learning_rate": 3.2172869147659066e-05,
+      "loss": 0.536,
+      "step": 8921
+    },
+    {
+      "epoch": 11.42016,
+      "grad_norm": 0.9187197685241699,
+      "learning_rate": 3.217086834733894e-05,
+      "loss": 0.5452,
+      "step": 8922
+    },
+    {
+      "epoch": 11.42144,
+      "grad_norm": 0.927579939365387,
+      "learning_rate": 3.216886754701881e-05,
+      "loss": 0.5534,
+      "step": 8923
+    },
+    {
+      "epoch": 11.42272,
+      "grad_norm": 0.9490095376968384,
+      "learning_rate": 3.216686674669868e-05,
+      "loss": 0.5418,
+      "step": 8924
+    },
+    {
+      "epoch": 11.424,
+      "grad_norm": 0.9742210507392883,
+      "learning_rate": 3.2164865946378553e-05,
+      "loss": 0.5727,
+      "step": 8925
+    },
+    {
+      "epoch": 11.42528,
+      "grad_norm": 0.9028719067573547,
+      "learning_rate": 3.2162865146058425e-05,
+      "loss": 0.5383,
+      "step": 8926
+    },
+    {
+      "epoch": 11.42656,
+      "grad_norm": 0.9693518280982971,
+      "learning_rate": 3.21608643457383e-05,
+      "loss": 0.6059,
+      "step": 8927
+    },
+    {
+      "epoch": 11.42784,
+      "grad_norm": 0.866688072681427,
+      "learning_rate": 3.215886354541817e-05,
+      "loss": 0.5364,
+      "step": 8928
+    },
+    {
+      "epoch": 11.42912,
+      "grad_norm": 0.9084386229515076,
+      "learning_rate": 3.215686274509804e-05,
+      "loss": 0.5616,
+      "step": 8929
+    },
+    {
+      "epoch": 11.4304,
+      "grad_norm": 0.9357685446739197,
+      "learning_rate": 3.215486194477791e-05,
+      "loss": 0.5707,
+      "step": 8930
+    },
+    {
+      "epoch": 11.43168,
+      "grad_norm": 0.8925808072090149,
+      "learning_rate": 3.2152861144457785e-05,
+      "loss": 0.5594,
+      "step": 8931
+    },
+    {
+      "epoch": 11.43296,
+      "grad_norm": 0.9576624035835266,
+      "learning_rate": 3.2150860344137656e-05,
+      "loss": 0.5892,
+      "step": 8932
+    },
+    {
+      "epoch": 11.43424,
+      "grad_norm": 0.8737156391143799,
+      "learning_rate": 3.214885954381753e-05,
+      "loss": 0.54,
+      "step": 8933
+    },
+    {
+      "epoch": 11.43552,
+      "grad_norm": 0.9288681745529175,
+      "learning_rate": 3.21468587434974e-05,
+      "loss": 0.5832,
+      "step": 8934
+    },
+    {
+      "epoch": 11.4368,
+      "grad_norm": 0.9747262597084045,
+      "learning_rate": 3.214485794317727e-05,
+      "loss": 0.5935,
+      "step": 8935
+    },
+    {
+      "epoch": 11.43808,
+      "grad_norm": 0.8848868012428284,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 0.5971,
+      "step": 8936
+    },
+    {
+      "epoch": 11.43936,
+      "grad_norm": 0.9349198341369629,
+      "learning_rate": 3.2140856342537016e-05,
+      "loss": 0.5941,
+      "step": 8937
+    },
+    {
+      "epoch": 11.44064,
+      "grad_norm": 0.8732218146324158,
+      "learning_rate": 3.213885554221689e-05,
+      "loss": 0.5509,
+      "step": 8938
+    },
+    {
+      "epoch": 11.44192,
+      "grad_norm": 0.8781768083572388,
+      "learning_rate": 3.213685474189676e-05,
+      "loss": 0.569,
+      "step": 8939
+    },
+    {
+      "epoch": 11.4432,
+      "grad_norm": 0.8788825869560242,
+      "learning_rate": 3.213485394157663e-05,
+      "loss": 0.5946,
+      "step": 8940
+    },
+    {
+      "epoch": 11.44448,
+      "grad_norm": 0.8523260354995728,
+      "learning_rate": 3.21328531412565e-05,
+      "loss": 0.5232,
+      "step": 8941
+    },
+    {
+      "epoch": 11.44576,
+      "grad_norm": 0.9410714507102966,
+      "learning_rate": 3.2130852340936375e-05,
+      "loss": 0.5787,
+      "step": 8942
+    },
+    {
+      "epoch": 11.44704,
+      "grad_norm": 0.9099283814430237,
+      "learning_rate": 3.2128851540616254e-05,
+      "loss": 0.5687,
+      "step": 8943
+    },
+    {
+      "epoch": 11.44832,
+      "grad_norm": 0.9799555540084839,
+      "learning_rate": 3.212685074029612e-05,
+      "loss": 0.586,
+      "step": 8944
+    },
+    {
+      "epoch": 11.4496,
+      "grad_norm": 0.983367383480072,
+      "learning_rate": 3.212484993997599e-05,
+      "loss": 0.5625,
+      "step": 8945
+    },
+    {
+      "epoch": 11.45088,
+      "grad_norm": 0.9114019870758057,
+      "learning_rate": 3.212284913965586e-05,
+      "loss": 0.5501,
+      "step": 8946
+    },
+    {
+      "epoch": 11.45216,
+      "grad_norm": 0.964597761631012,
+      "learning_rate": 3.2120848339335734e-05,
+      "loss": 0.584,
+      "step": 8947
+    },
+    {
+      "epoch": 11.45344,
+      "grad_norm": 0.8593063950538635,
+      "learning_rate": 3.2118847539015606e-05,
+      "loss": 0.5178,
+      "step": 8948
+    },
+    {
+      "epoch": 11.45472,
+      "grad_norm": 0.8868183493614197,
+      "learning_rate": 3.211684673869548e-05,
+      "loss": 0.5591,
+      "step": 8949
+    },
+    {
+      "epoch": 11.456,
+      "grad_norm": 0.8840005397796631,
+      "learning_rate": 3.211484593837536e-05,
+      "loss": 0.5101,
+      "step": 8950
+    },
+    {
+      "epoch": 11.45728,
+      "grad_norm": 0.9012945890426636,
+      "learning_rate": 3.211284513805523e-05,
+      "loss": 0.5854,
+      "step": 8951
+    },
+    {
+      "epoch": 11.45856,
+      "grad_norm": 0.8864527940750122,
+      "learning_rate": 3.2110844337735094e-05,
+      "loss": 0.5846,
+      "step": 8952
+    },
+    {
+      "epoch": 11.45984,
+      "grad_norm": 0.9185726642608643,
+      "learning_rate": 3.2108843537414965e-05,
+      "loss": 0.5771,
+      "step": 8953
+    },
+    {
+      "epoch": 11.46112,
+      "grad_norm": 0.8853892683982849,
+      "learning_rate": 3.210684273709484e-05,
+      "loss": 0.526,
+      "step": 8954
+    },
+    {
+      "epoch": 11.4624,
+      "grad_norm": 0.9288807511329651,
+      "learning_rate": 3.210484193677471e-05,
+      "loss": 0.5306,
+      "step": 8955
+    },
+    {
+      "epoch": 11.46368,
+      "grad_norm": 0.9466400742530823,
+      "learning_rate": 3.210284113645458e-05,
+      "loss": 0.606,
+      "step": 8956
+    },
+    {
+      "epoch": 11.46496,
+      "grad_norm": 0.9334494471549988,
+      "learning_rate": 3.210084033613446e-05,
+      "loss": 0.5746,
+      "step": 8957
+    },
+    {
+      "epoch": 11.466239999999999,
+      "grad_norm": 0.9179563522338867,
+      "learning_rate": 3.209883953581433e-05,
+      "loss": 0.5453,
+      "step": 8958
+    },
+    {
+      "epoch": 11.46752,
+      "grad_norm": 0.9171554446220398,
+      "learning_rate": 3.2096838735494203e-05,
+      "loss": 0.5683,
+      "step": 8959
+    },
+    {
+      "epoch": 11.4688,
+      "grad_norm": 0.9348716735839844,
+      "learning_rate": 3.209483793517407e-05,
+      "loss": 0.5502,
+      "step": 8960
+    },
+    {
+      "epoch": 11.47008,
+      "grad_norm": 0.9068107604980469,
+      "learning_rate": 3.209283713485394e-05,
+      "loss": 0.5924,
+      "step": 8961
+    },
+    {
+      "epoch": 11.47136,
+      "grad_norm": 0.9149397611618042,
+      "learning_rate": 3.209083633453381e-05,
+      "loss": 0.5681,
+      "step": 8962
+    },
+    {
+      "epoch": 11.47264,
+      "grad_norm": 0.876530110836029,
+      "learning_rate": 3.2088835534213684e-05,
+      "loss": 0.55,
+      "step": 8963
+    },
+    {
+      "epoch": 11.47392,
+      "grad_norm": 0.8814985752105713,
+      "learning_rate": 3.208683473389356e-05,
+      "loss": 0.5513,
+      "step": 8964
+    },
+    {
+      "epoch": 11.4752,
+      "grad_norm": 0.8350082039833069,
+      "learning_rate": 3.2084833933573435e-05,
+      "loss": 0.5217,
+      "step": 8965
+    },
+    {
+      "epoch": 11.47648,
+      "grad_norm": 0.8814594149589539,
+      "learning_rate": 3.2082833133253306e-05,
+      "loss": 0.5626,
+      "step": 8966
+    },
+    {
+      "epoch": 11.47776,
+      "grad_norm": 0.8578638434410095,
+      "learning_rate": 3.208083233293318e-05,
+      "loss": 0.5671,
+      "step": 8967
+    },
+    {
+      "epoch": 11.47904,
+      "grad_norm": 0.8845701217651367,
+      "learning_rate": 3.207883153261304e-05,
+      "loss": 0.5927,
+      "step": 8968
+    },
+    {
+      "epoch": 11.48032,
+      "grad_norm": 0.8991924524307251,
+      "learning_rate": 3.2076830732292915e-05,
+      "loss": 0.5528,
+      "step": 8969
+    },
+    {
+      "epoch": 11.4816,
+      "grad_norm": 0.898190975189209,
+      "learning_rate": 3.207482993197279e-05,
+      "loss": 0.5988,
+      "step": 8970
+    },
+    {
+      "epoch": 11.48288,
+      "grad_norm": 0.9360696077346802,
+      "learning_rate": 3.2072829131652666e-05,
+      "loss": 0.569,
+      "step": 8971
+    },
+    {
+      "epoch": 11.48416,
+      "grad_norm": 0.845306932926178,
+      "learning_rate": 3.207082833133254e-05,
+      "loss": 0.5489,
+      "step": 8972
+    },
+    {
+      "epoch": 11.48544,
+      "grad_norm": 0.9122028946876526,
+      "learning_rate": 3.206882753101241e-05,
+      "loss": 0.5735,
+      "step": 8973
+    },
+    {
+      "epoch": 11.48672,
+      "grad_norm": 0.9009816646575928,
+      "learning_rate": 3.206682673069228e-05,
+      "loss": 0.5192,
+      "step": 8974
+    },
+    {
+      "epoch": 11.488,
+      "grad_norm": 0.9005442261695862,
+      "learning_rate": 3.206482593037215e-05,
+      "loss": 0.6311,
+      "step": 8975
+    },
+    {
+      "epoch": 11.48928,
+      "grad_norm": 0.931171715259552,
+      "learning_rate": 3.206282513005202e-05,
+      "loss": 0.5831,
+      "step": 8976
+    },
+    {
+      "epoch": 11.49056,
+      "grad_norm": 0.902475893497467,
+      "learning_rate": 3.206082432973189e-05,
+      "loss": 0.5382,
+      "step": 8977
+    },
+    {
+      "epoch": 11.49184,
+      "grad_norm": 0.9351593852043152,
+      "learning_rate": 3.205882352941177e-05,
+      "loss": 0.5586,
+      "step": 8978
+    },
+    {
+      "epoch": 11.49312,
+      "grad_norm": 0.9076836705207825,
+      "learning_rate": 3.205682272909164e-05,
+      "loss": 0.5568,
+      "step": 8979
+    },
+    {
+      "epoch": 11.4944,
+      "grad_norm": 0.8612614274024963,
+      "learning_rate": 3.205482192877151e-05,
+      "loss": 0.531,
+      "step": 8980
+    },
+    {
+      "epoch": 11.49568,
+      "grad_norm": 0.8604639768600464,
+      "learning_rate": 3.2052821128451384e-05,
+      "loss": 0.5539,
+      "step": 8981
+    },
+    {
+      "epoch": 11.49696,
+      "grad_norm": 0.8722749352455139,
+      "learning_rate": 3.2050820328131256e-05,
+      "loss": 0.5436,
+      "step": 8982
+    },
+    {
+      "epoch": 11.49824,
+      "grad_norm": 0.8463775515556335,
+      "learning_rate": 3.204881952781113e-05,
+      "loss": 0.5086,
+      "step": 8983
+    },
+    {
+      "epoch": 11.49952,
+      "grad_norm": 0.9018126130104065,
+      "learning_rate": 3.204681872749099e-05,
+      "loss": 0.6194,
+      "step": 8984
+    },
+    {
+      "epoch": 11.5008,
+      "grad_norm": 0.9355940222740173,
+      "learning_rate": 3.204481792717087e-05,
+      "loss": 0.5751,
+      "step": 8985
+    },
+    {
+      "epoch": 11.50208,
+      "grad_norm": 0.8555790185928345,
+      "learning_rate": 3.2042817126850744e-05,
+      "loss": 0.5685,
+      "step": 8986
+    },
+    {
+      "epoch": 11.50336,
+      "grad_norm": 0.8574039936065674,
+      "learning_rate": 3.2040816326530615e-05,
+      "loss": 0.5265,
+      "step": 8987
+    },
+    {
+      "epoch": 11.50464,
+      "grad_norm": 0.898993730545044,
+      "learning_rate": 3.203881552621049e-05,
+      "loss": 0.5406,
+      "step": 8988
+    },
+    {
+      "epoch": 11.50592,
+      "grad_norm": 0.9504684805870056,
+      "learning_rate": 3.203681472589036e-05,
+      "loss": 0.6053,
+      "step": 8989
+    },
+    {
+      "epoch": 11.5072,
+      "grad_norm": 0.9039958119392395,
+      "learning_rate": 3.203481392557023e-05,
+      "loss": 0.577,
+      "step": 8990
+    },
+    {
+      "epoch": 11.50848,
+      "grad_norm": 0.9396102428436279,
+      "learning_rate": 3.20328131252501e-05,
+      "loss": 0.5763,
+      "step": 8991
+    },
+    {
+      "epoch": 11.50976,
+      "grad_norm": 0.8567662835121155,
+      "learning_rate": 3.203081232492997e-05,
+      "loss": 0.5162,
+      "step": 8992
+    },
+    {
+      "epoch": 11.51104,
+      "grad_norm": 0.9210956692695618,
+      "learning_rate": 3.2028811524609847e-05,
+      "loss": 0.579,
+      "step": 8993
+    },
+    {
+      "epoch": 11.51232,
+      "grad_norm": 0.8853841423988342,
+      "learning_rate": 3.202681072428972e-05,
+      "loss": 0.5236,
+      "step": 8994
+    },
+    {
+      "epoch": 11.5136,
+      "grad_norm": 0.9339121580123901,
+      "learning_rate": 3.202480992396959e-05,
+      "loss": 0.619,
+      "step": 8995
+    },
+    {
+      "epoch": 11.51488,
+      "grad_norm": 0.8573284149169922,
+      "learning_rate": 3.202280912364946e-05,
+      "loss": 0.5583,
+      "step": 8996
+    },
+    {
+      "epoch": 11.51616,
+      "grad_norm": 0.9069136381149292,
+      "learning_rate": 3.2020808323329334e-05,
+      "loss": 0.5256,
+      "step": 8997
+    },
+    {
+      "epoch": 11.51744,
+      "grad_norm": 0.9280446767807007,
+      "learning_rate": 3.2018807523009206e-05,
+      "loss": 0.6084,
+      "step": 8998
+    },
+    {
+      "epoch": 11.51872,
+      "grad_norm": 0.8712737560272217,
+      "learning_rate": 3.201680672268908e-05,
+      "loss": 0.5816,
+      "step": 8999
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 0.8745893836021423,
+      "learning_rate": 3.201480592236895e-05,
+      "loss": 0.5167,
+      "step": 9000
+    },
+    {
+      "epoch": 11.52128,
+      "grad_norm": 0.8918792605400085,
+      "learning_rate": 3.201280512204882e-05,
+      "loss": 0.587,
+      "step": 9001
+    },
+    {
+      "epoch": 11.52256,
+      "grad_norm": 0.9053367972373962,
+      "learning_rate": 3.201080432172869e-05,
+      "loss": 0.5899,
+      "step": 9002
+    },
+    {
+      "epoch": 11.52384,
+      "grad_norm": 0.8964258432388306,
+      "learning_rate": 3.2008803521408565e-05,
+      "loss": 0.5589,
+      "step": 9003
+    },
+    {
+      "epoch": 11.52512,
+      "grad_norm": 0.852827250957489,
+      "learning_rate": 3.200680272108844e-05,
+      "loss": 0.5024,
+      "step": 9004
+    },
+    {
+      "epoch": 11.5264,
+      "grad_norm": 0.8889139890670776,
+      "learning_rate": 3.200480192076831e-05,
+      "loss": 0.5795,
+      "step": 9005
+    },
+    {
+      "epoch": 11.52768,
+      "grad_norm": 0.8449302315711975,
+      "learning_rate": 3.200280112044818e-05,
+      "loss": 0.5439,
+      "step": 9006
+    },
+    {
+      "epoch": 11.52896,
+      "grad_norm": 0.9062175154685974,
+      "learning_rate": 3.200080032012805e-05,
+      "loss": 0.543,
+      "step": 9007
+    },
+    {
+      "epoch": 11.53024,
+      "grad_norm": 0.9764111042022705,
+      "learning_rate": 3.1998799519807924e-05,
+      "loss": 0.5982,
+      "step": 9008
+    },
+    {
+      "epoch": 11.53152,
+      "grad_norm": 0.8889600038528442,
+      "learning_rate": 3.1996798719487796e-05,
+      "loss": 0.5308,
+      "step": 9009
+    },
+    {
+      "epoch": 11.5328,
+      "grad_norm": 0.9224411845207214,
+      "learning_rate": 3.199479791916767e-05,
+      "loss": 0.5809,
+      "step": 9010
+    },
+    {
+      "epoch": 11.53408,
+      "grad_norm": 0.8506020307540894,
+      "learning_rate": 3.199279711884754e-05,
+      "loss": 0.5375,
+      "step": 9011
+    },
+    {
+      "epoch": 11.53536,
+      "grad_norm": 0.8242222666740417,
+      "learning_rate": 3.199079631852741e-05,
+      "loss": 0.5254,
+      "step": 9012
+    },
+    {
+      "epoch": 11.53664,
+      "grad_norm": 0.8847529888153076,
+      "learning_rate": 3.1988795518207284e-05,
+      "loss": 0.5867,
+      "step": 9013
+    },
+    {
+      "epoch": 11.53792,
+      "grad_norm": 0.8491979837417603,
+      "learning_rate": 3.1986794717887156e-05,
+      "loss": 0.5463,
+      "step": 9014
+    },
+    {
+      "epoch": 11.5392,
+      "grad_norm": 0.8366856575012207,
+      "learning_rate": 3.198479391756703e-05,
+      "loss": 0.5193,
+      "step": 9015
+    },
+    {
+      "epoch": 11.54048,
+      "grad_norm": 0.8921111822128296,
+      "learning_rate": 3.19827931172469e-05,
+      "loss": 0.5723,
+      "step": 9016
+    },
+    {
+      "epoch": 11.54176,
+      "grad_norm": 0.8784570097923279,
+      "learning_rate": 3.198079231692677e-05,
+      "loss": 0.5305,
+      "step": 9017
+    },
+    {
+      "epoch": 11.54304,
+      "grad_norm": 0.9135284423828125,
+      "learning_rate": 3.197879151660664e-05,
+      "loss": 0.5463,
+      "step": 9018
+    },
+    {
+      "epoch": 11.54432,
+      "grad_norm": 0.8987352848052979,
+      "learning_rate": 3.1976790716286515e-05,
+      "loss": 0.6218,
+      "step": 9019
+    },
+    {
+      "epoch": 11.5456,
+      "grad_norm": 0.857602059841156,
+      "learning_rate": 3.197478991596639e-05,
+      "loss": 0.5085,
+      "step": 9020
+    },
+    {
+      "epoch": 11.54688,
+      "grad_norm": 0.8649136424064636,
+      "learning_rate": 3.1972789115646265e-05,
+      "loss": 0.5577,
+      "step": 9021
+    },
+    {
+      "epoch": 11.54816,
+      "grad_norm": 0.8804261684417725,
+      "learning_rate": 3.197078831532613e-05,
+      "loss": 0.5755,
+      "step": 9022
+    },
+    {
+      "epoch": 11.54944,
+      "grad_norm": 0.8786762356758118,
+      "learning_rate": 3.1968787515006e-05,
+      "loss": 0.5368,
+      "step": 9023
+    },
+    {
+      "epoch": 11.55072,
+      "grad_norm": 0.8894196152687073,
+      "learning_rate": 3.1966786714685874e-05,
+      "loss": 0.5745,
+      "step": 9024
+    },
+    {
+      "epoch": 11.552,
+      "grad_norm": 0.9118727445602417,
+      "learning_rate": 3.1964785914365746e-05,
+      "loss": 0.5589,
+      "step": 9025
+    },
+    {
+      "epoch": 11.55328,
+      "grad_norm": 0.9220285415649414,
+      "learning_rate": 3.196278511404562e-05,
+      "loss": 0.5938,
+      "step": 9026
+    },
+    {
+      "epoch": 11.55456,
+      "grad_norm": 0.9039615392684937,
+      "learning_rate": 3.196078431372549e-05,
+      "loss": 0.5626,
+      "step": 9027
+    },
+    {
+      "epoch": 11.55584,
+      "grad_norm": 0.8908433318138123,
+      "learning_rate": 3.195878351340537e-05,
+      "loss": 0.6023,
+      "step": 9028
+    },
+    {
+      "epoch": 11.55712,
+      "grad_norm": 0.8978443741798401,
+      "learning_rate": 3.195678271308524e-05,
+      "loss": 0.5491,
+      "step": 9029
+    },
+    {
+      "epoch": 11.5584,
+      "grad_norm": 0.8988606929779053,
+      "learning_rate": 3.1954781912765105e-05,
+      "loss": 0.5366,
+      "step": 9030
+    },
+    {
+      "epoch": 11.55968,
+      "grad_norm": 0.9168279767036438,
+      "learning_rate": 3.195278111244498e-05,
+      "loss": 0.5961,
+      "step": 9031
+    },
+    {
+      "epoch": 11.56096,
+      "grad_norm": 0.8574641942977905,
+      "learning_rate": 3.195078031212485e-05,
+      "loss": 0.5461,
+      "step": 9032
+    },
+    {
+      "epoch": 11.56224,
+      "grad_norm": 0.9339374303817749,
+      "learning_rate": 3.194877951180472e-05,
+      "loss": 0.5732,
+      "step": 9033
+    },
+    {
+      "epoch": 11.56352,
+      "grad_norm": 0.9759158492088318,
+      "learning_rate": 3.194677871148459e-05,
+      "loss": 0.6018,
+      "step": 9034
+    },
+    {
+      "epoch": 11.5648,
+      "grad_norm": 0.9553015232086182,
+      "learning_rate": 3.194477791116447e-05,
+      "loss": 0.5936,
+      "step": 9035
+    },
+    {
+      "epoch": 11.56608,
+      "grad_norm": 0.9375491142272949,
+      "learning_rate": 3.194277711084434e-05,
+      "loss": 0.5606,
+      "step": 9036
+    },
+    {
+      "epoch": 11.56736,
+      "grad_norm": 0.9261005520820618,
+      "learning_rate": 3.1940776310524215e-05,
+      "loss": 0.61,
+      "step": 9037
+    },
+    {
+      "epoch": 11.56864,
+      "grad_norm": 0.9291247129440308,
+      "learning_rate": 3.193877551020408e-05,
+      "loss": 0.6361,
+      "step": 9038
+    },
+    {
+      "epoch": 11.56992,
+      "grad_norm": 0.9398210644721985,
+      "learning_rate": 3.193677470988395e-05,
+      "loss": 0.554,
+      "step": 9039
+    },
+    {
+      "epoch": 11.5712,
+      "grad_norm": 0.897376298904419,
+      "learning_rate": 3.1934773909563824e-05,
+      "loss": 0.5887,
+      "step": 9040
+    },
+    {
+      "epoch": 11.57248,
+      "grad_norm": 0.8442445397377014,
+      "learning_rate": 3.1932773109243696e-05,
+      "loss": 0.5555,
+      "step": 9041
+    },
+    {
+      "epoch": 11.57376,
+      "grad_norm": 0.8896632194519043,
+      "learning_rate": 3.1930772308923574e-05,
+      "loss": 0.5612,
+      "step": 9042
+    },
+    {
+      "epoch": 11.57504,
+      "grad_norm": 0.8563700914382935,
+      "learning_rate": 3.1928771508603446e-05,
+      "loss": 0.5588,
+      "step": 9043
+    },
+    {
+      "epoch": 11.57632,
+      "grad_norm": 0.8944402933120728,
+      "learning_rate": 3.192677070828332e-05,
+      "loss": 0.5691,
+      "step": 9044
+    },
+    {
+      "epoch": 11.5776,
+      "grad_norm": 0.9250562191009521,
+      "learning_rate": 3.192476990796319e-05,
+      "loss": 0.582,
+      "step": 9045
+    },
+    {
+      "epoch": 11.57888,
+      "grad_norm": 0.9460017085075378,
+      "learning_rate": 3.1922769107643055e-05,
+      "loss": 0.6296,
+      "step": 9046
+    },
+    {
+      "epoch": 11.58016,
+      "grad_norm": 0.9160764217376709,
+      "learning_rate": 3.192076830732293e-05,
+      "loss": 0.5749,
+      "step": 9047
+    },
+    {
+      "epoch": 11.58144,
+      "grad_norm": 0.8978300094604492,
+      "learning_rate": 3.19187675070028e-05,
+      "loss": 0.522,
+      "step": 9048
+    },
+    {
+      "epoch": 11.58272,
+      "grad_norm": 0.9328131675720215,
+      "learning_rate": 3.191676670668268e-05,
+      "loss": 0.6059,
+      "step": 9049
+    },
+    {
+      "epoch": 11.584,
+      "grad_norm": 0.9057248830795288,
+      "learning_rate": 3.191476590636255e-05,
+      "loss": 0.5399,
+      "step": 9050
+    },
+    {
+      "epoch": 11.585280000000001,
+      "grad_norm": 0.9383625984191895,
+      "learning_rate": 3.191276510604242e-05,
+      "loss": 0.6006,
+      "step": 9051
+    },
+    {
+      "epoch": 11.58656,
+      "grad_norm": 0.9605696797370911,
+      "learning_rate": 3.191076430572229e-05,
+      "loss": 0.6013,
+      "step": 9052
+    },
+    {
+      "epoch": 11.58784,
+      "grad_norm": 0.9236608743667603,
+      "learning_rate": 3.1908763505402165e-05,
+      "loss": 0.54,
+      "step": 9053
+    },
+    {
+      "epoch": 11.58912,
+      "grad_norm": 0.9085046052932739,
+      "learning_rate": 3.190676270508203e-05,
+      "loss": 0.5383,
+      "step": 9054
+    },
+    {
+      "epoch": 11.5904,
+      "grad_norm": 0.9364131689071655,
+      "learning_rate": 3.19047619047619e-05,
+      "loss": 0.5888,
+      "step": 9055
+    },
+    {
+      "epoch": 11.59168,
+      "grad_norm": 0.9124594330787659,
+      "learning_rate": 3.190276110444178e-05,
+      "loss": 0.5496,
+      "step": 9056
+    },
+    {
+      "epoch": 11.59296,
+      "grad_norm": 0.900393009185791,
+      "learning_rate": 3.190076030412165e-05,
+      "loss": 0.5176,
+      "step": 9057
+    },
+    {
+      "epoch": 11.59424,
+      "grad_norm": 0.8931260704994202,
+      "learning_rate": 3.1898759503801524e-05,
+      "loss": 0.5582,
+      "step": 9058
+    },
+    {
+      "epoch": 11.59552,
+      "grad_norm": 0.8596901297569275,
+      "learning_rate": 3.1896758703481396e-05,
+      "loss": 0.5446,
+      "step": 9059
+    },
+    {
+      "epoch": 11.5968,
+      "grad_norm": 0.8517556190490723,
+      "learning_rate": 3.189475790316127e-05,
+      "loss": 0.5244,
+      "step": 9060
+    },
+    {
+      "epoch": 11.59808,
+      "grad_norm": 0.9195119142532349,
+      "learning_rate": 3.189275710284114e-05,
+      "loss": 0.6166,
+      "step": 9061
+    },
+    {
+      "epoch": 11.59936,
+      "grad_norm": 0.8437806367874146,
+      "learning_rate": 3.1890756302521005e-05,
+      "loss": 0.5334,
+      "step": 9062
+    },
+    {
+      "epoch": 11.60064,
+      "grad_norm": 0.8890377283096313,
+      "learning_rate": 3.188875550220088e-05,
+      "loss": 0.5651,
+      "step": 9063
+    },
+    {
+      "epoch": 11.60192,
+      "grad_norm": 0.9175596833229065,
+      "learning_rate": 3.1886754701880755e-05,
+      "loss": 0.587,
+      "step": 9064
+    },
+    {
+      "epoch": 11.6032,
+      "grad_norm": 0.8746216893196106,
+      "learning_rate": 3.188475390156063e-05,
+      "loss": 0.562,
+      "step": 9065
+    },
+    {
+      "epoch": 11.60448,
+      "grad_norm": 0.9140630960464478,
+      "learning_rate": 3.18827531012405e-05,
+      "loss": 0.5781,
+      "step": 9066
+    },
+    {
+      "epoch": 11.60576,
+      "grad_norm": 0.8744653463363647,
+      "learning_rate": 3.188075230092037e-05,
+      "loss": 0.5093,
+      "step": 9067
+    },
+    {
+      "epoch": 11.60704,
+      "grad_norm": 0.8821300268173218,
+      "learning_rate": 3.187875150060024e-05,
+      "loss": 0.5314,
+      "step": 9068
+    },
+    {
+      "epoch": 11.608319999999999,
+      "grad_norm": 0.9396583437919617,
+      "learning_rate": 3.1876750700280114e-05,
+      "loss": 0.5425,
+      "step": 9069
+    },
+    {
+      "epoch": 11.6096,
+      "grad_norm": 0.9217917323112488,
+      "learning_rate": 3.1874749899959986e-05,
+      "loss": 0.5373,
+      "step": 9070
+    },
+    {
+      "epoch": 11.61088,
+      "grad_norm": 0.9106178283691406,
+      "learning_rate": 3.187274909963986e-05,
+      "loss": 0.5847,
+      "step": 9071
+    },
+    {
+      "epoch": 11.61216,
+      "grad_norm": 0.852333128452301,
+      "learning_rate": 3.187074829931973e-05,
+      "loss": 0.4806,
+      "step": 9072
+    },
+    {
+      "epoch": 11.61344,
+      "grad_norm": 0.8371844291687012,
+      "learning_rate": 3.18687474989996e-05,
+      "loss": 0.5455,
+      "step": 9073
+    },
+    {
+      "epoch": 11.61472,
+      "grad_norm": 0.8995749950408936,
+      "learning_rate": 3.1866746698679474e-05,
+      "loss": 0.5826,
+      "step": 9074
+    },
+    {
+      "epoch": 11.616,
+      "grad_norm": 0.8314163088798523,
+      "learning_rate": 3.1864745898359346e-05,
+      "loss": 0.5006,
+      "step": 9075
+    },
+    {
+      "epoch": 11.617280000000001,
+      "grad_norm": 0.920068085193634,
+      "learning_rate": 3.186274509803922e-05,
+      "loss": 0.5947,
+      "step": 9076
+    },
+    {
+      "epoch": 11.61856,
+      "grad_norm": 0.9251027703285217,
+      "learning_rate": 3.186074429771909e-05,
+      "loss": 0.5639,
+      "step": 9077
+    },
+    {
+      "epoch": 11.61984,
+      "grad_norm": 0.8840105533599854,
+      "learning_rate": 3.185874349739896e-05,
+      "loss": 0.543,
+      "step": 9078
+    },
+    {
+      "epoch": 11.62112,
+      "grad_norm": 0.8957288265228271,
+      "learning_rate": 3.185674269707883e-05,
+      "loss": 0.5646,
+      "step": 9079
+    },
+    {
+      "epoch": 11.6224,
+      "grad_norm": 0.8777086734771729,
+      "learning_rate": 3.1854741896758705e-05,
+      "loss": 0.554,
+      "step": 9080
+    },
+    {
+      "epoch": 11.62368,
+      "grad_norm": 0.8875241875648499,
+      "learning_rate": 3.185274109643858e-05,
+      "loss": 0.5173,
+      "step": 9081
+    },
+    {
+      "epoch": 11.62496,
+      "grad_norm": 0.9366897344589233,
+      "learning_rate": 3.185074029611845e-05,
+      "loss": 0.6264,
+      "step": 9082
+    },
+    {
+      "epoch": 11.62624,
+      "grad_norm": 0.9162141680717468,
+      "learning_rate": 3.184873949579832e-05,
+      "loss": 0.5632,
+      "step": 9083
+    },
+    {
+      "epoch": 11.62752,
+      "grad_norm": 0.9224411249160767,
+      "learning_rate": 3.184673869547819e-05,
+      "loss": 0.5485,
+      "step": 9084
+    },
+    {
+      "epoch": 11.6288,
+      "grad_norm": 0.8495880365371704,
+      "learning_rate": 3.1844737895158064e-05,
+      "loss": 0.4792,
+      "step": 9085
+    },
+    {
+      "epoch": 11.63008,
+      "grad_norm": 0.9102917313575745,
+      "learning_rate": 3.1842737094837936e-05,
+      "loss": 0.563,
+      "step": 9086
+    },
+    {
+      "epoch": 11.63136,
+      "grad_norm": 0.9147833585739136,
+      "learning_rate": 3.184073629451781e-05,
+      "loss": 0.5744,
+      "step": 9087
+    },
+    {
+      "epoch": 11.63264,
+      "grad_norm": 0.897863507270813,
+      "learning_rate": 3.183873549419768e-05,
+      "loss": 0.544,
+      "step": 9088
+    },
+    {
+      "epoch": 11.63392,
+      "grad_norm": 0.8683566451072693,
+      "learning_rate": 3.183673469387755e-05,
+      "loss": 0.5532,
+      "step": 9089
+    },
+    {
+      "epoch": 11.6352,
+      "grad_norm": 0.8494045734405518,
+      "learning_rate": 3.1834733893557423e-05,
+      "loss": 0.5413,
+      "step": 9090
+    },
+    {
+      "epoch": 11.63648,
+      "grad_norm": 0.8431652784347534,
+      "learning_rate": 3.18327330932373e-05,
+      "loss": 0.5293,
+      "step": 9091
+    },
+    {
+      "epoch": 11.63776,
+      "grad_norm": 0.8813535571098328,
+      "learning_rate": 3.183073229291717e-05,
+      "loss": 0.5671,
+      "step": 9092
+    },
+    {
+      "epoch": 11.63904,
+      "grad_norm": 0.9110351204872131,
+      "learning_rate": 3.182873149259704e-05,
+      "loss": 0.5919,
+      "step": 9093
+    },
+    {
+      "epoch": 11.64032,
+      "grad_norm": 0.9060587286949158,
+      "learning_rate": 3.182673069227691e-05,
+      "loss": 0.5494,
+      "step": 9094
+    },
+    {
+      "epoch": 11.6416,
+      "grad_norm": 0.8764411211013794,
+      "learning_rate": 3.182472989195678e-05,
+      "loss": 0.5471,
+      "step": 9095
+    },
+    {
+      "epoch": 11.64288,
+      "grad_norm": 0.8792155981063843,
+      "learning_rate": 3.1822729091636655e-05,
+      "loss": 0.581,
+      "step": 9096
+    },
+    {
+      "epoch": 11.64416,
+      "grad_norm": 0.8817364573478699,
+      "learning_rate": 3.1820728291316526e-05,
+      "loss": 0.5777,
+      "step": 9097
+    },
+    {
+      "epoch": 11.64544,
+      "grad_norm": 0.8980505466461182,
+      "learning_rate": 3.1818727490996405e-05,
+      "loss": 0.5925,
+      "step": 9098
+    },
+    {
+      "epoch": 11.64672,
+      "grad_norm": 0.9368979930877686,
+      "learning_rate": 3.181672669067628e-05,
+      "loss": 0.5798,
+      "step": 9099
+    },
+    {
+      "epoch": 11.648,
+      "grad_norm": 0.9280688762664795,
+      "learning_rate": 3.181472589035614e-05,
+      "loss": 0.559,
+      "step": 9100
+    },
+    {
+      "epoch": 11.64928,
+      "grad_norm": 0.8853485584259033,
+      "learning_rate": 3.1812725090036014e-05,
+      "loss": 0.5102,
+      "step": 9101
+    },
+    {
+      "epoch": 11.65056,
+      "grad_norm": 0.9357476234436035,
+      "learning_rate": 3.1810724289715886e-05,
+      "loss": 0.5616,
+      "step": 9102
+    },
+    {
+      "epoch": 11.65184,
+      "grad_norm": 0.8820586800575256,
+      "learning_rate": 3.180872348939576e-05,
+      "loss": 0.5462,
+      "step": 9103
+    },
+    {
+      "epoch": 11.65312,
+      "grad_norm": 0.9340028166770935,
+      "learning_rate": 3.180672268907563e-05,
+      "loss": 0.5758,
+      "step": 9104
+    },
+    {
+      "epoch": 11.6544,
+      "grad_norm": 0.9392242431640625,
+      "learning_rate": 3.18047218887555e-05,
+      "loss": 0.546,
+      "step": 9105
+    },
+    {
+      "epoch": 11.65568,
+      "grad_norm": 0.874309241771698,
+      "learning_rate": 3.180272108843538e-05,
+      "loss": 0.5407,
+      "step": 9106
+    },
+    {
+      "epoch": 11.65696,
+      "grad_norm": 0.9422518610954285,
+      "learning_rate": 3.180072028811525e-05,
+      "loss": 0.5634,
+      "step": 9107
+    },
+    {
+      "epoch": 11.65824,
+      "grad_norm": 0.8917849659919739,
+      "learning_rate": 3.179871948779512e-05,
+      "loss": 0.5674,
+      "step": 9108
+    },
+    {
+      "epoch": 11.65952,
+      "grad_norm": 0.9151961207389832,
+      "learning_rate": 3.179671868747499e-05,
+      "loss": 0.5681,
+      "step": 9109
+    },
+    {
+      "epoch": 11.6608,
+      "grad_norm": 0.9490141868591309,
+      "learning_rate": 3.179471788715486e-05,
+      "loss": 0.5851,
+      "step": 9110
+    },
+    {
+      "epoch": 11.66208,
+      "grad_norm": 0.9519876837730408,
+      "learning_rate": 3.179271708683473e-05,
+      "loss": 0.5503,
+      "step": 9111
+    },
+    {
+      "epoch": 11.66336,
+      "grad_norm": 0.9187465906143188,
+      "learning_rate": 3.1790716286514604e-05,
+      "loss": 0.5978,
+      "step": 9112
+    },
+    {
+      "epoch": 11.66464,
+      "grad_norm": 0.8703121542930603,
+      "learning_rate": 3.178871548619448e-05,
+      "loss": 0.5282,
+      "step": 9113
+    },
+    {
+      "epoch": 11.66592,
+      "grad_norm": 0.9381796717643738,
+      "learning_rate": 3.1786714685874355e-05,
+      "loss": 0.5909,
+      "step": 9114
+    },
+    {
+      "epoch": 11.6672,
+      "grad_norm": 0.8817717432975769,
+      "learning_rate": 3.1784713885554227e-05,
+      "loss": 0.5652,
+      "step": 9115
+    },
+    {
+      "epoch": 11.66848,
+      "grad_norm": 0.8601658940315247,
+      "learning_rate": 3.178271308523409e-05,
+      "loss": 0.5071,
+      "step": 9116
+    },
+    {
+      "epoch": 11.66976,
+      "grad_norm": 0.9270244836807251,
+      "learning_rate": 3.1780712284913964e-05,
+      "loss": 0.5354,
+      "step": 9117
+    },
+    {
+      "epoch": 11.67104,
+      "grad_norm": 0.885297954082489,
+      "learning_rate": 3.1778711484593835e-05,
+      "loss": 0.577,
+      "step": 9118
+    },
+    {
+      "epoch": 11.67232,
+      "grad_norm": 0.929111659526825,
+      "learning_rate": 3.177671068427371e-05,
+      "loss": 0.5876,
+      "step": 9119
+    },
+    {
+      "epoch": 11.6736,
+      "grad_norm": 0.8553233742713928,
+      "learning_rate": 3.1774709883953586e-05,
+      "loss": 0.5356,
+      "step": 9120
+    },
+    {
+      "epoch": 11.67488,
+      "grad_norm": 0.9026567339897156,
+      "learning_rate": 3.177270908363346e-05,
+      "loss": 0.5431,
+      "step": 9121
+    },
+    {
+      "epoch": 11.67616,
+      "grad_norm": 0.9046716690063477,
+      "learning_rate": 3.177070828331333e-05,
+      "loss": 0.5823,
+      "step": 9122
+    },
+    {
+      "epoch": 11.67744,
+      "grad_norm": 0.8545054197311401,
+      "learning_rate": 3.17687074829932e-05,
+      "loss": 0.5063,
+      "step": 9123
+    },
+    {
+      "epoch": 11.67872,
+      "grad_norm": 0.9412639737129211,
+      "learning_rate": 3.1766706682673067e-05,
+      "loss": 0.5906,
+      "step": 9124
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 0.9309195280075073,
+      "learning_rate": 3.176470588235294e-05,
+      "loss": 0.5749,
+      "step": 9125
+    },
+    {
+      "epoch": 11.68128,
+      "grad_norm": 0.9051558971405029,
+      "learning_rate": 3.176270508203281e-05,
+      "loss": 0.5678,
+      "step": 9126
+    },
+    {
+      "epoch": 11.68256,
+      "grad_norm": 0.927655816078186,
+      "learning_rate": 3.176070428171269e-05,
+      "loss": 0.6117,
+      "step": 9127
+    },
+    {
+      "epoch": 11.68384,
+      "grad_norm": 0.9260802865028381,
+      "learning_rate": 3.175870348139256e-05,
+      "loss": 0.5492,
+      "step": 9128
+    },
+    {
+      "epoch": 11.68512,
+      "grad_norm": 0.8989552855491638,
+      "learning_rate": 3.175670268107243e-05,
+      "loss": 0.5712,
+      "step": 9129
+    },
+    {
+      "epoch": 11.6864,
+      "grad_norm": 0.9030582308769226,
+      "learning_rate": 3.1754701880752304e-05,
+      "loss": 0.5752,
+      "step": 9130
+    },
+    {
+      "epoch": 11.68768,
+      "grad_norm": 0.9276950359344482,
+      "learning_rate": 3.1752701080432176e-05,
+      "loss": 0.6009,
+      "step": 9131
+    },
+    {
+      "epoch": 11.68896,
+      "grad_norm": 0.8523882031440735,
+      "learning_rate": 3.175070028011204e-05,
+      "loss": 0.5174,
+      "step": 9132
+    },
+    {
+      "epoch": 11.69024,
+      "grad_norm": 0.9310316443443298,
+      "learning_rate": 3.174869947979191e-05,
+      "loss": 0.6016,
+      "step": 9133
+    },
+    {
+      "epoch": 11.69152,
+      "grad_norm": 0.9005877375602722,
+      "learning_rate": 3.174669867947179e-05,
+      "loss": 0.5351,
+      "step": 9134
+    },
+    {
+      "epoch": 11.6928,
+      "grad_norm": 0.9501352906227112,
+      "learning_rate": 3.1744697879151664e-05,
+      "loss": 0.5938,
+      "step": 9135
+    },
+    {
+      "epoch": 11.69408,
+      "grad_norm": 0.9486740827560425,
+      "learning_rate": 3.1742697078831536e-05,
+      "loss": 0.5906,
+      "step": 9136
+    },
+    {
+      "epoch": 11.69536,
+      "grad_norm": 0.9294406771659851,
+      "learning_rate": 3.174069627851141e-05,
+      "loss": 0.592,
+      "step": 9137
+    },
+    {
+      "epoch": 11.69664,
+      "grad_norm": 0.8708564043045044,
+      "learning_rate": 3.173869547819128e-05,
+      "loss": 0.5135,
+      "step": 9138
+    },
+    {
+      "epoch": 11.69792,
+      "grad_norm": 1.0165308713912964,
+      "learning_rate": 3.173669467787115e-05,
+      "loss": 0.6272,
+      "step": 9139
+    },
+    {
+      "epoch": 11.6992,
+      "grad_norm": 0.8839307427406311,
+      "learning_rate": 3.1734693877551016e-05,
+      "loss": 0.5633,
+      "step": 9140
+    },
+    {
+      "epoch": 11.70048,
+      "grad_norm": 0.8960999846458435,
+      "learning_rate": 3.1732693077230895e-05,
+      "loss": 0.5565,
+      "step": 9141
+    },
+    {
+      "epoch": 11.70176,
+      "grad_norm": 0.9335712790489197,
+      "learning_rate": 3.173069227691077e-05,
+      "loss": 0.5516,
+      "step": 9142
+    },
+    {
+      "epoch": 11.70304,
+      "grad_norm": 0.8799536228179932,
+      "learning_rate": 3.172869147659064e-05,
+      "loss": 0.585,
+      "step": 9143
+    },
+    {
+      "epoch": 11.70432,
+      "grad_norm": 0.9027646780014038,
+      "learning_rate": 3.172669067627051e-05,
+      "loss": 0.5533,
+      "step": 9144
+    },
+    {
+      "epoch": 11.7056,
+      "grad_norm": 0.9302271008491516,
+      "learning_rate": 3.172468987595038e-05,
+      "loss": 0.6,
+      "step": 9145
+    },
+    {
+      "epoch": 11.70688,
+      "grad_norm": 0.8945030570030212,
+      "learning_rate": 3.1722689075630254e-05,
+      "loss": 0.5827,
+      "step": 9146
+    },
+    {
+      "epoch": 11.70816,
+      "grad_norm": 0.9455578327178955,
+      "learning_rate": 3.1720688275310126e-05,
+      "loss": 0.546,
+      "step": 9147
+    },
+    {
+      "epoch": 11.70944,
+      "grad_norm": 0.8383356332778931,
+      "learning_rate": 3.171868747499e-05,
+      "loss": 0.5304,
+      "step": 9148
+    },
+    {
+      "epoch": 11.71072,
+      "grad_norm": 0.8386917114257812,
+      "learning_rate": 3.171668667466987e-05,
+      "loss": 0.5498,
+      "step": 9149
+    },
+    {
+      "epoch": 11.712,
+      "grad_norm": 0.8906264901161194,
+      "learning_rate": 3.171468587434974e-05,
+      "loss": 0.5319,
+      "step": 9150
+    },
+    {
+      "epoch": 11.71328,
+      "grad_norm": 0.9018164873123169,
+      "learning_rate": 3.1712685074029613e-05,
+      "loss": 0.5535,
+      "step": 9151
+    },
+    {
+      "epoch": 11.71456,
+      "grad_norm": 0.9292755126953125,
+      "learning_rate": 3.1710684273709485e-05,
+      "loss": 0.5615,
+      "step": 9152
+    },
+    {
+      "epoch": 11.71584,
+      "grad_norm": 0.947083592414856,
+      "learning_rate": 3.170868347338936e-05,
+      "loss": 0.5326,
+      "step": 9153
+    },
+    {
+      "epoch": 11.71712,
+      "grad_norm": 0.872265100479126,
+      "learning_rate": 3.170668267306923e-05,
+      "loss": 0.5514,
+      "step": 9154
+    },
+    {
+      "epoch": 11.7184,
+      "grad_norm": 0.8972936272621155,
+      "learning_rate": 3.17046818727491e-05,
+      "loss": 0.555,
+      "step": 9155
+    },
+    {
+      "epoch": 11.71968,
+      "grad_norm": 0.8738512992858887,
+      "learning_rate": 3.170268107242897e-05,
+      "loss": 0.5585,
+      "step": 9156
+    },
+    {
+      "epoch": 11.72096,
+      "grad_norm": 0.8479512333869934,
+      "learning_rate": 3.1700680272108845e-05,
+      "loss": 0.5217,
+      "step": 9157
+    },
+    {
+      "epoch": 11.72224,
+      "grad_norm": 0.853814959526062,
+      "learning_rate": 3.1698679471788716e-05,
+      "loss": 0.5418,
+      "step": 9158
+    },
+    {
+      "epoch": 11.72352,
+      "grad_norm": 0.9030658006668091,
+      "learning_rate": 3.169667867146859e-05,
+      "loss": 0.5371,
+      "step": 9159
+    },
+    {
+      "epoch": 11.7248,
+      "grad_norm": 0.9065144062042236,
+      "learning_rate": 3.169467787114846e-05,
+      "loss": 0.555,
+      "step": 9160
+    },
+    {
+      "epoch": 11.72608,
+      "grad_norm": 0.8621678948402405,
+      "learning_rate": 3.169267707082833e-05,
+      "loss": 0.552,
+      "step": 9161
+    },
+    {
+      "epoch": 11.727360000000001,
+      "grad_norm": 0.8453906178474426,
+      "learning_rate": 3.1690676270508204e-05,
+      "loss": 0.5394,
+      "step": 9162
+    },
+    {
+      "epoch": 11.72864,
+      "grad_norm": 0.9425848722457886,
+      "learning_rate": 3.1688675470188076e-05,
+      "loss": 0.6404,
+      "step": 9163
+    },
+    {
+      "epoch": 11.72992,
+      "grad_norm": 0.9024143815040588,
+      "learning_rate": 3.168667466986795e-05,
+      "loss": 0.5704,
+      "step": 9164
+    },
+    {
+      "epoch": 11.7312,
+      "grad_norm": 0.89797443151474,
+      "learning_rate": 3.168467386954782e-05,
+      "loss": 0.5647,
+      "step": 9165
+    },
+    {
+      "epoch": 11.73248,
+      "grad_norm": 0.9100660085678101,
+      "learning_rate": 3.168267306922769e-05,
+      "loss": 0.553,
+      "step": 9166
+    },
+    {
+      "epoch": 11.73376,
+      "grad_norm": 0.8714706301689148,
+      "learning_rate": 3.168067226890756e-05,
+      "loss": 0.5353,
+      "step": 9167
+    },
+    {
+      "epoch": 11.73504,
+      "grad_norm": 0.8743889331817627,
+      "learning_rate": 3.1678671468587435e-05,
+      "loss": 0.5513,
+      "step": 9168
+    },
+    {
+      "epoch": 11.73632,
+      "grad_norm": 0.8862664699554443,
+      "learning_rate": 3.1676670668267314e-05,
+      "loss": 0.522,
+      "step": 9169
+    },
+    {
+      "epoch": 11.7376,
+      "grad_norm": 0.8375951647758484,
+      "learning_rate": 3.167466986794718e-05,
+      "loss": 0.5308,
+      "step": 9170
+    },
+    {
+      "epoch": 11.73888,
+      "grad_norm": 0.9067046642303467,
+      "learning_rate": 3.167266906762705e-05,
+      "loss": 0.5673,
+      "step": 9171
+    },
+    {
+      "epoch": 11.74016,
+      "grad_norm": 0.8670250177383423,
+      "learning_rate": 3.167066826730692e-05,
+      "loss": 0.519,
+      "step": 9172
+    },
+    {
+      "epoch": 11.74144,
+      "grad_norm": 0.8583195805549622,
+      "learning_rate": 3.1668667466986794e-05,
+      "loss": 0.532,
+      "step": 9173
+    },
+    {
+      "epoch": 11.74272,
+      "grad_norm": 0.9073027968406677,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.5339,
+      "step": 9174
+    },
+    {
+      "epoch": 11.744,
+      "grad_norm": 0.8694731593132019,
+      "learning_rate": 3.166466586634654e-05,
+      "loss": 0.5488,
+      "step": 9175
+    },
+    {
+      "epoch": 11.74528,
+      "grad_norm": 0.9093069434165955,
+      "learning_rate": 3.166266506602642e-05,
+      "loss": 0.5305,
+      "step": 9176
+    },
+    {
+      "epoch": 11.74656,
+      "grad_norm": 0.8719558119773865,
+      "learning_rate": 3.166066426570629e-05,
+      "loss": 0.5112,
+      "step": 9177
+    },
+    {
+      "epoch": 11.74784,
+      "grad_norm": 0.8447220325469971,
+      "learning_rate": 3.1658663465386154e-05,
+      "loss": 0.4831,
+      "step": 9178
+    },
+    {
+      "epoch": 11.74912,
+      "grad_norm": 0.911290168762207,
+      "learning_rate": 3.1656662665066025e-05,
+      "loss": 0.5559,
+      "step": 9179
+    },
+    {
+      "epoch": 11.750399999999999,
+      "grad_norm": 0.8428891897201538,
+      "learning_rate": 3.16546618647459e-05,
+      "loss": 0.5264,
+      "step": 9180
+    },
+    {
+      "epoch": 11.75168,
+      "grad_norm": 0.9175834059715271,
+      "learning_rate": 3.165266106442577e-05,
+      "loss": 0.5798,
+      "step": 9181
+    },
+    {
+      "epoch": 11.75296,
+      "grad_norm": 0.8715541958808899,
+      "learning_rate": 3.165066026410564e-05,
+      "loss": 0.5017,
+      "step": 9182
+    },
+    {
+      "epoch": 11.75424,
+      "grad_norm": 0.8270224928855896,
+      "learning_rate": 3.164865946378552e-05,
+      "loss": 0.5368,
+      "step": 9183
+    },
+    {
+      "epoch": 11.75552,
+      "grad_norm": 0.9201467037200928,
+      "learning_rate": 3.164665866346539e-05,
+      "loss": 0.5696,
+      "step": 9184
+    },
+    {
+      "epoch": 11.7568,
+      "grad_norm": 0.8654879331588745,
+      "learning_rate": 3.164465786314526e-05,
+      "loss": 0.5424,
+      "step": 9185
+    },
+    {
+      "epoch": 11.75808,
+      "grad_norm": 0.897590696811676,
+      "learning_rate": 3.164265706282513e-05,
+      "loss": 0.5809,
+      "step": 9186
+    },
+    {
+      "epoch": 11.759360000000001,
+      "grad_norm": 0.895999014377594,
+      "learning_rate": 3.1640656262505e-05,
+      "loss": 0.5432,
+      "step": 9187
+    },
+    {
+      "epoch": 11.76064,
+      "grad_norm": 0.9101575613021851,
+      "learning_rate": 3.163865546218487e-05,
+      "loss": 0.6348,
+      "step": 9188
+    },
+    {
+      "epoch": 11.76192,
+      "grad_norm": 0.8356999754905701,
+      "learning_rate": 3.1636654661864744e-05,
+      "loss": 0.4978,
+      "step": 9189
+    },
+    {
+      "epoch": 11.7632,
+      "grad_norm": 0.886934757232666,
+      "learning_rate": 3.163465386154462e-05,
+      "loss": 0.5392,
+      "step": 9190
+    },
+    {
+      "epoch": 11.76448,
+      "grad_norm": 0.8999463319778442,
+      "learning_rate": 3.1632653061224494e-05,
+      "loss": 0.5413,
+      "step": 9191
+    },
+    {
+      "epoch": 11.76576,
+      "grad_norm": 0.9087318778038025,
+      "learning_rate": 3.1630652260904366e-05,
+      "loss": 0.5776,
+      "step": 9192
+    },
+    {
+      "epoch": 11.76704,
+      "grad_norm": 0.8902879357337952,
+      "learning_rate": 3.162865146058424e-05,
+      "loss": 0.5987,
+      "step": 9193
+    },
+    {
+      "epoch": 11.76832,
+      "grad_norm": 0.8536181449890137,
+      "learning_rate": 3.16266506602641e-05,
+      "loss": 0.5449,
+      "step": 9194
+    },
+    {
+      "epoch": 11.7696,
+      "grad_norm": 0.8617259860038757,
+      "learning_rate": 3.1624649859943975e-05,
+      "loss": 0.5814,
+      "step": 9195
+    },
+    {
+      "epoch": 11.77088,
+      "grad_norm": 0.9288644790649414,
+      "learning_rate": 3.162264905962385e-05,
+      "loss": 0.5716,
+      "step": 9196
+    },
+    {
+      "epoch": 11.77216,
+      "grad_norm": 0.9424710869789124,
+      "learning_rate": 3.1620648259303726e-05,
+      "loss": 0.5987,
+      "step": 9197
+    },
+    {
+      "epoch": 11.77344,
+      "grad_norm": 0.8956670761108398,
+      "learning_rate": 3.16186474589836e-05,
+      "loss": 0.5165,
+      "step": 9198
+    },
+    {
+      "epoch": 11.77472,
+      "grad_norm": 0.9281831383705139,
+      "learning_rate": 3.161664665866347e-05,
+      "loss": 0.572,
+      "step": 9199
+    },
+    {
+      "epoch": 11.776,
+      "grad_norm": 0.9560943841934204,
+      "learning_rate": 3.161464585834334e-05,
+      "loss": 0.5877,
+      "step": 9200
+    },
+    {
+      "epoch": 11.77728,
+      "grad_norm": 0.8967903852462769,
+      "learning_rate": 3.161264505802321e-05,
+      "loss": 0.5637,
+      "step": 9201
+    },
+    {
+      "epoch": 11.77856,
+      "grad_norm": 0.8792325258255005,
+      "learning_rate": 3.161064425770308e-05,
+      "loss": 0.5224,
+      "step": 9202
+    },
+    {
+      "epoch": 11.77984,
+      "grad_norm": 0.910875678062439,
+      "learning_rate": 3.160864345738295e-05,
+      "loss": 0.5554,
+      "step": 9203
+    },
+    {
+      "epoch": 11.78112,
+      "grad_norm": 0.8686668872833252,
+      "learning_rate": 3.160664265706283e-05,
+      "loss": 0.5578,
+      "step": 9204
+    },
+    {
+      "epoch": 11.782399999999999,
+      "grad_norm": 0.8657653331756592,
+      "learning_rate": 3.16046418567427e-05,
+      "loss": 0.5172,
+      "step": 9205
+    },
+    {
+      "epoch": 11.78368,
+      "grad_norm": 0.8818771243095398,
+      "learning_rate": 3.160264105642257e-05,
+      "loss": 0.5662,
+      "step": 9206
+    },
+    {
+      "epoch": 11.78496,
+      "grad_norm": 0.9204904437065125,
+      "learning_rate": 3.1600640256102444e-05,
+      "loss": 0.5711,
+      "step": 9207
+    },
+    {
+      "epoch": 11.78624,
+      "grad_norm": 0.9225419759750366,
+      "learning_rate": 3.1598639455782316e-05,
+      "loss": 0.5677,
+      "step": 9208
+    },
+    {
+      "epoch": 11.78752,
+      "grad_norm": 0.8883199095726013,
+      "learning_rate": 3.159663865546219e-05,
+      "loss": 0.5218,
+      "step": 9209
+    },
+    {
+      "epoch": 11.7888,
+      "grad_norm": 0.9116802215576172,
+      "learning_rate": 3.159463785514205e-05,
+      "loss": 0.6062,
+      "step": 9210
+    },
+    {
+      "epoch": 11.79008,
+      "grad_norm": 0.8611482977867126,
+      "learning_rate": 3.159263705482193e-05,
+      "loss": 0.5445,
+      "step": 9211
+    },
+    {
+      "epoch": 11.79136,
+      "grad_norm": 0.8924483060836792,
+      "learning_rate": 3.1590636254501803e-05,
+      "loss": 0.602,
+      "step": 9212
+    },
+    {
+      "epoch": 11.79264,
+      "grad_norm": 0.8485545516014099,
+      "learning_rate": 3.1588635454181675e-05,
+      "loss": 0.5326,
+      "step": 9213
+    },
+    {
+      "epoch": 11.79392,
+      "grad_norm": 0.8666700124740601,
+      "learning_rate": 3.158663465386155e-05,
+      "loss": 0.5149,
+      "step": 9214
+    },
+    {
+      "epoch": 11.7952,
+      "grad_norm": 0.9414648413658142,
+      "learning_rate": 3.158463385354142e-05,
+      "loss": 0.5767,
+      "step": 9215
+    },
+    {
+      "epoch": 11.79648,
+      "grad_norm": 0.885181188583374,
+      "learning_rate": 3.158263305322129e-05,
+      "loss": 0.5112,
+      "step": 9216
+    },
+    {
+      "epoch": 11.79776,
+      "grad_norm": 0.8811646103858948,
+      "learning_rate": 3.158063225290116e-05,
+      "loss": 0.5482,
+      "step": 9217
+    },
+    {
+      "epoch": 11.79904,
+      "grad_norm": 0.905148983001709,
+      "learning_rate": 3.157863145258103e-05,
+      "loss": 0.6129,
+      "step": 9218
+    },
+    {
+      "epoch": 11.80032,
+      "grad_norm": 0.9198310971260071,
+      "learning_rate": 3.1576630652260906e-05,
+      "loss": 0.5667,
+      "step": 9219
+    },
+    {
+      "epoch": 11.8016,
+      "grad_norm": 0.9002284407615662,
+      "learning_rate": 3.157462985194078e-05,
+      "loss": 0.5478,
+      "step": 9220
+    },
+    {
+      "epoch": 11.80288,
+      "grad_norm": 0.8892956972122192,
+      "learning_rate": 3.157262905162065e-05,
+      "loss": 0.5446,
+      "step": 9221
+    },
+    {
+      "epoch": 11.80416,
+      "grad_norm": 0.8949340581893921,
+      "learning_rate": 3.157062825130052e-05,
+      "loss": 0.5443,
+      "step": 9222
+    },
+    {
+      "epoch": 11.80544,
+      "grad_norm": 0.869086742401123,
+      "learning_rate": 3.1568627450980394e-05,
+      "loss": 0.546,
+      "step": 9223
+    },
+    {
+      "epoch": 11.80672,
+      "grad_norm": 0.920300304889679,
+      "learning_rate": 3.1566626650660266e-05,
+      "loss": 0.5784,
+      "step": 9224
+    },
+    {
+      "epoch": 11.808,
+      "grad_norm": 0.8822394013404846,
+      "learning_rate": 3.156462585034014e-05,
+      "loss": 0.5174,
+      "step": 9225
+    },
+    {
+      "epoch": 11.80928,
+      "grad_norm": 0.9120069146156311,
+      "learning_rate": 3.156262505002001e-05,
+      "loss": 0.5333,
+      "step": 9226
+    },
+    {
+      "epoch": 11.81056,
+      "grad_norm": 0.8972925543785095,
+      "learning_rate": 3.156062424969988e-05,
+      "loss": 0.6007,
+      "step": 9227
+    },
+    {
+      "epoch": 11.81184,
+      "grad_norm": 0.8929551839828491,
+      "learning_rate": 3.155862344937975e-05,
+      "loss": 0.5792,
+      "step": 9228
+    },
+    {
+      "epoch": 11.81312,
+      "grad_norm": 0.8880111575126648,
+      "learning_rate": 3.1556622649059625e-05,
+      "loss": 0.5315,
+      "step": 9229
+    },
+    {
+      "epoch": 11.8144,
+      "grad_norm": 0.9577479362487793,
+      "learning_rate": 3.15546218487395e-05,
+      "loss": 0.6392,
+      "step": 9230
+    },
+    {
+      "epoch": 11.81568,
+      "grad_norm": 0.9124792814254761,
+      "learning_rate": 3.155262104841937e-05,
+      "loss": 0.5666,
+      "step": 9231
+    },
+    {
+      "epoch": 11.81696,
+      "grad_norm": 0.9073526263237,
+      "learning_rate": 3.155062024809924e-05,
+      "loss": 0.5656,
+      "step": 9232
+    },
+    {
+      "epoch": 11.81824,
+      "grad_norm": 0.8945934176445007,
+      "learning_rate": 3.154861944777911e-05,
+      "loss": 0.5644,
+      "step": 9233
+    },
+    {
+      "epoch": 11.81952,
+      "grad_norm": 0.8661502599716187,
+      "learning_rate": 3.1546618647458984e-05,
+      "loss": 0.5162,
+      "step": 9234
+    },
+    {
+      "epoch": 11.8208,
+      "grad_norm": 0.9258007407188416,
+      "learning_rate": 3.1544617847138856e-05,
+      "loss": 0.542,
+      "step": 9235
+    },
+    {
+      "epoch": 11.82208,
+      "grad_norm": 0.859990119934082,
+      "learning_rate": 3.154261704681873e-05,
+      "loss": 0.5686,
+      "step": 9236
+    },
+    {
+      "epoch": 11.82336,
+      "grad_norm": 0.8406195044517517,
+      "learning_rate": 3.15406162464986e-05,
+      "loss": 0.5378,
+      "step": 9237
+    },
+    {
+      "epoch": 11.82464,
+      "grad_norm": 0.8850319385528564,
+      "learning_rate": 3.153861544617847e-05,
+      "loss": 0.5726,
+      "step": 9238
+    },
+    {
+      "epoch": 11.82592,
+      "grad_norm": 0.9233469367027283,
+      "learning_rate": 3.1536614645858344e-05,
+      "loss": 0.5878,
+      "step": 9239
+    },
+    {
+      "epoch": 11.8272,
+      "grad_norm": 0.9111922979354858,
+      "learning_rate": 3.153461384553822e-05,
+      "loss": 0.5668,
+      "step": 9240
+    },
+    {
+      "epoch": 11.82848,
+      "grad_norm": 0.903139054775238,
+      "learning_rate": 3.153261304521809e-05,
+      "loss": 0.5535,
+      "step": 9241
+    },
+    {
+      "epoch": 11.82976,
+      "grad_norm": 0.8692033886909485,
+      "learning_rate": 3.153061224489796e-05,
+      "loss": 0.4958,
+      "step": 9242
+    },
+    {
+      "epoch": 11.83104,
+      "grad_norm": 0.8684501051902771,
+      "learning_rate": 3.152861144457783e-05,
+      "loss": 0.5101,
+      "step": 9243
+    },
+    {
+      "epoch": 11.83232,
+      "grad_norm": 0.8597356081008911,
+      "learning_rate": 3.15266106442577e-05,
+      "loss": 0.5575,
+      "step": 9244
+    },
+    {
+      "epoch": 11.8336,
+      "grad_norm": 0.8873345851898193,
+      "learning_rate": 3.1524609843937575e-05,
+      "loss": 0.5863,
+      "step": 9245
+    },
+    {
+      "epoch": 11.83488,
+      "grad_norm": 0.8959015011787415,
+      "learning_rate": 3.152260904361745e-05,
+      "loss": 0.5667,
+      "step": 9246
+    },
+    {
+      "epoch": 11.83616,
+      "grad_norm": 0.9511164426803589,
+      "learning_rate": 3.1520608243297325e-05,
+      "loss": 0.5755,
+      "step": 9247
+    },
+    {
+      "epoch": 11.83744,
+      "grad_norm": 0.9274942278862,
+      "learning_rate": 3.15186074429772e-05,
+      "loss": 0.5661,
+      "step": 9248
+    },
+    {
+      "epoch": 11.83872,
+      "grad_norm": 0.893008291721344,
+      "learning_rate": 3.151660664265706e-05,
+      "loss": 0.5641,
+      "step": 9249
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 0.9084400534629822,
+      "learning_rate": 3.1514605842336934e-05,
+      "loss": 0.533,
+      "step": 9250
+    },
+    {
+      "epoch": 11.84128,
+      "grad_norm": 0.9191239476203918,
+      "learning_rate": 3.1512605042016806e-05,
+      "loss": 0.557,
+      "step": 9251
+    },
+    {
+      "epoch": 11.84256,
+      "grad_norm": 0.933152973651886,
+      "learning_rate": 3.151060424169668e-05,
+      "loss": 0.5457,
+      "step": 9252
+    },
+    {
+      "epoch": 11.84384,
+      "grad_norm": 0.9432767629623413,
+      "learning_rate": 3.150860344137655e-05,
+      "loss": 0.6098,
+      "step": 9253
+    },
+    {
+      "epoch": 11.84512,
+      "grad_norm": 0.8910885453224182,
+      "learning_rate": 3.150660264105643e-05,
+      "loss": 0.5155,
+      "step": 9254
+    },
+    {
+      "epoch": 11.8464,
+      "grad_norm": 0.8926212787628174,
+      "learning_rate": 3.15046018407363e-05,
+      "loss": 0.5236,
+      "step": 9255
+    },
+    {
+      "epoch": 11.84768,
+      "grad_norm": 0.8529669046401978,
+      "learning_rate": 3.150260104041617e-05,
+      "loss": 0.5063,
+      "step": 9256
+    },
+    {
+      "epoch": 11.84896,
+      "grad_norm": 0.8807107210159302,
+      "learning_rate": 3.150060024009604e-05,
+      "loss": 0.5411,
+      "step": 9257
+    },
+    {
+      "epoch": 11.85024,
+      "grad_norm": 0.8540694117546082,
+      "learning_rate": 3.149859943977591e-05,
+      "loss": 0.5482,
+      "step": 9258
+    },
+    {
+      "epoch": 11.85152,
+      "grad_norm": 0.8822780847549438,
+      "learning_rate": 3.149659863945578e-05,
+      "loss": 0.5256,
+      "step": 9259
+    },
+    {
+      "epoch": 11.8528,
+      "grad_norm": 0.8807215094566345,
+      "learning_rate": 3.149459783913565e-05,
+      "loss": 0.5734,
+      "step": 9260
+    },
+    {
+      "epoch": 11.85408,
+      "grad_norm": 0.8632503151893616,
+      "learning_rate": 3.149259703881553e-05,
+      "loss": 0.5134,
+      "step": 9261
+    },
+    {
+      "epoch": 11.85536,
+      "grad_norm": 0.9131238460540771,
+      "learning_rate": 3.14905962384954e-05,
+      "loss": 0.5878,
+      "step": 9262
+    },
+    {
+      "epoch": 11.85664,
+      "grad_norm": 0.862756609916687,
+      "learning_rate": 3.1488595438175275e-05,
+      "loss": 0.5613,
+      "step": 9263
+    },
+    {
+      "epoch": 11.85792,
+      "grad_norm": 0.8660969734191895,
+      "learning_rate": 3.148659463785515e-05,
+      "loss": 0.5352,
+      "step": 9264
+    },
+    {
+      "epoch": 11.8592,
+      "grad_norm": 0.9220228791236877,
+      "learning_rate": 3.148459383753501e-05,
+      "loss": 0.5738,
+      "step": 9265
+    },
+    {
+      "epoch": 11.86048,
+      "grad_norm": 0.872033953666687,
+      "learning_rate": 3.1482593037214884e-05,
+      "loss": 0.5792,
+      "step": 9266
+    },
+    {
+      "epoch": 11.86176,
+      "grad_norm": 0.8677129745483398,
+      "learning_rate": 3.1480592236894756e-05,
+      "loss": 0.5334,
+      "step": 9267
+    },
+    {
+      "epoch": 11.86304,
+      "grad_norm": 0.8851068615913391,
+      "learning_rate": 3.1478591436574634e-05,
+      "loss": 0.5471,
+      "step": 9268
+    },
+    {
+      "epoch": 11.86432,
+      "grad_norm": 0.921118438243866,
+      "learning_rate": 3.1476590636254506e-05,
+      "loss": 0.562,
+      "step": 9269
+    },
+    {
+      "epoch": 11.8656,
+      "grad_norm": 0.868963897228241,
+      "learning_rate": 3.147458983593438e-05,
+      "loss": 0.5733,
+      "step": 9270
+    },
+    {
+      "epoch": 11.86688,
+      "grad_norm": 0.9390560388565063,
+      "learning_rate": 3.147258903561425e-05,
+      "loss": 0.5904,
+      "step": 9271
+    },
+    {
+      "epoch": 11.86816,
+      "grad_norm": 0.9005419611930847,
+      "learning_rate": 3.147058823529412e-05,
+      "loss": 0.5529,
+      "step": 9272
+    },
+    {
+      "epoch": 11.86944,
+      "grad_norm": 0.8554105758666992,
+      "learning_rate": 3.146858743497399e-05,
+      "loss": 0.6002,
+      "step": 9273
+    },
+    {
+      "epoch": 11.87072,
+      "grad_norm": 0.9244778752326965,
+      "learning_rate": 3.146658663465386e-05,
+      "loss": 0.5658,
+      "step": 9274
+    },
+    {
+      "epoch": 11.872,
+      "grad_norm": 0.8940372467041016,
+      "learning_rate": 3.146458583433374e-05,
+      "loss": 0.5647,
+      "step": 9275
+    },
+    {
+      "epoch": 11.87328,
+      "grad_norm": 0.998906672000885,
+      "learning_rate": 3.146258503401361e-05,
+      "loss": 0.6089,
+      "step": 9276
+    },
+    {
+      "epoch": 11.87456,
+      "grad_norm": 0.915664792060852,
+      "learning_rate": 3.146058423369348e-05,
+      "loss": 0.5579,
+      "step": 9277
+    },
+    {
+      "epoch": 11.87584,
+      "grad_norm": 0.8526196479797363,
+      "learning_rate": 3.145858343337335e-05,
+      "loss": 0.5319,
+      "step": 9278
+    },
+    {
+      "epoch": 11.87712,
+      "grad_norm": 0.8531317710876465,
+      "learning_rate": 3.1456582633053225e-05,
+      "loss": 0.5384,
+      "step": 9279
+    },
+    {
+      "epoch": 11.8784,
+      "grad_norm": 0.876233696937561,
+      "learning_rate": 3.1454581832733097e-05,
+      "loss": 0.5413,
+      "step": 9280
+    },
+    {
+      "epoch": 11.87968,
+      "grad_norm": 0.8565805554389954,
+      "learning_rate": 3.145258103241296e-05,
+      "loss": 0.5441,
+      "step": 9281
+    },
+    {
+      "epoch": 11.88096,
+      "grad_norm": 0.8764281272888184,
+      "learning_rate": 3.145058023209284e-05,
+      "loss": 0.5208,
+      "step": 9282
+    },
+    {
+      "epoch": 11.88224,
+      "grad_norm": 0.837112545967102,
+      "learning_rate": 3.144857943177271e-05,
+      "loss": 0.5515,
+      "step": 9283
+    },
+    {
+      "epoch": 11.88352,
+      "grad_norm": 0.8274785280227661,
+      "learning_rate": 3.1446578631452584e-05,
+      "loss": 0.5302,
+      "step": 9284
+    },
+    {
+      "epoch": 11.8848,
+      "grad_norm": 0.8528085350990295,
+      "learning_rate": 3.1444577831132456e-05,
+      "loss": 0.5414,
+      "step": 9285
+    },
+    {
+      "epoch": 11.88608,
+      "grad_norm": 0.8616606593132019,
+      "learning_rate": 3.144257703081233e-05,
+      "loss": 0.5315,
+      "step": 9286
+    },
+    {
+      "epoch": 11.88736,
+      "grad_norm": 0.9000251293182373,
+      "learning_rate": 3.14405762304922e-05,
+      "loss": 0.6074,
+      "step": 9287
+    },
+    {
+      "epoch": 11.88864,
+      "grad_norm": 0.9343587756156921,
+      "learning_rate": 3.143857543017207e-05,
+      "loss": 0.5863,
+      "step": 9288
+    },
+    {
+      "epoch": 11.88992,
+      "grad_norm": 0.9544128179550171,
+      "learning_rate": 3.143657462985194e-05,
+      "loss": 0.5909,
+      "step": 9289
+    },
+    {
+      "epoch": 11.8912,
+      "grad_norm": 0.8408989310264587,
+      "learning_rate": 3.1434573829531815e-05,
+      "loss": 0.5349,
+      "step": 9290
+    },
+    {
+      "epoch": 11.89248,
+      "grad_norm": 0.8732945919036865,
+      "learning_rate": 3.143257302921169e-05,
+      "loss": 0.5135,
+      "step": 9291
+    },
+    {
+      "epoch": 11.89376,
+      "grad_norm": 0.8831563591957092,
+      "learning_rate": 3.143057222889156e-05,
+      "loss": 0.5941,
+      "step": 9292
+    },
+    {
+      "epoch": 11.89504,
+      "grad_norm": 0.7958306074142456,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.5004,
+      "step": 9293
+    },
+    {
+      "epoch": 11.89632,
+      "grad_norm": 0.9264546632766724,
+      "learning_rate": 3.14265706282513e-05,
+      "loss": 0.5965,
+      "step": 9294
+    },
+    {
+      "epoch": 11.8976,
+      "grad_norm": 0.9189742803573608,
+      "learning_rate": 3.1424569827931174e-05,
+      "loss": 0.5785,
+      "step": 9295
+    },
+    {
+      "epoch": 11.89888,
+      "grad_norm": 0.9090092778205872,
+      "learning_rate": 3.1422569027611046e-05,
+      "loss": 0.5715,
+      "step": 9296
+    },
+    {
+      "epoch": 11.90016,
+      "grad_norm": 0.9040752053260803,
+      "learning_rate": 3.142056822729092e-05,
+      "loss": 0.5748,
+      "step": 9297
+    },
+    {
+      "epoch": 11.901440000000001,
+      "grad_norm": 0.8540357947349548,
+      "learning_rate": 3.141856742697079e-05,
+      "loss": 0.5637,
+      "step": 9298
+    },
+    {
+      "epoch": 11.90272,
+      "grad_norm": 0.8528608083724976,
+      "learning_rate": 3.141656662665066e-05,
+      "loss": 0.5238,
+      "step": 9299
+    },
+    {
+      "epoch": 11.904,
+      "grad_norm": 0.8939346075057983,
+      "learning_rate": 3.1414565826330534e-05,
+      "loss": 0.5637,
+      "step": 9300
+    },
+    {
+      "epoch": 11.90528,
+      "grad_norm": 0.8866257071495056,
+      "learning_rate": 3.1412565026010406e-05,
+      "loss": 0.6088,
+      "step": 9301
+    },
+    {
+      "epoch": 11.90656,
+      "grad_norm": 0.9136728644371033,
+      "learning_rate": 3.141056422569028e-05,
+      "loss": 0.5708,
+      "step": 9302
+    },
+    {
+      "epoch": 11.90784,
+      "grad_norm": 0.908406138420105,
+      "learning_rate": 3.140856342537015e-05,
+      "loss": 0.5593,
+      "step": 9303
+    },
+    {
+      "epoch": 11.90912,
+      "grad_norm": 0.9199392199516296,
+      "learning_rate": 3.140656262505002e-05,
+      "loss": 0.5653,
+      "step": 9304
+    },
+    {
+      "epoch": 11.9104,
+      "grad_norm": 0.9563983082771301,
+      "learning_rate": 3.140456182472989e-05,
+      "loss": 0.56,
+      "step": 9305
+    },
+    {
+      "epoch": 11.91168,
+      "grad_norm": 0.8945441842079163,
+      "learning_rate": 3.1402561024409765e-05,
+      "loss": 0.5475,
+      "step": 9306
+    },
+    {
+      "epoch": 11.91296,
+      "grad_norm": 0.9007996320724487,
+      "learning_rate": 3.140056022408964e-05,
+      "loss": 0.6062,
+      "step": 9307
+    },
+    {
+      "epoch": 11.91424,
+      "grad_norm": 0.9043374061584473,
+      "learning_rate": 3.139855942376951e-05,
+      "loss": 0.5396,
+      "step": 9308
+    },
+    {
+      "epoch": 11.91552,
+      "grad_norm": 0.8611972332000732,
+      "learning_rate": 3.139655862344938e-05,
+      "loss": 0.5255,
+      "step": 9309
+    },
+    {
+      "epoch": 11.9168,
+      "grad_norm": 0.9283512234687805,
+      "learning_rate": 3.139455782312926e-05,
+      "loss": 0.5369,
+      "step": 9310
+    },
+    {
+      "epoch": 11.91808,
+      "grad_norm": 0.9080418944358826,
+      "learning_rate": 3.1392557022809124e-05,
+      "loss": 0.5843,
+      "step": 9311
+    },
+    {
+      "epoch": 11.91936,
+      "grad_norm": 0.9004126191139221,
+      "learning_rate": 3.1390556222488996e-05,
+      "loss": 0.5956,
+      "step": 9312
+    },
+    {
+      "epoch": 11.92064,
+      "grad_norm": 0.9140704274177551,
+      "learning_rate": 3.138855542216887e-05,
+      "loss": 0.5571,
+      "step": 9313
+    },
+    {
+      "epoch": 11.92192,
+      "grad_norm": 0.8815629482269287,
+      "learning_rate": 3.138655462184874e-05,
+      "loss": 0.5591,
+      "step": 9314
+    },
+    {
+      "epoch": 11.9232,
+      "grad_norm": 0.9093433618545532,
+      "learning_rate": 3.138455382152861e-05,
+      "loss": 0.5775,
+      "step": 9315
+    },
+    {
+      "epoch": 11.924479999999999,
+      "grad_norm": 0.8649457693099976,
+      "learning_rate": 3.138255302120848e-05,
+      "loss": 0.547,
+      "step": 9316
+    },
+    {
+      "epoch": 11.92576,
+      "grad_norm": 0.8733769059181213,
+      "learning_rate": 3.138055222088836e-05,
+      "loss": 0.5395,
+      "step": 9317
+    },
+    {
+      "epoch": 11.92704,
+      "grad_norm": 0.8359034061431885,
+      "learning_rate": 3.1378551420568234e-05,
+      "loss": 0.5197,
+      "step": 9318
+    },
+    {
+      "epoch": 11.92832,
+      "grad_norm": 0.9323531985282898,
+      "learning_rate": 3.13765506202481e-05,
+      "loss": 0.5518,
+      "step": 9319
+    },
+    {
+      "epoch": 11.9296,
+      "grad_norm": 0.8930750489234924,
+      "learning_rate": 3.137454981992797e-05,
+      "loss": 0.576,
+      "step": 9320
+    },
+    {
+      "epoch": 11.93088,
+      "grad_norm": 0.9025473594665527,
+      "learning_rate": 3.137254901960784e-05,
+      "loss": 0.6194,
+      "step": 9321
+    },
+    {
+      "epoch": 11.93216,
+      "grad_norm": 0.8879018425941467,
+      "learning_rate": 3.1370548219287715e-05,
+      "loss": 0.5253,
+      "step": 9322
+    },
+    {
+      "epoch": 11.933440000000001,
+      "grad_norm": 0.9026443958282471,
+      "learning_rate": 3.1368547418967586e-05,
+      "loss": 0.61,
+      "step": 9323
+    },
+    {
+      "epoch": 11.93472,
+      "grad_norm": 0.9492444396018982,
+      "learning_rate": 3.1366546618647465e-05,
+      "loss": 0.5656,
+      "step": 9324
+    },
+    {
+      "epoch": 11.936,
+      "grad_norm": 0.913387656211853,
+      "learning_rate": 3.136454581832734e-05,
+      "loss": 0.533,
+      "step": 9325
+    },
+    {
+      "epoch": 11.93728,
+      "grad_norm": 0.8534862995147705,
+      "learning_rate": 3.136254501800721e-05,
+      "loss": 0.5176,
+      "step": 9326
+    },
+    {
+      "epoch": 11.93856,
+      "grad_norm": 0.8602653741836548,
+      "learning_rate": 3.1360544217687074e-05,
+      "loss": 0.5341,
+      "step": 9327
+    },
+    {
+      "epoch": 11.93984,
+      "grad_norm": 0.9074930548667908,
+      "learning_rate": 3.1358543417366946e-05,
+      "loss": 0.539,
+      "step": 9328
+    },
+    {
+      "epoch": 11.94112,
+      "grad_norm": 0.9066956639289856,
+      "learning_rate": 3.135654261704682e-05,
+      "loss": 0.5496,
+      "step": 9329
+    },
+    {
+      "epoch": 11.9424,
+      "grad_norm": 0.9023146033287048,
+      "learning_rate": 3.135454181672669e-05,
+      "loss": 0.5679,
+      "step": 9330
+    },
+    {
+      "epoch": 11.94368,
+      "grad_norm": 0.9154323935508728,
+      "learning_rate": 3.135254101640656e-05,
+      "loss": 0.539,
+      "step": 9331
+    },
+    {
+      "epoch": 11.94496,
+      "grad_norm": 0.9193128347396851,
+      "learning_rate": 3.135054021608644e-05,
+      "loss": 0.5444,
+      "step": 9332
+    },
+    {
+      "epoch": 11.94624,
+      "grad_norm": 0.9258593916893005,
+      "learning_rate": 3.134853941576631e-05,
+      "loss": 0.5432,
+      "step": 9333
+    },
+    {
+      "epoch": 11.94752,
+      "grad_norm": 0.9019486904144287,
+      "learning_rate": 3.1346538615446184e-05,
+      "loss": 0.5888,
+      "step": 9334
+    },
+    {
+      "epoch": 11.9488,
+      "grad_norm": 0.9071494340896606,
+      "learning_rate": 3.134453781512605e-05,
+      "loss": 0.5334,
+      "step": 9335
+    },
+    {
+      "epoch": 11.95008,
+      "grad_norm": 0.8721327185630798,
+      "learning_rate": 3.134253701480592e-05,
+      "loss": 0.5885,
+      "step": 9336
+    },
+    {
+      "epoch": 11.95136,
+      "grad_norm": 0.9365826845169067,
+      "learning_rate": 3.134053621448579e-05,
+      "loss": 0.5799,
+      "step": 9337
+    },
+    {
+      "epoch": 11.95264,
+      "grad_norm": 0.9165831804275513,
+      "learning_rate": 3.1338535414165664e-05,
+      "loss": 0.5104,
+      "step": 9338
+    },
+    {
+      "epoch": 11.95392,
+      "grad_norm": 0.8874984979629517,
+      "learning_rate": 3.133653461384554e-05,
+      "loss": 0.5204,
+      "step": 9339
+    },
+    {
+      "epoch": 11.9552,
+      "grad_norm": 0.8862015604972839,
+      "learning_rate": 3.1334533813525415e-05,
+      "loss": 0.5502,
+      "step": 9340
+    },
+    {
+      "epoch": 11.956479999999999,
+      "grad_norm": 0.9116657376289368,
+      "learning_rate": 3.1332533013205287e-05,
+      "loss": 0.5619,
+      "step": 9341
+    },
+    {
+      "epoch": 11.95776,
+      "grad_norm": 0.8775522112846375,
+      "learning_rate": 3.133053221288516e-05,
+      "loss": 0.526,
+      "step": 9342
+    },
+    {
+      "epoch": 11.95904,
+      "grad_norm": 0.8832023739814758,
+      "learning_rate": 3.1328531412565023e-05,
+      "loss": 0.5192,
+      "step": 9343
+    },
+    {
+      "epoch": 11.96032,
+      "grad_norm": 0.8574889302253723,
+      "learning_rate": 3.1326530612244895e-05,
+      "loss": 0.5594,
+      "step": 9344
+    },
+    {
+      "epoch": 11.9616,
+      "grad_norm": 0.8446320295333862,
+      "learning_rate": 3.132452981192477e-05,
+      "loss": 0.5316,
+      "step": 9345
+    },
+    {
+      "epoch": 11.96288,
+      "grad_norm": 0.8501649498939514,
+      "learning_rate": 3.1322529011604646e-05,
+      "loss": 0.5666,
+      "step": 9346
+    },
+    {
+      "epoch": 11.96416,
+      "grad_norm": 0.9033554196357727,
+      "learning_rate": 3.132052821128452e-05,
+      "loss": 0.5631,
+      "step": 9347
+    },
+    {
+      "epoch": 11.96544,
+      "grad_norm": 0.9225219488143921,
+      "learning_rate": 3.131852741096439e-05,
+      "loss": 0.573,
+      "step": 9348
+    },
+    {
+      "epoch": 11.96672,
+      "grad_norm": 0.9399309754371643,
+      "learning_rate": 3.131652661064426e-05,
+      "loss": 0.6026,
+      "step": 9349
+    },
+    {
+      "epoch": 11.968,
+      "grad_norm": 0.8722439408302307,
+      "learning_rate": 3.131452581032413e-05,
+      "loss": 0.5392,
+      "step": 9350
+    },
+    {
+      "epoch": 11.96928,
+      "grad_norm": 0.8984782099723816,
+      "learning_rate": 3.1312525010004e-05,
+      "loss": 0.5778,
+      "step": 9351
+    },
+    {
+      "epoch": 11.97056,
+      "grad_norm": 0.9469655156135559,
+      "learning_rate": 3.131052420968387e-05,
+      "loss": 0.6142,
+      "step": 9352
+    },
+    {
+      "epoch": 11.97184,
+      "grad_norm": 0.8549695014953613,
+      "learning_rate": 3.130852340936375e-05,
+      "loss": 0.5452,
+      "step": 9353
+    },
+    {
+      "epoch": 11.97312,
+      "grad_norm": 0.913131058216095,
+      "learning_rate": 3.130652260904362e-05,
+      "loss": 0.5255,
+      "step": 9354
+    },
+    {
+      "epoch": 11.9744,
+      "grad_norm": 0.8826735615730286,
+      "learning_rate": 3.130452180872349e-05,
+      "loss": 0.5603,
+      "step": 9355
+    },
+    {
+      "epoch": 11.97568,
+      "grad_norm": 0.8853660225868225,
+      "learning_rate": 3.1302521008403364e-05,
+      "loss": 0.5223,
+      "step": 9356
+    },
+    {
+      "epoch": 11.97696,
+      "grad_norm": 0.9496334791183472,
+      "learning_rate": 3.1300520208083236e-05,
+      "loss": 0.6112,
+      "step": 9357
+    },
+    {
+      "epoch": 11.97824,
+      "grad_norm": 0.9710426926612854,
+      "learning_rate": 3.129851940776311e-05,
+      "loss": 0.5996,
+      "step": 9358
+    },
+    {
+      "epoch": 11.97952,
+      "grad_norm": 0.9180809855461121,
+      "learning_rate": 3.129651860744297e-05,
+      "loss": 0.5501,
+      "step": 9359
+    },
+    {
+      "epoch": 11.9808,
+      "grad_norm": 0.875312089920044,
+      "learning_rate": 3.129451780712285e-05,
+      "loss": 0.5474,
+      "step": 9360
+    },
+    {
+      "epoch": 11.98208,
+      "grad_norm": 0.9497184753417969,
+      "learning_rate": 3.1292517006802724e-05,
+      "loss": 0.6125,
+      "step": 9361
+    },
+    {
+      "epoch": 11.98336,
+      "grad_norm": 0.9112341403961182,
+      "learning_rate": 3.1290516206482596e-05,
+      "loss": 0.5461,
+      "step": 9362
+    },
+    {
+      "epoch": 11.98464,
+      "grad_norm": 0.8672278523445129,
+      "learning_rate": 3.128851540616247e-05,
+      "loss": 0.5453,
+      "step": 9363
+    },
+    {
+      "epoch": 11.98592,
+      "grad_norm": 0.853042721748352,
+      "learning_rate": 3.128651460584234e-05,
+      "loss": 0.5445,
+      "step": 9364
+    },
+    {
+      "epoch": 11.9872,
+      "grad_norm": 0.936596155166626,
+      "learning_rate": 3.128451380552221e-05,
+      "loss": 0.5467,
+      "step": 9365
+    },
+    {
+      "epoch": 11.98848,
+      "grad_norm": 0.901436984539032,
+      "learning_rate": 3.128251300520208e-05,
+      "loss": 0.5904,
+      "step": 9366
+    },
+    {
+      "epoch": 11.98976,
+      "grad_norm": 0.9036868214607239,
+      "learning_rate": 3.1280512204881955e-05,
+      "loss": 0.5724,
+      "step": 9367
+    },
+    {
+      "epoch": 11.99104,
+      "grad_norm": 0.8857167959213257,
+      "learning_rate": 3.127851140456183e-05,
+      "loss": 0.5707,
+      "step": 9368
+    },
+    {
+      "epoch": 11.99232,
+      "grad_norm": 0.8939414024353027,
+      "learning_rate": 3.12765106042417e-05,
+      "loss": 0.6032,
+      "step": 9369
+    },
+    {
+      "epoch": 11.9936,
+      "grad_norm": 0.9171572327613831,
+      "learning_rate": 3.127450980392157e-05,
+      "loss": 0.5319,
+      "step": 9370
+    },
+    {
+      "epoch": 11.99488,
+      "grad_norm": 0.92975914478302,
+      "learning_rate": 3.127250900360144e-05,
+      "loss": 0.6055,
+      "step": 9371
+    },
+    {
+      "epoch": 11.99616,
+      "grad_norm": 0.9182210564613342,
+      "learning_rate": 3.1270508203281314e-05,
+      "loss": 0.6125,
+      "step": 9372
+    },
+    {
+      "epoch": 11.99744,
+      "grad_norm": 0.8801096081733704,
+      "learning_rate": 3.1268507402961186e-05,
+      "loss": 0.5197,
+      "step": 9373
+    },
+    {
+      "epoch": 11.99872,
+      "grad_norm": 0.9054288268089294,
+      "learning_rate": 3.126650660264106e-05,
+      "loss": 0.5574,
+      "step": 9374
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": Infinity,
+      "learning_rate": 3.126650660264106e-05,
+      "loss": 0.9918,
+      "step": 9375
+    },
+    {
+      "epoch": 12.00128,
+      "grad_norm": 0.8894946575164795,
+      "learning_rate": 3.126450580232093e-05,
+      "loss": 0.5222,
+      "step": 9376
+    },
+    {
+      "epoch": 12.00256,
+      "grad_norm": 0.8508769273757935,
+      "learning_rate": 3.12625050020008e-05,
+      "loss": 0.521,
+      "step": 9377
+    },
+    {
+      "epoch": 12.00384,
+      "grad_norm": 0.8415517807006836,
+      "learning_rate": 3.1260504201680673e-05,
+      "loss": 0.5538,
+      "step": 9378
+    },
+    {
+      "epoch": 12.00512,
+      "grad_norm": 0.8784507513046265,
+      "learning_rate": 3.1258503401360545e-05,
+      "loss": 0.5281,
+      "step": 9379
+    },
+    {
+      "epoch": 12.0064,
+      "grad_norm": 0.8953071236610413,
+      "learning_rate": 3.125650260104042e-05,
+      "loss": 0.5587,
+      "step": 9380
+    },
+    {
+      "epoch": 12.00768,
+      "grad_norm": 0.9295496344566345,
+      "learning_rate": 3.125450180072029e-05,
+      "loss": 0.5007,
+      "step": 9381
+    },
+    {
+      "epoch": 12.00896,
+      "grad_norm": 0.8452316522598267,
+      "learning_rate": 3.125250100040016e-05,
+      "loss": 0.5127,
+      "step": 9382
+    },
+    {
+      "epoch": 12.01024,
+      "grad_norm": 0.9824538826942444,
+      "learning_rate": 3.125050020008003e-05,
+      "loss": 0.6222,
+      "step": 9383
+    },
+    {
+      "epoch": 12.01152,
+      "grad_norm": 0.8931326270103455,
+      "learning_rate": 3.1248499399759905e-05,
+      "loss": 0.5504,
+      "step": 9384
+    },
+    {
+      "epoch": 12.0128,
+      "grad_norm": 0.8844808340072632,
+      "learning_rate": 3.1246498599439776e-05,
+      "loss": 0.5371,
+      "step": 9385
+    },
+    {
+      "epoch": 12.01408,
+      "grad_norm": 0.8983806371688843,
+      "learning_rate": 3.124449779911965e-05,
+      "loss": 0.5548,
+      "step": 9386
+    },
+    {
+      "epoch": 12.01536,
+      "grad_norm": 0.9335402250289917,
+      "learning_rate": 3.124249699879952e-05,
+      "loss": 0.5677,
+      "step": 9387
+    },
+    {
+      "epoch": 12.01664,
+      "grad_norm": 0.928325355052948,
+      "learning_rate": 3.124049619847939e-05,
+      "loss": 0.5343,
+      "step": 9388
+    },
+    {
+      "epoch": 12.01792,
+      "grad_norm": 0.9786158800125122,
+      "learning_rate": 3.123849539815927e-05,
+      "loss": 0.5708,
+      "step": 9389
+    },
+    {
+      "epoch": 12.0192,
+      "grad_norm": 0.8691074252128601,
+      "learning_rate": 3.1236494597839136e-05,
+      "loss": 0.5301,
+      "step": 9390
+    },
+    {
+      "epoch": 12.02048,
+      "grad_norm": 0.9225201606750488,
+      "learning_rate": 3.123449379751901e-05,
+      "loss": 0.5683,
+      "step": 9391
+    },
+    {
+      "epoch": 12.02176,
+      "grad_norm": 0.9116083383560181,
+      "learning_rate": 3.123249299719888e-05,
+      "loss": 0.5318,
+      "step": 9392
+    },
+    {
+      "epoch": 12.02304,
+      "grad_norm": 0.9338945150375366,
+      "learning_rate": 3.123049219687875e-05,
+      "loss": 0.5751,
+      "step": 9393
+    },
+    {
+      "epoch": 12.02432,
+      "grad_norm": 0.8985954523086548,
+      "learning_rate": 3.122849139655862e-05,
+      "loss": 0.5471,
+      "step": 9394
+    },
+    {
+      "epoch": 12.0256,
+      "grad_norm": 0.9236059188842773,
+      "learning_rate": 3.1226490596238495e-05,
+      "loss": 0.54,
+      "step": 9395
+    },
+    {
+      "epoch": 12.02688,
+      "grad_norm": 0.902243971824646,
+      "learning_rate": 3.1224489795918374e-05,
+      "loss": 0.5632,
+      "step": 9396
+    },
+    {
+      "epoch": 12.02816,
+      "grad_norm": 0.9559075832366943,
+      "learning_rate": 3.1222488995598245e-05,
+      "loss": 0.5201,
+      "step": 9397
+    },
+    {
+      "epoch": 12.02944,
+      "grad_norm": 0.8943519592285156,
+      "learning_rate": 3.122048819527811e-05,
+      "loss": 0.5407,
+      "step": 9398
+    },
+    {
+      "epoch": 12.03072,
+      "grad_norm": 0.8215572834014893,
+      "learning_rate": 3.121848739495798e-05,
+      "loss": 0.4806,
+      "step": 9399
+    },
+    {
+      "epoch": 12.032,
+      "grad_norm": 0.9161409139633179,
+      "learning_rate": 3.1216486594637854e-05,
+      "loss": 0.5971,
+      "step": 9400
+    },
+    {
+      "epoch": 12.03328,
+      "grad_norm": 0.9485968947410583,
+      "learning_rate": 3.1214485794317726e-05,
+      "loss": 0.5852,
+      "step": 9401
+    },
+    {
+      "epoch": 12.03456,
+      "grad_norm": 0.8830835223197937,
+      "learning_rate": 3.12124849939976e-05,
+      "loss": 0.475,
+      "step": 9402
+    },
+    {
+      "epoch": 12.03584,
+      "grad_norm": 0.9634863138198853,
+      "learning_rate": 3.1210484193677477e-05,
+      "loss": 0.5458,
+      "step": 9403
+    },
+    {
+      "epoch": 12.03712,
+      "grad_norm": 0.9121254086494446,
+      "learning_rate": 3.120848339335735e-05,
+      "loss": 0.5025,
+      "step": 9404
+    },
+    {
+      "epoch": 12.0384,
+      "grad_norm": 1.0343906879425049,
+      "learning_rate": 3.120648259303722e-05,
+      "loss": 0.5407,
+      "step": 9405
+    },
+    {
+      "epoch": 12.03968,
+      "grad_norm": 0.892846941947937,
+      "learning_rate": 3.1204481792717085e-05,
+      "loss": 0.5607,
+      "step": 9406
+    },
+    {
+      "epoch": 12.04096,
+      "grad_norm": 0.8945642113685608,
+      "learning_rate": 3.120248099239696e-05,
+      "loss": 0.5562,
+      "step": 9407
+    },
+    {
+      "epoch": 12.04224,
+      "grad_norm": 0.8921273350715637,
+      "learning_rate": 3.120048019207683e-05,
+      "loss": 0.5033,
+      "step": 9408
+    },
+    {
+      "epoch": 12.043520000000001,
+      "grad_norm": 0.8849148154258728,
+      "learning_rate": 3.11984793917567e-05,
+      "loss": 0.558,
+      "step": 9409
+    },
+    {
+      "epoch": 12.0448,
+      "grad_norm": 0.8761171698570251,
+      "learning_rate": 3.119647859143658e-05,
+      "loss": 0.5437,
+      "step": 9410
+    },
+    {
+      "epoch": 12.04608,
+      "grad_norm": 0.9475454092025757,
+      "learning_rate": 3.119447779111645e-05,
+      "loss": 0.5873,
+      "step": 9411
+    },
+    {
+      "epoch": 12.04736,
+      "grad_norm": 0.9096229076385498,
+      "learning_rate": 3.119247699079632e-05,
+      "loss": 0.5356,
+      "step": 9412
+    },
+    {
+      "epoch": 12.04864,
+      "grad_norm": 0.8909059166908264,
+      "learning_rate": 3.1190476190476195e-05,
+      "loss": 0.5258,
+      "step": 9413
+    },
+    {
+      "epoch": 12.04992,
+      "grad_norm": 0.8790292739868164,
+      "learning_rate": 3.118847539015606e-05,
+      "loss": 0.5287,
+      "step": 9414
+    },
+    {
+      "epoch": 12.0512,
+      "grad_norm": 0.9126578569412231,
+      "learning_rate": 3.118647458983593e-05,
+      "loss": 0.5081,
+      "step": 9415
+    },
+    {
+      "epoch": 12.05248,
+      "grad_norm": 0.9787217378616333,
+      "learning_rate": 3.1184473789515804e-05,
+      "loss": 0.5798,
+      "step": 9416
+    },
+    {
+      "epoch": 12.05376,
+      "grad_norm": 0.8833532333374023,
+      "learning_rate": 3.118247298919568e-05,
+      "loss": 0.5202,
+      "step": 9417
+    },
+    {
+      "epoch": 12.05504,
+      "grad_norm": 0.8925086855888367,
+      "learning_rate": 3.1180472188875554e-05,
+      "loss": 0.5235,
+      "step": 9418
+    },
+    {
+      "epoch": 12.05632,
+      "grad_norm": 0.9011042714118958,
+      "learning_rate": 3.1178471388555426e-05,
+      "loss": 0.54,
+      "step": 9419
+    },
+    {
+      "epoch": 12.0576,
+      "grad_norm": 0.8985966444015503,
+      "learning_rate": 3.11764705882353e-05,
+      "loss": 0.5078,
+      "step": 9420
+    },
+    {
+      "epoch": 12.05888,
+      "grad_norm": 0.9557191729545593,
+      "learning_rate": 3.117446978791517e-05,
+      "loss": 0.5882,
+      "step": 9421
+    },
+    {
+      "epoch": 12.06016,
+      "grad_norm": 0.9084435701370239,
+      "learning_rate": 3.1172468987595035e-05,
+      "loss": 0.5736,
+      "step": 9422
+    },
+    {
+      "epoch": 12.06144,
+      "grad_norm": 0.8469879627227783,
+      "learning_rate": 3.117046818727491e-05,
+      "loss": 0.5194,
+      "step": 9423
+    },
+    {
+      "epoch": 12.06272,
+      "grad_norm": 0.9563634991645813,
+      "learning_rate": 3.1168467386954786e-05,
+      "loss": 0.5507,
+      "step": 9424
+    },
+    {
+      "epoch": 12.064,
+      "grad_norm": 0.9139657020568848,
+      "learning_rate": 3.116646658663466e-05,
+      "loss": 0.5972,
+      "step": 9425
+    },
+    {
+      "epoch": 12.06528,
+      "grad_norm": 0.8986368775367737,
+      "learning_rate": 3.116446578631453e-05,
+      "loss": 0.5257,
+      "step": 9426
+    },
+    {
+      "epoch": 12.06656,
+      "grad_norm": 0.9286632537841797,
+      "learning_rate": 3.11624649859944e-05,
+      "loss": 0.5412,
+      "step": 9427
+    },
+    {
+      "epoch": 12.06784,
+      "grad_norm": 0.9259238243103027,
+      "learning_rate": 3.116046418567427e-05,
+      "loss": 0.5729,
+      "step": 9428
+    },
+    {
+      "epoch": 12.06912,
+      "grad_norm": 0.8716070652008057,
+      "learning_rate": 3.1158463385354145e-05,
+      "loss": 0.5373,
+      "step": 9429
+    },
+    {
+      "epoch": 12.0704,
+      "grad_norm": 0.937944233417511,
+      "learning_rate": 3.115646258503401e-05,
+      "loss": 0.5433,
+      "step": 9430
+    },
+    {
+      "epoch": 12.07168,
+      "grad_norm": 0.8506153225898743,
+      "learning_rate": 3.115446178471389e-05,
+      "loss": 0.4806,
+      "step": 9431
+    },
+    {
+      "epoch": 12.07296,
+      "grad_norm": 0.9485256671905518,
+      "learning_rate": 3.115246098439376e-05,
+      "loss": 0.5236,
+      "step": 9432
+    },
+    {
+      "epoch": 12.07424,
+      "grad_norm": 0.9169779419898987,
+      "learning_rate": 3.115046018407363e-05,
+      "loss": 0.5357,
+      "step": 9433
+    },
+    {
+      "epoch": 12.07552,
+      "grad_norm": 0.9298108816146851,
+      "learning_rate": 3.1148459383753504e-05,
+      "loss": 0.5527,
+      "step": 9434
+    },
+    {
+      "epoch": 12.0768,
+      "grad_norm": 0.8851594924926758,
+      "learning_rate": 3.1146458583433376e-05,
+      "loss": 0.4717,
+      "step": 9435
+    },
+    {
+      "epoch": 12.07808,
+      "grad_norm": 0.9402860999107361,
+      "learning_rate": 3.114445778311325e-05,
+      "loss": 0.5484,
+      "step": 9436
+    },
+    {
+      "epoch": 12.07936,
+      "grad_norm": 0.9027056097984314,
+      "learning_rate": 3.114245698279312e-05,
+      "loss": 0.5506,
+      "step": 9437
+    },
+    {
+      "epoch": 12.08064,
+      "grad_norm": 0.9592556953430176,
+      "learning_rate": 3.114045618247299e-05,
+      "loss": 0.5557,
+      "step": 9438
+    },
+    {
+      "epoch": 12.08192,
+      "grad_norm": 0.9053078293800354,
+      "learning_rate": 3.1138455382152863e-05,
+      "loss": 0.5468,
+      "step": 9439
+    },
+    {
+      "epoch": 12.0832,
+      "grad_norm": 0.941765546798706,
+      "learning_rate": 3.1136454581832735e-05,
+      "loss": 0.5678,
+      "step": 9440
+    },
+    {
+      "epoch": 12.08448,
+      "grad_norm": 0.9676002264022827,
+      "learning_rate": 3.113445378151261e-05,
+      "loss": 0.5451,
+      "step": 9441
+    },
+    {
+      "epoch": 12.08576,
+      "grad_norm": 0.9788287878036499,
+      "learning_rate": 3.113245298119248e-05,
+      "loss": 0.6131,
+      "step": 9442
+    },
+    {
+      "epoch": 12.08704,
+      "grad_norm": 0.9841639399528503,
+      "learning_rate": 3.113045218087235e-05,
+      "loss": 0.5438,
+      "step": 9443
+    },
+    {
+      "epoch": 12.08832,
+      "grad_norm": 0.955268144607544,
+      "learning_rate": 3.112845138055222e-05,
+      "loss": 0.5622,
+      "step": 9444
+    },
+    {
+      "epoch": 12.0896,
+      "grad_norm": 0.905160129070282,
+      "learning_rate": 3.1126450580232095e-05,
+      "loss": 0.5138,
+      "step": 9445
+    },
+    {
+      "epoch": 12.09088,
+      "grad_norm": 0.9529837369918823,
+      "learning_rate": 3.1124449779911966e-05,
+      "loss": 0.5891,
+      "step": 9446
+    },
+    {
+      "epoch": 12.09216,
+      "grad_norm": 1.0116820335388184,
+      "learning_rate": 3.112244897959184e-05,
+      "loss": 0.5347,
+      "step": 9447
+    },
+    {
+      "epoch": 12.09344,
+      "grad_norm": 0.9079636335372925,
+      "learning_rate": 3.112044817927171e-05,
+      "loss": 0.5178,
+      "step": 9448
+    },
+    {
+      "epoch": 12.09472,
+      "grad_norm": 0.9194412231445312,
+      "learning_rate": 3.111844737895158e-05,
+      "loss": 0.5488,
+      "step": 9449
+    },
+    {
+      "epoch": 12.096,
+      "grad_norm": 0.9513563513755798,
+      "learning_rate": 3.1116446578631454e-05,
+      "loss": 0.5472,
+      "step": 9450
+    },
+    {
+      "epoch": 12.09728,
+      "grad_norm": 0.9122597575187683,
+      "learning_rate": 3.1114445778311326e-05,
+      "loss": 0.5887,
+      "step": 9451
+    },
+    {
+      "epoch": 12.09856,
+      "grad_norm": 0.9185177683830261,
+      "learning_rate": 3.11124449779912e-05,
+      "loss": 0.5636,
+      "step": 9452
+    },
+    {
+      "epoch": 12.09984,
+      "grad_norm": 0.923226535320282,
+      "learning_rate": 3.111044417767107e-05,
+      "loss": 0.5302,
+      "step": 9453
+    },
+    {
+      "epoch": 12.10112,
+      "grad_norm": 0.8814284801483154,
+      "learning_rate": 3.110844337735094e-05,
+      "loss": 0.522,
+      "step": 9454
+    },
+    {
+      "epoch": 12.1024,
+      "grad_norm": 0.9242017865180969,
+      "learning_rate": 3.110644257703081e-05,
+      "loss": 0.5458,
+      "step": 9455
+    },
+    {
+      "epoch": 12.10368,
+      "grad_norm": 0.9658417701721191,
+      "learning_rate": 3.1104441776710685e-05,
+      "loss": 0.5912,
+      "step": 9456
+    },
+    {
+      "epoch": 12.10496,
+      "grad_norm": 0.90215003490448,
+      "learning_rate": 3.110244097639056e-05,
+      "loss": 0.5586,
+      "step": 9457
+    },
+    {
+      "epoch": 12.10624,
+      "grad_norm": 0.9708035588264465,
+      "learning_rate": 3.110044017607043e-05,
+      "loss": 0.5718,
+      "step": 9458
+    },
+    {
+      "epoch": 12.10752,
+      "grad_norm": 0.8893763422966003,
+      "learning_rate": 3.10984393757503e-05,
+      "loss": 0.532,
+      "step": 9459
+    },
+    {
+      "epoch": 12.1088,
+      "grad_norm": 0.9912137389183044,
+      "learning_rate": 3.109643857543017e-05,
+      "loss": 0.5976,
+      "step": 9460
+    },
+    {
+      "epoch": 12.11008,
+      "grad_norm": 0.89457768201828,
+      "learning_rate": 3.1094437775110044e-05,
+      "loss": 0.5394,
+      "step": 9461
+    },
+    {
+      "epoch": 12.11136,
+      "grad_norm": 0.9052734971046448,
+      "learning_rate": 3.1092436974789916e-05,
+      "loss": 0.5318,
+      "step": 9462
+    },
+    {
+      "epoch": 12.11264,
+      "grad_norm": 0.9517974257469177,
+      "learning_rate": 3.109043617446979e-05,
+      "loss": 0.5494,
+      "step": 9463
+    },
+    {
+      "epoch": 12.11392,
+      "grad_norm": 0.921795666217804,
+      "learning_rate": 3.108843537414966e-05,
+      "loss": 0.4871,
+      "step": 9464
+    },
+    {
+      "epoch": 12.1152,
+      "grad_norm": 0.8893502950668335,
+      "learning_rate": 3.108643457382953e-05,
+      "loss": 0.5022,
+      "step": 9465
+    },
+    {
+      "epoch": 12.11648,
+      "grad_norm": 1.020261287689209,
+      "learning_rate": 3.1084433773509404e-05,
+      "loss": 0.6282,
+      "step": 9466
+    },
+    {
+      "epoch": 12.11776,
+      "grad_norm": 0.9455602169036865,
+      "learning_rate": 3.108243297318928e-05,
+      "loss": 0.5625,
+      "step": 9467
+    },
+    {
+      "epoch": 12.11904,
+      "grad_norm": 0.9229424595832825,
+      "learning_rate": 3.108043217286915e-05,
+      "loss": 0.5478,
+      "step": 9468
+    },
+    {
+      "epoch": 12.12032,
+      "grad_norm": 0.9324727654457092,
+      "learning_rate": 3.107843137254902e-05,
+      "loss": 0.5842,
+      "step": 9469
+    },
+    {
+      "epoch": 12.1216,
+      "grad_norm": 0.8508306741714478,
+      "learning_rate": 3.107643057222889e-05,
+      "loss": 0.4829,
+      "step": 9470
+    },
+    {
+      "epoch": 12.12288,
+      "grad_norm": 0.9504947662353516,
+      "learning_rate": 3.107442977190876e-05,
+      "loss": 0.5494,
+      "step": 9471
+    },
+    {
+      "epoch": 12.12416,
+      "grad_norm": 0.8809190988540649,
+      "learning_rate": 3.1072428971588635e-05,
+      "loss": 0.518,
+      "step": 9472
+    },
+    {
+      "epoch": 12.12544,
+      "grad_norm": 0.9718751311302185,
+      "learning_rate": 3.1070428171268507e-05,
+      "loss": 0.5083,
+      "step": 9473
+    },
+    {
+      "epoch": 12.12672,
+      "grad_norm": 0.9734033346176147,
+      "learning_rate": 3.1068427370948385e-05,
+      "loss": 0.5977,
+      "step": 9474
+    },
+    {
+      "epoch": 12.128,
+      "grad_norm": 0.892812967300415,
+      "learning_rate": 3.106642657062826e-05,
+      "loss": 0.5214,
+      "step": 9475
+    },
+    {
+      "epoch": 12.12928,
+      "grad_norm": 0.9172582626342773,
+      "learning_rate": 3.106442577030812e-05,
+      "loss": 0.5311,
+      "step": 9476
+    },
+    {
+      "epoch": 12.13056,
+      "grad_norm": 0.8705976009368896,
+      "learning_rate": 3.1062424969987994e-05,
+      "loss": 0.4711,
+      "step": 9477
+    },
+    {
+      "epoch": 12.13184,
+      "grad_norm": 0.9229205250740051,
+      "learning_rate": 3.1060424169667866e-05,
+      "loss": 0.5198,
+      "step": 9478
+    },
+    {
+      "epoch": 12.13312,
+      "grad_norm": 0.9089353680610657,
+      "learning_rate": 3.105842336934774e-05,
+      "loss": 0.5156,
+      "step": 9479
+    },
+    {
+      "epoch": 12.1344,
+      "grad_norm": 0.9263928532600403,
+      "learning_rate": 3.105642256902761e-05,
+      "loss": 0.4731,
+      "step": 9480
+    },
+    {
+      "epoch": 12.13568,
+      "grad_norm": 0.8989574313163757,
+      "learning_rate": 3.105442176870749e-05,
+      "loss": 0.5337,
+      "step": 9481
+    },
+    {
+      "epoch": 12.13696,
+      "grad_norm": 0.9506709575653076,
+      "learning_rate": 3.105242096838736e-05,
+      "loss": 0.5357,
+      "step": 9482
+    },
+    {
+      "epoch": 12.13824,
+      "grad_norm": 0.9234217405319214,
+      "learning_rate": 3.105042016806723e-05,
+      "loss": 0.5471,
+      "step": 9483
+    },
+    {
+      "epoch": 12.13952,
+      "grad_norm": 0.8666446805000305,
+      "learning_rate": 3.10484193677471e-05,
+      "loss": 0.5298,
+      "step": 9484
+    },
+    {
+      "epoch": 12.1408,
+      "grad_norm": 0.9469799399375916,
+      "learning_rate": 3.104641856742697e-05,
+      "loss": 0.546,
+      "step": 9485
+    },
+    {
+      "epoch": 12.14208,
+      "grad_norm": 0.8840956091880798,
+      "learning_rate": 3.104441776710684e-05,
+      "loss": 0.5417,
+      "step": 9486
+    },
+    {
+      "epoch": 12.14336,
+      "grad_norm": 0.8798984885215759,
+      "learning_rate": 3.104241696678671e-05,
+      "loss": 0.5403,
+      "step": 9487
+    },
+    {
+      "epoch": 12.14464,
+      "grad_norm": 0.9367090463638306,
+      "learning_rate": 3.104041616646659e-05,
+      "loss": 0.562,
+      "step": 9488
+    },
+    {
+      "epoch": 12.14592,
+      "grad_norm": 0.8626176118850708,
+      "learning_rate": 3.103841536614646e-05,
+      "loss": 0.5044,
+      "step": 9489
+    },
+    {
+      "epoch": 12.1472,
+      "grad_norm": 0.953429102897644,
+      "learning_rate": 3.1036414565826335e-05,
+      "loss": 0.5363,
+      "step": 9490
+    },
+    {
+      "epoch": 12.14848,
+      "grad_norm": 0.847478449344635,
+      "learning_rate": 3.103441376550621e-05,
+      "loss": 0.4921,
+      "step": 9491
+    },
+    {
+      "epoch": 12.14976,
+      "grad_norm": 0.8839147090911865,
+      "learning_rate": 3.103241296518607e-05,
+      "loss": 0.5174,
+      "step": 9492
+    },
+    {
+      "epoch": 12.15104,
+      "grad_norm": 0.952755868434906,
+      "learning_rate": 3.1030412164865944e-05,
+      "loss": 0.5461,
+      "step": 9493
+    },
+    {
+      "epoch": 12.15232,
+      "grad_norm": 0.8509531021118164,
+      "learning_rate": 3.1028411364545816e-05,
+      "loss": 0.5218,
+      "step": 9494
+    },
+    {
+      "epoch": 12.1536,
+      "grad_norm": 0.9300630688667297,
+      "learning_rate": 3.1026410564225694e-05,
+      "loss": 0.5625,
+      "step": 9495
+    },
+    {
+      "epoch": 12.15488,
+      "grad_norm": 0.9531006217002869,
+      "learning_rate": 3.1024409763905566e-05,
+      "loss": 0.5831,
+      "step": 9496
+    },
+    {
+      "epoch": 12.15616,
+      "grad_norm": 0.9329797625541687,
+      "learning_rate": 3.102240896358544e-05,
+      "loss": 0.539,
+      "step": 9497
+    },
+    {
+      "epoch": 12.15744,
+      "grad_norm": 0.855829656124115,
+      "learning_rate": 3.102040816326531e-05,
+      "loss": 0.4812,
+      "step": 9498
+    },
+    {
+      "epoch": 12.15872,
+      "grad_norm": 0.8939371705055237,
+      "learning_rate": 3.101840736294518e-05,
+      "loss": 0.5208,
+      "step": 9499
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 0.9188044667243958,
+      "learning_rate": 3.101640656262505e-05,
+      "loss": 0.545,
+      "step": 9500
+    },
+    {
+      "epoch": 12.16128,
+      "grad_norm": 0.8524420261383057,
+      "learning_rate": 3.101440576230492e-05,
+      "loss": 0.5028,
+      "step": 9501
+    },
+    {
+      "epoch": 12.16256,
+      "grad_norm": 0.906829297542572,
+      "learning_rate": 3.10124049619848e-05,
+      "loss": 0.5557,
+      "step": 9502
+    },
+    {
+      "epoch": 12.16384,
+      "grad_norm": 0.904629647731781,
+      "learning_rate": 3.101040416166467e-05,
+      "loss": 0.5265,
+      "step": 9503
+    },
+    {
+      "epoch": 12.16512,
+      "grad_norm": 0.930142343044281,
+      "learning_rate": 3.100840336134454e-05,
+      "loss": 0.5353,
+      "step": 9504
+    },
+    {
+      "epoch": 12.1664,
+      "grad_norm": 0.9185201525688171,
+      "learning_rate": 3.100640256102441e-05,
+      "loss": 0.5448,
+      "step": 9505
+    },
+    {
+      "epoch": 12.16768,
+      "grad_norm": 0.9382092952728271,
+      "learning_rate": 3.1004401760704285e-05,
+      "loss": 0.5498,
+      "step": 9506
+    },
+    {
+      "epoch": 12.16896,
+      "grad_norm": 0.9041383862495422,
+      "learning_rate": 3.1002400960384156e-05,
+      "loss": 0.4895,
+      "step": 9507
+    },
+    {
+      "epoch": 12.17024,
+      "grad_norm": 0.921959638595581,
+      "learning_rate": 3.100040016006402e-05,
+      "loss": 0.5513,
+      "step": 9508
+    },
+    {
+      "epoch": 12.17152,
+      "grad_norm": 0.9161378145217896,
+      "learning_rate": 3.09983993597439e-05,
+      "loss": 0.5745,
+      "step": 9509
+    },
+    {
+      "epoch": 12.1728,
+      "grad_norm": 0.9752117991447449,
+      "learning_rate": 3.099639855942377e-05,
+      "loss": 0.5436,
+      "step": 9510
+    },
+    {
+      "epoch": 12.17408,
+      "grad_norm": 0.8904070854187012,
+      "learning_rate": 3.0994397759103644e-05,
+      "loss": 0.5351,
+      "step": 9511
+    },
+    {
+      "epoch": 12.17536,
+      "grad_norm": 0.8877695202827454,
+      "learning_rate": 3.0992396958783516e-05,
+      "loss": 0.5604,
+      "step": 9512
+    },
+    {
+      "epoch": 12.17664,
+      "grad_norm": 0.9066382050514221,
+      "learning_rate": 3.099039615846339e-05,
+      "loss": 0.5335,
+      "step": 9513
+    },
+    {
+      "epoch": 12.17792,
+      "grad_norm": 0.9197863936424255,
+      "learning_rate": 3.098839535814326e-05,
+      "loss": 0.5328,
+      "step": 9514
+    },
+    {
+      "epoch": 12.1792,
+      "grad_norm": 0.9085410237312317,
+      "learning_rate": 3.098639455782313e-05,
+      "loss": 0.4907,
+      "step": 9515
+    },
+    {
+      "epoch": 12.18048,
+      "grad_norm": 0.9633046984672546,
+      "learning_rate": 3.0984393757503e-05,
+      "loss": 0.5437,
+      "step": 9516
+    },
+    {
+      "epoch": 12.18176,
+      "grad_norm": 0.9514236450195312,
+      "learning_rate": 3.0982392957182875e-05,
+      "loss": 0.5653,
+      "step": 9517
+    },
+    {
+      "epoch": 12.18304,
+      "grad_norm": 0.9399169087409973,
+      "learning_rate": 3.098039215686275e-05,
+      "loss": 0.521,
+      "step": 9518
+    },
+    {
+      "epoch": 12.18432,
+      "grad_norm": 0.9197844862937927,
+      "learning_rate": 3.097839135654262e-05,
+      "loss": 0.5644,
+      "step": 9519
+    },
+    {
+      "epoch": 12.1856,
+      "grad_norm": 0.8995923399925232,
+      "learning_rate": 3.097639055622249e-05,
+      "loss": 0.5357,
+      "step": 9520
+    },
+    {
+      "epoch": 12.18688,
+      "grad_norm": 0.9238200783729553,
+      "learning_rate": 3.097438975590236e-05,
+      "loss": 0.5414,
+      "step": 9521
+    },
+    {
+      "epoch": 12.18816,
+      "grad_norm": 0.9110772609710693,
+      "learning_rate": 3.0972388955582234e-05,
+      "loss": 0.5164,
+      "step": 9522
+    },
+    {
+      "epoch": 12.18944,
+      "grad_norm": 0.9395493865013123,
+      "learning_rate": 3.0970388155262106e-05,
+      "loss": 0.5542,
+      "step": 9523
+    },
+    {
+      "epoch": 12.19072,
+      "grad_norm": 0.8886218667030334,
+      "learning_rate": 3.096838735494198e-05,
+      "loss": 0.4889,
+      "step": 9524
+    },
+    {
+      "epoch": 12.192,
+      "grad_norm": 0.9301427006721497,
+      "learning_rate": 3.096638655462185e-05,
+      "loss": 0.5434,
+      "step": 9525
+    },
+    {
+      "epoch": 12.19328,
+      "grad_norm": 0.956852912902832,
+      "learning_rate": 3.096438575430172e-05,
+      "loss": 0.5286,
+      "step": 9526
+    },
+    {
+      "epoch": 12.19456,
+      "grad_norm": 0.9380202293395996,
+      "learning_rate": 3.0962384953981594e-05,
+      "loss": 0.5866,
+      "step": 9527
+    },
+    {
+      "epoch": 12.19584,
+      "grad_norm": 0.9663145542144775,
+      "learning_rate": 3.0960384153661465e-05,
+      "loss": 0.5784,
+      "step": 9528
+    },
+    {
+      "epoch": 12.19712,
+      "grad_norm": 0.9319117069244385,
+      "learning_rate": 3.095838335334134e-05,
+      "loss": 0.5328,
+      "step": 9529
+    },
+    {
+      "epoch": 12.1984,
+      "grad_norm": 0.9537164568901062,
+      "learning_rate": 3.0956382553021216e-05,
+      "loss": 0.5598,
+      "step": 9530
+    },
+    {
+      "epoch": 12.19968,
+      "grad_norm": 0.9373032450675964,
+      "learning_rate": 3.095438175270108e-05,
+      "loss": 0.5457,
+      "step": 9531
+    },
+    {
+      "epoch": 12.20096,
+      "grad_norm": 0.9559725522994995,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.5518,
+      "step": 9532
+    },
+    {
+      "epoch": 12.20224,
+      "grad_norm": 0.9501890540122986,
+      "learning_rate": 3.0950380152060825e-05,
+      "loss": 0.5554,
+      "step": 9533
+    },
+    {
+      "epoch": 12.20352,
+      "grad_norm": 0.9486271142959595,
+      "learning_rate": 3.0948379351740697e-05,
+      "loss": 0.5774,
+      "step": 9534
+    },
+    {
+      "epoch": 12.2048,
+      "grad_norm": 0.8942713737487793,
+      "learning_rate": 3.094637855142057e-05,
+      "loss": 0.539,
+      "step": 9535
+    },
+    {
+      "epoch": 12.20608,
+      "grad_norm": 0.9453210234642029,
+      "learning_rate": 3.094437775110044e-05,
+      "loss": 0.5477,
+      "step": 9536
+    },
+    {
+      "epoch": 12.20736,
+      "grad_norm": 0.9913842678070068,
+      "learning_rate": 3.094237695078032e-05,
+      "loss": 0.5536,
+      "step": 9537
+    },
+    {
+      "epoch": 12.20864,
+      "grad_norm": 0.9225572347640991,
+      "learning_rate": 3.094037615046019e-05,
+      "loss": 0.5175,
+      "step": 9538
+    },
+    {
+      "epoch": 12.20992,
+      "grad_norm": 0.8077876567840576,
+      "learning_rate": 3.0938375350140056e-05,
+      "loss": 0.5227,
+      "step": 9539
+    },
+    {
+      "epoch": 12.2112,
+      "grad_norm": 0.9560815691947937,
+      "learning_rate": 3.093637454981993e-05,
+      "loss": 0.5408,
+      "step": 9540
+    },
+    {
+      "epoch": 12.21248,
+      "grad_norm": 0.9433364868164062,
+      "learning_rate": 3.09343737494998e-05,
+      "loss": 0.5622,
+      "step": 9541
+    },
+    {
+      "epoch": 12.21376,
+      "grad_norm": 0.934070885181427,
+      "learning_rate": 3.093237294917967e-05,
+      "loss": 0.5873,
+      "step": 9542
+    },
+    {
+      "epoch": 12.21504,
+      "grad_norm": 0.8797191977500916,
+      "learning_rate": 3.093037214885954e-05,
+      "loss": 0.5258,
+      "step": 9543
+    },
+    {
+      "epoch": 12.21632,
+      "grad_norm": 0.9075624346733093,
+      "learning_rate": 3.092837134853942e-05,
+      "loss": 0.5438,
+      "step": 9544
+    },
+    {
+      "epoch": 12.2176,
+      "grad_norm": 0.9165863990783691,
+      "learning_rate": 3.0926370548219294e-05,
+      "loss": 0.5279,
+      "step": 9545
+    },
+    {
+      "epoch": 12.21888,
+      "grad_norm": 0.9102526307106018,
+      "learning_rate": 3.0924369747899166e-05,
+      "loss": 0.5665,
+      "step": 9546
+    },
+    {
+      "epoch": 12.22016,
+      "grad_norm": 0.9222195744514465,
+      "learning_rate": 3.092236894757903e-05,
+      "loss": 0.5457,
+      "step": 9547
+    },
+    {
+      "epoch": 12.22144,
+      "grad_norm": 0.9091429710388184,
+      "learning_rate": 3.09203681472589e-05,
+      "loss": 0.5565,
+      "step": 9548
+    },
+    {
+      "epoch": 12.22272,
+      "grad_norm": 0.8707777857780457,
+      "learning_rate": 3.0918367346938774e-05,
+      "loss": 0.5086,
+      "step": 9549
+    },
+    {
+      "epoch": 12.224,
+      "grad_norm": 0.8751879930496216,
+      "learning_rate": 3.0916366546618646e-05,
+      "loss": 0.4871,
+      "step": 9550
+    },
+    {
+      "epoch": 12.22528,
+      "grad_norm": 0.9578078389167786,
+      "learning_rate": 3.091436574629852e-05,
+      "loss": 0.5734,
+      "step": 9551
+    },
+    {
+      "epoch": 12.22656,
+      "grad_norm": 0.9274276494979858,
+      "learning_rate": 3.09123649459784e-05,
+      "loss": 0.5629,
+      "step": 9552
+    },
+    {
+      "epoch": 12.22784,
+      "grad_norm": 0.9546663761138916,
+      "learning_rate": 3.091036414565827e-05,
+      "loss": 0.5466,
+      "step": 9553
+    },
+    {
+      "epoch": 12.22912,
+      "grad_norm": 0.9720085859298706,
+      "learning_rate": 3.090836334533814e-05,
+      "loss": 0.5884,
+      "step": 9554
+    },
+    {
+      "epoch": 12.2304,
+      "grad_norm": 0.9668819308280945,
+      "learning_rate": 3.0906362545018006e-05,
+      "loss": 0.5534,
+      "step": 9555
+    },
+    {
+      "epoch": 12.23168,
+      "grad_norm": 0.9304522275924683,
+      "learning_rate": 3.090436174469788e-05,
+      "loss": 0.5697,
+      "step": 9556
+    },
+    {
+      "epoch": 12.23296,
+      "grad_norm": 0.915131688117981,
+      "learning_rate": 3.090236094437775e-05,
+      "loss": 0.5135,
+      "step": 9557
+    },
+    {
+      "epoch": 12.23424,
+      "grad_norm": 0.8442711234092712,
+      "learning_rate": 3.090036014405762e-05,
+      "loss": 0.5244,
+      "step": 9558
+    },
+    {
+      "epoch": 12.23552,
+      "grad_norm": 0.8993996381759644,
+      "learning_rate": 3.08983593437375e-05,
+      "loss": 0.5362,
+      "step": 9559
+    },
+    {
+      "epoch": 12.2368,
+      "grad_norm": 0.9662063121795654,
+      "learning_rate": 3.089635854341737e-05,
+      "loss": 0.5618,
+      "step": 9560
+    },
+    {
+      "epoch": 12.23808,
+      "grad_norm": 0.9213639497756958,
+      "learning_rate": 3.0894357743097244e-05,
+      "loss": 0.5064,
+      "step": 9561
+    },
+    {
+      "epoch": 12.23936,
+      "grad_norm": 0.9037312269210815,
+      "learning_rate": 3.0892356942777115e-05,
+      "loss": 0.5408,
+      "step": 9562
+    },
+    {
+      "epoch": 12.24064,
+      "grad_norm": 0.9527158141136169,
+      "learning_rate": 3.089035614245698e-05,
+      "loss": 0.5756,
+      "step": 9563
+    },
+    {
+      "epoch": 12.24192,
+      "grad_norm": 0.886060357093811,
+      "learning_rate": 3.088835534213685e-05,
+      "loss": 0.5351,
+      "step": 9564
+    },
+    {
+      "epoch": 12.2432,
+      "grad_norm": 0.9264008402824402,
+      "learning_rate": 3.0886354541816724e-05,
+      "loss": 0.5426,
+      "step": 9565
+    },
+    {
+      "epoch": 12.24448,
+      "grad_norm": 0.9923083782196045,
+      "learning_rate": 3.08843537414966e-05,
+      "loss": 0.556,
+      "step": 9566
+    },
+    {
+      "epoch": 12.24576,
+      "grad_norm": 0.9164190888404846,
+      "learning_rate": 3.0882352941176475e-05,
+      "loss": 0.5639,
+      "step": 9567
+    },
+    {
+      "epoch": 12.24704,
+      "grad_norm": 0.9647570252418518,
+      "learning_rate": 3.0880352140856347e-05,
+      "loss": 0.5686,
+      "step": 9568
+    },
+    {
+      "epoch": 12.24832,
+      "grad_norm": 0.9134481549263,
+      "learning_rate": 3.087835134053622e-05,
+      "loss": 0.5295,
+      "step": 9569
+    },
+    {
+      "epoch": 12.2496,
+      "grad_norm": 0.8650757670402527,
+      "learning_rate": 3.087635054021609e-05,
+      "loss": 0.5215,
+      "step": 9570
+    },
+    {
+      "epoch": 12.25088,
+      "grad_norm": 0.9101705551147461,
+      "learning_rate": 3.0874349739895955e-05,
+      "loss": 0.5681,
+      "step": 9571
+    },
+    {
+      "epoch": 12.25216,
+      "grad_norm": 0.9110287427902222,
+      "learning_rate": 3.087234893957583e-05,
+      "loss": 0.5318,
+      "step": 9572
+    },
+    {
+      "epoch": 12.25344,
+      "grad_norm": 0.8674918413162231,
+      "learning_rate": 3.0870348139255706e-05,
+      "loss": 0.5185,
+      "step": 9573
+    },
+    {
+      "epoch": 12.25472,
+      "grad_norm": 0.9074335694313049,
+      "learning_rate": 3.086834733893558e-05,
+      "loss": 0.5352,
+      "step": 9574
+    },
+    {
+      "epoch": 12.256,
+      "grad_norm": 0.9423937797546387,
+      "learning_rate": 3.086634653861545e-05,
+      "loss": 0.4859,
+      "step": 9575
+    },
+    {
+      "epoch": 12.25728,
+      "grad_norm": 0.8732250928878784,
+      "learning_rate": 3.086434573829532e-05,
+      "loss": 0.5237,
+      "step": 9576
+    },
+    {
+      "epoch": 12.25856,
+      "grad_norm": 0.9658347964286804,
+      "learning_rate": 3.086234493797519e-05,
+      "loss": 0.5853,
+      "step": 9577
+    },
+    {
+      "epoch": 12.25984,
+      "grad_norm": 0.9143103957176208,
+      "learning_rate": 3.0860344137655065e-05,
+      "loss": 0.5277,
+      "step": 9578
+    },
+    {
+      "epoch": 12.26112,
+      "grad_norm": 0.8892953395843506,
+      "learning_rate": 3.085834333733493e-05,
+      "loss": 0.5178,
+      "step": 9579
+    },
+    {
+      "epoch": 12.2624,
+      "grad_norm": 0.8988597393035889,
+      "learning_rate": 3.085634253701481e-05,
+      "loss": 0.5443,
+      "step": 9580
+    },
+    {
+      "epoch": 12.26368,
+      "grad_norm": 0.9245293736457825,
+      "learning_rate": 3.085434173669468e-05,
+      "loss": 0.5107,
+      "step": 9581
+    },
+    {
+      "epoch": 12.26496,
+      "grad_norm": 0.9400307536125183,
+      "learning_rate": 3.085234093637455e-05,
+      "loss": 0.5741,
+      "step": 9582
+    },
+    {
+      "epoch": 12.26624,
+      "grad_norm": 0.9172420501708984,
+      "learning_rate": 3.0850340136054424e-05,
+      "loss": 0.5751,
+      "step": 9583
+    },
+    {
+      "epoch": 12.26752,
+      "grad_norm": 0.8972305059432983,
+      "learning_rate": 3.0848339335734296e-05,
+      "loss": 0.4933,
+      "step": 9584
+    },
+    {
+      "epoch": 12.2688,
+      "grad_norm": 0.8898636698722839,
+      "learning_rate": 3.084633853541417e-05,
+      "loss": 0.5589,
+      "step": 9585
+    },
+    {
+      "epoch": 12.27008,
+      "grad_norm": 0.9300894141197205,
+      "learning_rate": 3.084433773509404e-05,
+      "loss": 0.5789,
+      "step": 9586
+    },
+    {
+      "epoch": 12.27136,
+      "grad_norm": 0.8965767621994019,
+      "learning_rate": 3.084233693477391e-05,
+      "loss": 0.5071,
+      "step": 9587
+    },
+    {
+      "epoch": 12.272639999999999,
+      "grad_norm": 0.9093717932701111,
+      "learning_rate": 3.0840336134453784e-05,
+      "loss": 0.5453,
+      "step": 9588
+    },
+    {
+      "epoch": 12.27392,
+      "grad_norm": 0.9376620650291443,
+      "learning_rate": 3.0838335334133656e-05,
+      "loss": 0.5392,
+      "step": 9589
+    },
+    {
+      "epoch": 12.2752,
+      "grad_norm": 0.9461066126823425,
+      "learning_rate": 3.083633453381353e-05,
+      "loss": 0.5636,
+      "step": 9590
+    },
+    {
+      "epoch": 12.27648,
+      "grad_norm": 0.8881168365478516,
+      "learning_rate": 3.08343337334934e-05,
+      "loss": 0.5378,
+      "step": 9591
+    },
+    {
+      "epoch": 12.27776,
+      "grad_norm": 0.9719573855400085,
+      "learning_rate": 3.083233293317327e-05,
+      "loss": 0.6019,
+      "step": 9592
+    },
+    {
+      "epoch": 12.27904,
+      "grad_norm": 0.9166926741600037,
+      "learning_rate": 3.083033213285314e-05,
+      "loss": 0.579,
+      "step": 9593
+    },
+    {
+      "epoch": 12.28032,
+      "grad_norm": 0.9417465925216675,
+      "learning_rate": 3.0828331332533015e-05,
+      "loss": 0.5209,
+      "step": 9594
+    },
+    {
+      "epoch": 12.2816,
+      "grad_norm": 0.9969757199287415,
+      "learning_rate": 3.082633053221289e-05,
+      "loss": 0.5814,
+      "step": 9595
+    },
+    {
+      "epoch": 12.28288,
+      "grad_norm": 0.9610475897789001,
+      "learning_rate": 3.082432973189276e-05,
+      "loss": 0.5997,
+      "step": 9596
+    },
+    {
+      "epoch": 12.28416,
+      "grad_norm": 0.8945482969284058,
+      "learning_rate": 3.082232893157263e-05,
+      "loss": 0.5166,
+      "step": 9597
+    },
+    {
+      "epoch": 12.28544,
+      "grad_norm": 0.9552536606788635,
+      "learning_rate": 3.08203281312525e-05,
+      "loss": 0.5762,
+      "step": 9598
+    },
+    {
+      "epoch": 12.28672,
+      "grad_norm": 0.9500235915184021,
+      "learning_rate": 3.0818327330932374e-05,
+      "loss": 0.5129,
+      "step": 9599
+    },
+    {
+      "epoch": 12.288,
+      "grad_norm": 0.8789950609207153,
+      "learning_rate": 3.0816326530612246e-05,
+      "loss": 0.5308,
+      "step": 9600
+    },
+    {
+      "epoch": 12.28928,
+      "grad_norm": 0.8765109181404114,
+      "learning_rate": 3.081432573029212e-05,
+      "loss": 0.4918,
+      "step": 9601
+    },
+    {
+      "epoch": 12.29056,
+      "grad_norm": 0.9502473473548889,
+      "learning_rate": 3.081232492997199e-05,
+      "loss": 0.5485,
+      "step": 9602
+    },
+    {
+      "epoch": 12.29184,
+      "grad_norm": 0.9585152864456177,
+      "learning_rate": 3.081032412965186e-05,
+      "loss": 0.5748,
+      "step": 9603
+    },
+    {
+      "epoch": 12.29312,
+      "grad_norm": 0.9246171116828918,
+      "learning_rate": 3.080832332933173e-05,
+      "loss": 0.5394,
+      "step": 9604
+    },
+    {
+      "epoch": 12.2944,
+      "grad_norm": 0.9247891902923584,
+      "learning_rate": 3.0806322529011605e-05,
+      "loss": 0.5278,
+      "step": 9605
+    },
+    {
+      "epoch": 12.29568,
+      "grad_norm": 0.9395011067390442,
+      "learning_rate": 3.080432172869148e-05,
+      "loss": 0.555,
+      "step": 9606
+    },
+    {
+      "epoch": 12.29696,
+      "grad_norm": 0.9174982905387878,
+      "learning_rate": 3.080232092837135e-05,
+      "loss": 0.51,
+      "step": 9607
+    },
+    {
+      "epoch": 12.29824,
+      "grad_norm": 0.8428568840026855,
+      "learning_rate": 3.080032012805123e-05,
+      "loss": 0.5107,
+      "step": 9608
+    },
+    {
+      "epoch": 12.29952,
+      "grad_norm": 0.9542026519775391,
+      "learning_rate": 3.079831932773109e-05,
+      "loss": 0.5491,
+      "step": 9609
+    },
+    {
+      "epoch": 12.3008,
+      "grad_norm": 0.9842756390571594,
+      "learning_rate": 3.0796318527410964e-05,
+      "loss": 0.5392,
+      "step": 9610
+    },
+    {
+      "epoch": 12.30208,
+      "grad_norm": 0.9100461006164551,
+      "learning_rate": 3.0794317727090836e-05,
+      "loss": 0.5535,
+      "step": 9611
+    },
+    {
+      "epoch": 12.30336,
+      "grad_norm": 0.8975451588630676,
+      "learning_rate": 3.079231692677071e-05,
+      "loss": 0.5449,
+      "step": 9612
+    },
+    {
+      "epoch": 12.30464,
+      "grad_norm": 0.9020168781280518,
+      "learning_rate": 3.079031612645058e-05,
+      "loss": 0.5142,
+      "step": 9613
+    },
+    {
+      "epoch": 12.30592,
+      "grad_norm": 0.9032678604125977,
+      "learning_rate": 3.078831532613045e-05,
+      "loss": 0.5045,
+      "step": 9614
+    },
+    {
+      "epoch": 12.3072,
+      "grad_norm": 0.9697889089584351,
+      "learning_rate": 3.078631452581033e-05,
+      "loss": 0.5418,
+      "step": 9615
+    },
+    {
+      "epoch": 12.30848,
+      "grad_norm": 0.9151008725166321,
+      "learning_rate": 3.07843137254902e-05,
+      "loss": 0.5524,
+      "step": 9616
+    },
+    {
+      "epoch": 12.30976,
+      "grad_norm": 0.9026890993118286,
+      "learning_rate": 3.078231292517007e-05,
+      "loss": 0.5102,
+      "step": 9617
+    },
+    {
+      "epoch": 12.31104,
+      "grad_norm": 0.9061116576194763,
+      "learning_rate": 3.078031212484994e-05,
+      "loss": 0.5607,
+      "step": 9618
+    },
+    {
+      "epoch": 12.31232,
+      "grad_norm": 0.893916666507721,
+      "learning_rate": 3.077831132452981e-05,
+      "loss": 0.4953,
+      "step": 9619
+    },
+    {
+      "epoch": 12.3136,
+      "grad_norm": 0.9334031343460083,
+      "learning_rate": 3.077631052420968e-05,
+      "loss": 0.5265,
+      "step": 9620
+    },
+    {
+      "epoch": 12.31488,
+      "grad_norm": 0.8903616070747375,
+      "learning_rate": 3.0774309723889555e-05,
+      "loss": 0.5208,
+      "step": 9621
+    },
+    {
+      "epoch": 12.31616,
+      "grad_norm": 0.9477413296699524,
+      "learning_rate": 3.0772308923569434e-05,
+      "loss": 0.5759,
+      "step": 9622
+    },
+    {
+      "epoch": 12.31744,
+      "grad_norm": 0.881305456161499,
+      "learning_rate": 3.0770308123249305e-05,
+      "loss": 0.531,
+      "step": 9623
+    },
+    {
+      "epoch": 12.31872,
+      "grad_norm": 0.9640594124794006,
+      "learning_rate": 3.076830732292918e-05,
+      "loss": 0.5694,
+      "step": 9624
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 0.888465404510498,
+      "learning_rate": 3.076630652260904e-05,
+      "loss": 0.4961,
+      "step": 9625
+    },
+    {
+      "epoch": 12.32128,
+      "grad_norm": 0.8741172552108765,
+      "learning_rate": 3.0764305722288914e-05,
+      "loss": 0.5258,
+      "step": 9626
+    },
+    {
+      "epoch": 12.32256,
+      "grad_norm": 0.9086135029792786,
+      "learning_rate": 3.0762304921968786e-05,
+      "loss": 0.545,
+      "step": 9627
+    },
+    {
+      "epoch": 12.32384,
+      "grad_norm": 0.8762311935424805,
+      "learning_rate": 3.076030412164866e-05,
+      "loss": 0.538,
+      "step": 9628
+    },
+    {
+      "epoch": 12.32512,
+      "grad_norm": 0.9073178172111511,
+      "learning_rate": 3.0758303321328537e-05,
+      "loss": 0.5466,
+      "step": 9629
+    },
+    {
+      "epoch": 12.3264,
+      "grad_norm": 0.866665780544281,
+      "learning_rate": 3.075630252100841e-05,
+      "loss": 0.545,
+      "step": 9630
+    },
+    {
+      "epoch": 12.32768,
+      "grad_norm": 0.9432623386383057,
+      "learning_rate": 3.075430172068828e-05,
+      "loss": 0.5641,
+      "step": 9631
+    },
+    {
+      "epoch": 12.32896,
+      "grad_norm": 0.9183183312416077,
+      "learning_rate": 3.075230092036815e-05,
+      "loss": 0.5232,
+      "step": 9632
+    },
+    {
+      "epoch": 12.33024,
+      "grad_norm": 0.9821698069572449,
+      "learning_rate": 3.075030012004802e-05,
+      "loss": 0.5631,
+      "step": 9633
+    },
+    {
+      "epoch": 12.33152,
+      "grad_norm": 0.9003911018371582,
+      "learning_rate": 3.074829931972789e-05,
+      "loss": 0.5493,
+      "step": 9634
+    },
+    {
+      "epoch": 12.3328,
+      "grad_norm": 0.9370520114898682,
+      "learning_rate": 3.074629851940776e-05,
+      "loss": 0.556,
+      "step": 9635
+    },
+    {
+      "epoch": 12.33408,
+      "grad_norm": 0.8950033187866211,
+      "learning_rate": 3.074429771908764e-05,
+      "loss": 0.5258,
+      "step": 9636
+    },
+    {
+      "epoch": 12.33536,
+      "grad_norm": 0.9155859351158142,
+      "learning_rate": 3.074229691876751e-05,
+      "loss": 0.5623,
+      "step": 9637
+    },
+    {
+      "epoch": 12.33664,
+      "grad_norm": 0.929165244102478,
+      "learning_rate": 3.074029611844738e-05,
+      "loss": 0.5535,
+      "step": 9638
+    },
+    {
+      "epoch": 12.33792,
+      "grad_norm": 0.9462549686431885,
+      "learning_rate": 3.0738295318127255e-05,
+      "loss": 0.5687,
+      "step": 9639
+    },
+    {
+      "epoch": 12.3392,
+      "grad_norm": 0.9272847175598145,
+      "learning_rate": 3.073629451780713e-05,
+      "loss": 0.5393,
+      "step": 9640
+    },
+    {
+      "epoch": 12.34048,
+      "grad_norm": 0.8985689282417297,
+      "learning_rate": 3.073429371748699e-05,
+      "loss": 0.5385,
+      "step": 9641
+    },
+    {
+      "epoch": 12.34176,
+      "grad_norm": 0.9289854764938354,
+      "learning_rate": 3.0732292917166864e-05,
+      "loss": 0.5499,
+      "step": 9642
+    },
+    {
+      "epoch": 12.34304,
+      "grad_norm": 0.9333150386810303,
+      "learning_rate": 3.073029211684674e-05,
+      "loss": 0.5588,
+      "step": 9643
+    },
+    {
+      "epoch": 12.34432,
+      "grad_norm": 0.9284766316413879,
+      "learning_rate": 3.0728291316526614e-05,
+      "loss": 0.5602,
+      "step": 9644
+    },
+    {
+      "epoch": 12.3456,
+      "grad_norm": 0.9135120511054993,
+      "learning_rate": 3.0726290516206486e-05,
+      "loss": 0.5444,
+      "step": 9645
+    },
+    {
+      "epoch": 12.34688,
+      "grad_norm": 0.8909054398536682,
+      "learning_rate": 3.072428971588636e-05,
+      "loss": 0.553,
+      "step": 9646
+    },
+    {
+      "epoch": 12.34816,
+      "grad_norm": 0.9332317113876343,
+      "learning_rate": 3.072228891556623e-05,
+      "loss": 0.558,
+      "step": 9647
+    },
+    {
+      "epoch": 12.34944,
+      "grad_norm": 0.8675878643989563,
+      "learning_rate": 3.07202881152461e-05,
+      "loss": 0.5093,
+      "step": 9648
+    },
+    {
+      "epoch": 12.35072,
+      "grad_norm": 0.9375959634780884,
+      "learning_rate": 3.071828731492597e-05,
+      "loss": 0.5277,
+      "step": 9649
+    },
+    {
+      "epoch": 12.352,
+      "grad_norm": 0.9183427691459656,
+      "learning_rate": 3.0716286514605846e-05,
+      "loss": 0.5171,
+      "step": 9650
+    },
+    {
+      "epoch": 12.35328,
+      "grad_norm": 0.9373577237129211,
+      "learning_rate": 3.071428571428572e-05,
+      "loss": 0.5495,
+      "step": 9651
+    },
+    {
+      "epoch": 12.35456,
+      "grad_norm": 0.9106853604316711,
+      "learning_rate": 3.071228491396559e-05,
+      "loss": 0.512,
+      "step": 9652
+    },
+    {
+      "epoch": 12.35584,
+      "grad_norm": 0.9207229614257812,
+      "learning_rate": 3.071028411364546e-05,
+      "loss": 0.538,
+      "step": 9653
+    },
+    {
+      "epoch": 12.35712,
+      "grad_norm": 0.9344127178192139,
+      "learning_rate": 3.070828331332533e-05,
+      "loss": 0.5688,
+      "step": 9654
+    },
+    {
+      "epoch": 12.3584,
+      "grad_norm": 0.9381983280181885,
+      "learning_rate": 3.0706282513005205e-05,
+      "loss": 0.5687,
+      "step": 9655
+    },
+    {
+      "epoch": 12.35968,
+      "grad_norm": 0.9426577687263489,
+      "learning_rate": 3.070428171268508e-05,
+      "loss": 0.5498,
+      "step": 9656
+    },
+    {
+      "epoch": 12.36096,
+      "grad_norm": 0.9524323344230652,
+      "learning_rate": 3.070228091236495e-05,
+      "loss": 0.5394,
+      "step": 9657
+    },
+    {
+      "epoch": 12.36224,
+      "grad_norm": 0.9800800681114197,
+      "learning_rate": 3.070028011204482e-05,
+      "loss": 0.5799,
+      "step": 9658
+    },
+    {
+      "epoch": 12.36352,
+      "grad_norm": 0.9301595687866211,
+      "learning_rate": 3.069827931172469e-05,
+      "loss": 0.551,
+      "step": 9659
+    },
+    {
+      "epoch": 12.3648,
+      "grad_norm": 0.9066020250320435,
+      "learning_rate": 3.0696278511404564e-05,
+      "loss": 0.5535,
+      "step": 9660
+    },
+    {
+      "epoch": 12.36608,
+      "grad_norm": 0.9361892938613892,
+      "learning_rate": 3.0694277711084436e-05,
+      "loss": 0.5224,
+      "step": 9661
+    },
+    {
+      "epoch": 12.36736,
+      "grad_norm": 0.9013199806213379,
+      "learning_rate": 3.069227691076431e-05,
+      "loss": 0.5531,
+      "step": 9662
+    },
+    {
+      "epoch": 12.36864,
+      "grad_norm": 0.9619473814964294,
+      "learning_rate": 3.069027611044418e-05,
+      "loss": 0.5865,
+      "step": 9663
+    },
+    {
+      "epoch": 12.36992,
+      "grad_norm": 0.9332448244094849,
+      "learning_rate": 3.068827531012405e-05,
+      "loss": 0.528,
+      "step": 9664
+    },
+    {
+      "epoch": 12.3712,
+      "grad_norm": 0.9028695821762085,
+      "learning_rate": 3.0686274509803923e-05,
+      "loss": 0.5707,
+      "step": 9665
+    },
+    {
+      "epoch": 12.37248,
+      "grad_norm": 0.9249699115753174,
+      "learning_rate": 3.0684273709483795e-05,
+      "loss": 0.5385,
+      "step": 9666
+    },
+    {
+      "epoch": 12.37376,
+      "grad_norm": 0.8583320379257202,
+      "learning_rate": 3.068227290916367e-05,
+      "loss": 0.5264,
+      "step": 9667
+    },
+    {
+      "epoch": 12.37504,
+      "grad_norm": 0.8901596069335938,
+      "learning_rate": 3.068027210884354e-05,
+      "loss": 0.5339,
+      "step": 9668
+    },
+    {
+      "epoch": 12.37632,
+      "grad_norm": 0.9462403059005737,
+      "learning_rate": 3.067827130852341e-05,
+      "loss": 0.5604,
+      "step": 9669
+    },
+    {
+      "epoch": 12.3776,
+      "grad_norm": 0.9234113693237305,
+      "learning_rate": 3.067627050820328e-05,
+      "loss": 0.5289,
+      "step": 9670
+    },
+    {
+      "epoch": 12.37888,
+      "grad_norm": 0.9540828466415405,
+      "learning_rate": 3.0674269707883155e-05,
+      "loss": 0.5867,
+      "step": 9671
+    },
+    {
+      "epoch": 12.38016,
+      "grad_norm": 0.936007559299469,
+      "learning_rate": 3.0672268907563026e-05,
+      "loss": 0.5348,
+      "step": 9672
+    },
+    {
+      "epoch": 12.38144,
+      "grad_norm": 0.9366984367370605,
+      "learning_rate": 3.06702681072429e-05,
+      "loss": 0.52,
+      "step": 9673
+    },
+    {
+      "epoch": 12.38272,
+      "grad_norm": 0.9363959431648254,
+      "learning_rate": 3.066826730692277e-05,
+      "loss": 0.5781,
+      "step": 9674
+    },
+    {
+      "epoch": 12.384,
+      "grad_norm": 0.947193443775177,
+      "learning_rate": 3.066626650660264e-05,
+      "loss": 0.5687,
+      "step": 9675
+    },
+    {
+      "epoch": 12.38528,
+      "grad_norm": 0.8775532841682434,
+      "learning_rate": 3.0664265706282514e-05,
+      "loss": 0.5269,
+      "step": 9676
+    },
+    {
+      "epoch": 12.38656,
+      "grad_norm": 0.8915206789970398,
+      "learning_rate": 3.0662264905962386e-05,
+      "loss": 0.5509,
+      "step": 9677
+    },
+    {
+      "epoch": 12.38784,
+      "grad_norm": 0.9614311456680298,
+      "learning_rate": 3.066026410564226e-05,
+      "loss": 0.6082,
+      "step": 9678
+    },
+    {
+      "epoch": 12.38912,
+      "grad_norm": 0.9689681529998779,
+      "learning_rate": 3.065826330532213e-05,
+      "loss": 0.5464,
+      "step": 9679
+    },
+    {
+      "epoch": 12.3904,
+      "grad_norm": 0.9640049338340759,
+      "learning_rate": 3.0656262505002e-05,
+      "loss": 0.5523,
+      "step": 9680
+    },
+    {
+      "epoch": 12.39168,
+      "grad_norm": 0.8928719758987427,
+      "learning_rate": 3.065426170468187e-05,
+      "loss": 0.568,
+      "step": 9681
+    },
+    {
+      "epoch": 12.39296,
+      "grad_norm": 0.9739533066749573,
+      "learning_rate": 3.0652260904361745e-05,
+      "loss": 0.5606,
+      "step": 9682
+    },
+    {
+      "epoch": 12.39424,
+      "grad_norm": 0.894177258014679,
+      "learning_rate": 3.065026010404162e-05,
+      "loss": 0.5587,
+      "step": 9683
+    },
+    {
+      "epoch": 12.39552,
+      "grad_norm": 0.9096843600273132,
+      "learning_rate": 3.064825930372149e-05,
+      "loss": 0.5484,
+      "step": 9684
+    },
+    {
+      "epoch": 12.3968,
+      "grad_norm": 0.862076997756958,
+      "learning_rate": 3.064625850340136e-05,
+      "loss": 0.4924,
+      "step": 9685
+    },
+    {
+      "epoch": 12.39808,
+      "grad_norm": 0.8842014670372009,
+      "learning_rate": 3.064425770308124e-05,
+      "loss": 0.5408,
+      "step": 9686
+    },
+    {
+      "epoch": 12.39936,
+      "grad_norm": 0.9338904023170471,
+      "learning_rate": 3.0642256902761104e-05,
+      "loss": 0.556,
+      "step": 9687
+    },
+    {
+      "epoch": 12.40064,
+      "grad_norm": 0.9398663640022278,
+      "learning_rate": 3.0640256102440976e-05,
+      "loss": 0.5535,
+      "step": 9688
+    },
+    {
+      "epoch": 12.40192,
+      "grad_norm": 0.944598913192749,
+      "learning_rate": 3.063825530212085e-05,
+      "loss": 0.5534,
+      "step": 9689
+    },
+    {
+      "epoch": 12.4032,
+      "grad_norm": 0.9525898098945618,
+      "learning_rate": 3.063625450180072e-05,
+      "loss": 0.5571,
+      "step": 9690
+    },
+    {
+      "epoch": 12.40448,
+      "grad_norm": 0.9750270247459412,
+      "learning_rate": 3.063425370148059e-05,
+      "loss": 0.6101,
+      "step": 9691
+    },
+    {
+      "epoch": 12.40576,
+      "grad_norm": 0.9584131240844727,
+      "learning_rate": 3.0632252901160464e-05,
+      "loss": 0.5429,
+      "step": 9692
+    },
+    {
+      "epoch": 12.40704,
+      "grad_norm": 0.8970118165016174,
+      "learning_rate": 3.063025210084034e-05,
+      "loss": 0.537,
+      "step": 9693
+    },
+    {
+      "epoch": 12.40832,
+      "grad_norm": 0.9430438876152039,
+      "learning_rate": 3.0628251300520214e-05,
+      "loss": 0.5364,
+      "step": 9694
+    },
+    {
+      "epoch": 12.4096,
+      "grad_norm": 0.9645063281059265,
+      "learning_rate": 3.062625050020008e-05,
+      "loss": 0.5354,
+      "step": 9695
+    },
+    {
+      "epoch": 12.41088,
+      "grad_norm": 0.9782135486602783,
+      "learning_rate": 3.062424969987995e-05,
+      "loss": 0.551,
+      "step": 9696
+    },
+    {
+      "epoch": 12.41216,
+      "grad_norm": 0.9144942760467529,
+      "learning_rate": 3.062224889955982e-05,
+      "loss": 0.5719,
+      "step": 9697
+    },
+    {
+      "epoch": 12.41344,
+      "grad_norm": 0.9149807691574097,
+      "learning_rate": 3.0620248099239695e-05,
+      "loss": 0.4937,
+      "step": 9698
+    },
+    {
+      "epoch": 12.414719999999999,
+      "grad_norm": 0.9331946969032288,
+      "learning_rate": 3.0618247298919567e-05,
+      "loss": 0.5472,
+      "step": 9699
+    },
+    {
+      "epoch": 12.416,
+      "grad_norm": 0.9157038927078247,
+      "learning_rate": 3.0616246498599445e-05,
+      "loss": 0.5839,
+      "step": 9700
+    },
+    {
+      "epoch": 12.41728,
+      "grad_norm": 0.955386757850647,
+      "learning_rate": 3.061424569827932e-05,
+      "loss": 0.5573,
+      "step": 9701
+    },
+    {
+      "epoch": 12.41856,
+      "grad_norm": 0.9963140487670898,
+      "learning_rate": 3.061224489795919e-05,
+      "loss": 0.5524,
+      "step": 9702
+    },
+    {
+      "epoch": 12.41984,
+      "grad_norm": 0.9035356044769287,
+      "learning_rate": 3.0610244097639054e-05,
+      "loss": 0.5279,
+      "step": 9703
+    },
+    {
+      "epoch": 12.42112,
+      "grad_norm": 0.885988712310791,
+      "learning_rate": 3.0608243297318926e-05,
+      "loss": 0.5387,
+      "step": 9704
+    },
+    {
+      "epoch": 12.4224,
+      "grad_norm": 0.9754456281661987,
+      "learning_rate": 3.06062424969988e-05,
+      "loss": 0.5724,
+      "step": 9705
+    },
+    {
+      "epoch": 12.42368,
+      "grad_norm": 0.8800594806671143,
+      "learning_rate": 3.060424169667867e-05,
+      "loss": 0.503,
+      "step": 9706
+    },
+    {
+      "epoch": 12.42496,
+      "grad_norm": 0.9732749462127686,
+      "learning_rate": 3.060224089635855e-05,
+      "loss": 0.5478,
+      "step": 9707
+    },
+    {
+      "epoch": 12.42624,
+      "grad_norm": 0.8836948871612549,
+      "learning_rate": 3.060024009603842e-05,
+      "loss": 0.5481,
+      "step": 9708
+    },
+    {
+      "epoch": 12.42752,
+      "grad_norm": 0.8984959721565247,
+      "learning_rate": 3.059823929571829e-05,
+      "loss": 0.57,
+      "step": 9709
+    },
+    {
+      "epoch": 12.4288,
+      "grad_norm": 0.9155954122543335,
+      "learning_rate": 3.0596238495398164e-05,
+      "loss": 0.5659,
+      "step": 9710
+    },
+    {
+      "epoch": 12.43008,
+      "grad_norm": 0.897638201713562,
+      "learning_rate": 3.059423769507803e-05,
+      "loss": 0.5256,
+      "step": 9711
+    },
+    {
+      "epoch": 12.43136,
+      "grad_norm": 0.9188939929008484,
+      "learning_rate": 3.05922368947579e-05,
+      "loss": 0.575,
+      "step": 9712
+    },
+    {
+      "epoch": 12.43264,
+      "grad_norm": 0.9399385452270508,
+      "learning_rate": 3.059023609443777e-05,
+      "loss": 0.569,
+      "step": 9713
+    },
+    {
+      "epoch": 12.43392,
+      "grad_norm": 0.9348775148391724,
+      "learning_rate": 3.058823529411765e-05,
+      "loss": 0.5243,
+      "step": 9714
+    },
+    {
+      "epoch": 12.4352,
+      "grad_norm": 0.9413948655128479,
+      "learning_rate": 3.058623449379752e-05,
+      "loss": 0.5641,
+      "step": 9715
+    },
+    {
+      "epoch": 12.43648,
+      "grad_norm": 0.9737404584884644,
+      "learning_rate": 3.0584233693477395e-05,
+      "loss": 0.5496,
+      "step": 9716
+    },
+    {
+      "epoch": 12.43776,
+      "grad_norm": 0.8885311484336853,
+      "learning_rate": 3.058223289315727e-05,
+      "loss": 0.5279,
+      "step": 9717
+    },
+    {
+      "epoch": 12.43904,
+      "grad_norm": 0.9236472249031067,
+      "learning_rate": 3.058023209283714e-05,
+      "loss": 0.5602,
+      "step": 9718
+    },
+    {
+      "epoch": 12.44032,
+      "grad_norm": 0.9534575343132019,
+      "learning_rate": 3.0578231292517004e-05,
+      "loss": 0.5359,
+      "step": 9719
+    },
+    {
+      "epoch": 12.4416,
+      "grad_norm": 0.9509499073028564,
+      "learning_rate": 3.0576230492196876e-05,
+      "loss": 0.5428,
+      "step": 9720
+    },
+    {
+      "epoch": 12.44288,
+      "grad_norm": 0.9832822680473328,
+      "learning_rate": 3.0574229691876754e-05,
+      "loss": 0.5792,
+      "step": 9721
+    },
+    {
+      "epoch": 12.44416,
+      "grad_norm": 0.8691754937171936,
+      "learning_rate": 3.0572228891556626e-05,
+      "loss": 0.4652,
+      "step": 9722
+    },
+    {
+      "epoch": 12.44544,
+      "grad_norm": 0.9270925521850586,
+      "learning_rate": 3.05702280912365e-05,
+      "loss": 0.5903,
+      "step": 9723
+    },
+    {
+      "epoch": 12.44672,
+      "grad_norm": 0.9018378853797913,
+      "learning_rate": 3.056822729091637e-05,
+      "loss": 0.5811,
+      "step": 9724
+    },
+    {
+      "epoch": 12.448,
+      "grad_norm": 0.8946050405502319,
+      "learning_rate": 3.056622649059624e-05,
+      "loss": 0.5234,
+      "step": 9725
+    },
+    {
+      "epoch": 12.44928,
+      "grad_norm": 0.9452985525131226,
+      "learning_rate": 3.0564225690276113e-05,
+      "loss": 0.5267,
+      "step": 9726
+    },
+    {
+      "epoch": 12.45056,
+      "grad_norm": 0.9652531147003174,
+      "learning_rate": 3.056222488995598e-05,
+      "loss": 0.582,
+      "step": 9727
+    },
+    {
+      "epoch": 12.45184,
+      "grad_norm": 0.9452758431434631,
+      "learning_rate": 3.056022408963586e-05,
+      "loss": 0.6246,
+      "step": 9728
+    },
+    {
+      "epoch": 12.45312,
+      "grad_norm": 0.9329007863998413,
+      "learning_rate": 3.055822328931573e-05,
+      "loss": 0.535,
+      "step": 9729
+    },
+    {
+      "epoch": 12.4544,
+      "grad_norm": 0.9430539011955261,
+      "learning_rate": 3.05562224889956e-05,
+      "loss": 0.5769,
+      "step": 9730
+    },
+    {
+      "epoch": 12.45568,
+      "grad_norm": 0.9002211689949036,
+      "learning_rate": 3.055422168867547e-05,
+      "loss": 0.5709,
+      "step": 9731
+    },
+    {
+      "epoch": 12.45696,
+      "grad_norm": 0.8887854814529419,
+      "learning_rate": 3.0552220888355345e-05,
+      "loss": 0.5366,
+      "step": 9732
+    },
+    {
+      "epoch": 12.45824,
+      "grad_norm": 0.8703662753105164,
+      "learning_rate": 3.0550220088035216e-05,
+      "loss": 0.4981,
+      "step": 9733
+    },
+    {
+      "epoch": 12.45952,
+      "grad_norm": 0.9427342414855957,
+      "learning_rate": 3.054821928771509e-05,
+      "loss": 0.5584,
+      "step": 9734
+    },
+    {
+      "epoch": 12.4608,
+      "grad_norm": 0.9092150926589966,
+      "learning_rate": 3.054621848739496e-05,
+      "loss": 0.5098,
+      "step": 9735
+    },
+    {
+      "epoch": 12.46208,
+      "grad_norm": 0.9247660636901855,
+      "learning_rate": 3.054421768707483e-05,
+      "loss": 0.5255,
+      "step": 9736
+    },
+    {
+      "epoch": 12.46336,
+      "grad_norm": 0.8831589221954346,
+      "learning_rate": 3.0542216886754704e-05,
+      "loss": 0.5409,
+      "step": 9737
+    },
+    {
+      "epoch": 12.46464,
+      "grad_norm": 0.9176967740058899,
+      "learning_rate": 3.0540216086434576e-05,
+      "loss": 0.5467,
+      "step": 9738
+    },
+    {
+      "epoch": 12.46592,
+      "grad_norm": 0.9632626175880432,
+      "learning_rate": 3.053821528611445e-05,
+      "loss": 0.5785,
+      "step": 9739
+    },
+    {
+      "epoch": 12.4672,
+      "grad_norm": 0.9407864212989807,
+      "learning_rate": 3.053621448579432e-05,
+      "loss": 0.5921,
+      "step": 9740
+    },
+    {
+      "epoch": 12.46848,
+      "grad_norm": 0.9391652941703796,
+      "learning_rate": 3.053421368547419e-05,
+      "loss": 0.5707,
+      "step": 9741
+    },
+    {
+      "epoch": 12.46976,
+      "grad_norm": 0.9201903939247131,
+      "learning_rate": 3.053221288515406e-05,
+      "loss": 0.5635,
+      "step": 9742
+    },
+    {
+      "epoch": 12.47104,
+      "grad_norm": 0.9038977026939392,
+      "learning_rate": 3.0530212084833935e-05,
+      "loss": 0.5281,
+      "step": 9743
+    },
+    {
+      "epoch": 12.47232,
+      "grad_norm": 0.8993562459945679,
+      "learning_rate": 3.052821128451381e-05,
+      "loss": 0.5565,
+      "step": 9744
+    },
+    {
+      "epoch": 12.4736,
+      "grad_norm": 0.9167428016662598,
+      "learning_rate": 3.052621048419368e-05,
+      "loss": 0.6111,
+      "step": 9745
+    },
+    {
+      "epoch": 12.47488,
+      "grad_norm": 0.9848437905311584,
+      "learning_rate": 3.052420968387355e-05,
+      "loss": 0.54,
+      "step": 9746
+    },
+    {
+      "epoch": 12.47616,
+      "grad_norm": 0.937060534954071,
+      "learning_rate": 3.052220888355342e-05,
+      "loss": 0.5115,
+      "step": 9747
+    },
+    {
+      "epoch": 12.47744,
+      "grad_norm": 0.9132764935493469,
+      "learning_rate": 3.0520208083233294e-05,
+      "loss": 0.5782,
+      "step": 9748
+    },
+    {
+      "epoch": 12.47872,
+      "grad_norm": 0.89909428358078,
+      "learning_rate": 3.0518207282913166e-05,
+      "loss": 0.5177,
+      "step": 9749
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 1.0129623413085938,
+      "learning_rate": 3.051620648259304e-05,
+      "loss": 0.6329,
+      "step": 9750
+    },
+    {
+      "epoch": 12.48128,
+      "grad_norm": 0.9289687871932983,
+      "learning_rate": 3.0514205682272913e-05,
+      "loss": 0.4931,
+      "step": 9751
+    },
+    {
+      "epoch": 12.48256,
+      "grad_norm": 0.8912503123283386,
+      "learning_rate": 3.0512204881952782e-05,
+      "loss": 0.4795,
+      "step": 9752
+    },
+    {
+      "epoch": 12.48384,
+      "grad_norm": 0.9053838849067688,
+      "learning_rate": 3.0510204081632654e-05,
+      "loss": 0.5524,
+      "step": 9753
+    },
+    {
+      "epoch": 12.48512,
+      "grad_norm": 0.9058827757835388,
+      "learning_rate": 3.0508203281312525e-05,
+      "loss": 0.5323,
+      "step": 9754
+    },
+    {
+      "epoch": 12.4864,
+      "grad_norm": 0.9562934041023254,
+      "learning_rate": 3.0506202480992397e-05,
+      "loss": 0.5378,
+      "step": 9755
+    },
+    {
+      "epoch": 12.48768,
+      "grad_norm": 0.8920333981513977,
+      "learning_rate": 3.0504201680672273e-05,
+      "loss": 0.516,
+      "step": 9756
+    },
+    {
+      "epoch": 12.48896,
+      "grad_norm": 0.909192681312561,
+      "learning_rate": 3.0502200880352144e-05,
+      "loss": 0.5491,
+      "step": 9757
+    },
+    {
+      "epoch": 12.49024,
+      "grad_norm": 0.8839040398597717,
+      "learning_rate": 3.0500200080032016e-05,
+      "loss": 0.5202,
+      "step": 9758
+    },
+    {
+      "epoch": 12.49152,
+      "grad_norm": 0.9545120000839233,
+      "learning_rate": 3.0498199279711888e-05,
+      "loss": 0.5876,
+      "step": 9759
+    },
+    {
+      "epoch": 12.4928,
+      "grad_norm": 0.885185182094574,
+      "learning_rate": 3.0496198479391757e-05,
+      "loss": 0.5223,
+      "step": 9760
+    },
+    {
+      "epoch": 12.49408,
+      "grad_norm": 0.9367235898971558,
+      "learning_rate": 3.049419767907163e-05,
+      "loss": 0.5421,
+      "step": 9761
+    },
+    {
+      "epoch": 12.49536,
+      "grad_norm": 0.9383306503295898,
+      "learning_rate": 3.04921968787515e-05,
+      "loss": 0.5907,
+      "step": 9762
+    },
+    {
+      "epoch": 12.49664,
+      "grad_norm": 0.869220495223999,
+      "learning_rate": 3.0490196078431376e-05,
+      "loss": 0.5031,
+      "step": 9763
+    },
+    {
+      "epoch": 12.49792,
+      "grad_norm": 0.9352115392684937,
+      "learning_rate": 3.0488195278111247e-05,
+      "loss": 0.559,
+      "step": 9764
+    },
+    {
+      "epoch": 12.4992,
+      "grad_norm": 0.8659606575965881,
+      "learning_rate": 3.048619447779112e-05,
+      "loss": 0.5303,
+      "step": 9765
+    },
+    {
+      "epoch": 12.50048,
+      "grad_norm": 0.921378493309021,
+      "learning_rate": 3.048419367747099e-05,
+      "loss": 0.5447,
+      "step": 9766
+    },
+    {
+      "epoch": 12.50176,
+      "grad_norm": 0.8634151816368103,
+      "learning_rate": 3.0482192877150863e-05,
+      "loss": 0.5637,
+      "step": 9767
+    },
+    {
+      "epoch": 12.50304,
+      "grad_norm": 0.9479573965072632,
+      "learning_rate": 3.048019207683073e-05,
+      "loss": 0.5185,
+      "step": 9768
+    },
+    {
+      "epoch": 12.50432,
+      "grad_norm": 0.9313710927963257,
+      "learning_rate": 3.0478191276510603e-05,
+      "loss": 0.5721,
+      "step": 9769
+    },
+    {
+      "epoch": 12.5056,
+      "grad_norm": 0.9208516478538513,
+      "learning_rate": 3.0476190476190482e-05,
+      "loss": 0.5565,
+      "step": 9770
+    },
+    {
+      "epoch": 12.50688,
+      "grad_norm": 0.9547169804573059,
+      "learning_rate": 3.047418967587035e-05,
+      "loss": 0.5426,
+      "step": 9771
+    },
+    {
+      "epoch": 12.50816,
+      "grad_norm": 0.8872602581977844,
+      "learning_rate": 3.0472188875550222e-05,
+      "loss": 0.5087,
+      "step": 9772
+    },
+    {
+      "epoch": 12.50944,
+      "grad_norm": 0.8986743092536926,
+      "learning_rate": 3.0470188075230094e-05,
+      "loss": 0.5241,
+      "step": 9773
+    },
+    {
+      "epoch": 12.51072,
+      "grad_norm": 0.9405331015586853,
+      "learning_rate": 3.0468187274909966e-05,
+      "loss": 0.5672,
+      "step": 9774
+    },
+    {
+      "epoch": 12.512,
+      "grad_norm": 0.9390310645103455,
+      "learning_rate": 3.0466186474589838e-05,
+      "loss": 0.5732,
+      "step": 9775
+    },
+    {
+      "epoch": 12.51328,
+      "grad_norm": 0.994657039642334,
+      "learning_rate": 3.0464185674269706e-05,
+      "loss": 0.5688,
+      "step": 9776
+    },
+    {
+      "epoch": 12.51456,
+      "grad_norm": 0.9204804301261902,
+      "learning_rate": 3.0462184873949578e-05,
+      "loss": 0.6019,
+      "step": 9777
+    },
+    {
+      "epoch": 12.51584,
+      "grad_norm": 0.9733226299285889,
+      "learning_rate": 3.0460184073629457e-05,
+      "loss": 0.5185,
+      "step": 9778
+    },
+    {
+      "epoch": 12.51712,
+      "grad_norm": 0.9087737202644348,
+      "learning_rate": 3.0458183273309325e-05,
+      "loss": 0.5242,
+      "step": 9779
+    },
+    {
+      "epoch": 12.5184,
+      "grad_norm": 0.9519286751747131,
+      "learning_rate": 3.0456182472989197e-05,
+      "loss": 0.5675,
+      "step": 9780
+    },
+    {
+      "epoch": 12.51968,
+      "grad_norm": 0.899489164352417,
+      "learning_rate": 3.045418167266907e-05,
+      "loss": 0.5307,
+      "step": 9781
+    },
+    {
+      "epoch": 12.52096,
+      "grad_norm": 0.926542341709137,
+      "learning_rate": 3.045218087234894e-05,
+      "loss": 0.5442,
+      "step": 9782
+    },
+    {
+      "epoch": 12.52224,
+      "grad_norm": 0.8883079290390015,
+      "learning_rate": 3.0450180072028813e-05,
+      "loss": 0.5385,
+      "step": 9783
+    },
+    {
+      "epoch": 12.52352,
+      "grad_norm": 0.9173272252082825,
+      "learning_rate": 3.044817927170868e-05,
+      "loss": 0.5661,
+      "step": 9784
+    },
+    {
+      "epoch": 12.5248,
+      "grad_norm": 1.0070183277130127,
+      "learning_rate": 3.044617847138856e-05,
+      "loss": 0.5957,
+      "step": 9785
+    },
+    {
+      "epoch": 12.52608,
+      "grad_norm": 0.8930588364601135,
+      "learning_rate": 3.044417767106843e-05,
+      "loss": 0.5212,
+      "step": 9786
+    },
+    {
+      "epoch": 12.52736,
+      "grad_norm": 0.883655846118927,
+      "learning_rate": 3.04421768707483e-05,
+      "loss": 0.547,
+      "step": 9787
+    },
+    {
+      "epoch": 12.52864,
+      "grad_norm": 0.917133092880249,
+      "learning_rate": 3.0440176070428172e-05,
+      "loss": 0.4803,
+      "step": 9788
+    },
+    {
+      "epoch": 12.52992,
+      "grad_norm": 0.9094095826148987,
+      "learning_rate": 3.0438175270108044e-05,
+      "loss": 0.544,
+      "step": 9789
+    },
+    {
+      "epoch": 12.5312,
+      "grad_norm": 0.977846622467041,
+      "learning_rate": 3.0436174469787916e-05,
+      "loss": 0.6179,
+      "step": 9790
+    },
+    {
+      "epoch": 12.53248,
+      "grad_norm": 0.9368249177932739,
+      "learning_rate": 3.0434173669467788e-05,
+      "loss": 0.5934,
+      "step": 9791
+    },
+    {
+      "epoch": 12.533760000000001,
+      "grad_norm": 0.8966599106788635,
+      "learning_rate": 3.0432172869147663e-05,
+      "loss": 0.55,
+      "step": 9792
+    },
+    {
+      "epoch": 12.53504,
+      "grad_norm": 0.9707033038139343,
+      "learning_rate": 3.0430172068827535e-05,
+      "loss": 0.5788,
+      "step": 9793
+    },
+    {
+      "epoch": 12.53632,
+      "grad_norm": 0.9984316825866699,
+      "learning_rate": 3.0428171268507406e-05,
+      "loss": 0.6074,
+      "step": 9794
+    },
+    {
+      "epoch": 12.5376,
+      "grad_norm": 0.9540078043937683,
+      "learning_rate": 3.0426170468187275e-05,
+      "loss": 0.5769,
+      "step": 9795
+    },
+    {
+      "epoch": 12.53888,
+      "grad_norm": 0.9404662847518921,
+      "learning_rate": 3.0424169667867147e-05,
+      "loss": 0.5381,
+      "step": 9796
+    },
+    {
+      "epoch": 12.54016,
+      "grad_norm": 0.8832229375839233,
+      "learning_rate": 3.042216886754702e-05,
+      "loss": 0.5524,
+      "step": 9797
+    },
+    {
+      "epoch": 12.54144,
+      "grad_norm": 0.9047516584396362,
+      "learning_rate": 3.042016806722689e-05,
+      "loss": 0.5836,
+      "step": 9798
+    },
+    {
+      "epoch": 12.54272,
+      "grad_norm": 0.9186232089996338,
+      "learning_rate": 3.0418167266906766e-05,
+      "loss": 0.5923,
+      "step": 9799
+    },
+    {
+      "epoch": 12.544,
+      "grad_norm": 1.0132439136505127,
+      "learning_rate": 3.0416166466586638e-05,
+      "loss": 0.5755,
+      "step": 9800
+    },
+    {
+      "epoch": 12.54528,
+      "grad_norm": 0.9297081232070923,
+      "learning_rate": 3.041416566626651e-05,
+      "loss": 0.5604,
+      "step": 9801
+    },
+    {
+      "epoch": 12.54656,
+      "grad_norm": 0.9263479113578796,
+      "learning_rate": 3.041216486594638e-05,
+      "loss": 0.5238,
+      "step": 9802
+    },
+    {
+      "epoch": 12.54784,
+      "grad_norm": 0.964608371257782,
+      "learning_rate": 3.041016406562625e-05,
+      "loss": 0.5523,
+      "step": 9803
+    },
+    {
+      "epoch": 12.54912,
+      "grad_norm": 0.8875142931938171,
+      "learning_rate": 3.040816326530612e-05,
+      "loss": 0.4931,
+      "step": 9804
+    },
+    {
+      "epoch": 12.5504,
+      "grad_norm": 0.9739357829093933,
+      "learning_rate": 3.0406162464985994e-05,
+      "loss": 0.5274,
+      "step": 9805
+    },
+    {
+      "epoch": 12.55168,
+      "grad_norm": 0.8982617855072021,
+      "learning_rate": 3.040416166466587e-05,
+      "loss": 0.5359,
+      "step": 9806
+    },
+    {
+      "epoch": 12.55296,
+      "grad_norm": 0.9386458396911621,
+      "learning_rate": 3.040216086434574e-05,
+      "loss": 0.5505,
+      "step": 9807
+    },
+    {
+      "epoch": 12.55424,
+      "grad_norm": 0.9533564448356628,
+      "learning_rate": 3.0400160064025612e-05,
+      "loss": 0.6093,
+      "step": 9808
+    },
+    {
+      "epoch": 12.55552,
+      "grad_norm": 0.9058338403701782,
+      "learning_rate": 3.0398159263705484e-05,
+      "loss": 0.5394,
+      "step": 9809
+    },
+    {
+      "epoch": 12.556799999999999,
+      "grad_norm": 1.01373291015625,
+      "learning_rate": 3.0396158463385356e-05,
+      "loss": 0.5627,
+      "step": 9810
+    },
+    {
+      "epoch": 12.55808,
+      "grad_norm": 1.0229073762893677,
+      "learning_rate": 3.0394157663065225e-05,
+      "loss": 0.5781,
+      "step": 9811
+    },
+    {
+      "epoch": 12.55936,
+      "grad_norm": 0.9454696178436279,
+      "learning_rate": 3.0392156862745097e-05,
+      "loss": 0.5293,
+      "step": 9812
+    },
+    {
+      "epoch": 12.56064,
+      "grad_norm": 0.9132083058357239,
+      "learning_rate": 3.0390156062424975e-05,
+      "loss": 0.546,
+      "step": 9813
+    },
+    {
+      "epoch": 12.56192,
+      "grad_norm": 0.90849369764328,
+      "learning_rate": 3.0388155262104844e-05,
+      "loss": 0.5221,
+      "step": 9814
+    },
+    {
+      "epoch": 12.5632,
+      "grad_norm": 0.9330136179924011,
+      "learning_rate": 3.0386154461784715e-05,
+      "loss": 0.5212,
+      "step": 9815
+    },
+    {
+      "epoch": 12.56448,
+      "grad_norm": 0.9285899996757507,
+      "learning_rate": 3.0384153661464587e-05,
+      "loss": 0.6005,
+      "step": 9816
+    },
+    {
+      "epoch": 12.565760000000001,
+      "grad_norm": 0.8947268724441528,
+      "learning_rate": 3.038215286114446e-05,
+      "loss": 0.5294,
+      "step": 9817
+    },
+    {
+      "epoch": 12.56704,
+      "grad_norm": 0.9373891949653625,
+      "learning_rate": 3.038015206082433e-05,
+      "loss": 0.5463,
+      "step": 9818
+    },
+    {
+      "epoch": 12.56832,
+      "grad_norm": 0.9649326205253601,
+      "learning_rate": 3.03781512605042e-05,
+      "loss": 0.5701,
+      "step": 9819
+    },
+    {
+      "epoch": 12.5696,
+      "grad_norm": 0.9598551392555237,
+      "learning_rate": 3.0376150460184078e-05,
+      "loss": 0.6032,
+      "step": 9820
+    },
+    {
+      "epoch": 12.57088,
+      "grad_norm": 1.0046522617340088,
+      "learning_rate": 3.037414965986395e-05,
+      "loss": 0.6048,
+      "step": 9821
+    },
+    {
+      "epoch": 12.57216,
+      "grad_norm": 0.9179285168647766,
+      "learning_rate": 3.037214885954382e-05,
+      "loss": 0.5717,
+      "step": 9822
+    },
+    {
+      "epoch": 12.57344,
+      "grad_norm": 0.9317951798439026,
+      "learning_rate": 3.037014805922369e-05,
+      "loss": 0.5137,
+      "step": 9823
+    },
+    {
+      "epoch": 12.57472,
+      "grad_norm": 0.9623469114303589,
+      "learning_rate": 3.0368147258903562e-05,
+      "loss": 0.564,
+      "step": 9824
+    },
+    {
+      "epoch": 12.576,
+      "grad_norm": 0.9758321046829224,
+      "learning_rate": 3.0366146458583434e-05,
+      "loss": 0.586,
+      "step": 9825
+    },
+    {
+      "epoch": 12.57728,
+      "grad_norm": 0.9277058839797974,
+      "learning_rate": 3.0364145658263306e-05,
+      "loss": 0.5431,
+      "step": 9826
+    },
+    {
+      "epoch": 12.57856,
+      "grad_norm": 0.9288407564163208,
+      "learning_rate": 3.036214485794318e-05,
+      "loss": 0.5268,
+      "step": 9827
+    },
+    {
+      "epoch": 12.57984,
+      "grad_norm": 0.9227582812309265,
+      "learning_rate": 3.0360144057623053e-05,
+      "loss": 0.5604,
+      "step": 9828
+    },
+    {
+      "epoch": 12.58112,
+      "grad_norm": 0.8422914743423462,
+      "learning_rate": 3.0358143257302925e-05,
+      "loss": 0.5244,
+      "step": 9829
+    },
+    {
+      "epoch": 12.5824,
+      "grad_norm": 0.9041396975517273,
+      "learning_rate": 3.0356142456982793e-05,
+      "loss": 0.558,
+      "step": 9830
+    },
+    {
+      "epoch": 12.58368,
+      "grad_norm": 0.9546091556549072,
+      "learning_rate": 3.0354141656662665e-05,
+      "loss": 0.5733,
+      "step": 9831
+    },
+    {
+      "epoch": 12.58496,
+      "grad_norm": 0.9388670325279236,
+      "learning_rate": 3.0352140856342537e-05,
+      "loss": 0.542,
+      "step": 9832
+    },
+    {
+      "epoch": 12.58624,
+      "grad_norm": 0.9128877520561218,
+      "learning_rate": 3.035014005602241e-05,
+      "loss": 0.5452,
+      "step": 9833
+    },
+    {
+      "epoch": 12.58752,
+      "grad_norm": 0.9351105690002441,
+      "learning_rate": 3.0348139255702284e-05,
+      "loss": 0.5283,
+      "step": 9834
+    },
+    {
+      "epoch": 12.588799999999999,
+      "grad_norm": 0.9028121829032898,
+      "learning_rate": 3.0346138455382156e-05,
+      "loss": 0.5479,
+      "step": 9835
+    },
+    {
+      "epoch": 12.59008,
+      "grad_norm": 0.9574125409126282,
+      "learning_rate": 3.0344137655062028e-05,
+      "loss": 0.5786,
+      "step": 9836
+    },
+    {
+      "epoch": 12.59136,
+      "grad_norm": 0.9389735460281372,
+      "learning_rate": 3.03421368547419e-05,
+      "loss": 0.5688,
+      "step": 9837
+    },
+    {
+      "epoch": 12.59264,
+      "grad_norm": 0.889968752861023,
+      "learning_rate": 3.0340136054421768e-05,
+      "loss": 0.5263,
+      "step": 9838
+    },
+    {
+      "epoch": 12.59392,
+      "grad_norm": 0.8959866762161255,
+      "learning_rate": 3.033813525410164e-05,
+      "loss": 0.5766,
+      "step": 9839
+    },
+    {
+      "epoch": 12.5952,
+      "grad_norm": 0.9044232964515686,
+      "learning_rate": 3.0336134453781512e-05,
+      "loss": 0.536,
+      "step": 9840
+    },
+    {
+      "epoch": 12.59648,
+      "grad_norm": 0.896634042263031,
+      "learning_rate": 3.0334133653461387e-05,
+      "loss": 0.5389,
+      "step": 9841
+    },
+    {
+      "epoch": 12.59776,
+      "grad_norm": 0.9344484806060791,
+      "learning_rate": 3.033213285314126e-05,
+      "loss": 0.552,
+      "step": 9842
+    },
+    {
+      "epoch": 12.59904,
+      "grad_norm": 1.0194035768508911,
+      "learning_rate": 3.033013205282113e-05,
+      "loss": 0.5967,
+      "step": 9843
+    },
+    {
+      "epoch": 12.60032,
+      "grad_norm": 0.9549538493156433,
+      "learning_rate": 3.0328131252501003e-05,
+      "loss": 0.5707,
+      "step": 9844
+    },
+    {
+      "epoch": 12.6016,
+      "grad_norm": 0.9823207259178162,
+      "learning_rate": 3.0326130452180875e-05,
+      "loss": 0.5702,
+      "step": 9845
+    },
+    {
+      "epoch": 12.60288,
+      "grad_norm": 0.9882277846336365,
+      "learning_rate": 3.0324129651860743e-05,
+      "loss": 0.6005,
+      "step": 9846
+    },
+    {
+      "epoch": 12.60416,
+      "grad_norm": 0.8940028548240662,
+      "learning_rate": 3.0322128851540615e-05,
+      "loss": 0.5271,
+      "step": 9847
+    },
+    {
+      "epoch": 12.60544,
+      "grad_norm": 0.9345920085906982,
+      "learning_rate": 3.0320128051220494e-05,
+      "loss": 0.5547,
+      "step": 9848
+    },
+    {
+      "epoch": 12.60672,
+      "grad_norm": 0.8393700122833252,
+      "learning_rate": 3.0318127250900362e-05,
+      "loss": 0.5294,
+      "step": 9849
+    },
+    {
+      "epoch": 12.608,
+      "grad_norm": 0.9185911417007446,
+      "learning_rate": 3.0316126450580234e-05,
+      "loss": 0.5218,
+      "step": 9850
+    },
+    {
+      "epoch": 12.60928,
+      "grad_norm": 0.9599105715751648,
+      "learning_rate": 3.0314125650260106e-05,
+      "loss": 0.5365,
+      "step": 9851
+    },
+    {
+      "epoch": 12.61056,
+      "grad_norm": 0.9423932433128357,
+      "learning_rate": 3.0312124849939978e-05,
+      "loss": 0.5799,
+      "step": 9852
+    },
+    {
+      "epoch": 12.61184,
+      "grad_norm": 0.9011184573173523,
+      "learning_rate": 3.031012404961985e-05,
+      "loss": 0.5154,
+      "step": 9853
+    },
+    {
+      "epoch": 12.61312,
+      "grad_norm": 0.9672764539718628,
+      "learning_rate": 3.0308123249299718e-05,
+      "loss": 0.5381,
+      "step": 9854
+    },
+    {
+      "epoch": 12.6144,
+      "grad_norm": 0.9262174963951111,
+      "learning_rate": 3.0306122448979597e-05,
+      "loss": 0.5639,
+      "step": 9855
+    },
+    {
+      "epoch": 12.61568,
+      "grad_norm": 0.9052280783653259,
+      "learning_rate": 3.030412164865947e-05,
+      "loss": 0.5398,
+      "step": 9856
+    },
+    {
+      "epoch": 12.61696,
+      "grad_norm": 0.9824522733688354,
+      "learning_rate": 3.0302120848339337e-05,
+      "loss": 0.5554,
+      "step": 9857
+    },
+    {
+      "epoch": 12.61824,
+      "grad_norm": 0.9158837795257568,
+      "learning_rate": 3.030012004801921e-05,
+      "loss": 0.5455,
+      "step": 9858
+    },
+    {
+      "epoch": 12.61952,
+      "grad_norm": 0.8802503943443298,
+      "learning_rate": 3.029811924769908e-05,
+      "loss": 0.5605,
+      "step": 9859
+    },
+    {
+      "epoch": 12.6208,
+      "grad_norm": 0.9751753807067871,
+      "learning_rate": 3.0296118447378952e-05,
+      "loss": 0.6144,
+      "step": 9860
+    },
+    {
+      "epoch": 12.62208,
+      "grad_norm": 0.9232403039932251,
+      "learning_rate": 3.0294117647058824e-05,
+      "loss": 0.5928,
+      "step": 9861
+    },
+    {
+      "epoch": 12.62336,
+      "grad_norm": 0.9599583148956299,
+      "learning_rate": 3.02921168467387e-05,
+      "loss": 0.5603,
+      "step": 9862
+    },
+    {
+      "epoch": 12.62464,
+      "grad_norm": 0.8870642185211182,
+      "learning_rate": 3.029011604641857e-05,
+      "loss": 0.5307,
+      "step": 9863
+    },
+    {
+      "epoch": 12.62592,
+      "grad_norm": 0.9634974598884583,
+      "learning_rate": 3.0288115246098443e-05,
+      "loss": 0.5891,
+      "step": 9864
+    },
+    {
+      "epoch": 12.6272,
+      "grad_norm": 0.8991132974624634,
+      "learning_rate": 3.028611444577831e-05,
+      "loss": 0.5648,
+      "step": 9865
+    },
+    {
+      "epoch": 12.62848,
+      "grad_norm": 0.8831954002380371,
+      "learning_rate": 3.0284113645458184e-05,
+      "loss": 0.5817,
+      "step": 9866
+    },
+    {
+      "epoch": 12.62976,
+      "grad_norm": 0.9551595449447632,
+      "learning_rate": 3.0282112845138055e-05,
+      "loss": 0.5596,
+      "step": 9867
+    },
+    {
+      "epoch": 12.63104,
+      "grad_norm": 0.9024450182914734,
+      "learning_rate": 3.0280112044817927e-05,
+      "loss": 0.5449,
+      "step": 9868
+    },
+    {
+      "epoch": 12.63232,
+      "grad_norm": 0.8698076009750366,
+      "learning_rate": 3.0278111244497803e-05,
+      "loss": 0.5622,
+      "step": 9869
+    },
+    {
+      "epoch": 12.6336,
+      "grad_norm": 0.9642755389213562,
+      "learning_rate": 3.0276110444177674e-05,
+      "loss": 0.5489,
+      "step": 9870
+    },
+    {
+      "epoch": 12.63488,
+      "grad_norm": 0.8824167251586914,
+      "learning_rate": 3.0274109643857546e-05,
+      "loss": 0.5393,
+      "step": 9871
+    },
+    {
+      "epoch": 12.63616,
+      "grad_norm": 0.9382432103157043,
+      "learning_rate": 3.0272108843537418e-05,
+      "loss": 0.5089,
+      "step": 9872
+    },
+    {
+      "epoch": 12.63744,
+      "grad_norm": 0.9504085779190063,
+      "learning_rate": 3.0270108043217287e-05,
+      "loss": 0.5866,
+      "step": 9873
+    },
+    {
+      "epoch": 12.63872,
+      "grad_norm": 0.9067440032958984,
+      "learning_rate": 3.026810724289716e-05,
+      "loss": 0.5128,
+      "step": 9874
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 0.8995640277862549,
+      "learning_rate": 3.026610644257703e-05,
+      "loss": 0.5669,
+      "step": 9875
+    },
+    {
+      "epoch": 12.64128,
+      "grad_norm": 0.919801652431488,
+      "learning_rate": 3.0264105642256906e-05,
+      "loss": 0.5359,
+      "step": 9876
+    },
+    {
+      "epoch": 12.64256,
+      "grad_norm": 0.8864008188247681,
+      "learning_rate": 3.0262104841936777e-05,
+      "loss": 0.5566,
+      "step": 9877
+    },
+    {
+      "epoch": 12.64384,
+      "grad_norm": 0.940040111541748,
+      "learning_rate": 3.026010404161665e-05,
+      "loss": 0.4943,
+      "step": 9878
+    },
+    {
+      "epoch": 12.64512,
+      "grad_norm": 0.9585533738136292,
+      "learning_rate": 3.025810324129652e-05,
+      "loss": 0.6033,
+      "step": 9879
+    },
+    {
+      "epoch": 12.6464,
+      "grad_norm": 0.9533222317695618,
+      "learning_rate": 3.0256102440976393e-05,
+      "loss": 0.5542,
+      "step": 9880
+    },
+    {
+      "epoch": 12.64768,
+      "grad_norm": 0.9033933877944946,
+      "learning_rate": 3.025410164065626e-05,
+      "loss": 0.5347,
+      "step": 9881
+    },
+    {
+      "epoch": 12.64896,
+      "grad_norm": 0.8955773115158081,
+      "learning_rate": 3.0252100840336133e-05,
+      "loss": 0.5131,
+      "step": 9882
+    },
+    {
+      "epoch": 12.65024,
+      "grad_norm": 0.9527932405471802,
+      "learning_rate": 3.0250100040016012e-05,
+      "loss": 0.5438,
+      "step": 9883
+    },
+    {
+      "epoch": 12.65152,
+      "grad_norm": 0.9614503979682922,
+      "learning_rate": 3.024809923969588e-05,
+      "loss": 0.5858,
+      "step": 9884
+    },
+    {
+      "epoch": 12.6528,
+      "grad_norm": 0.9274621605873108,
+      "learning_rate": 3.0246098439375752e-05,
+      "loss": 0.4716,
+      "step": 9885
+    },
+    {
+      "epoch": 12.65408,
+      "grad_norm": 0.8693538904190063,
+      "learning_rate": 3.0244097639055624e-05,
+      "loss": 0.5251,
+      "step": 9886
+    },
+    {
+      "epoch": 12.65536,
+      "grad_norm": 0.9951110482215881,
+      "learning_rate": 3.0242096838735496e-05,
+      "loss": 0.5828,
+      "step": 9887
+    },
+    {
+      "epoch": 12.65664,
+      "grad_norm": 0.9591448903083801,
+      "learning_rate": 3.0240096038415368e-05,
+      "loss": 0.5891,
+      "step": 9888
+    },
+    {
+      "epoch": 12.65792,
+      "grad_norm": 0.9329800009727478,
+      "learning_rate": 3.0238095238095236e-05,
+      "loss": 0.5501,
+      "step": 9889
+    },
+    {
+      "epoch": 12.6592,
+      "grad_norm": 0.911115288734436,
+      "learning_rate": 3.0236094437775108e-05,
+      "loss": 0.5655,
+      "step": 9890
+    },
+    {
+      "epoch": 12.66048,
+      "grad_norm": 0.9046706557273865,
+      "learning_rate": 3.0234093637454987e-05,
+      "loss": 0.5582,
+      "step": 9891
+    },
+    {
+      "epoch": 12.66176,
+      "grad_norm": 0.9165447950363159,
+      "learning_rate": 3.0232092837134855e-05,
+      "loss": 0.5773,
+      "step": 9892
+    },
+    {
+      "epoch": 12.66304,
+      "grad_norm": 0.9891864657402039,
+      "learning_rate": 3.0230092036814727e-05,
+      "loss": 0.6351,
+      "step": 9893
+    },
+    {
+      "epoch": 12.66432,
+      "grad_norm": 0.8324431777000427,
+      "learning_rate": 3.02280912364946e-05,
+      "loss": 0.5198,
+      "step": 9894
+    },
+    {
+      "epoch": 12.6656,
+      "grad_norm": 0.8809410929679871,
+      "learning_rate": 3.022609043617447e-05,
+      "loss": 0.5294,
+      "step": 9895
+    },
+    {
+      "epoch": 12.66688,
+      "grad_norm": 0.8862389922142029,
+      "learning_rate": 3.0224089635854343e-05,
+      "loss": 0.5305,
+      "step": 9896
+    },
+    {
+      "epoch": 12.66816,
+      "grad_norm": 0.8923200368881226,
+      "learning_rate": 3.022208883553421e-05,
+      "loss": 0.5304,
+      "step": 9897
+    },
+    {
+      "epoch": 12.66944,
+      "grad_norm": 0.9289109110832214,
+      "learning_rate": 3.022008803521409e-05,
+      "loss": 0.5732,
+      "step": 9898
+    },
+    {
+      "epoch": 12.67072,
+      "grad_norm": 0.9402267336845398,
+      "learning_rate": 3.021808723489396e-05,
+      "loss": 0.5936,
+      "step": 9899
+    },
+    {
+      "epoch": 12.672,
+      "grad_norm": 0.9303426146507263,
+      "learning_rate": 3.021608643457383e-05,
+      "loss": 0.555,
+      "step": 9900
+    },
+    {
+      "epoch": 12.67328,
+      "grad_norm": 0.8610079288482666,
+      "learning_rate": 3.0214085634253702e-05,
+      "loss": 0.5454,
+      "step": 9901
+    },
+    {
+      "epoch": 12.67456,
+      "grad_norm": 0.9670707583427429,
+      "learning_rate": 3.0212084833933574e-05,
+      "loss": 0.5369,
+      "step": 9902
+    },
+    {
+      "epoch": 12.67584,
+      "grad_norm": 0.9356277585029602,
+      "learning_rate": 3.0210084033613446e-05,
+      "loss": 0.556,
+      "step": 9903
+    },
+    {
+      "epoch": 12.67712,
+      "grad_norm": 0.8627066016197205,
+      "learning_rate": 3.0208083233293317e-05,
+      "loss": 0.5081,
+      "step": 9904
+    },
+    {
+      "epoch": 12.6784,
+      "grad_norm": 0.918406069278717,
+      "learning_rate": 3.0206082432973193e-05,
+      "loss": 0.5187,
+      "step": 9905
+    },
+    {
+      "epoch": 12.67968,
+      "grad_norm": 0.9237809777259827,
+      "learning_rate": 3.0204081632653065e-05,
+      "loss": 0.5165,
+      "step": 9906
+    },
+    {
+      "epoch": 12.68096,
+      "grad_norm": 0.8767214417457581,
+      "learning_rate": 3.0202080832332936e-05,
+      "loss": 0.4903,
+      "step": 9907
+    },
+    {
+      "epoch": 12.68224,
+      "grad_norm": 1.0054161548614502,
+      "learning_rate": 3.0200080032012805e-05,
+      "loss": 0.5947,
+      "step": 9908
+    },
+    {
+      "epoch": 12.68352,
+      "grad_norm": 0.9165403246879578,
+      "learning_rate": 3.0198079231692677e-05,
+      "loss": 0.5614,
+      "step": 9909
+    },
+    {
+      "epoch": 12.6848,
+      "grad_norm": 0.9228652715682983,
+      "learning_rate": 3.019607843137255e-05,
+      "loss": 0.5589,
+      "step": 9910
+    },
+    {
+      "epoch": 12.68608,
+      "grad_norm": 0.8921694755554199,
+      "learning_rate": 3.019407763105242e-05,
+      "loss": 0.5131,
+      "step": 9911
+    },
+    {
+      "epoch": 12.68736,
+      "grad_norm": 0.9007654786109924,
+      "learning_rate": 3.0192076830732296e-05,
+      "loss": 0.5555,
+      "step": 9912
+    },
+    {
+      "epoch": 12.68864,
+      "grad_norm": 0.8586267828941345,
+      "learning_rate": 3.0190076030412168e-05,
+      "loss": 0.5562,
+      "step": 9913
+    },
+    {
+      "epoch": 12.68992,
+      "grad_norm": 0.9020005464553833,
+      "learning_rate": 3.018807523009204e-05,
+      "loss": 0.5793,
+      "step": 9914
+    },
+    {
+      "epoch": 12.6912,
+      "grad_norm": 0.887444019317627,
+      "learning_rate": 3.018607442977191e-05,
+      "loss": 0.572,
+      "step": 9915
+    },
+    {
+      "epoch": 12.69248,
+      "grad_norm": 0.9526282548904419,
+      "learning_rate": 3.018407362945178e-05,
+      "loss": 0.574,
+      "step": 9916
+    },
+    {
+      "epoch": 12.69376,
+      "grad_norm": 0.9392738938331604,
+      "learning_rate": 3.018207282913165e-05,
+      "loss": 0.5512,
+      "step": 9917
+    },
+    {
+      "epoch": 12.69504,
+      "grad_norm": 0.938337504863739,
+      "learning_rate": 3.0180072028811523e-05,
+      "loss": 0.5903,
+      "step": 9918
+    },
+    {
+      "epoch": 12.69632,
+      "grad_norm": 0.9258535504341125,
+      "learning_rate": 3.01780712284914e-05,
+      "loss": 0.5464,
+      "step": 9919
+    },
+    {
+      "epoch": 12.6976,
+      "grad_norm": 0.9325132369995117,
+      "learning_rate": 3.017607042817127e-05,
+      "loss": 0.5417,
+      "step": 9920
+    },
+    {
+      "epoch": 12.698879999999999,
+      "grad_norm": 0.8271166682243347,
+      "learning_rate": 3.0174069627851142e-05,
+      "loss": 0.4818,
+      "step": 9921
+    },
+    {
+      "epoch": 12.70016,
+      "grad_norm": 0.9156856536865234,
+      "learning_rate": 3.0172068827531014e-05,
+      "loss": 0.508,
+      "step": 9922
+    },
+    {
+      "epoch": 12.70144,
+      "grad_norm": 0.9150842428207397,
+      "learning_rate": 3.0170068027210886e-05,
+      "loss": 0.5311,
+      "step": 9923
+    },
+    {
+      "epoch": 12.70272,
+      "grad_norm": 0.9068151116371155,
+      "learning_rate": 3.0168067226890755e-05,
+      "loss": 0.5557,
+      "step": 9924
+    },
+    {
+      "epoch": 12.704,
+      "grad_norm": 0.9375526309013367,
+      "learning_rate": 3.0166066426570626e-05,
+      "loss": 0.5309,
+      "step": 9925
+    },
+    {
+      "epoch": 12.70528,
+      "grad_norm": 0.9023036360740662,
+      "learning_rate": 3.0164065626250505e-05,
+      "loss": 0.5057,
+      "step": 9926
+    },
+    {
+      "epoch": 12.70656,
+      "grad_norm": 0.9717710614204407,
+      "learning_rate": 3.0162064825930374e-05,
+      "loss": 0.5915,
+      "step": 9927
+    },
+    {
+      "epoch": 12.707840000000001,
+      "grad_norm": 0.8739220499992371,
+      "learning_rate": 3.0160064025610245e-05,
+      "loss": 0.5367,
+      "step": 9928
+    },
+    {
+      "epoch": 12.70912,
+      "grad_norm": 0.903511643409729,
+      "learning_rate": 3.0158063225290117e-05,
+      "loss": 0.515,
+      "step": 9929
+    },
+    {
+      "epoch": 12.7104,
+      "grad_norm": 0.9045807719230652,
+      "learning_rate": 3.015606242496999e-05,
+      "loss": 0.5255,
+      "step": 9930
+    },
+    {
+      "epoch": 12.71168,
+      "grad_norm": 0.8613312244415283,
+      "learning_rate": 3.015406162464986e-05,
+      "loss": 0.5145,
+      "step": 9931
+    },
+    {
+      "epoch": 12.71296,
+      "grad_norm": 0.9359320998191833,
+      "learning_rate": 3.015206082432973e-05,
+      "loss": 0.5632,
+      "step": 9932
+    },
+    {
+      "epoch": 12.71424,
+      "grad_norm": 0.8847656846046448,
+      "learning_rate": 3.0150060024009608e-05,
+      "loss": 0.5334,
+      "step": 9933
+    },
+    {
+      "epoch": 12.71552,
+      "grad_norm": 0.9487404823303223,
+      "learning_rate": 3.014805922368948e-05,
+      "loss": 0.5566,
+      "step": 9934
+    },
+    {
+      "epoch": 12.7168,
+      "grad_norm": 1.0114420652389526,
+      "learning_rate": 3.014605842336935e-05,
+      "loss": 0.5976,
+      "step": 9935
+    },
+    {
+      "epoch": 12.71808,
+      "grad_norm": 0.9208101034164429,
+      "learning_rate": 3.014405762304922e-05,
+      "loss": 0.5365,
+      "step": 9936
+    },
+    {
+      "epoch": 12.71936,
+      "grad_norm": 0.8874728083610535,
+      "learning_rate": 3.0142056822729092e-05,
+      "loss": 0.5601,
+      "step": 9937
+    },
+    {
+      "epoch": 12.72064,
+      "grad_norm": 0.8904001116752625,
+      "learning_rate": 3.0140056022408964e-05,
+      "loss": 0.534,
+      "step": 9938
+    },
+    {
+      "epoch": 12.72192,
+      "grad_norm": 0.9263580441474915,
+      "learning_rate": 3.0138055222088836e-05,
+      "loss": 0.5514,
+      "step": 9939
+    },
+    {
+      "epoch": 12.7232,
+      "grad_norm": 0.9768319725990295,
+      "learning_rate": 3.013605442176871e-05,
+      "loss": 0.556,
+      "step": 9940
+    },
+    {
+      "epoch": 12.72448,
+      "grad_norm": 0.8745740652084351,
+      "learning_rate": 3.0134053621448583e-05,
+      "loss": 0.559,
+      "step": 9941
+    },
+    {
+      "epoch": 12.72576,
+      "grad_norm": 0.963864266872406,
+      "learning_rate": 3.0132052821128455e-05,
+      "loss": 0.5594,
+      "step": 9942
+    },
+    {
+      "epoch": 12.72704,
+      "grad_norm": 0.9123207330703735,
+      "learning_rate": 3.0130052020808323e-05,
+      "loss": 0.5439,
+      "step": 9943
+    },
+    {
+      "epoch": 12.72832,
+      "grad_norm": 0.8876965045928955,
+      "learning_rate": 3.0128051220488195e-05,
+      "loss": 0.5076,
+      "step": 9944
+    },
+    {
+      "epoch": 12.7296,
+      "grad_norm": 0.8987967371940613,
+      "learning_rate": 3.0126050420168067e-05,
+      "loss": 0.5557,
+      "step": 9945
+    },
+    {
+      "epoch": 12.730879999999999,
+      "grad_norm": 0.9241291880607605,
+      "learning_rate": 3.012404961984794e-05,
+      "loss": 0.5262,
+      "step": 9946
+    },
+    {
+      "epoch": 12.73216,
+      "grad_norm": 0.9232166409492493,
+      "learning_rate": 3.0122048819527814e-05,
+      "loss": 0.5576,
+      "step": 9947
+    },
+    {
+      "epoch": 12.73344,
+      "grad_norm": 0.8862061500549316,
+      "learning_rate": 3.0120048019207686e-05,
+      "loss": 0.5504,
+      "step": 9948
+    },
+    {
+      "epoch": 12.73472,
+      "grad_norm": 0.9697844982147217,
+      "learning_rate": 3.0118047218887558e-05,
+      "loss": 0.578,
+      "step": 9949
+    },
+    {
+      "epoch": 12.736,
+      "grad_norm": 0.9346733093261719,
+      "learning_rate": 3.011604641856743e-05,
+      "loss": 0.5529,
+      "step": 9950
+    },
+    {
+      "epoch": 12.73728,
+      "grad_norm": 0.9218923449516296,
+      "learning_rate": 3.0114045618247298e-05,
+      "loss": 0.5634,
+      "step": 9951
+    },
+    {
+      "epoch": 12.73856,
+      "grad_norm": 1.0053513050079346,
+      "learning_rate": 3.011204481792717e-05,
+      "loss": 0.5313,
+      "step": 9952
+    },
+    {
+      "epoch": 12.739840000000001,
+      "grad_norm": 0.9116829633712769,
+      "learning_rate": 3.0110044017607042e-05,
+      "loss": 0.565,
+      "step": 9953
+    },
+    {
+      "epoch": 12.74112,
+      "grad_norm": 0.8650526404380798,
+      "learning_rate": 3.0108043217286917e-05,
+      "loss": 0.4948,
+      "step": 9954
+    },
+    {
+      "epoch": 12.7424,
+      "grad_norm": 0.9802688360214233,
+      "learning_rate": 3.010604241696679e-05,
+      "loss": 0.5878,
+      "step": 9955
+    },
+    {
+      "epoch": 12.74368,
+      "grad_norm": 0.9233496785163879,
+      "learning_rate": 3.010404161664666e-05,
+      "loss": 0.5235,
+      "step": 9956
+    },
+    {
+      "epoch": 12.74496,
+      "grad_norm": 0.9107963442802429,
+      "learning_rate": 3.0102040816326533e-05,
+      "loss": 0.5203,
+      "step": 9957
+    },
+    {
+      "epoch": 12.74624,
+      "grad_norm": 0.9463147521018982,
+      "learning_rate": 3.0100040016006405e-05,
+      "loss": 0.5231,
+      "step": 9958
+    },
+    {
+      "epoch": 12.74752,
+      "grad_norm": 0.9093472957611084,
+      "learning_rate": 3.0098039215686273e-05,
+      "loss": 0.5534,
+      "step": 9959
+    },
+    {
+      "epoch": 12.7488,
+      "grad_norm": 1.0168708562850952,
+      "learning_rate": 3.0096038415366145e-05,
+      "loss": 0.6179,
+      "step": 9960
+    },
+    {
+      "epoch": 12.75008,
+      "grad_norm": 0.8969560265541077,
+      "learning_rate": 3.0094037615046023e-05,
+      "loss": 0.5302,
+      "step": 9961
+    },
+    {
+      "epoch": 12.75136,
+      "grad_norm": 0.9327144026756287,
+      "learning_rate": 3.0092036814725892e-05,
+      "loss": 0.5877,
+      "step": 9962
+    },
+    {
+      "epoch": 12.75264,
+      "grad_norm": 0.9616484642028809,
+      "learning_rate": 3.0090036014405764e-05,
+      "loss": 0.5779,
+      "step": 9963
+    },
+    {
+      "epoch": 12.75392,
+      "grad_norm": 0.9223829507827759,
+      "learning_rate": 3.0088035214085636e-05,
+      "loss": 0.5649,
+      "step": 9964
+    },
+    {
+      "epoch": 12.7552,
+      "grad_norm": 0.8914867639541626,
+      "learning_rate": 3.0086034413765508e-05,
+      "loss": 0.5435,
+      "step": 9965
+    },
+    {
+      "epoch": 12.75648,
+      "grad_norm": 0.8898464441299438,
+      "learning_rate": 3.008403361344538e-05,
+      "loss": 0.5918,
+      "step": 9966
+    },
+    {
+      "epoch": 12.75776,
+      "grad_norm": 0.8924893140792847,
+      "learning_rate": 3.0082032813125248e-05,
+      "loss": 0.5312,
+      "step": 9967
+    },
+    {
+      "epoch": 12.75904,
+      "grad_norm": 0.9874442219734192,
+      "learning_rate": 3.0080032012805126e-05,
+      "loss": 0.5591,
+      "step": 9968
+    },
+    {
+      "epoch": 12.76032,
+      "grad_norm": 0.8837193846702576,
+      "learning_rate": 3.0078031212485e-05,
+      "loss": 0.5174,
+      "step": 9969
+    },
+    {
+      "epoch": 12.7616,
+      "grad_norm": 0.9593360424041748,
+      "learning_rate": 3.0076030412164867e-05,
+      "loss": 0.5871,
+      "step": 9970
+    },
+    {
+      "epoch": 12.76288,
+      "grad_norm": 0.9044248461723328,
+      "learning_rate": 3.007402961184474e-05,
+      "loss": 0.5099,
+      "step": 9971
+    },
+    {
+      "epoch": 12.76416,
+      "grad_norm": 0.9392541646957397,
+      "learning_rate": 3.007202881152461e-05,
+      "loss": 0.5553,
+      "step": 9972
+    },
+    {
+      "epoch": 12.76544,
+      "grad_norm": 0.9666686058044434,
+      "learning_rate": 3.0070028011204482e-05,
+      "loss": 0.54,
+      "step": 9973
+    },
+    {
+      "epoch": 12.76672,
+      "grad_norm": 0.9441251754760742,
+      "learning_rate": 3.0068027210884354e-05,
+      "loss": 0.6014,
+      "step": 9974
+    },
+    {
+      "epoch": 12.768,
+      "grad_norm": 0.8636193871498108,
+      "learning_rate": 3.006602641056423e-05,
+      "loss": 0.5608,
+      "step": 9975
+    },
+    {
+      "epoch": 12.76928,
+      "grad_norm": 0.8688570261001587,
+      "learning_rate": 3.00640256102441e-05,
+      "loss": 0.5199,
+      "step": 9976
+    },
+    {
+      "epoch": 12.77056,
+      "grad_norm": 0.9947240352630615,
+      "learning_rate": 3.0062024809923973e-05,
+      "loss": 0.5931,
+      "step": 9977
+    },
+    {
+      "epoch": 12.77184,
+      "grad_norm": 0.9212228059768677,
+      "learning_rate": 3.006002400960384e-05,
+      "loss": 0.4972,
+      "step": 9978
+    },
+    {
+      "epoch": 12.77312,
+      "grad_norm": 0.9512316584587097,
+      "learning_rate": 3.0058023209283714e-05,
+      "loss": 0.5568,
+      "step": 9979
+    },
+    {
+      "epoch": 12.7744,
+      "grad_norm": 0.9175422191619873,
+      "learning_rate": 3.0056022408963585e-05,
+      "loss": 0.5375,
+      "step": 9980
+    },
+    {
+      "epoch": 12.77568,
+      "grad_norm": 0.918135404586792,
+      "learning_rate": 3.0054021608643457e-05,
+      "loss": 0.5353,
+      "step": 9981
+    },
+    {
+      "epoch": 12.77696,
+      "grad_norm": 0.937879204750061,
+      "learning_rate": 3.0052020808323332e-05,
+      "loss": 0.5707,
+      "step": 9982
+    },
+    {
+      "epoch": 12.77824,
+      "grad_norm": 0.8635044693946838,
+      "learning_rate": 3.0050020008003204e-05,
+      "loss": 0.5118,
+      "step": 9983
+    },
+    {
+      "epoch": 12.77952,
+      "grad_norm": 0.8819299936294556,
+      "learning_rate": 3.0048019207683076e-05,
+      "loss": 0.4941,
+      "step": 9984
+    },
+    {
+      "epoch": 12.7808,
+      "grad_norm": 0.9287380576133728,
+      "learning_rate": 3.0046018407362948e-05,
+      "loss": 0.5739,
+      "step": 9985
+    },
+    {
+      "epoch": 12.78208,
+      "grad_norm": 0.8525633811950684,
+      "learning_rate": 3.0044017607042817e-05,
+      "loss": 0.5316,
+      "step": 9986
+    },
+    {
+      "epoch": 12.78336,
+      "grad_norm": 0.8659927845001221,
+      "learning_rate": 3.004201680672269e-05,
+      "loss": 0.5096,
+      "step": 9987
+    },
+    {
+      "epoch": 12.78464,
+      "grad_norm": 0.876082181930542,
+      "learning_rate": 3.004001600640256e-05,
+      "loss": 0.5481,
+      "step": 9988
+    },
+    {
+      "epoch": 12.78592,
+      "grad_norm": 0.8915519118309021,
+      "learning_rate": 3.0038015206082435e-05,
+      "loss": 0.5542,
+      "step": 9989
+    },
+    {
+      "epoch": 12.7872,
+      "grad_norm": 0.961881697177887,
+      "learning_rate": 3.0036014405762307e-05,
+      "loss": 0.541,
+      "step": 9990
+    },
+    {
+      "epoch": 12.78848,
+      "grad_norm": 0.9022373557090759,
+      "learning_rate": 3.003401360544218e-05,
+      "loss": 0.5454,
+      "step": 9991
+    },
+    {
+      "epoch": 12.78976,
+      "grad_norm": 0.9297406077384949,
+      "learning_rate": 3.003201280512205e-05,
+      "loss": 0.5537,
+      "step": 9992
+    },
+    {
+      "epoch": 12.79104,
+      "grad_norm": 0.9228201508522034,
+      "learning_rate": 3.0030012004801923e-05,
+      "loss": 0.5281,
+      "step": 9993
+    },
+    {
+      "epoch": 12.79232,
+      "grad_norm": 0.8959901928901672,
+      "learning_rate": 3.002801120448179e-05,
+      "loss": 0.5428,
+      "step": 9994
+    },
+    {
+      "epoch": 12.7936,
+      "grad_norm": 0.903972864151001,
+      "learning_rate": 3.0026010404161663e-05,
+      "loss": 0.5513,
+      "step": 9995
+    },
+    {
+      "epoch": 12.79488,
+      "grad_norm": 0.9172789454460144,
+      "learning_rate": 3.0024009603841542e-05,
+      "loss": 0.5723,
+      "step": 9996
+    },
+    {
+      "epoch": 12.79616,
+      "grad_norm": 0.872391402721405,
+      "learning_rate": 3.002200880352141e-05,
+      "loss": 0.5302,
+      "step": 9997
+    },
+    {
+      "epoch": 12.79744,
+      "grad_norm": 0.9103299379348755,
+      "learning_rate": 3.0020008003201282e-05,
+      "loss": 0.5548,
+      "step": 9998
+    },
+    {
+      "epoch": 12.79872,
+      "grad_norm": 0.9463931322097778,
+      "learning_rate": 3.0018007202881154e-05,
+      "loss": 0.5299,
+      "step": 9999
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 0.9099661707878113,
+      "learning_rate": 3.0016006402561026e-05,
+      "loss": 0.5733,
+      "step": 10000
+    },
+    {
+      "epoch": 12.80128,
+      "grad_norm": 0.8782345652580261,
+      "learning_rate": 3.0014005602240898e-05,
+      "loss": 0.5495,
+      "step": 10001
+    },
+    {
+      "epoch": 12.80256,
+      "grad_norm": 0.9504924416542053,
+      "learning_rate": 3.0012004801920766e-05,
+      "loss": 0.5548,
+      "step": 10002
+    },
+    {
+      "epoch": 12.80384,
+      "grad_norm": 0.9535840749740601,
+      "learning_rate": 3.0010004001600638e-05,
+      "loss": 0.6183,
+      "step": 10003
+    },
+    {
+      "epoch": 12.80512,
+      "grad_norm": 0.9448462724685669,
+      "learning_rate": 3.0008003201280517e-05,
+      "loss": 0.5871,
+      "step": 10004
+    },
+    {
+      "epoch": 12.8064,
+      "grad_norm": 0.856593906879425,
+      "learning_rate": 3.0006002400960385e-05,
+      "loss": 0.5501,
+      "step": 10005
+    },
+    {
+      "epoch": 12.80768,
+      "grad_norm": 0.9613591432571411,
+      "learning_rate": 3.0004001600640257e-05,
+      "loss": 0.5669,
+      "step": 10006
+    },
+    {
+      "epoch": 12.80896,
+      "grad_norm": 0.9325043559074402,
+      "learning_rate": 3.000200080032013e-05,
+      "loss": 0.5625,
+      "step": 10007
+    },
+    {
+      "epoch": 12.81024,
+      "grad_norm": 0.8977808356285095,
+      "learning_rate": 3e-05,
+      "loss": 0.5378,
+      "step": 10008
+    },
+    {
+      "epoch": 12.81152,
+      "grad_norm": 0.8371548652648926,
+      "learning_rate": 2.9997999199679873e-05,
+      "loss": 0.5012,
+      "step": 10009
+    },
+    {
+      "epoch": 12.8128,
+      "grad_norm": 0.9407792091369629,
+      "learning_rate": 2.999599839935974e-05,
+      "loss": 0.6057,
+      "step": 10010
+    },
+    {
+      "epoch": 12.81408,
+      "grad_norm": 0.911292552947998,
+      "learning_rate": 2.999399759903962e-05,
+      "loss": 0.5342,
+      "step": 10011
+    },
+    {
+      "epoch": 12.81536,
+      "grad_norm": 0.947562038898468,
+      "learning_rate": 2.999199679871949e-05,
+      "loss": 0.597,
+      "step": 10012
+    },
+    {
+      "epoch": 12.81664,
+      "grad_norm": 0.8605309724807739,
+      "learning_rate": 2.998999599839936e-05,
+      "loss": 0.5011,
+      "step": 10013
+    },
+    {
+      "epoch": 12.81792,
+      "grad_norm": 0.9538720846176147,
+      "learning_rate": 2.9987995198079232e-05,
+      "loss": 0.5383,
+      "step": 10014
+    },
+    {
+      "epoch": 12.8192,
+      "grad_norm": 0.8674808144569397,
+      "learning_rate": 2.9985994397759104e-05,
+      "loss": 0.5635,
+      "step": 10015
+    },
+    {
+      "epoch": 12.82048,
+      "grad_norm": 0.8953600525856018,
+      "learning_rate": 2.9983993597438976e-05,
+      "loss": 0.5338,
+      "step": 10016
+    },
+    {
+      "epoch": 12.82176,
+      "grad_norm": 0.9249882102012634,
+      "learning_rate": 2.9981992797118847e-05,
+      "loss": 0.5529,
+      "step": 10017
+    },
+    {
+      "epoch": 12.82304,
+      "grad_norm": 0.9073203802108765,
+      "learning_rate": 2.9979991996798723e-05,
+      "loss": 0.5515,
+      "step": 10018
+    },
+    {
+      "epoch": 12.82432,
+      "grad_norm": 0.8862814903259277,
+      "learning_rate": 2.9977991196478595e-05,
+      "loss": 0.5639,
+      "step": 10019
+    },
+    {
+      "epoch": 12.8256,
+      "grad_norm": 0.9115248918533325,
+      "learning_rate": 2.9975990396158466e-05,
+      "loss": 0.5338,
+      "step": 10020
+    },
+    {
+      "epoch": 12.82688,
+      "grad_norm": 0.9155441522598267,
+      "learning_rate": 2.9973989595838335e-05,
+      "loss": 0.5904,
+      "step": 10021
+    },
+    {
+      "epoch": 12.82816,
+      "grad_norm": 0.936619222164154,
+      "learning_rate": 2.9971988795518207e-05,
+      "loss": 0.5606,
+      "step": 10022
+    },
+    {
+      "epoch": 12.82944,
+      "grad_norm": 0.8957839012145996,
+      "learning_rate": 2.996998799519808e-05,
+      "loss": 0.5432,
+      "step": 10023
+    },
+    {
+      "epoch": 12.83072,
+      "grad_norm": 0.8455110192298889,
+      "learning_rate": 2.996798719487795e-05,
+      "loss": 0.5095,
+      "step": 10024
+    },
+    {
+      "epoch": 12.832,
+      "grad_norm": 0.8867446184158325,
+      "learning_rate": 2.9965986394557826e-05,
+      "loss": 0.5319,
+      "step": 10025
+    },
+    {
+      "epoch": 12.83328,
+      "grad_norm": 0.8974578976631165,
+      "learning_rate": 2.9963985594237698e-05,
+      "loss": 0.5514,
+      "step": 10026
+    },
+    {
+      "epoch": 12.83456,
+      "grad_norm": 0.8834570050239563,
+      "learning_rate": 2.996198479391757e-05,
+      "loss": 0.5682,
+      "step": 10027
+    },
+    {
+      "epoch": 12.83584,
+      "grad_norm": 0.9071859121322632,
+      "learning_rate": 2.995998399359744e-05,
+      "loss": 0.5334,
+      "step": 10028
+    },
+    {
+      "epoch": 12.83712,
+      "grad_norm": 0.9252317547798157,
+      "learning_rate": 2.995798319327731e-05,
+      "loss": 0.5062,
+      "step": 10029
+    },
+    {
+      "epoch": 12.8384,
+      "grad_norm": 0.9470645189285278,
+      "learning_rate": 2.995598239295718e-05,
+      "loss": 0.5328,
+      "step": 10030
+    },
+    {
+      "epoch": 12.83968,
+      "grad_norm": 0.9187636375427246,
+      "learning_rate": 2.9953981592637053e-05,
+      "loss": 0.6012,
+      "step": 10031
+    },
+    {
+      "epoch": 12.84096,
+      "grad_norm": 0.9187378883361816,
+      "learning_rate": 2.9951980792316932e-05,
+      "loss": 0.582,
+      "step": 10032
+    },
+    {
+      "epoch": 12.84224,
+      "grad_norm": 0.9218907952308655,
+      "learning_rate": 2.99499799919968e-05,
+      "loss": 0.5591,
+      "step": 10033
+    },
+    {
+      "epoch": 12.84352,
+      "grad_norm": 0.8669567704200745,
+      "learning_rate": 2.9947979191676672e-05,
+      "loss": 0.5021,
+      "step": 10034
+    },
+    {
+      "epoch": 12.8448,
+      "grad_norm": 0.9718922972679138,
+      "learning_rate": 2.9945978391356544e-05,
+      "loss": 0.5636,
+      "step": 10035
+    },
+    {
+      "epoch": 12.84608,
+      "grad_norm": 0.9400469064712524,
+      "learning_rate": 2.9943977591036416e-05,
+      "loss": 0.5565,
+      "step": 10036
+    },
+    {
+      "epoch": 12.84736,
+      "grad_norm": 0.9250262975692749,
+      "learning_rate": 2.9941976790716285e-05,
+      "loss": 0.5336,
+      "step": 10037
+    },
+    {
+      "epoch": 12.84864,
+      "grad_norm": 0.9361352324485779,
+      "learning_rate": 2.9939975990396156e-05,
+      "loss": 0.5438,
+      "step": 10038
+    },
+    {
+      "epoch": 12.849920000000001,
+      "grad_norm": 0.9436172246932983,
+      "learning_rate": 2.9937975190076035e-05,
+      "loss": 0.5784,
+      "step": 10039
+    },
+    {
+      "epoch": 12.8512,
+      "grad_norm": 0.9591764807701111,
+      "learning_rate": 2.9935974389755907e-05,
+      "loss": 0.5426,
+      "step": 10040
+    },
+    {
+      "epoch": 12.85248,
+      "grad_norm": 0.9026290774345398,
+      "learning_rate": 2.9933973589435775e-05,
+      "loss": 0.5524,
+      "step": 10041
+    },
+    {
+      "epoch": 12.85376,
+      "grad_norm": 0.8631820678710938,
+      "learning_rate": 2.9931972789115647e-05,
+      "loss": 0.5365,
+      "step": 10042
+    },
+    {
+      "epoch": 12.85504,
+      "grad_norm": 0.8909380435943604,
+      "learning_rate": 2.992997198879552e-05,
+      "loss": 0.4973,
+      "step": 10043
+    },
+    {
+      "epoch": 12.85632,
+      "grad_norm": 0.9117045998573303,
+      "learning_rate": 2.992797118847539e-05,
+      "loss": 0.5777,
+      "step": 10044
+    },
+    {
+      "epoch": 12.8576,
+      "grad_norm": 0.9356653094291687,
+      "learning_rate": 2.992597038815526e-05,
+      "loss": 0.5703,
+      "step": 10045
+    },
+    {
+      "epoch": 12.85888,
+      "grad_norm": 0.9161821007728577,
+      "learning_rate": 2.9923969587835138e-05,
+      "loss": 0.5378,
+      "step": 10046
+    },
+    {
+      "epoch": 12.86016,
+      "grad_norm": 0.9785245060920715,
+      "learning_rate": 2.992196878751501e-05,
+      "loss": 0.6163,
+      "step": 10047
+    },
+    {
+      "epoch": 12.86144,
+      "grad_norm": 0.896397590637207,
+      "learning_rate": 2.9919967987194882e-05,
+      "loss": 0.5743,
+      "step": 10048
+    },
+    {
+      "epoch": 12.86272,
+      "grad_norm": 0.8948002457618713,
+      "learning_rate": 2.991796718687475e-05,
+      "loss": 0.5688,
+      "step": 10049
+    },
+    {
+      "epoch": 12.864,
+      "grad_norm": 0.9673354625701904,
+      "learning_rate": 2.9915966386554622e-05,
+      "loss": 0.572,
+      "step": 10050
+    },
+    {
+      "epoch": 12.86528,
+      "grad_norm": 0.9555987119674683,
+      "learning_rate": 2.9913965586234494e-05,
+      "loss": 0.5516,
+      "step": 10051
+    },
+    {
+      "epoch": 12.86656,
+      "grad_norm": 0.908279299736023,
+      "learning_rate": 2.9911964785914366e-05,
+      "loss": 0.5513,
+      "step": 10052
+    },
+    {
+      "epoch": 12.86784,
+      "grad_norm": 0.8543174862861633,
+      "learning_rate": 2.990996398559424e-05,
+      "loss": 0.5194,
+      "step": 10053
+    },
+    {
+      "epoch": 12.86912,
+      "grad_norm": 0.9750853180885315,
+      "learning_rate": 2.9907963185274113e-05,
+      "loss": 0.5985,
+      "step": 10054
+    },
+    {
+      "epoch": 12.8704,
+      "grad_norm": 0.8315509557723999,
+      "learning_rate": 2.9905962384953985e-05,
+      "loss": 0.5259,
+      "step": 10055
+    },
+    {
+      "epoch": 12.87168,
+      "grad_norm": 0.8759975433349609,
+      "learning_rate": 2.9903961584633857e-05,
+      "loss": 0.5003,
+      "step": 10056
+    },
+    {
+      "epoch": 12.872959999999999,
+      "grad_norm": 0.8982505798339844,
+      "learning_rate": 2.9901960784313725e-05,
+      "loss": 0.5745,
+      "step": 10057
+    },
+    {
+      "epoch": 12.87424,
+      "grad_norm": 0.9328362941741943,
+      "learning_rate": 2.9899959983993597e-05,
+      "loss": 0.5945,
+      "step": 10058
+    },
+    {
+      "epoch": 12.87552,
+      "grad_norm": 0.9085741639137268,
+      "learning_rate": 2.989795918367347e-05,
+      "loss": 0.5479,
+      "step": 10059
+    },
+    {
+      "epoch": 12.8768,
+      "grad_norm": 0.879937469959259,
+      "learning_rate": 2.9895958383353344e-05,
+      "loss": 0.5385,
+      "step": 10060
+    },
+    {
+      "epoch": 12.87808,
+      "grad_norm": 0.932284951210022,
+      "learning_rate": 2.9893957583033216e-05,
+      "loss": 0.5911,
+      "step": 10061
+    },
+    {
+      "epoch": 12.87936,
+      "grad_norm": 0.9298887252807617,
+      "learning_rate": 2.9891956782713088e-05,
+      "loss": 0.5586,
+      "step": 10062
+    },
+    {
+      "epoch": 12.88064,
+      "grad_norm": 0.9269364476203918,
+      "learning_rate": 2.988995598239296e-05,
+      "loss": 0.5432,
+      "step": 10063
+    },
+    {
+      "epoch": 12.881920000000001,
+      "grad_norm": 0.9070879817008972,
+      "learning_rate": 2.988795518207283e-05,
+      "loss": 0.5722,
+      "step": 10064
+    },
+    {
+      "epoch": 12.8832,
+      "grad_norm": 0.9287840723991394,
+      "learning_rate": 2.98859543817527e-05,
+      "loss": 0.5642,
+      "step": 10065
+    },
+    {
+      "epoch": 12.88448,
+      "grad_norm": 0.9351973533630371,
+      "learning_rate": 2.9883953581432572e-05,
+      "loss": 0.5648,
+      "step": 10066
+    },
+    {
+      "epoch": 12.88576,
+      "grad_norm": 0.9211728572845459,
+      "learning_rate": 2.988195278111245e-05,
+      "loss": 0.5144,
+      "step": 10067
+    },
+    {
+      "epoch": 12.88704,
+      "grad_norm": 0.9010429382324219,
+      "learning_rate": 2.987995198079232e-05,
+      "loss": 0.536,
+      "step": 10068
+    },
+    {
+      "epoch": 12.88832,
+      "grad_norm": 0.9658145904541016,
+      "learning_rate": 2.987795118047219e-05,
+      "loss": 0.5977,
+      "step": 10069
+    },
+    {
+      "epoch": 12.8896,
+      "grad_norm": 1.0316338539123535,
+      "learning_rate": 2.9875950380152063e-05,
+      "loss": 0.5963,
+      "step": 10070
+    },
+    {
+      "epoch": 12.89088,
+      "grad_norm": 0.8775344491004944,
+      "learning_rate": 2.9873949579831935e-05,
+      "loss": 0.525,
+      "step": 10071
+    },
+    {
+      "epoch": 12.89216,
+      "grad_norm": 0.9124401807785034,
+      "learning_rate": 2.9871948779511806e-05,
+      "loss": 0.5544,
+      "step": 10072
+    },
+    {
+      "epoch": 12.89344,
+      "grad_norm": 0.9111108183860779,
+      "learning_rate": 2.9869947979191675e-05,
+      "loss": 0.574,
+      "step": 10073
+    },
+    {
+      "epoch": 12.89472,
+      "grad_norm": 0.9600886702537537,
+      "learning_rate": 2.9867947178871553e-05,
+      "loss": 0.5347,
+      "step": 10074
+    },
+    {
+      "epoch": 12.896,
+      "grad_norm": 0.967065691947937,
+      "learning_rate": 2.9865946378551425e-05,
+      "loss": 0.5945,
+      "step": 10075
+    },
+    {
+      "epoch": 12.89728,
+      "grad_norm": 0.8715311288833618,
+      "learning_rate": 2.9863945578231294e-05,
+      "loss": 0.5239,
+      "step": 10076
+    },
+    {
+      "epoch": 12.89856,
+      "grad_norm": 0.8990671634674072,
+      "learning_rate": 2.9861944777911166e-05,
+      "loss": 0.5526,
+      "step": 10077
+    },
+    {
+      "epoch": 12.89984,
+      "grad_norm": 0.8786954283714294,
+      "learning_rate": 2.9859943977591038e-05,
+      "loss": 0.5653,
+      "step": 10078
+    },
+    {
+      "epoch": 12.90112,
+      "grad_norm": 0.9412379860877991,
+      "learning_rate": 2.985794317727091e-05,
+      "loss": 0.5592,
+      "step": 10079
+    },
+    {
+      "epoch": 12.9024,
+      "grad_norm": 0.9109273552894592,
+      "learning_rate": 2.985594237695078e-05,
+      "loss": 0.5307,
+      "step": 10080
+    },
+    {
+      "epoch": 12.90368,
+      "grad_norm": 0.9346706867218018,
+      "learning_rate": 2.9853941576630656e-05,
+      "loss": 0.5661,
+      "step": 10081
+    },
+    {
+      "epoch": 12.904959999999999,
+      "grad_norm": 0.9366570711135864,
+      "learning_rate": 2.985194077631053e-05,
+      "loss": 0.5806,
+      "step": 10082
+    },
+    {
+      "epoch": 12.90624,
+      "grad_norm": 0.906199038028717,
+      "learning_rate": 2.98499399759904e-05,
+      "loss": 0.5577,
+      "step": 10083
+    },
+    {
+      "epoch": 12.90752,
+      "grad_norm": 0.9052384495735168,
+      "learning_rate": 2.984793917567027e-05,
+      "loss": 0.5144,
+      "step": 10084
+    },
+    {
+      "epoch": 12.9088,
+      "grad_norm": 0.8487529754638672,
+      "learning_rate": 2.984593837535014e-05,
+      "loss": 0.5046,
+      "step": 10085
+    },
+    {
+      "epoch": 12.91008,
+      "grad_norm": 0.9538546204566956,
+      "learning_rate": 2.9843937575030012e-05,
+      "loss": 0.5881,
+      "step": 10086
+    },
+    {
+      "epoch": 12.91136,
+      "grad_norm": 0.9074375033378601,
+      "learning_rate": 2.9841936774709884e-05,
+      "loss": 0.5077,
+      "step": 10087
+    },
+    {
+      "epoch": 12.91264,
+      "grad_norm": 0.8918171525001526,
+      "learning_rate": 2.983993597438976e-05,
+      "loss": 0.547,
+      "step": 10088
+    },
+    {
+      "epoch": 12.91392,
+      "grad_norm": 0.877153217792511,
+      "learning_rate": 2.983793517406963e-05,
+      "loss": 0.5116,
+      "step": 10089
+    },
+    {
+      "epoch": 12.9152,
+      "grad_norm": 0.9327219724655151,
+      "learning_rate": 2.9835934373749503e-05,
+      "loss": 0.5949,
+      "step": 10090
+    },
+    {
+      "epoch": 12.91648,
+      "grad_norm": 0.8527021408081055,
+      "learning_rate": 2.9833933573429375e-05,
+      "loss": 0.5414,
+      "step": 10091
+    },
+    {
+      "epoch": 12.91776,
+      "grad_norm": 0.9309681057929993,
+      "learning_rate": 2.9831932773109244e-05,
+      "loss": 0.5714,
+      "step": 10092
+    },
+    {
+      "epoch": 12.91904,
+      "grad_norm": 1.0311362743377686,
+      "learning_rate": 2.9829931972789115e-05,
+      "loss": 0.6181,
+      "step": 10093
+    },
+    {
+      "epoch": 12.92032,
+      "grad_norm": 0.9770155549049377,
+      "learning_rate": 2.9827931172468987e-05,
+      "loss": 0.6033,
+      "step": 10094
+    },
+    {
+      "epoch": 12.9216,
+      "grad_norm": 0.9168463945388794,
+      "learning_rate": 2.9825930372148862e-05,
+      "loss": 0.5045,
+      "step": 10095
+    },
+    {
+      "epoch": 12.92288,
+      "grad_norm": 0.9091753959655762,
+      "learning_rate": 2.9823929571828734e-05,
+      "loss": 0.5097,
+      "step": 10096
+    },
+    {
+      "epoch": 12.92416,
+      "grad_norm": 0.9451500177383423,
+      "learning_rate": 2.9821928771508606e-05,
+      "loss": 0.5358,
+      "step": 10097
+    },
+    {
+      "epoch": 12.92544,
+      "grad_norm": 0.8995253443717957,
+      "learning_rate": 2.9819927971188478e-05,
+      "loss": 0.5671,
+      "step": 10098
+    },
+    {
+      "epoch": 12.92672,
+      "grad_norm": 0.9600387215614319,
+      "learning_rate": 2.981792717086835e-05,
+      "loss": 0.6202,
+      "step": 10099
+    },
+    {
+      "epoch": 12.928,
+      "grad_norm": 0.9363338351249695,
+      "learning_rate": 2.981592637054822e-05,
+      "loss": 0.5377,
+      "step": 10100
+    },
+    {
+      "epoch": 12.92928,
+      "grad_norm": 0.9418403506278992,
+      "learning_rate": 2.981392557022809e-05,
+      "loss": 0.5899,
+      "step": 10101
+    },
+    {
+      "epoch": 12.93056,
+      "grad_norm": 0.9346731305122375,
+      "learning_rate": 2.981192476990797e-05,
+      "loss": 0.5907,
+      "step": 10102
+    },
+    {
+      "epoch": 12.93184,
+      "grad_norm": 0.8982060551643372,
+      "learning_rate": 2.9809923969587837e-05,
+      "loss": 0.5422,
+      "step": 10103
+    },
+    {
+      "epoch": 12.93312,
+      "grad_norm": 0.980532705783844,
+      "learning_rate": 2.980792316926771e-05,
+      "loss": 0.6078,
+      "step": 10104
+    },
+    {
+      "epoch": 12.9344,
+      "grad_norm": 0.9058598875999451,
+      "learning_rate": 2.980592236894758e-05,
+      "loss": 0.5936,
+      "step": 10105
+    },
+    {
+      "epoch": 12.93568,
+      "grad_norm": 0.81413733959198,
+      "learning_rate": 2.9803921568627453e-05,
+      "loss": 0.4857,
+      "step": 10106
+    },
+    {
+      "epoch": 12.93696,
+      "grad_norm": 0.9065173864364624,
+      "learning_rate": 2.9801920768307325e-05,
+      "loss": 0.5294,
+      "step": 10107
+    },
+    {
+      "epoch": 12.93824,
+      "grad_norm": 0.9533013701438904,
+      "learning_rate": 2.9799919967987193e-05,
+      "loss": 0.5727,
+      "step": 10108
+    },
+    {
+      "epoch": 12.93952,
+      "grad_norm": 0.9087998867034912,
+      "learning_rate": 2.9797919167667065e-05,
+      "loss": 0.5334,
+      "step": 10109
+    },
+    {
+      "epoch": 12.9408,
+      "grad_norm": 0.905795156955719,
+      "learning_rate": 2.9795918367346944e-05,
+      "loss": 0.5665,
+      "step": 10110
+    },
+    {
+      "epoch": 12.94208,
+      "grad_norm": 0.9425820708274841,
+      "learning_rate": 2.9793917567026812e-05,
+      "loss": 0.5665,
+      "step": 10111
+    },
+    {
+      "epoch": 12.94336,
+      "grad_norm": 0.8737591505050659,
+      "learning_rate": 2.9791916766706684e-05,
+      "loss": 0.5013,
+      "step": 10112
+    },
+    {
+      "epoch": 12.94464,
+      "grad_norm": 0.9114744067192078,
+      "learning_rate": 2.9789915966386556e-05,
+      "loss": 0.5359,
+      "step": 10113
+    },
+    {
+      "epoch": 12.94592,
+      "grad_norm": 0.9109938144683838,
+      "learning_rate": 2.9787915166066428e-05,
+      "loss": 0.525,
+      "step": 10114
+    },
+    {
+      "epoch": 12.9472,
+      "grad_norm": 1.0331107378005981,
+      "learning_rate": 2.97859143657463e-05,
+      "loss": 0.6145,
+      "step": 10115
+    },
+    {
+      "epoch": 12.94848,
+      "grad_norm": 0.9183046221733093,
+      "learning_rate": 2.9783913565426168e-05,
+      "loss": 0.5607,
+      "step": 10116
+    },
+    {
+      "epoch": 12.94976,
+      "grad_norm": 0.8793911337852478,
+      "learning_rate": 2.9781912765106047e-05,
+      "loss": 0.5091,
+      "step": 10117
+    },
+    {
+      "epoch": 12.95104,
+      "grad_norm": 0.8715007305145264,
+      "learning_rate": 2.977991196478592e-05,
+      "loss": 0.5342,
+      "step": 10118
+    },
+    {
+      "epoch": 12.95232,
+      "grad_norm": 0.9130722284317017,
+      "learning_rate": 2.9777911164465787e-05,
+      "loss": 0.569,
+      "step": 10119
+    },
+    {
+      "epoch": 12.9536,
+      "grad_norm": 0.8778414130210876,
+      "learning_rate": 2.977591036414566e-05,
+      "loss": 0.5715,
+      "step": 10120
+    },
+    {
+      "epoch": 12.95488,
+      "grad_norm": 0.8723252415657043,
+      "learning_rate": 2.977390956382553e-05,
+      "loss": 0.5556,
+      "step": 10121
+    },
+    {
+      "epoch": 12.95616,
+      "grad_norm": 0.9245275259017944,
+      "learning_rate": 2.9771908763505403e-05,
+      "loss": 0.5706,
+      "step": 10122
+    },
+    {
+      "epoch": 12.95744,
+      "grad_norm": 0.9028117656707764,
+      "learning_rate": 2.9769907963185274e-05,
+      "loss": 0.5435,
+      "step": 10123
+    },
+    {
+      "epoch": 12.95872,
+      "grad_norm": 0.8952034115791321,
+      "learning_rate": 2.976790716286515e-05,
+      "loss": 0.5545,
+      "step": 10124
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 0.889771580696106,
+      "learning_rate": 2.976590636254502e-05,
+      "loss": 0.5124,
+      "step": 10125
+    },
+    {
+      "epoch": 12.96128,
+      "grad_norm": 0.9668723940849304,
+      "learning_rate": 2.9763905562224893e-05,
+      "loss": 0.537,
+      "step": 10126
+    },
+    {
+      "epoch": 12.96256,
+      "grad_norm": 0.8818986415863037,
+      "learning_rate": 2.9761904761904762e-05,
+      "loss": 0.5641,
+      "step": 10127
+    },
+    {
+      "epoch": 12.96384,
+      "grad_norm": 0.8750602006912231,
+      "learning_rate": 2.9759903961584634e-05,
+      "loss": 0.5355,
+      "step": 10128
+    },
+    {
+      "epoch": 12.96512,
+      "grad_norm": 0.9368957877159119,
+      "learning_rate": 2.9757903161264506e-05,
+      "loss": 0.5498,
+      "step": 10129
+    },
+    {
+      "epoch": 12.9664,
+      "grad_norm": 0.9112243056297302,
+      "learning_rate": 2.9755902360944377e-05,
+      "loss": 0.5243,
+      "step": 10130
+    },
+    {
+      "epoch": 12.96768,
+      "grad_norm": 0.9702836871147156,
+      "learning_rate": 2.9753901560624253e-05,
+      "loss": 0.5832,
+      "step": 10131
+    },
+    {
+      "epoch": 12.96896,
+      "grad_norm": 0.9459221363067627,
+      "learning_rate": 2.9751900760304125e-05,
+      "loss": 0.5452,
+      "step": 10132
+    },
+    {
+      "epoch": 12.97024,
+      "grad_norm": 0.8926064968109131,
+      "learning_rate": 2.9749899959983996e-05,
+      "loss": 0.5324,
+      "step": 10133
+    },
+    {
+      "epoch": 12.97152,
+      "grad_norm": 0.9436557292938232,
+      "learning_rate": 2.9747899159663868e-05,
+      "loss": 0.5705,
+      "step": 10134
+    },
+    {
+      "epoch": 12.9728,
+      "grad_norm": 0.8792611956596375,
+      "learning_rate": 2.9745898359343737e-05,
+      "loss": 0.5387,
+      "step": 10135
+    },
+    {
+      "epoch": 12.97408,
+      "grad_norm": 0.8789375424385071,
+      "learning_rate": 2.974389755902361e-05,
+      "loss": 0.4897,
+      "step": 10136
+    },
+    {
+      "epoch": 12.97536,
+      "grad_norm": 0.9235910773277283,
+      "learning_rate": 2.974189675870348e-05,
+      "loss": 0.5667,
+      "step": 10137
+    },
+    {
+      "epoch": 12.97664,
+      "grad_norm": 0.9125930070877075,
+      "learning_rate": 2.9739895958383356e-05,
+      "loss": 0.5704,
+      "step": 10138
+    },
+    {
+      "epoch": 12.97792,
+      "grad_norm": 0.9023631811141968,
+      "learning_rate": 2.9737895158063228e-05,
+      "loss": 0.5521,
+      "step": 10139
+    },
+    {
+      "epoch": 12.9792,
+      "grad_norm": 0.9152993559837341,
+      "learning_rate": 2.97358943577431e-05,
+      "loss": 0.4928,
+      "step": 10140
+    },
+    {
+      "epoch": 12.98048,
+      "grad_norm": 0.9630506038665771,
+      "learning_rate": 2.973389355742297e-05,
+      "loss": 0.5829,
+      "step": 10141
+    },
+    {
+      "epoch": 12.98176,
+      "grad_norm": 0.9099013209342957,
+      "learning_rate": 2.9731892757102843e-05,
+      "loss": 0.5566,
+      "step": 10142
+    },
+    {
+      "epoch": 12.98304,
+      "grad_norm": 0.8991549015045166,
+      "learning_rate": 2.972989195678271e-05,
+      "loss": 0.5459,
+      "step": 10143
+    },
+    {
+      "epoch": 12.98432,
+      "grad_norm": 0.9547131657600403,
+      "learning_rate": 2.9727891156462583e-05,
+      "loss": 0.5874,
+      "step": 10144
+    },
+    {
+      "epoch": 12.9856,
+      "grad_norm": 0.9026567339897156,
+      "learning_rate": 2.9725890356142462e-05,
+      "loss": 0.547,
+      "step": 10145
+    },
+    {
+      "epoch": 12.98688,
+      "grad_norm": 0.9058331847190857,
+      "learning_rate": 2.972388955582233e-05,
+      "loss": 0.52,
+      "step": 10146
+    },
+    {
+      "epoch": 12.98816,
+      "grad_norm": 0.9603937864303589,
+      "learning_rate": 2.9721888755502202e-05,
+      "loss": 0.615,
+      "step": 10147
+    },
+    {
+      "epoch": 12.98944,
+      "grad_norm": 0.9125460982322693,
+      "learning_rate": 2.9719887955182074e-05,
+      "loss": 0.5463,
+      "step": 10148
+    },
+    {
+      "epoch": 12.99072,
+      "grad_norm": 0.9278149008750916,
+      "learning_rate": 2.9717887154861946e-05,
+      "loss": 0.5386,
+      "step": 10149
+    },
+    {
+      "epoch": 12.992,
+      "grad_norm": 0.88197922706604,
+      "learning_rate": 2.9715886354541818e-05,
+      "loss": 0.5443,
+      "step": 10150
+    },
+    {
+      "epoch": 12.99328,
+      "grad_norm": 0.9040613174438477,
+      "learning_rate": 2.9713885554221686e-05,
+      "loss": 0.5515,
+      "step": 10151
+    },
+    {
+      "epoch": 12.99456,
+      "grad_norm": 0.9090338945388794,
+      "learning_rate": 2.9711884753901565e-05,
+      "loss": 0.5483,
+      "step": 10152
+    },
+    {
+      "epoch": 12.99584,
+      "grad_norm": 0.9126954674720764,
+      "learning_rate": 2.9709883953581437e-05,
+      "loss": 0.6022,
+      "step": 10153
+    },
+    {
+      "epoch": 12.99712,
+      "grad_norm": 0.9582985043525696,
+      "learning_rate": 2.9707883153261305e-05,
+      "loss": 0.5414,
+      "step": 10154
+    },
+    {
+      "epoch": 12.9984,
+      "grad_norm": 0.9858940839767456,
+      "learning_rate": 2.9705882352941177e-05,
+      "loss": 0.5773,
+      "step": 10155
+    },
+    {
+      "epoch": 12.99968,
+      "grad_norm": 0.9113115072250366,
+      "learning_rate": 2.970388155262105e-05,
+      "loss": 0.5455,
+      "step": 10156
+    },
+    {
+      "epoch": 13.00096,
+      "grad_norm": 1.8516324758529663,
+      "learning_rate": 2.970188075230092e-05,
+      "loss": 0.9295,
+      "step": 10157
+    },
+    {
+      "epoch": 13.00224,
+      "grad_norm": 0.8609241843223572,
+      "learning_rate": 2.9699879951980793e-05,
+      "loss": 0.4725,
+      "step": 10158
+    },
+    {
+      "epoch": 13.00352,
+      "grad_norm": 0.8947498202323914,
+      "learning_rate": 2.9697879151660668e-05,
+      "loss": 0.5303,
+      "step": 10159
+    },
+    {
+      "epoch": 13.0048,
+      "grad_norm": 0.9149647951126099,
+      "learning_rate": 2.969587835134054e-05,
+      "loss": 0.5236,
+      "step": 10160
+    },
+    {
+      "epoch": 13.00608,
+      "grad_norm": 0.9216925501823425,
+      "learning_rate": 2.9693877551020412e-05,
+      "loss": 0.5461,
+      "step": 10161
+    },
+    {
+      "epoch": 13.00736,
+      "grad_norm": 0.9156137108802795,
+      "learning_rate": 2.969187675070028e-05,
+      "loss": 0.5692,
+      "step": 10162
+    },
+    {
+      "epoch": 13.00864,
+      "grad_norm": 0.9287101030349731,
+      "learning_rate": 2.9689875950380152e-05,
+      "loss": 0.5375,
+      "step": 10163
+    },
+    {
+      "epoch": 13.00992,
+      "grad_norm": 0.9227731227874756,
+      "learning_rate": 2.9687875150060024e-05,
+      "loss": 0.5584,
+      "step": 10164
+    },
+    {
+      "epoch": 13.0112,
+      "grad_norm": 0.9401906132698059,
+      "learning_rate": 2.9685874349739896e-05,
+      "loss": 0.5354,
+      "step": 10165
+    },
+    {
+      "epoch": 13.01248,
+      "grad_norm": 0.9380677938461304,
+      "learning_rate": 2.968387354941977e-05,
+      "loss": 0.5367,
+      "step": 10166
+    },
+    {
+      "epoch": 13.01376,
+      "grad_norm": 0.9107190370559692,
+      "learning_rate": 2.9681872749099643e-05,
+      "loss": 0.515,
+      "step": 10167
+    },
+    {
+      "epoch": 13.01504,
+      "grad_norm": 0.9699810147285461,
+      "learning_rate": 2.9679871948779515e-05,
+      "loss": 0.5413,
+      "step": 10168
+    },
+    {
+      "epoch": 13.01632,
+      "grad_norm": 0.9228163361549377,
+      "learning_rate": 2.9677871148459387e-05,
+      "loss": 0.5589,
+      "step": 10169
+    },
+    {
+      "epoch": 13.0176,
+      "grad_norm": 0.8890709280967712,
+      "learning_rate": 2.9675870348139255e-05,
+      "loss": 0.5434,
+      "step": 10170
+    },
+    {
+      "epoch": 13.01888,
+      "grad_norm": 0.8726478815078735,
+      "learning_rate": 2.9673869547819127e-05,
+      "loss": 0.5035,
+      "step": 10171
+    },
+    {
+      "epoch": 13.02016,
+      "grad_norm": 0.9581865072250366,
+      "learning_rate": 2.9671868747499e-05,
+      "loss": 0.5897,
+      "step": 10172
+    },
+    {
+      "epoch": 13.02144,
+      "grad_norm": 0.9103356599807739,
+      "learning_rate": 2.9669867947178874e-05,
+      "loss": 0.5305,
+      "step": 10173
+    },
+    {
+      "epoch": 13.02272,
+      "grad_norm": 0.9056451916694641,
+      "learning_rate": 2.9667867146858746e-05,
+      "loss": 0.5249,
+      "step": 10174
+    },
+    {
+      "epoch": 13.024,
+      "grad_norm": 0.9359303712844849,
+      "learning_rate": 2.9665866346538618e-05,
+      "loss": 0.552,
+      "step": 10175
+    },
+    {
+      "epoch": 13.02528,
+      "grad_norm": 0.9274691343307495,
+      "learning_rate": 2.966386554621849e-05,
+      "loss": 0.5194,
+      "step": 10176
+    },
+    {
+      "epoch": 13.02656,
+      "grad_norm": 0.8577048182487488,
+      "learning_rate": 2.966186474589836e-05,
+      "loss": 0.488,
+      "step": 10177
+    },
+    {
+      "epoch": 13.02784,
+      "grad_norm": 0.8899839520454407,
+      "learning_rate": 2.965986394557823e-05,
+      "loss": 0.5371,
+      "step": 10178
+    },
+    {
+      "epoch": 13.02912,
+      "grad_norm": 0.8708844184875488,
+      "learning_rate": 2.9657863145258102e-05,
+      "loss": 0.5118,
+      "step": 10179
+    },
+    {
+      "epoch": 13.0304,
+      "grad_norm": 0.9025720357894897,
+      "learning_rate": 2.965586234493798e-05,
+      "loss": 0.5332,
+      "step": 10180
+    },
+    {
+      "epoch": 13.03168,
+      "grad_norm": 0.9257388710975647,
+      "learning_rate": 2.965386154461785e-05,
+      "loss": 0.6321,
+      "step": 10181
+    },
+    {
+      "epoch": 13.03296,
+      "grad_norm": 0.8940935730934143,
+      "learning_rate": 2.965186074429772e-05,
+      "loss": 0.5406,
+      "step": 10182
+    },
+    {
+      "epoch": 13.03424,
+      "grad_norm": 0.9493997693061829,
+      "learning_rate": 2.9649859943977593e-05,
+      "loss": 0.5364,
+      "step": 10183
+    },
+    {
+      "epoch": 13.03552,
+      "grad_norm": 0.8869070410728455,
+      "learning_rate": 2.9647859143657464e-05,
+      "loss": 0.5076,
+      "step": 10184
+    },
+    {
+      "epoch": 13.0368,
+      "grad_norm": 0.9130921363830566,
+      "learning_rate": 2.9645858343337336e-05,
+      "loss": 0.553,
+      "step": 10185
+    },
+    {
+      "epoch": 13.03808,
+      "grad_norm": 0.8858230113983154,
+      "learning_rate": 2.9643857543017205e-05,
+      "loss": 0.471,
+      "step": 10186
+    },
+    {
+      "epoch": 13.03936,
+      "grad_norm": 0.8924762010574341,
+      "learning_rate": 2.9641856742697083e-05,
+      "loss": 0.5224,
+      "step": 10187
+    },
+    {
+      "epoch": 13.04064,
+      "grad_norm": 0.9254968762397766,
+      "learning_rate": 2.9639855942376955e-05,
+      "loss": 0.5203,
+      "step": 10188
+    },
+    {
+      "epoch": 13.04192,
+      "grad_norm": 0.9559713006019592,
+      "learning_rate": 2.9637855142056824e-05,
+      "loss": 0.5552,
+      "step": 10189
+    },
+    {
+      "epoch": 13.0432,
+      "grad_norm": 0.9039125442504883,
+      "learning_rate": 2.9635854341736696e-05,
+      "loss": 0.5119,
+      "step": 10190
+    },
+    {
+      "epoch": 13.04448,
+      "grad_norm": 0.9470018744468689,
+      "learning_rate": 2.9633853541416567e-05,
+      "loss": 0.5233,
+      "step": 10191
+    },
+    {
+      "epoch": 13.04576,
+      "grad_norm": 0.991151750087738,
+      "learning_rate": 2.963185274109644e-05,
+      "loss": 0.5534,
+      "step": 10192
+    },
+    {
+      "epoch": 13.04704,
+      "grad_norm": 0.9508792161941528,
+      "learning_rate": 2.962985194077631e-05,
+      "loss": 0.4863,
+      "step": 10193
+    },
+    {
+      "epoch": 13.04832,
+      "grad_norm": 0.9612112045288086,
+      "learning_rate": 2.9627851140456186e-05,
+      "loss": 0.5613,
+      "step": 10194
+    },
+    {
+      "epoch": 13.0496,
+      "grad_norm": 0.9475077986717224,
+      "learning_rate": 2.9625850340136058e-05,
+      "loss": 0.5695,
+      "step": 10195
+    },
+    {
+      "epoch": 13.05088,
+      "grad_norm": 0.9162499308586121,
+      "learning_rate": 2.962384953981593e-05,
+      "loss": 0.5628,
+      "step": 10196
+    },
+    {
+      "epoch": 13.05216,
+      "grad_norm": 0.9611510038375854,
+      "learning_rate": 2.96218487394958e-05,
+      "loss": 0.5155,
+      "step": 10197
+    },
+    {
+      "epoch": 13.05344,
+      "grad_norm": 0.9086456298828125,
+      "learning_rate": 2.961984793917567e-05,
+      "loss": 0.5197,
+      "step": 10198
+    },
+    {
+      "epoch": 13.05472,
+      "grad_norm": 0.9529975652694702,
+      "learning_rate": 2.9617847138855542e-05,
+      "loss": 0.5468,
+      "step": 10199
+    },
+    {
+      "epoch": 13.056,
+      "grad_norm": 0.9286540746688843,
+      "learning_rate": 2.9615846338535414e-05,
+      "loss": 0.5342,
+      "step": 10200
+    },
+    {
+      "epoch": 13.05728,
+      "grad_norm": 0.9417791366577148,
+      "learning_rate": 2.961384553821529e-05,
+      "loss": 0.5515,
+      "step": 10201
+    },
+    {
+      "epoch": 13.05856,
+      "grad_norm": 0.9122342467308044,
+      "learning_rate": 2.961184473789516e-05,
+      "loss": 0.4782,
+      "step": 10202
+    },
+    {
+      "epoch": 13.05984,
+      "grad_norm": 0.9562501311302185,
+      "learning_rate": 2.9609843937575033e-05,
+      "loss": 0.5577,
+      "step": 10203
+    },
+    {
+      "epoch": 13.06112,
+      "grad_norm": 0.9384812116622925,
+      "learning_rate": 2.9607843137254905e-05,
+      "loss": 0.5087,
+      "step": 10204
+    },
+    {
+      "epoch": 13.0624,
+      "grad_norm": 0.9272902607917786,
+      "learning_rate": 2.9605842336934773e-05,
+      "loss": 0.4929,
+      "step": 10205
+    },
+    {
+      "epoch": 13.06368,
+      "grad_norm": 0.9776669144630432,
+      "learning_rate": 2.9603841536614645e-05,
+      "loss": 0.5292,
+      "step": 10206
+    },
+    {
+      "epoch": 13.06496,
+      "grad_norm": 0.9316274523735046,
+      "learning_rate": 2.9601840736294517e-05,
+      "loss": 0.5501,
+      "step": 10207
+    },
+    {
+      "epoch": 13.06624,
+      "grad_norm": 0.9690887928009033,
+      "learning_rate": 2.9599839935974392e-05,
+      "loss": 0.5342,
+      "step": 10208
+    },
+    {
+      "epoch": 13.06752,
+      "grad_norm": 0.8935588002204895,
+      "learning_rate": 2.9597839135654264e-05,
+      "loss": 0.5046,
+      "step": 10209
+    },
+    {
+      "epoch": 13.0688,
+      "grad_norm": 0.9328128099441528,
+      "learning_rate": 2.9595838335334136e-05,
+      "loss": 0.5601,
+      "step": 10210
+    },
+    {
+      "epoch": 13.07008,
+      "grad_norm": 0.9481015801429749,
+      "learning_rate": 2.9593837535014008e-05,
+      "loss": 0.5336,
+      "step": 10211
+    },
+    {
+      "epoch": 13.07136,
+      "grad_norm": 0.9740045070648193,
+      "learning_rate": 2.959183673469388e-05,
+      "loss": 0.5383,
+      "step": 10212
+    },
+    {
+      "epoch": 13.07264,
+      "grad_norm": 0.9591369032859802,
+      "learning_rate": 2.958983593437375e-05,
+      "loss": 0.5688,
+      "step": 10213
+    },
+    {
+      "epoch": 13.07392,
+      "grad_norm": 0.8902487754821777,
+      "learning_rate": 2.958783513405362e-05,
+      "loss": 0.5243,
+      "step": 10214
+    },
+    {
+      "epoch": 13.0752,
+      "grad_norm": 0.8981714844703674,
+      "learning_rate": 2.95858343337335e-05,
+      "loss": 0.504,
+      "step": 10215
+    },
+    {
+      "epoch": 13.07648,
+      "grad_norm": 0.857538640499115,
+      "learning_rate": 2.9583833533413367e-05,
+      "loss": 0.4907,
+      "step": 10216
+    },
+    {
+      "epoch": 13.07776,
+      "grad_norm": 0.9131780862808228,
+      "learning_rate": 2.958183273309324e-05,
+      "loss": 0.5018,
+      "step": 10217
+    },
+    {
+      "epoch": 13.079039999999999,
+      "grad_norm": 0.9661425352096558,
+      "learning_rate": 2.957983193277311e-05,
+      "loss": 0.5639,
+      "step": 10218
+    },
+    {
+      "epoch": 13.08032,
+      "grad_norm": 0.9604724645614624,
+      "learning_rate": 2.9577831132452983e-05,
+      "loss": 0.5433,
+      "step": 10219
+    },
+    {
+      "epoch": 13.0816,
+      "grad_norm": 0.9069200158119202,
+      "learning_rate": 2.9575830332132855e-05,
+      "loss": 0.5409,
+      "step": 10220
+    },
+    {
+      "epoch": 13.08288,
+      "grad_norm": 0.9007838368415833,
+      "learning_rate": 2.9573829531812723e-05,
+      "loss": 0.5377,
+      "step": 10221
+    },
+    {
+      "epoch": 13.08416,
+      "grad_norm": 0.9173440337181091,
+      "learning_rate": 2.9571828731492595e-05,
+      "loss": 0.5275,
+      "step": 10222
+    },
+    {
+      "epoch": 13.08544,
+      "grad_norm": 0.936603844165802,
+      "learning_rate": 2.9569827931172474e-05,
+      "loss": 0.5257,
+      "step": 10223
+    },
+    {
+      "epoch": 13.08672,
+      "grad_norm": 0.9958267211914062,
+      "learning_rate": 2.9567827130852342e-05,
+      "loss": 0.5632,
+      "step": 10224
+    },
+    {
+      "epoch": 13.088,
+      "grad_norm": 0.9502796530723572,
+      "learning_rate": 2.9565826330532214e-05,
+      "loss": 0.5396,
+      "step": 10225
+    },
+    {
+      "epoch": 13.08928,
+      "grad_norm": 0.8708899617195129,
+      "learning_rate": 2.9563825530212086e-05,
+      "loss": 0.4855,
+      "step": 10226
+    },
+    {
+      "epoch": 13.09056,
+      "grad_norm": 0.9281712770462036,
+      "learning_rate": 2.9561824729891958e-05,
+      "loss": 0.4983,
+      "step": 10227
+    },
+    {
+      "epoch": 13.09184,
+      "grad_norm": 0.943922221660614,
+      "learning_rate": 2.955982392957183e-05,
+      "loss": 0.5545,
+      "step": 10228
+    },
+    {
+      "epoch": 13.09312,
+      "grad_norm": 0.9381184577941895,
+      "learning_rate": 2.9557823129251698e-05,
+      "loss": 0.5406,
+      "step": 10229
+    },
+    {
+      "epoch": 13.0944,
+      "grad_norm": 0.9126680493354797,
+      "learning_rate": 2.9555822328931577e-05,
+      "loss": 0.5364,
+      "step": 10230
+    },
+    {
+      "epoch": 13.09568,
+      "grad_norm": 0.9556214809417725,
+      "learning_rate": 2.955382152861145e-05,
+      "loss": 0.5298,
+      "step": 10231
+    },
+    {
+      "epoch": 13.09696,
+      "grad_norm": 0.9941635727882385,
+      "learning_rate": 2.9551820728291317e-05,
+      "loss": 0.5511,
+      "step": 10232
+    },
+    {
+      "epoch": 13.09824,
+      "grad_norm": 0.9520924687385559,
+      "learning_rate": 2.954981992797119e-05,
+      "loss": 0.5176,
+      "step": 10233
+    },
+    {
+      "epoch": 13.09952,
+      "grad_norm": 1.0025166273117065,
+      "learning_rate": 2.954781912765106e-05,
+      "loss": 0.5434,
+      "step": 10234
+    },
+    {
+      "epoch": 13.1008,
+      "grad_norm": 0.9303215146064758,
+      "learning_rate": 2.9545818327330933e-05,
+      "loss": 0.5081,
+      "step": 10235
+    },
+    {
+      "epoch": 13.10208,
+      "grad_norm": 0.9125444293022156,
+      "learning_rate": 2.9543817527010804e-05,
+      "loss": 0.5134,
+      "step": 10236
+    },
+    {
+      "epoch": 13.10336,
+      "grad_norm": 0.9002273678779602,
+      "learning_rate": 2.954181672669068e-05,
+      "loss": 0.4885,
+      "step": 10237
+    },
+    {
+      "epoch": 13.10464,
+      "grad_norm": 0.9505068063735962,
+      "learning_rate": 2.953981592637055e-05,
+      "loss": 0.5645,
+      "step": 10238
+    },
+    {
+      "epoch": 13.10592,
+      "grad_norm": 0.9050679206848145,
+      "learning_rate": 2.9537815126050423e-05,
+      "loss": 0.5322,
+      "step": 10239
+    },
+    {
+      "epoch": 13.1072,
+      "grad_norm": 0.9206121563911438,
+      "learning_rate": 2.9535814325730292e-05,
+      "loss": 0.5077,
+      "step": 10240
+    },
+    {
+      "epoch": 13.10848,
+      "grad_norm": 0.9037232398986816,
+      "learning_rate": 2.9533813525410164e-05,
+      "loss": 0.5202,
+      "step": 10241
+    },
+    {
+      "epoch": 13.10976,
+      "grad_norm": 0.9261408448219299,
+      "learning_rate": 2.9531812725090036e-05,
+      "loss": 0.5514,
+      "step": 10242
+    },
+    {
+      "epoch": 13.11104,
+      "grad_norm": 0.9304765462875366,
+      "learning_rate": 2.9529811924769907e-05,
+      "loss": 0.5671,
+      "step": 10243
+    },
+    {
+      "epoch": 13.11232,
+      "grad_norm": 0.8699920773506165,
+      "learning_rate": 2.9527811124449783e-05,
+      "loss": 0.5076,
+      "step": 10244
+    },
+    {
+      "epoch": 13.1136,
+      "grad_norm": 0.9810715913772583,
+      "learning_rate": 2.9525810324129655e-05,
+      "loss": 0.6187,
+      "step": 10245
+    },
+    {
+      "epoch": 13.11488,
+      "grad_norm": 0.9661498665809631,
+      "learning_rate": 2.9523809523809526e-05,
+      "loss": 0.575,
+      "step": 10246
+    },
+    {
+      "epoch": 13.11616,
+      "grad_norm": 0.9530900120735168,
+      "learning_rate": 2.9521808723489398e-05,
+      "loss": 0.5207,
+      "step": 10247
+    },
+    {
+      "epoch": 13.11744,
+      "grad_norm": 0.8980867266654968,
+      "learning_rate": 2.9519807923169267e-05,
+      "loss": 0.48,
+      "step": 10248
+    },
+    {
+      "epoch": 13.11872,
+      "grad_norm": 1.0098669528961182,
+      "learning_rate": 2.951780712284914e-05,
+      "loss": 0.5863,
+      "step": 10249
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 0.8966968655586243,
+      "learning_rate": 2.951580632252901e-05,
+      "loss": 0.5161,
+      "step": 10250
+    },
+    {
+      "epoch": 13.12128,
+      "grad_norm": 0.8702352046966553,
+      "learning_rate": 2.9513805522208886e-05,
+      "loss": 0.4706,
+      "step": 10251
+    },
+    {
+      "epoch": 13.12256,
+      "grad_norm": 0.923116147518158,
+      "learning_rate": 2.9511804721888758e-05,
+      "loss": 0.5097,
+      "step": 10252
+    },
+    {
+      "epoch": 13.12384,
+      "grad_norm": 0.9864121079444885,
+      "learning_rate": 2.950980392156863e-05,
+      "loss": 0.5453,
+      "step": 10253
+    },
+    {
+      "epoch": 13.12512,
+      "grad_norm": 0.9209245443344116,
+      "learning_rate": 2.95078031212485e-05,
+      "loss": 0.5178,
+      "step": 10254
+    },
+    {
+      "epoch": 13.1264,
+      "grad_norm": 0.9390770792961121,
+      "learning_rate": 2.9505802320928373e-05,
+      "loss": 0.5423,
+      "step": 10255
+    },
+    {
+      "epoch": 13.12768,
+      "grad_norm": 0.9811953902244568,
+      "learning_rate": 2.950380152060824e-05,
+      "loss": 0.5938,
+      "step": 10256
+    },
+    {
+      "epoch": 13.12896,
+      "grad_norm": 0.9948422908782959,
+      "learning_rate": 2.9501800720288113e-05,
+      "loss": 0.6015,
+      "step": 10257
+    },
+    {
+      "epoch": 13.13024,
+      "grad_norm": 0.9334530830383301,
+      "learning_rate": 2.9499799919967992e-05,
+      "loss": 0.5042,
+      "step": 10258
+    },
+    {
+      "epoch": 13.13152,
+      "grad_norm": 0.887524425983429,
+      "learning_rate": 2.949779911964786e-05,
+      "loss": 0.5115,
+      "step": 10259
+    },
+    {
+      "epoch": 13.1328,
+      "grad_norm": 0.9282869696617126,
+      "learning_rate": 2.9495798319327732e-05,
+      "loss": 0.5348,
+      "step": 10260
+    },
+    {
+      "epoch": 13.13408,
+      "grad_norm": 0.8602845072746277,
+      "learning_rate": 2.9493797519007604e-05,
+      "loss": 0.477,
+      "step": 10261
+    },
+    {
+      "epoch": 13.13536,
+      "grad_norm": 0.9811587333679199,
+      "learning_rate": 2.9491796718687476e-05,
+      "loss": 0.6148,
+      "step": 10262
+    },
+    {
+      "epoch": 13.13664,
+      "grad_norm": 0.938101589679718,
+      "learning_rate": 2.9489795918367348e-05,
+      "loss": 0.5469,
+      "step": 10263
+    },
+    {
+      "epoch": 13.13792,
+      "grad_norm": 0.938423216342926,
+      "learning_rate": 2.9487795118047216e-05,
+      "loss": 0.5415,
+      "step": 10264
+    },
+    {
+      "epoch": 13.1392,
+      "grad_norm": 0.9609931111335754,
+      "learning_rate": 2.9485794317727095e-05,
+      "loss": 0.5779,
+      "step": 10265
+    },
+    {
+      "epoch": 13.14048,
+      "grad_norm": 0.9402124881744385,
+      "learning_rate": 2.9483793517406967e-05,
+      "loss": 0.5688,
+      "step": 10266
+    },
+    {
+      "epoch": 13.14176,
+      "grad_norm": 0.9588215351104736,
+      "learning_rate": 2.9481792717086835e-05,
+      "loss": 0.5559,
+      "step": 10267
+    },
+    {
+      "epoch": 13.14304,
+      "grad_norm": 0.9343993663787842,
+      "learning_rate": 2.9479791916766707e-05,
+      "loss": 0.5233,
+      "step": 10268
+    },
+    {
+      "epoch": 13.14432,
+      "grad_norm": 0.8951725363731384,
+      "learning_rate": 2.947779111644658e-05,
+      "loss": 0.4691,
+      "step": 10269
+    },
+    {
+      "epoch": 13.1456,
+      "grad_norm": 0.9678958654403687,
+      "learning_rate": 2.947579031612645e-05,
+      "loss": 0.5648,
+      "step": 10270
+    },
+    {
+      "epoch": 13.14688,
+      "grad_norm": 0.973098874092102,
+      "learning_rate": 2.9473789515806323e-05,
+      "loss": 0.527,
+      "step": 10271
+    },
+    {
+      "epoch": 13.14816,
+      "grad_norm": 0.9279984831809998,
+      "learning_rate": 2.9471788715486198e-05,
+      "loss": 0.4758,
+      "step": 10272
+    },
+    {
+      "epoch": 13.14944,
+      "grad_norm": 0.9784411787986755,
+      "learning_rate": 2.946978791516607e-05,
+      "loss": 0.5491,
+      "step": 10273
+    },
+    {
+      "epoch": 13.15072,
+      "grad_norm": 0.9150949120521545,
+      "learning_rate": 2.9467787114845942e-05,
+      "loss": 0.5418,
+      "step": 10274
+    },
+    {
+      "epoch": 13.152,
+      "grad_norm": 0.8972795605659485,
+      "learning_rate": 2.946578631452581e-05,
+      "loss": 0.5065,
+      "step": 10275
+    },
+    {
+      "epoch": 13.15328,
+      "grad_norm": 0.8859349489212036,
+      "learning_rate": 2.9463785514205682e-05,
+      "loss": 0.5211,
+      "step": 10276
+    },
+    {
+      "epoch": 13.15456,
+      "grad_norm": 0.9786564707756042,
+      "learning_rate": 2.9461784713885554e-05,
+      "loss": 0.5468,
+      "step": 10277
+    },
+    {
+      "epoch": 13.15584,
+      "grad_norm": 0.9281904101371765,
+      "learning_rate": 2.9459783913565426e-05,
+      "loss": 0.5303,
+      "step": 10278
+    },
+    {
+      "epoch": 13.15712,
+      "grad_norm": 0.8598402142524719,
+      "learning_rate": 2.94577831132453e-05,
+      "loss": 0.4675,
+      "step": 10279
+    },
+    {
+      "epoch": 13.1584,
+      "grad_norm": 0.881607174873352,
+      "learning_rate": 2.9455782312925173e-05,
+      "loss": 0.4969,
+      "step": 10280
+    },
+    {
+      "epoch": 13.15968,
+      "grad_norm": 0.8797462582588196,
+      "learning_rate": 2.9453781512605045e-05,
+      "loss": 0.5216,
+      "step": 10281
+    },
+    {
+      "epoch": 13.16096,
+      "grad_norm": 0.9314226508140564,
+      "learning_rate": 2.9451780712284917e-05,
+      "loss": 0.5117,
+      "step": 10282
+    },
+    {
+      "epoch": 13.16224,
+      "grad_norm": 0.9078558683395386,
+      "learning_rate": 2.9449779911964785e-05,
+      "loss": 0.5344,
+      "step": 10283
+    },
+    {
+      "epoch": 13.16352,
+      "grad_norm": 0.972720742225647,
+      "learning_rate": 2.9447779111644657e-05,
+      "loss": 0.5651,
+      "step": 10284
+    },
+    {
+      "epoch": 13.1648,
+      "grad_norm": 0.8826903104782104,
+      "learning_rate": 2.944577831132453e-05,
+      "loss": 0.4932,
+      "step": 10285
+    },
+    {
+      "epoch": 13.166080000000001,
+      "grad_norm": 0.9253196120262146,
+      "learning_rate": 2.9443777511004404e-05,
+      "loss": 0.5725,
+      "step": 10286
+    },
+    {
+      "epoch": 13.16736,
+      "grad_norm": 0.9346468448638916,
+      "learning_rate": 2.9441776710684276e-05,
+      "loss": 0.5665,
+      "step": 10287
+    },
+    {
+      "epoch": 13.16864,
+      "grad_norm": 0.86590176820755,
+      "learning_rate": 2.9439775910364148e-05,
+      "loss": 0.4922,
+      "step": 10288
+    },
+    {
+      "epoch": 13.16992,
+      "grad_norm": 0.8929559588432312,
+      "learning_rate": 2.943777511004402e-05,
+      "loss": 0.5051,
+      "step": 10289
+    },
+    {
+      "epoch": 13.1712,
+      "grad_norm": 0.9785338044166565,
+      "learning_rate": 2.943577430972389e-05,
+      "loss": 0.565,
+      "step": 10290
+    },
+    {
+      "epoch": 13.17248,
+      "grad_norm": 0.9786810278892517,
+      "learning_rate": 2.943377350940376e-05,
+      "loss": 0.5285,
+      "step": 10291
+    },
+    {
+      "epoch": 13.17376,
+      "grad_norm": 1.0224378108978271,
+      "learning_rate": 2.9431772709083632e-05,
+      "loss": 0.6223,
+      "step": 10292
+    },
+    {
+      "epoch": 13.17504,
+      "grad_norm": 0.8732693195343018,
+      "learning_rate": 2.942977190876351e-05,
+      "loss": 0.4957,
+      "step": 10293
+    },
+    {
+      "epoch": 13.17632,
+      "grad_norm": 0.8983555436134338,
+      "learning_rate": 2.942777110844338e-05,
+      "loss": 0.5085,
+      "step": 10294
+    },
+    {
+      "epoch": 13.1776,
+      "grad_norm": 0.9694886803627014,
+      "learning_rate": 2.942577030812325e-05,
+      "loss": 0.534,
+      "step": 10295
+    },
+    {
+      "epoch": 13.17888,
+      "grad_norm": 0.9326276183128357,
+      "learning_rate": 2.9423769507803123e-05,
+      "loss": 0.5244,
+      "step": 10296
+    },
+    {
+      "epoch": 13.18016,
+      "grad_norm": 0.9629067778587341,
+      "learning_rate": 2.9421768707482994e-05,
+      "loss": 0.5082,
+      "step": 10297
+    },
+    {
+      "epoch": 13.18144,
+      "grad_norm": 0.8940865397453308,
+      "learning_rate": 2.9419767907162866e-05,
+      "loss": 0.4947,
+      "step": 10298
+    },
+    {
+      "epoch": 13.18272,
+      "grad_norm": 0.9360054135322571,
+      "learning_rate": 2.9417767106842735e-05,
+      "loss": 0.5556,
+      "step": 10299
+    },
+    {
+      "epoch": 13.184,
+      "grad_norm": 0.8930971026420593,
+      "learning_rate": 2.9415766306522613e-05,
+      "loss": 0.4866,
+      "step": 10300
+    },
+    {
+      "epoch": 13.18528,
+      "grad_norm": 0.9753515720367432,
+      "learning_rate": 2.9413765506202485e-05,
+      "loss": 0.4929,
+      "step": 10301
+    },
+    {
+      "epoch": 13.18656,
+      "grad_norm": 0.8833822011947632,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.5081,
+      "step": 10302
+    },
+    {
+      "epoch": 13.18784,
+      "grad_norm": 0.9951672554016113,
+      "learning_rate": 2.9409763905562226e-05,
+      "loss": 0.5509,
+      "step": 10303
+    },
+    {
+      "epoch": 13.18912,
+      "grad_norm": 0.9447710514068604,
+      "learning_rate": 2.9407763105242097e-05,
+      "loss": 0.5132,
+      "step": 10304
+    },
+    {
+      "epoch": 13.1904,
+      "grad_norm": 0.960518479347229,
+      "learning_rate": 2.940576230492197e-05,
+      "loss": 0.5609,
+      "step": 10305
+    },
+    {
+      "epoch": 13.19168,
+      "grad_norm": 0.9514713883399963,
+      "learning_rate": 2.940376150460184e-05,
+      "loss": 0.5378,
+      "step": 10306
+    },
+    {
+      "epoch": 13.19296,
+      "grad_norm": 0.9678971767425537,
+      "learning_rate": 2.9401760704281716e-05,
+      "loss": 0.5464,
+      "step": 10307
+    },
+    {
+      "epoch": 13.19424,
+      "grad_norm": 0.9750454425811768,
+      "learning_rate": 2.9399759903961588e-05,
+      "loss": 0.5489,
+      "step": 10308
+    },
+    {
+      "epoch": 13.19552,
+      "grad_norm": 0.9416602849960327,
+      "learning_rate": 2.939775910364146e-05,
+      "loss": 0.5295,
+      "step": 10309
+    },
+    {
+      "epoch": 13.1968,
+      "grad_norm": 0.9217522144317627,
+      "learning_rate": 2.939575830332133e-05,
+      "loss": 0.562,
+      "step": 10310
+    },
+    {
+      "epoch": 13.19808,
+      "grad_norm": 0.8989900946617126,
+      "learning_rate": 2.93937575030012e-05,
+      "loss": 0.5067,
+      "step": 10311
+    },
+    {
+      "epoch": 13.19936,
+      "grad_norm": 0.9681139588356018,
+      "learning_rate": 2.9391756702681072e-05,
+      "loss": 0.5817,
+      "step": 10312
+    },
+    {
+      "epoch": 13.20064,
+      "grad_norm": 0.9325006604194641,
+      "learning_rate": 2.9389755902360944e-05,
+      "loss": 0.5379,
+      "step": 10313
+    },
+    {
+      "epoch": 13.20192,
+      "grad_norm": 0.9002286195755005,
+      "learning_rate": 2.938775510204082e-05,
+      "loss": 0.5274,
+      "step": 10314
+    },
+    {
+      "epoch": 13.2032,
+      "grad_norm": 0.9191522598266602,
+      "learning_rate": 2.938575430172069e-05,
+      "loss": 0.5265,
+      "step": 10315
+    },
+    {
+      "epoch": 13.20448,
+      "grad_norm": 0.9868195056915283,
+      "learning_rate": 2.9383753501400563e-05,
+      "loss": 0.5672,
+      "step": 10316
+    },
+    {
+      "epoch": 13.20576,
+      "grad_norm": 0.935142457485199,
+      "learning_rate": 2.9381752701080435e-05,
+      "loss": 0.559,
+      "step": 10317
+    },
+    {
+      "epoch": 13.20704,
+      "grad_norm": 0.9310742616653442,
+      "learning_rate": 2.9379751900760303e-05,
+      "loss": 0.5183,
+      "step": 10318
+    },
+    {
+      "epoch": 13.20832,
+      "grad_norm": 0.893144428730011,
+      "learning_rate": 2.9377751100440175e-05,
+      "loss": 0.5092,
+      "step": 10319
+    },
+    {
+      "epoch": 13.2096,
+      "grad_norm": 0.9234650135040283,
+      "learning_rate": 2.9375750300120047e-05,
+      "loss": 0.5185,
+      "step": 10320
+    },
+    {
+      "epoch": 13.21088,
+      "grad_norm": 0.868157684803009,
+      "learning_rate": 2.9373749499799922e-05,
+      "loss": 0.4812,
+      "step": 10321
+    },
+    {
+      "epoch": 13.21216,
+      "grad_norm": 0.900688886642456,
+      "learning_rate": 2.9371748699479794e-05,
+      "loss": 0.5181,
+      "step": 10322
+    },
+    {
+      "epoch": 13.21344,
+      "grad_norm": 0.9493388533592224,
+      "learning_rate": 2.9369747899159666e-05,
+      "loss": 0.5613,
+      "step": 10323
+    },
+    {
+      "epoch": 13.21472,
+      "grad_norm": 0.9593266844749451,
+      "learning_rate": 2.9367747098839538e-05,
+      "loss": 0.5462,
+      "step": 10324
+    },
+    {
+      "epoch": 13.216,
+      "grad_norm": 1.0145825147628784,
+      "learning_rate": 2.936574629851941e-05,
+      "loss": 0.5891,
+      "step": 10325
+    },
+    {
+      "epoch": 13.21728,
+      "grad_norm": 0.9658340811729431,
+      "learning_rate": 2.936374549819928e-05,
+      "loss": 0.533,
+      "step": 10326
+    },
+    {
+      "epoch": 13.21856,
+      "grad_norm": 0.8957772850990295,
+      "learning_rate": 2.936174469787915e-05,
+      "loss": 0.5246,
+      "step": 10327
+    },
+    {
+      "epoch": 13.21984,
+      "grad_norm": 0.9555619955062866,
+      "learning_rate": 2.935974389755903e-05,
+      "loss": 0.5181,
+      "step": 10328
+    },
+    {
+      "epoch": 13.22112,
+      "grad_norm": 0.9531823396682739,
+      "learning_rate": 2.9357743097238897e-05,
+      "loss": 0.5459,
+      "step": 10329
+    },
+    {
+      "epoch": 13.2224,
+      "grad_norm": 0.8920832276344299,
+      "learning_rate": 2.935574229691877e-05,
+      "loss": 0.5796,
+      "step": 10330
+    },
+    {
+      "epoch": 13.22368,
+      "grad_norm": 0.9434327483177185,
+      "learning_rate": 2.935374149659864e-05,
+      "loss": 0.5694,
+      "step": 10331
+    },
+    {
+      "epoch": 13.22496,
+      "grad_norm": 0.9526799321174622,
+      "learning_rate": 2.9351740696278513e-05,
+      "loss": 0.5646,
+      "step": 10332
+    },
+    {
+      "epoch": 13.22624,
+      "grad_norm": 1.0042227506637573,
+      "learning_rate": 2.9349739895958385e-05,
+      "loss": 0.5437,
+      "step": 10333
+    },
+    {
+      "epoch": 13.22752,
+      "grad_norm": 0.974702775478363,
+      "learning_rate": 2.9347739095638253e-05,
+      "loss": 0.5569,
+      "step": 10334
+    },
+    {
+      "epoch": 13.2288,
+      "grad_norm": 0.9160645604133606,
+      "learning_rate": 2.9345738295318125e-05,
+      "loss": 0.5174,
+      "step": 10335
+    },
+    {
+      "epoch": 13.23008,
+      "grad_norm": 0.9802337884902954,
+      "learning_rate": 2.9343737494998004e-05,
+      "loss": 0.578,
+      "step": 10336
+    },
+    {
+      "epoch": 13.23136,
+      "grad_norm": 0.9759039878845215,
+      "learning_rate": 2.9341736694677872e-05,
+      "loss": 0.5744,
+      "step": 10337
+    },
+    {
+      "epoch": 13.23264,
+      "grad_norm": 0.9348410367965698,
+      "learning_rate": 2.9339735894357744e-05,
+      "loss": 0.5304,
+      "step": 10338
+    },
+    {
+      "epoch": 13.23392,
+      "grad_norm": 0.9038177728652954,
+      "learning_rate": 2.9337735094037616e-05,
+      "loss": 0.5301,
+      "step": 10339
+    },
+    {
+      "epoch": 13.2352,
+      "grad_norm": 0.880155086517334,
+      "learning_rate": 2.9335734293717488e-05,
+      "loss": 0.496,
+      "step": 10340
+    },
+    {
+      "epoch": 13.23648,
+      "grad_norm": 0.9181184768676758,
+      "learning_rate": 2.933373349339736e-05,
+      "loss": 0.4825,
+      "step": 10341
+    },
+    {
+      "epoch": 13.23776,
+      "grad_norm": 0.9265780448913574,
+      "learning_rate": 2.9331732693077228e-05,
+      "loss": 0.5313,
+      "step": 10342
+    },
+    {
+      "epoch": 13.23904,
+      "grad_norm": 0.9894605875015259,
+      "learning_rate": 2.9329731892757107e-05,
+      "loss": 0.5617,
+      "step": 10343
+    },
+    {
+      "epoch": 13.24032,
+      "grad_norm": 0.9149695634841919,
+      "learning_rate": 2.932773109243698e-05,
+      "loss": 0.5262,
+      "step": 10344
+    },
+    {
+      "epoch": 13.2416,
+      "grad_norm": 0.9303181171417236,
+      "learning_rate": 2.9325730292116847e-05,
+      "loss": 0.542,
+      "step": 10345
+    },
+    {
+      "epoch": 13.24288,
+      "grad_norm": 0.958263099193573,
+      "learning_rate": 2.932372949179672e-05,
+      "loss": 0.5283,
+      "step": 10346
+    },
+    {
+      "epoch": 13.24416,
+      "grad_norm": 0.9637152552604675,
+      "learning_rate": 2.932172869147659e-05,
+      "loss": 0.5236,
+      "step": 10347
+    },
+    {
+      "epoch": 13.24544,
+      "grad_norm": 0.9949811100959778,
+      "learning_rate": 2.9319727891156463e-05,
+      "loss": 0.5897,
+      "step": 10348
+    },
+    {
+      "epoch": 13.24672,
+      "grad_norm": 0.988006055355072,
+      "learning_rate": 2.9317727090836334e-05,
+      "loss": 0.4802,
+      "step": 10349
+    },
+    {
+      "epoch": 13.248,
+      "grad_norm": 0.9795376062393188,
+      "learning_rate": 2.931572629051621e-05,
+      "loss": 0.5514,
+      "step": 10350
+    },
+    {
+      "epoch": 13.24928,
+      "grad_norm": 0.9923257827758789,
+      "learning_rate": 2.931372549019608e-05,
+      "loss": 0.5244,
+      "step": 10351
+    },
+    {
+      "epoch": 13.25056,
+      "grad_norm": 0.9084767699241638,
+      "learning_rate": 2.9311724689875953e-05,
+      "loss": 0.5395,
+      "step": 10352
+    },
+    {
+      "epoch": 13.25184,
+      "grad_norm": 0.9532386660575867,
+      "learning_rate": 2.9309723889555822e-05,
+      "loss": 0.5635,
+      "step": 10353
+    },
+    {
+      "epoch": 13.25312,
+      "grad_norm": 0.9513797163963318,
+      "learning_rate": 2.9307723089235694e-05,
+      "loss": 0.5299,
+      "step": 10354
+    },
+    {
+      "epoch": 13.2544,
+      "grad_norm": 0.9385073184967041,
+      "learning_rate": 2.9305722288915566e-05,
+      "loss": 0.5027,
+      "step": 10355
+    },
+    {
+      "epoch": 13.25568,
+      "grad_norm": 0.9745035767555237,
+      "learning_rate": 2.9303721488595437e-05,
+      "loss": 0.4923,
+      "step": 10356
+    },
+    {
+      "epoch": 13.25696,
+      "grad_norm": 0.9193017482757568,
+      "learning_rate": 2.9301720688275313e-05,
+      "loss": 0.5418,
+      "step": 10357
+    },
+    {
+      "epoch": 13.25824,
+      "grad_norm": 0.9343931674957275,
+      "learning_rate": 2.9299719887955185e-05,
+      "loss": 0.5197,
+      "step": 10358
+    },
+    {
+      "epoch": 13.25952,
+      "grad_norm": 0.954754114151001,
+      "learning_rate": 2.9297719087635056e-05,
+      "loss": 0.5203,
+      "step": 10359
+    },
+    {
+      "epoch": 13.2608,
+      "grad_norm": 0.9113470911979675,
+      "learning_rate": 2.9295718287314928e-05,
+      "loss": 0.525,
+      "step": 10360
+    },
+    {
+      "epoch": 13.26208,
+      "grad_norm": 0.94349205493927,
+      "learning_rate": 2.9293717486994797e-05,
+      "loss": 0.5263,
+      "step": 10361
+    },
+    {
+      "epoch": 13.26336,
+      "grad_norm": 0.980129063129425,
+      "learning_rate": 2.929171668667467e-05,
+      "loss": 0.5881,
+      "step": 10362
+    },
+    {
+      "epoch": 13.26464,
+      "grad_norm": 0.9003955125808716,
+      "learning_rate": 2.928971588635454e-05,
+      "loss": 0.5014,
+      "step": 10363
+    },
+    {
+      "epoch": 13.26592,
+      "grad_norm": 0.9492081999778748,
+      "learning_rate": 2.928771508603442e-05,
+      "loss": 0.5945,
+      "step": 10364
+    },
+    {
+      "epoch": 13.2672,
+      "grad_norm": 0.9139607548713684,
+      "learning_rate": 2.9285714285714288e-05,
+      "loss": 0.4997,
+      "step": 10365
+    },
+    {
+      "epoch": 13.26848,
+      "grad_norm": 0.9153253436088562,
+      "learning_rate": 2.928371348539416e-05,
+      "loss": 0.5104,
+      "step": 10366
+    },
+    {
+      "epoch": 13.26976,
+      "grad_norm": 0.944437563419342,
+      "learning_rate": 2.928171268507403e-05,
+      "loss": 0.5496,
+      "step": 10367
+    },
+    {
+      "epoch": 13.27104,
+      "grad_norm": 0.9628861546516418,
+      "learning_rate": 2.9279711884753903e-05,
+      "loss": 0.5243,
+      "step": 10368
+    },
+    {
+      "epoch": 13.27232,
+      "grad_norm": 0.9358019232749939,
+      "learning_rate": 2.927771108443377e-05,
+      "loss": 0.5218,
+      "step": 10369
+    },
+    {
+      "epoch": 13.2736,
+      "grad_norm": 0.9231796860694885,
+      "learning_rate": 2.9275710284113643e-05,
+      "loss": 0.5187,
+      "step": 10370
+    },
+    {
+      "epoch": 13.27488,
+      "grad_norm": 0.9316679239273071,
+      "learning_rate": 2.9273709483793522e-05,
+      "loss": 0.5469,
+      "step": 10371
+    },
+    {
+      "epoch": 13.27616,
+      "grad_norm": 0.8943392634391785,
+      "learning_rate": 2.9271708683473394e-05,
+      "loss": 0.4963,
+      "step": 10372
+    },
+    {
+      "epoch": 13.27744,
+      "grad_norm": 0.9293007850646973,
+      "learning_rate": 2.9269707883153262e-05,
+      "loss": 0.5534,
+      "step": 10373
+    },
+    {
+      "epoch": 13.27872,
+      "grad_norm": 0.9566161632537842,
+      "learning_rate": 2.9267707082833134e-05,
+      "loss": 0.6032,
+      "step": 10374
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 0.945538341999054,
+      "learning_rate": 2.9265706282513006e-05,
+      "loss": 0.525,
+      "step": 10375
+    },
+    {
+      "epoch": 13.28128,
+      "grad_norm": 0.9262788891792297,
+      "learning_rate": 2.9263705482192878e-05,
+      "loss": 0.5087,
+      "step": 10376
+    },
+    {
+      "epoch": 13.28256,
+      "grad_norm": 0.9590131044387817,
+      "learning_rate": 2.9261704681872746e-05,
+      "loss": 0.4981,
+      "step": 10377
+    },
+    {
+      "epoch": 13.28384,
+      "grad_norm": 0.9247324466705322,
+      "learning_rate": 2.9259703881552625e-05,
+      "loss": 0.5045,
+      "step": 10378
+    },
+    {
+      "epoch": 13.28512,
+      "grad_norm": 0.9148200154304504,
+      "learning_rate": 2.9257703081232497e-05,
+      "loss": 0.4692,
+      "step": 10379
+    },
+    {
+      "epoch": 13.2864,
+      "grad_norm": 0.9700202941894531,
+      "learning_rate": 2.925570228091237e-05,
+      "loss": 0.5181,
+      "step": 10380
+    },
+    {
+      "epoch": 13.28768,
+      "grad_norm": 0.9987017512321472,
+      "learning_rate": 2.9253701480592237e-05,
+      "loss": 0.5395,
+      "step": 10381
+    },
+    {
+      "epoch": 13.28896,
+      "grad_norm": 1.037541151046753,
+      "learning_rate": 2.925170068027211e-05,
+      "loss": 0.5464,
+      "step": 10382
+    },
+    {
+      "epoch": 13.29024,
+      "grad_norm": 0.979844868183136,
+      "learning_rate": 2.924969987995198e-05,
+      "loss": 0.5191,
+      "step": 10383
+    },
+    {
+      "epoch": 13.29152,
+      "grad_norm": 0.9726676344871521,
+      "learning_rate": 2.9247699079631853e-05,
+      "loss": 0.5464,
+      "step": 10384
+    },
+    {
+      "epoch": 13.2928,
+      "grad_norm": 0.9922785758972168,
+      "learning_rate": 2.9245698279311728e-05,
+      "loss": 0.5719,
+      "step": 10385
+    },
+    {
+      "epoch": 13.29408,
+      "grad_norm": 0.9066249132156372,
+      "learning_rate": 2.92436974789916e-05,
+      "loss": 0.5037,
+      "step": 10386
+    },
+    {
+      "epoch": 13.29536,
+      "grad_norm": 0.9383981823921204,
+      "learning_rate": 2.9241696678671472e-05,
+      "loss": 0.5398,
+      "step": 10387
+    },
+    {
+      "epoch": 13.29664,
+      "grad_norm": 0.9474811553955078,
+      "learning_rate": 2.9239695878351344e-05,
+      "loss": 0.5275,
+      "step": 10388
+    },
+    {
+      "epoch": 13.29792,
+      "grad_norm": 0.9555049538612366,
+      "learning_rate": 2.9237695078031212e-05,
+      "loss": 0.5737,
+      "step": 10389
+    },
+    {
+      "epoch": 13.2992,
+      "grad_norm": 0.9900108575820923,
+      "learning_rate": 2.9235694277711084e-05,
+      "loss": 0.5522,
+      "step": 10390
+    },
+    {
+      "epoch": 13.30048,
+      "grad_norm": 0.9271486401557922,
+      "learning_rate": 2.9233693477390956e-05,
+      "loss": 0.5284,
+      "step": 10391
+    },
+    {
+      "epoch": 13.30176,
+      "grad_norm": 0.9608096480369568,
+      "learning_rate": 2.923169267707083e-05,
+      "loss": 0.5538,
+      "step": 10392
+    },
+    {
+      "epoch": 13.30304,
+      "grad_norm": 0.9278596639633179,
+      "learning_rate": 2.9229691876750703e-05,
+      "loss": 0.5174,
+      "step": 10393
+    },
+    {
+      "epoch": 13.30432,
+      "grad_norm": 0.9708346128463745,
+      "learning_rate": 2.9227691076430575e-05,
+      "loss": 0.5499,
+      "step": 10394
+    },
+    {
+      "epoch": 13.3056,
+      "grad_norm": 0.9422946572303772,
+      "learning_rate": 2.9225690276110447e-05,
+      "loss": 0.5444,
+      "step": 10395
+    },
+    {
+      "epoch": 13.30688,
+      "grad_norm": 0.9658584594726562,
+      "learning_rate": 2.922368947579032e-05,
+      "loss": 0.4903,
+      "step": 10396
+    },
+    {
+      "epoch": 13.30816,
+      "grad_norm": 0.978703498840332,
+      "learning_rate": 2.9221688675470187e-05,
+      "loss": 0.5569,
+      "step": 10397
+    },
+    {
+      "epoch": 13.30944,
+      "grad_norm": 0.9500980973243713,
+      "learning_rate": 2.921968787515006e-05,
+      "loss": 0.5384,
+      "step": 10398
+    },
+    {
+      "epoch": 13.31072,
+      "grad_norm": 0.9386541247367859,
+      "learning_rate": 2.9217687074829937e-05,
+      "loss": 0.5474,
+      "step": 10399
+    },
+    {
+      "epoch": 13.312,
+      "grad_norm": 0.9124325513839722,
+      "learning_rate": 2.9215686274509806e-05,
+      "loss": 0.5232,
+      "step": 10400
+    },
+    {
+      "epoch": 13.31328,
+      "grad_norm": 0.9420004487037659,
+      "learning_rate": 2.9213685474189678e-05,
+      "loss": 0.5154,
+      "step": 10401
+    },
+    {
+      "epoch": 13.31456,
+      "grad_norm": 1.029594898223877,
+      "learning_rate": 2.921168467386955e-05,
+      "loss": 0.6302,
+      "step": 10402
+    },
+    {
+      "epoch": 13.31584,
+      "grad_norm": 0.9675827622413635,
+      "learning_rate": 2.920968387354942e-05,
+      "loss": 0.5515,
+      "step": 10403
+    },
+    {
+      "epoch": 13.31712,
+      "grad_norm": 0.9070100784301758,
+      "learning_rate": 2.9207683073229293e-05,
+      "loss": 0.5061,
+      "step": 10404
+    },
+    {
+      "epoch": 13.3184,
+      "grad_norm": 0.9273656606674194,
+      "learning_rate": 2.9205682272909162e-05,
+      "loss": 0.5412,
+      "step": 10405
+    },
+    {
+      "epoch": 13.31968,
+      "grad_norm": 0.9532672762870789,
+      "learning_rate": 2.920368147258904e-05,
+      "loss": 0.5259,
+      "step": 10406
+    },
+    {
+      "epoch": 13.32096,
+      "grad_norm": 0.9627009630203247,
+      "learning_rate": 2.9201680672268912e-05,
+      "loss": 0.5173,
+      "step": 10407
+    },
+    {
+      "epoch": 13.32224,
+      "grad_norm": 0.9319729208946228,
+      "learning_rate": 2.919967987194878e-05,
+      "loss": 0.5869,
+      "step": 10408
+    },
+    {
+      "epoch": 13.32352,
+      "grad_norm": 0.873471200466156,
+      "learning_rate": 2.9197679071628653e-05,
+      "loss": 0.4851,
+      "step": 10409
+    },
+    {
+      "epoch": 13.3248,
+      "grad_norm": 0.9522954821586609,
+      "learning_rate": 2.9195678271308524e-05,
+      "loss": 0.5164,
+      "step": 10410
+    },
+    {
+      "epoch": 13.32608,
+      "grad_norm": 0.961445689201355,
+      "learning_rate": 2.9193677470988396e-05,
+      "loss": 0.5083,
+      "step": 10411
+    },
+    {
+      "epoch": 13.32736,
+      "grad_norm": 0.9519370794296265,
+      "learning_rate": 2.9191676670668268e-05,
+      "loss": 0.5352,
+      "step": 10412
+    },
+    {
+      "epoch": 13.32864,
+      "grad_norm": 0.9867494702339172,
+      "learning_rate": 2.9189675870348143e-05,
+      "loss": 0.5495,
+      "step": 10413
+    },
+    {
+      "epoch": 13.32992,
+      "grad_norm": 0.8867868185043335,
+      "learning_rate": 2.9187675070028015e-05,
+      "loss": 0.5229,
+      "step": 10414
+    },
+    {
+      "epoch": 13.3312,
+      "grad_norm": 0.9686993360519409,
+      "learning_rate": 2.9185674269707887e-05,
+      "loss": 0.5375,
+      "step": 10415
+    },
+    {
+      "epoch": 13.33248,
+      "grad_norm": 0.9888986945152283,
+      "learning_rate": 2.9183673469387756e-05,
+      "loss": 0.6055,
+      "step": 10416
+    },
+    {
+      "epoch": 13.33376,
+      "grad_norm": 0.9201542735099792,
+      "learning_rate": 2.9181672669067627e-05,
+      "loss": 0.5122,
+      "step": 10417
+    },
+    {
+      "epoch": 13.33504,
+      "grad_norm": 0.9173356294631958,
+      "learning_rate": 2.91796718687475e-05,
+      "loss": 0.5429,
+      "step": 10418
+    },
+    {
+      "epoch": 13.33632,
+      "grad_norm": 0.9418560862541199,
+      "learning_rate": 2.917767106842737e-05,
+      "loss": 0.5358,
+      "step": 10419
+    },
+    {
+      "epoch": 13.3376,
+      "grad_norm": 0.9006747603416443,
+      "learning_rate": 2.9175670268107246e-05,
+      "loss": 0.4745,
+      "step": 10420
+    },
+    {
+      "epoch": 13.33888,
+      "grad_norm": 0.9309629201889038,
+      "learning_rate": 2.9173669467787118e-05,
+      "loss": 0.5002,
+      "step": 10421
+    },
+    {
+      "epoch": 13.340160000000001,
+      "grad_norm": 0.9739012718200684,
+      "learning_rate": 2.917166866746699e-05,
+      "loss": 0.5861,
+      "step": 10422
+    },
+    {
+      "epoch": 13.34144,
+      "grad_norm": 0.9974890351295471,
+      "learning_rate": 2.9169667867146862e-05,
+      "loss": 0.5475,
+      "step": 10423
+    },
+    {
+      "epoch": 13.34272,
+      "grad_norm": 0.9011924862861633,
+      "learning_rate": 2.916766706682673e-05,
+      "loss": 0.5111,
+      "step": 10424
+    },
+    {
+      "epoch": 13.344,
+      "grad_norm": 1.0001959800720215,
+      "learning_rate": 2.9165666266506602e-05,
+      "loss": 0.5563,
+      "step": 10425
+    },
+    {
+      "epoch": 13.34528,
+      "grad_norm": 1.0525249242782593,
+      "learning_rate": 2.9163665466186474e-05,
+      "loss": 0.5812,
+      "step": 10426
+    },
+    {
+      "epoch": 13.34656,
+      "grad_norm": 0.9634531140327454,
+      "learning_rate": 2.916166466586635e-05,
+      "loss": 0.5414,
+      "step": 10427
+    },
+    {
+      "epoch": 13.34784,
+      "grad_norm": 0.9581708908081055,
+      "learning_rate": 2.915966386554622e-05,
+      "loss": 0.5498,
+      "step": 10428
+    },
+    {
+      "epoch": 13.34912,
+      "grad_norm": 1.0304726362228394,
+      "learning_rate": 2.9157663065226093e-05,
+      "loss": 0.543,
+      "step": 10429
+    },
+    {
+      "epoch": 13.3504,
+      "grad_norm": 0.9573806524276733,
+      "learning_rate": 2.9155662264905965e-05,
+      "loss": 0.5478,
+      "step": 10430
+    },
+    {
+      "epoch": 13.35168,
+      "grad_norm": 0.8998541831970215,
+      "learning_rate": 2.9153661464585837e-05,
+      "loss": 0.5521,
+      "step": 10431
+    },
+    {
+      "epoch": 13.35296,
+      "grad_norm": 0.9216572046279907,
+      "learning_rate": 2.9151660664265705e-05,
+      "loss": 0.5198,
+      "step": 10432
+    },
+    {
+      "epoch": 13.35424,
+      "grad_norm": 0.9832442402839661,
+      "learning_rate": 2.9149659863945577e-05,
+      "loss": 0.5542,
+      "step": 10433
+    },
+    {
+      "epoch": 13.35552,
+      "grad_norm": 0.9601088166236877,
+      "learning_rate": 2.9147659063625456e-05,
+      "loss": 0.5416,
+      "step": 10434
+    },
+    {
+      "epoch": 13.3568,
+      "grad_norm": 0.8933698534965515,
+      "learning_rate": 2.9145658263305324e-05,
+      "loss": 0.5418,
+      "step": 10435
+    },
+    {
+      "epoch": 13.35808,
+      "grad_norm": 0.9842942357063293,
+      "learning_rate": 2.9143657462985196e-05,
+      "loss": 0.5526,
+      "step": 10436
+    },
+    {
+      "epoch": 13.35936,
+      "grad_norm": 0.9776619076728821,
+      "learning_rate": 2.9141656662665068e-05,
+      "loss": 0.5864,
+      "step": 10437
+    },
+    {
+      "epoch": 13.36064,
+      "grad_norm": 0.9885453581809998,
+      "learning_rate": 2.913965586234494e-05,
+      "loss": 0.584,
+      "step": 10438
+    },
+    {
+      "epoch": 13.36192,
+      "grad_norm": 0.9770693182945251,
+      "learning_rate": 2.913765506202481e-05,
+      "loss": 0.5273,
+      "step": 10439
+    },
+    {
+      "epoch": 13.3632,
+      "grad_norm": 0.9777031540870667,
+      "learning_rate": 2.913565426170468e-05,
+      "loss": 0.5872,
+      "step": 10440
+    },
+    {
+      "epoch": 13.36448,
+      "grad_norm": 0.9541405439376831,
+      "learning_rate": 2.913365346138456e-05,
+      "loss": 0.5395,
+      "step": 10441
+    },
+    {
+      "epoch": 13.36576,
+      "grad_norm": 0.9622824192047119,
+      "learning_rate": 2.913165266106443e-05,
+      "loss": 0.5505,
+      "step": 10442
+    },
+    {
+      "epoch": 13.36704,
+      "grad_norm": 0.9015635848045349,
+      "learning_rate": 2.91296518607443e-05,
+      "loss": 0.5042,
+      "step": 10443
+    },
+    {
+      "epoch": 13.36832,
+      "grad_norm": 0.9411506056785583,
+      "learning_rate": 2.912765106042417e-05,
+      "loss": 0.5998,
+      "step": 10444
+    },
+    {
+      "epoch": 13.3696,
+      "grad_norm": 0.9454710483551025,
+      "learning_rate": 2.9125650260104043e-05,
+      "loss": 0.5047,
+      "step": 10445
+    },
+    {
+      "epoch": 13.37088,
+      "grad_norm": 0.9664983153343201,
+      "learning_rate": 2.9123649459783915e-05,
+      "loss": 0.5501,
+      "step": 10446
+    },
+    {
+      "epoch": 13.37216,
+      "grad_norm": 0.9950606226921082,
+      "learning_rate": 2.9121648659463787e-05,
+      "loss": 0.548,
+      "step": 10447
+    },
+    {
+      "epoch": 13.37344,
+      "grad_norm": 0.9428080916404724,
+      "learning_rate": 2.9119647859143655e-05,
+      "loss": 0.515,
+      "step": 10448
+    },
+    {
+      "epoch": 13.37472,
+      "grad_norm": 1.0005260705947876,
+      "learning_rate": 2.9117647058823534e-05,
+      "loss": 0.6053,
+      "step": 10449
+    },
+    {
+      "epoch": 13.376,
+      "grad_norm": 0.9256664514541626,
+      "learning_rate": 2.9115646258503405e-05,
+      "loss": 0.5313,
+      "step": 10450
+    },
+    {
+      "epoch": 13.37728,
+      "grad_norm": 0.9487593173980713,
+      "learning_rate": 2.9113645458183274e-05,
+      "loss": 0.5372,
+      "step": 10451
+    },
+    {
+      "epoch": 13.37856,
+      "grad_norm": 0.9541116952896118,
+      "learning_rate": 2.9111644657863146e-05,
+      "loss": 0.565,
+      "step": 10452
+    },
+    {
+      "epoch": 13.37984,
+      "grad_norm": 0.9423818588256836,
+      "learning_rate": 2.9109643857543018e-05,
+      "loss": 0.5485,
+      "step": 10453
+    },
+    {
+      "epoch": 13.38112,
+      "grad_norm": 0.9149635434150696,
+      "learning_rate": 2.910764305722289e-05,
+      "loss": 0.4844,
+      "step": 10454
+    },
+    {
+      "epoch": 13.3824,
+      "grad_norm": 1.0056577920913696,
+      "learning_rate": 2.910564225690276e-05,
+      "loss": 0.5753,
+      "step": 10455
+    },
+    {
+      "epoch": 13.38368,
+      "grad_norm": 1.0535948276519775,
+      "learning_rate": 2.9103641456582637e-05,
+      "loss": 0.5831,
+      "step": 10456
+    },
+    {
+      "epoch": 13.38496,
+      "grad_norm": 0.9600538015365601,
+      "learning_rate": 2.910164065626251e-05,
+      "loss": 0.5554,
+      "step": 10457
+    },
+    {
+      "epoch": 13.38624,
+      "grad_norm": 0.9667772054672241,
+      "learning_rate": 2.909963985594238e-05,
+      "loss": 0.5197,
+      "step": 10458
+    },
+    {
+      "epoch": 13.38752,
+      "grad_norm": 0.949582040309906,
+      "learning_rate": 2.909763905562225e-05,
+      "loss": 0.5152,
+      "step": 10459
+    },
+    {
+      "epoch": 13.3888,
+      "grad_norm": 0.9107711315155029,
+      "learning_rate": 2.909563825530212e-05,
+      "loss": 0.4774,
+      "step": 10460
+    },
+    {
+      "epoch": 13.39008,
+      "grad_norm": 1.0431016683578491,
+      "learning_rate": 2.9093637454981993e-05,
+      "loss": 0.5594,
+      "step": 10461
+    },
+    {
+      "epoch": 13.39136,
+      "grad_norm": 0.9607868194580078,
+      "learning_rate": 2.9091636654661864e-05,
+      "loss": 0.5385,
+      "step": 10462
+    },
+    {
+      "epoch": 13.39264,
+      "grad_norm": 0.9767603278160095,
+      "learning_rate": 2.908963585434174e-05,
+      "loss": 0.5555,
+      "step": 10463
+    },
+    {
+      "epoch": 13.39392,
+      "grad_norm": 0.920050859451294,
+      "learning_rate": 2.908763505402161e-05,
+      "loss": 0.5062,
+      "step": 10464
+    },
+    {
+      "epoch": 13.395199999999999,
+      "grad_norm": 0.9358279705047607,
+      "learning_rate": 2.9085634253701483e-05,
+      "loss": 0.5381,
+      "step": 10465
+    },
+    {
+      "epoch": 13.39648,
+      "grad_norm": 0.9233344793319702,
+      "learning_rate": 2.9083633453381355e-05,
+      "loss": 0.5331,
+      "step": 10466
+    },
+    {
+      "epoch": 13.39776,
+      "grad_norm": 0.9128267168998718,
+      "learning_rate": 2.9081632653061224e-05,
+      "loss": 0.5363,
+      "step": 10467
+    },
+    {
+      "epoch": 13.39904,
+      "grad_norm": 1.0040749311447144,
+      "learning_rate": 2.9079631852741096e-05,
+      "loss": 0.5948,
+      "step": 10468
+    },
+    {
+      "epoch": 13.40032,
+      "grad_norm": 0.9266423583030701,
+      "learning_rate": 2.9077631052420967e-05,
+      "loss": 0.5511,
+      "step": 10469
+    },
+    {
+      "epoch": 13.4016,
+      "grad_norm": 0.8982942700386047,
+      "learning_rate": 2.9075630252100843e-05,
+      "loss": 0.5446,
+      "step": 10470
+    },
+    {
+      "epoch": 13.40288,
+      "grad_norm": 0.9734070301055908,
+      "learning_rate": 2.9073629451780714e-05,
+      "loss": 0.5567,
+      "step": 10471
+    },
+    {
+      "epoch": 13.40416,
+      "grad_norm": 0.9396576285362244,
+      "learning_rate": 2.9071628651460586e-05,
+      "loss": 0.5146,
+      "step": 10472
+    },
+    {
+      "epoch": 13.40544,
+      "grad_norm": 0.9016363620758057,
+      "learning_rate": 2.9069627851140458e-05,
+      "loss": 0.5465,
+      "step": 10473
+    },
+    {
+      "epoch": 13.40672,
+      "grad_norm": 0.950202465057373,
+      "learning_rate": 2.906762705082033e-05,
+      "loss": 0.5659,
+      "step": 10474
+    },
+    {
+      "epoch": 13.408,
+      "grad_norm": 0.9889881610870361,
+      "learning_rate": 2.90656262505002e-05,
+      "loss": 0.5675,
+      "step": 10475
+    },
+    {
+      "epoch": 13.40928,
+      "grad_norm": 0.9744821190834045,
+      "learning_rate": 2.906362545018007e-05,
+      "loss": 0.5625,
+      "step": 10476
+    },
+    {
+      "epoch": 13.41056,
+      "grad_norm": 0.9143860340118408,
+      "learning_rate": 2.906162464985995e-05,
+      "loss": 0.522,
+      "step": 10477
+    },
+    {
+      "epoch": 13.41184,
+      "grad_norm": 0.9224212169647217,
+      "learning_rate": 2.9059623849539817e-05,
+      "loss": 0.4886,
+      "step": 10478
+    },
+    {
+      "epoch": 13.41312,
+      "grad_norm": 0.9844509959220886,
+      "learning_rate": 2.905762304921969e-05,
+      "loss": 0.559,
+      "step": 10479
+    },
+    {
+      "epoch": 13.4144,
+      "grad_norm": 0.919024646282196,
+      "learning_rate": 2.905562224889956e-05,
+      "loss": 0.5372,
+      "step": 10480
+    },
+    {
+      "epoch": 13.41568,
+      "grad_norm": 0.9639186859130859,
+      "learning_rate": 2.9053621448579433e-05,
+      "loss": 0.5797,
+      "step": 10481
+    },
+    {
+      "epoch": 13.41696,
+      "grad_norm": 0.926937460899353,
+      "learning_rate": 2.9051620648259305e-05,
+      "loss": 0.5645,
+      "step": 10482
+    },
+    {
+      "epoch": 13.41824,
+      "grad_norm": 0.9183882474899292,
+      "learning_rate": 2.9049619847939173e-05,
+      "loss": 0.5077,
+      "step": 10483
+    },
+    {
+      "epoch": 13.41952,
+      "grad_norm": 0.9052950739860535,
+      "learning_rate": 2.9047619047619052e-05,
+      "loss": 0.5042,
+      "step": 10484
+    },
+    {
+      "epoch": 13.4208,
+      "grad_norm": 0.9210243225097656,
+      "learning_rate": 2.9045618247298924e-05,
+      "loss": 0.5045,
+      "step": 10485
+    },
+    {
+      "epoch": 13.42208,
+      "grad_norm": 0.9332923293113708,
+      "learning_rate": 2.9043617446978792e-05,
+      "loss": 0.556,
+      "step": 10486
+    },
+    {
+      "epoch": 13.42336,
+      "grad_norm": 0.8918114304542542,
+      "learning_rate": 2.9041616646658664e-05,
+      "loss": 0.5455,
+      "step": 10487
+    },
+    {
+      "epoch": 13.42464,
+      "grad_norm": 1.0036391019821167,
+      "learning_rate": 2.9039615846338536e-05,
+      "loss": 0.6048,
+      "step": 10488
+    },
+    {
+      "epoch": 13.42592,
+      "grad_norm": 0.9452251195907593,
+      "learning_rate": 2.9037615046018408e-05,
+      "loss": 0.5311,
+      "step": 10489
+    },
+    {
+      "epoch": 13.4272,
+      "grad_norm": 0.9231464266777039,
+      "learning_rate": 2.903561424569828e-05,
+      "loss": 0.5153,
+      "step": 10490
+    },
+    {
+      "epoch": 13.42848,
+      "grad_norm": 0.9702645540237427,
+      "learning_rate": 2.9033613445378155e-05,
+      "loss": 0.549,
+      "step": 10491
+    },
+    {
+      "epoch": 13.42976,
+      "grad_norm": 0.9461646676063538,
+      "learning_rate": 2.9031612645058027e-05,
+      "loss": 0.5309,
+      "step": 10492
+    },
+    {
+      "epoch": 13.43104,
+      "grad_norm": 0.9646087288856506,
+      "learning_rate": 2.90296118447379e-05,
+      "loss": 0.5819,
+      "step": 10493
+    },
+    {
+      "epoch": 13.43232,
+      "grad_norm": 0.9191693067550659,
+      "learning_rate": 2.9027611044417767e-05,
+      "loss": 0.5552,
+      "step": 10494
+    },
+    {
+      "epoch": 13.4336,
+      "grad_norm": 0.9861147999763489,
+      "learning_rate": 2.902561024409764e-05,
+      "loss": 0.5568,
+      "step": 10495
+    },
+    {
+      "epoch": 13.43488,
+      "grad_norm": 0.9543395042419434,
+      "learning_rate": 2.902360944377751e-05,
+      "loss": 0.5337,
+      "step": 10496
+    },
+    {
+      "epoch": 13.43616,
+      "grad_norm": 0.9656491875648499,
+      "learning_rate": 2.9021608643457383e-05,
+      "loss": 0.5232,
+      "step": 10497
+    },
+    {
+      "epoch": 13.43744,
+      "grad_norm": 0.9861603379249573,
+      "learning_rate": 2.9019607843137258e-05,
+      "loss": 0.5745,
+      "step": 10498
+    },
+    {
+      "epoch": 13.43872,
+      "grad_norm": 0.9877046942710876,
+      "learning_rate": 2.901760704281713e-05,
+      "loss": 0.584,
+      "step": 10499
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 0.9454039931297302,
+      "learning_rate": 2.9015606242497002e-05,
+      "loss": 0.5021,
+      "step": 10500
+    },
+    {
+      "epoch": 13.44128,
+      "grad_norm": 0.9773223996162415,
+      "learning_rate": 2.9013605442176874e-05,
+      "loss": 0.5659,
+      "step": 10501
+    },
+    {
+      "epoch": 13.44256,
+      "grad_norm": 0.9586243033409119,
+      "learning_rate": 2.9011604641856742e-05,
+      "loss": 0.5496,
+      "step": 10502
+    },
+    {
+      "epoch": 13.44384,
+      "grad_norm": 0.9668572545051575,
+      "learning_rate": 2.9009603841536614e-05,
+      "loss": 0.5088,
+      "step": 10503
+    },
+    {
+      "epoch": 13.44512,
+      "grad_norm": 0.9630528092384338,
+      "learning_rate": 2.9007603041216486e-05,
+      "loss": 0.5007,
+      "step": 10504
+    },
+    {
+      "epoch": 13.4464,
+      "grad_norm": 0.984404444694519,
+      "learning_rate": 2.900560224089636e-05,
+      "loss": 0.5157,
+      "step": 10505
+    },
+    {
+      "epoch": 13.44768,
+      "grad_norm": 0.9502925276756287,
+      "learning_rate": 2.9003601440576233e-05,
+      "loss": 0.4632,
+      "step": 10506
+    },
+    {
+      "epoch": 13.44896,
+      "grad_norm": 0.9846747517585754,
+      "learning_rate": 2.9001600640256105e-05,
+      "loss": 0.5392,
+      "step": 10507
+    },
+    {
+      "epoch": 13.45024,
+      "grad_norm": 0.9732792377471924,
+      "learning_rate": 2.8999599839935977e-05,
+      "loss": 0.5066,
+      "step": 10508
+    },
+    {
+      "epoch": 13.45152,
+      "grad_norm": 0.9376108050346375,
+      "learning_rate": 2.899759903961585e-05,
+      "loss": 0.568,
+      "step": 10509
+    },
+    {
+      "epoch": 13.4528,
+      "grad_norm": 0.9324885010719299,
+      "learning_rate": 2.8995598239295717e-05,
+      "loss": 0.5635,
+      "step": 10510
+    },
+    {
+      "epoch": 13.45408,
+      "grad_norm": 0.9335364699363708,
+      "learning_rate": 2.899359743897559e-05,
+      "loss": 0.5284,
+      "step": 10511
+    },
+    {
+      "epoch": 13.45536,
+      "grad_norm": 0.9017517566680908,
+      "learning_rate": 2.8991596638655467e-05,
+      "loss": 0.4938,
+      "step": 10512
+    },
+    {
+      "epoch": 13.45664,
+      "grad_norm": 0.9306479096412659,
+      "learning_rate": 2.8989595838335336e-05,
+      "loss": 0.5521,
+      "step": 10513
+    },
+    {
+      "epoch": 13.45792,
+      "grad_norm": 0.9630993008613586,
+      "learning_rate": 2.8987595038015208e-05,
+      "loss": 0.5371,
+      "step": 10514
+    },
+    {
+      "epoch": 13.4592,
+      "grad_norm": 0.9604276418685913,
+      "learning_rate": 2.898559423769508e-05,
+      "loss": 0.5543,
+      "step": 10515
+    },
+    {
+      "epoch": 13.46048,
+      "grad_norm": 0.9642881155014038,
+      "learning_rate": 2.898359343737495e-05,
+      "loss": 0.525,
+      "step": 10516
+    },
+    {
+      "epoch": 13.46176,
+      "grad_norm": 0.9426066279411316,
+      "learning_rate": 2.8981592637054823e-05,
+      "loss": 0.5535,
+      "step": 10517
+    },
+    {
+      "epoch": 13.46304,
+      "grad_norm": 0.8915838003158569,
+      "learning_rate": 2.8979591836734692e-05,
+      "loss": 0.5155,
+      "step": 10518
+    },
+    {
+      "epoch": 13.46432,
+      "grad_norm": 0.9411042332649231,
+      "learning_rate": 2.897759103641457e-05,
+      "loss": 0.5197,
+      "step": 10519
+    },
+    {
+      "epoch": 13.4656,
+      "grad_norm": 0.9480369091033936,
+      "learning_rate": 2.8975590236094442e-05,
+      "loss": 0.5425,
+      "step": 10520
+    },
+    {
+      "epoch": 13.46688,
+      "grad_norm": 0.9183311462402344,
+      "learning_rate": 2.897358943577431e-05,
+      "loss": 0.5394,
+      "step": 10521
+    },
+    {
+      "epoch": 13.46816,
+      "grad_norm": 0.8983965516090393,
+      "learning_rate": 2.8971588635454183e-05,
+      "loss": 0.5216,
+      "step": 10522
+    },
+    {
+      "epoch": 13.46944,
+      "grad_norm": 0.943598210811615,
+      "learning_rate": 2.8969587835134054e-05,
+      "loss": 0.5223,
+      "step": 10523
+    },
+    {
+      "epoch": 13.47072,
+      "grad_norm": 0.9010968804359436,
+      "learning_rate": 2.8967587034813926e-05,
+      "loss": 0.5572,
+      "step": 10524
+    },
+    {
+      "epoch": 13.472,
+      "grad_norm": 0.9356712698936462,
+      "learning_rate": 2.8965586234493798e-05,
+      "loss": 0.5327,
+      "step": 10525
+    },
+    {
+      "epoch": 13.47328,
+      "grad_norm": 0.9779828786849976,
+      "learning_rate": 2.8963585434173673e-05,
+      "loss": 0.5572,
+      "step": 10526
+    },
+    {
+      "epoch": 13.47456,
+      "grad_norm": 0.9834756851196289,
+      "learning_rate": 2.8961584633853545e-05,
+      "loss": 0.5512,
+      "step": 10527
+    },
+    {
+      "epoch": 13.47584,
+      "grad_norm": 1.0116183757781982,
+      "learning_rate": 2.8959583833533417e-05,
+      "loss": 0.5726,
+      "step": 10528
+    },
+    {
+      "epoch": 13.47712,
+      "grad_norm": 0.9891726970672607,
+      "learning_rate": 2.8957583033213286e-05,
+      "loss": 0.5587,
+      "step": 10529
+    },
+    {
+      "epoch": 13.4784,
+      "grad_norm": 0.9156029224395752,
+      "learning_rate": 2.8955582232893157e-05,
+      "loss": 0.5175,
+      "step": 10530
+    },
+    {
+      "epoch": 13.47968,
+      "grad_norm": 0.9811277389526367,
+      "learning_rate": 2.895358143257303e-05,
+      "loss": 0.5572,
+      "step": 10531
+    },
+    {
+      "epoch": 13.48096,
+      "grad_norm": 0.9236703515052795,
+      "learning_rate": 2.89515806322529e-05,
+      "loss": 0.5119,
+      "step": 10532
+    },
+    {
+      "epoch": 13.482240000000001,
+      "grad_norm": 0.9552645087242126,
+      "learning_rate": 2.8949579831932776e-05,
+      "loss": 0.544,
+      "step": 10533
+    },
+    {
+      "epoch": 13.48352,
+      "grad_norm": 0.9037032127380371,
+      "learning_rate": 2.8947579031612648e-05,
+      "loss": 0.5298,
+      "step": 10534
+    },
+    {
+      "epoch": 13.4848,
+      "grad_norm": 0.947192370891571,
+      "learning_rate": 2.894557823129252e-05,
+      "loss": 0.5685,
+      "step": 10535
+    },
+    {
+      "epoch": 13.48608,
+      "grad_norm": 0.9519645571708679,
+      "learning_rate": 2.8943577430972392e-05,
+      "loss": 0.5339,
+      "step": 10536
+    },
+    {
+      "epoch": 13.48736,
+      "grad_norm": 0.9448265433311462,
+      "learning_rate": 2.894157663065226e-05,
+      "loss": 0.5473,
+      "step": 10537
+    },
+    {
+      "epoch": 13.48864,
+      "grad_norm": 0.9873289465904236,
+      "learning_rate": 2.8939575830332132e-05,
+      "loss": 0.5253,
+      "step": 10538
+    },
+    {
+      "epoch": 13.48992,
+      "grad_norm": 0.9552075862884521,
+      "learning_rate": 2.8937575030012004e-05,
+      "loss": 0.5561,
+      "step": 10539
+    },
+    {
+      "epoch": 13.4912,
+      "grad_norm": 0.9306079745292664,
+      "learning_rate": 2.893557422969188e-05,
+      "loss": 0.5058,
+      "step": 10540
+    },
+    {
+      "epoch": 13.49248,
+      "grad_norm": 0.9096477627754211,
+      "learning_rate": 2.893357342937175e-05,
+      "loss": 0.4799,
+      "step": 10541
+    },
+    {
+      "epoch": 13.49376,
+      "grad_norm": 1.00719153881073,
+      "learning_rate": 2.8931572629051623e-05,
+      "loss": 0.5475,
+      "step": 10542
+    },
+    {
+      "epoch": 13.49504,
+      "grad_norm": 1.0264086723327637,
+      "learning_rate": 2.8929571828731495e-05,
+      "loss": 0.5664,
+      "step": 10543
+    },
+    {
+      "epoch": 13.49632,
+      "grad_norm": 0.9574489593505859,
+      "learning_rate": 2.8927571028411367e-05,
+      "loss": 0.5239,
+      "step": 10544
+    },
+    {
+      "epoch": 13.4976,
+      "grad_norm": 0.9818712472915649,
+      "learning_rate": 2.8925570228091235e-05,
+      "loss": 0.5373,
+      "step": 10545
+    },
+    {
+      "epoch": 13.49888,
+      "grad_norm": 0.9273495674133301,
+      "learning_rate": 2.8923569427771107e-05,
+      "loss": 0.5535,
+      "step": 10546
+    },
+    {
+      "epoch": 13.50016,
+      "grad_norm": 0.9357905983924866,
+      "learning_rate": 2.8921568627450986e-05,
+      "loss": 0.5536,
+      "step": 10547
+    },
+    {
+      "epoch": 13.50144,
+      "grad_norm": 0.9313623309135437,
+      "learning_rate": 2.8919567827130854e-05,
+      "loss": 0.5436,
+      "step": 10548
+    },
+    {
+      "epoch": 13.50272,
+      "grad_norm": 0.9302566051483154,
+      "learning_rate": 2.8917567026810726e-05,
+      "loss": 0.503,
+      "step": 10549
+    },
+    {
+      "epoch": 13.504,
+      "grad_norm": 0.9522665143013,
+      "learning_rate": 2.8915566226490598e-05,
+      "loss": 0.5565,
+      "step": 10550
+    },
+    {
+      "epoch": 13.505279999999999,
+      "grad_norm": 0.9212354421615601,
+      "learning_rate": 2.891356542617047e-05,
+      "loss": 0.5448,
+      "step": 10551
+    },
+    {
+      "epoch": 13.50656,
+      "grad_norm": 0.9201790690422058,
+      "learning_rate": 2.891156462585034e-05,
+      "loss": 0.509,
+      "step": 10552
+    },
+    {
+      "epoch": 13.50784,
+      "grad_norm": 0.91869056224823,
+      "learning_rate": 2.890956382553021e-05,
+      "loss": 0.5319,
+      "step": 10553
+    },
+    {
+      "epoch": 13.50912,
+      "grad_norm": 0.9275939464569092,
+      "learning_rate": 2.890756302521009e-05,
+      "loss": 0.5429,
+      "step": 10554
+    },
+    {
+      "epoch": 13.5104,
+      "grad_norm": 0.9292113780975342,
+      "learning_rate": 2.890556222488996e-05,
+      "loss": 0.5673,
+      "step": 10555
+    },
+    {
+      "epoch": 13.51168,
+      "grad_norm": 0.9870272278785706,
+      "learning_rate": 2.890356142456983e-05,
+      "loss": 0.5295,
+      "step": 10556
+    },
+    {
+      "epoch": 13.51296,
+      "grad_norm": 0.9538768529891968,
+      "learning_rate": 2.89015606242497e-05,
+      "loss": 0.5421,
+      "step": 10557
+    },
+    {
+      "epoch": 13.514240000000001,
+      "grad_norm": 0.9605304598808289,
+      "learning_rate": 2.8899559823929573e-05,
+      "loss": 0.5666,
+      "step": 10558
+    },
+    {
+      "epoch": 13.51552,
+      "grad_norm": 0.9257391691207886,
+      "learning_rate": 2.8897559023609445e-05,
+      "loss": 0.5048,
+      "step": 10559
+    },
+    {
+      "epoch": 13.5168,
+      "grad_norm": 0.924060046672821,
+      "learning_rate": 2.8895558223289317e-05,
+      "loss": 0.5625,
+      "step": 10560
+    },
+    {
+      "epoch": 13.51808,
+      "grad_norm": 0.9286275506019592,
+      "learning_rate": 2.8893557422969185e-05,
+      "loss": 0.4978,
+      "step": 10561
+    },
+    {
+      "epoch": 13.51936,
+      "grad_norm": 1.009763240814209,
+      "learning_rate": 2.8891556622649064e-05,
+      "loss": 0.5748,
+      "step": 10562
+    },
+    {
+      "epoch": 13.52064,
+      "grad_norm": 0.9351606369018555,
+      "learning_rate": 2.8889555822328935e-05,
+      "loss": 0.4949,
+      "step": 10563
+    },
+    {
+      "epoch": 13.52192,
+      "grad_norm": 0.9865330457687378,
+      "learning_rate": 2.8887555022008804e-05,
+      "loss": 0.5523,
+      "step": 10564
+    },
+    {
+      "epoch": 13.5232,
+      "grad_norm": 0.9141054749488831,
+      "learning_rate": 2.8885554221688676e-05,
+      "loss": 0.535,
+      "step": 10565
+    },
+    {
+      "epoch": 13.52448,
+      "grad_norm": 0.8817773461341858,
+      "learning_rate": 2.8883553421368548e-05,
+      "loss": 0.5218,
+      "step": 10566
+    },
+    {
+      "epoch": 13.52576,
+      "grad_norm": 0.918411910533905,
+      "learning_rate": 2.888155262104842e-05,
+      "loss": 0.5497,
+      "step": 10567
+    },
+    {
+      "epoch": 13.52704,
+      "grad_norm": 0.9279268383979797,
+      "learning_rate": 2.887955182072829e-05,
+      "loss": 0.5519,
+      "step": 10568
+    },
+    {
+      "epoch": 13.52832,
+      "grad_norm": 0.9723789095878601,
+      "learning_rate": 2.8877551020408167e-05,
+      "loss": 0.5554,
+      "step": 10569
+    },
+    {
+      "epoch": 13.5296,
+      "grad_norm": 0.9536760449409485,
+      "learning_rate": 2.887555022008804e-05,
+      "loss": 0.5561,
+      "step": 10570
+    },
+    {
+      "epoch": 13.53088,
+      "grad_norm": 0.9966936111450195,
+      "learning_rate": 2.887354941976791e-05,
+      "loss": 0.5453,
+      "step": 10571
+    },
+    {
+      "epoch": 13.53216,
+      "grad_norm": 0.9413495659828186,
+      "learning_rate": 2.887154861944778e-05,
+      "loss": 0.5242,
+      "step": 10572
+    },
+    {
+      "epoch": 13.53344,
+      "grad_norm": 0.9562821388244629,
+      "learning_rate": 2.886954781912765e-05,
+      "loss": 0.5672,
+      "step": 10573
+    },
+    {
+      "epoch": 13.53472,
+      "grad_norm": 0.9556413292884827,
+      "learning_rate": 2.8867547018807523e-05,
+      "loss": 0.5895,
+      "step": 10574
+    },
+    {
+      "epoch": 13.536,
+      "grad_norm": 0.9961545467376709,
+      "learning_rate": 2.8865546218487394e-05,
+      "loss": 0.5486,
+      "step": 10575
+    },
+    {
+      "epoch": 13.537279999999999,
+      "grad_norm": 0.9472994208335876,
+      "learning_rate": 2.886354541816727e-05,
+      "loss": 0.5914,
+      "step": 10576
+    },
+    {
+      "epoch": 13.53856,
+      "grad_norm": 0.937106728553772,
+      "learning_rate": 2.886154461784714e-05,
+      "loss": 0.528,
+      "step": 10577
+    },
+    {
+      "epoch": 13.53984,
+      "grad_norm": 0.9636639952659607,
+      "learning_rate": 2.8859543817527013e-05,
+      "loss": 0.5109,
+      "step": 10578
+    },
+    {
+      "epoch": 13.54112,
+      "grad_norm": 0.9437137246131897,
+      "learning_rate": 2.8857543017206885e-05,
+      "loss": 0.5464,
+      "step": 10579
+    },
+    {
+      "epoch": 13.5424,
+      "grad_norm": 1.0153517723083496,
+      "learning_rate": 2.8855542216886754e-05,
+      "loss": 0.5756,
+      "step": 10580
+    },
+    {
+      "epoch": 13.54368,
+      "grad_norm": 0.9016063213348389,
+      "learning_rate": 2.8853541416566626e-05,
+      "loss": 0.4751,
+      "step": 10581
+    },
+    {
+      "epoch": 13.54496,
+      "grad_norm": 0.9339427947998047,
+      "learning_rate": 2.8851540616246497e-05,
+      "loss": 0.5462,
+      "step": 10582
+    },
+    {
+      "epoch": 13.54624,
+      "grad_norm": 0.9382839202880859,
+      "learning_rate": 2.8849539815926373e-05,
+      "loss": 0.5564,
+      "step": 10583
+    },
+    {
+      "epoch": 13.54752,
+      "grad_norm": 0.9853460192680359,
+      "learning_rate": 2.8847539015606244e-05,
+      "loss": 0.5785,
+      "step": 10584
+    },
+    {
+      "epoch": 13.5488,
+      "grad_norm": 0.8800140023231506,
+      "learning_rate": 2.8845538215286116e-05,
+      "loss": 0.5303,
+      "step": 10585
+    },
+    {
+      "epoch": 13.55008,
+      "grad_norm": 0.9256978631019592,
+      "learning_rate": 2.8843537414965988e-05,
+      "loss": 0.5192,
+      "step": 10586
+    },
+    {
+      "epoch": 13.55136,
+      "grad_norm": 0.9170429706573486,
+      "learning_rate": 2.884153661464586e-05,
+      "loss": 0.5438,
+      "step": 10587
+    },
+    {
+      "epoch": 13.55264,
+      "grad_norm": 0.9927389025688171,
+      "learning_rate": 2.883953581432573e-05,
+      "loss": 0.5558,
+      "step": 10588
+    },
+    {
+      "epoch": 13.55392,
+      "grad_norm": 0.9742659330368042,
+      "learning_rate": 2.88375350140056e-05,
+      "loss": 0.5768,
+      "step": 10589
+    },
+    {
+      "epoch": 13.5552,
+      "grad_norm": 1.032018780708313,
+      "learning_rate": 2.883553421368548e-05,
+      "loss": 0.5579,
+      "step": 10590
+    },
+    {
+      "epoch": 13.55648,
+      "grad_norm": 0.972882091999054,
+      "learning_rate": 2.8833533413365347e-05,
+      "loss": 0.5939,
+      "step": 10591
+    },
+    {
+      "epoch": 13.55776,
+      "grad_norm": 0.8545750975608826,
+      "learning_rate": 2.883153261304522e-05,
+      "loss": 0.5027,
+      "step": 10592
+    },
+    {
+      "epoch": 13.55904,
+      "grad_norm": 0.896020233631134,
+      "learning_rate": 2.882953181272509e-05,
+      "loss": 0.472,
+      "step": 10593
+    },
+    {
+      "epoch": 13.56032,
+      "grad_norm": 0.8964064717292786,
+      "learning_rate": 2.8827531012404963e-05,
+      "loss": 0.4956,
+      "step": 10594
+    },
+    {
+      "epoch": 13.5616,
+      "grad_norm": 0.939621090888977,
+      "learning_rate": 2.8825530212084835e-05,
+      "loss": 0.5365,
+      "step": 10595
+    },
+    {
+      "epoch": 13.56288,
+      "grad_norm": 0.9430614709854126,
+      "learning_rate": 2.8823529411764703e-05,
+      "loss": 0.5228,
+      "step": 10596
+    },
+    {
+      "epoch": 13.56416,
+      "grad_norm": 0.91892409324646,
+      "learning_rate": 2.8821528611444582e-05,
+      "loss": 0.5294,
+      "step": 10597
+    },
+    {
+      "epoch": 13.56544,
+      "grad_norm": 0.9037036895751953,
+      "learning_rate": 2.8819527811124454e-05,
+      "loss": 0.5209,
+      "step": 10598
+    },
+    {
+      "epoch": 13.56672,
+      "grad_norm": 0.9548211097717285,
+      "learning_rate": 2.8817527010804322e-05,
+      "loss": 0.5449,
+      "step": 10599
+    },
+    {
+      "epoch": 13.568,
+      "grad_norm": 0.9372276067733765,
+      "learning_rate": 2.8815526210484194e-05,
+      "loss": 0.5343,
+      "step": 10600
+    },
+    {
+      "epoch": 13.56928,
+      "grad_norm": 0.9246383905410767,
+      "learning_rate": 2.8813525410164066e-05,
+      "loss": 0.5319,
+      "step": 10601
+    },
+    {
+      "epoch": 13.57056,
+      "grad_norm": 0.947321355342865,
+      "learning_rate": 2.8811524609843938e-05,
+      "loss": 0.5573,
+      "step": 10602
+    },
+    {
+      "epoch": 13.57184,
+      "grad_norm": 0.9556154012680054,
+      "learning_rate": 2.880952380952381e-05,
+      "loss": 0.5551,
+      "step": 10603
+    },
+    {
+      "epoch": 13.57312,
+      "grad_norm": 0.9768837094306946,
+      "learning_rate": 2.8807523009203685e-05,
+      "loss": 0.5631,
+      "step": 10604
+    },
+    {
+      "epoch": 13.5744,
+      "grad_norm": 0.9723296761512756,
+      "learning_rate": 2.8805522208883557e-05,
+      "loss": 0.5453,
+      "step": 10605
+    },
+    {
+      "epoch": 13.57568,
+      "grad_norm": 0.9709224104881287,
+      "learning_rate": 2.880352140856343e-05,
+      "loss": 0.5324,
+      "step": 10606
+    },
+    {
+      "epoch": 13.57696,
+      "grad_norm": 0.9383196830749512,
+      "learning_rate": 2.8801520608243297e-05,
+      "loss": 0.5561,
+      "step": 10607
+    },
+    {
+      "epoch": 13.57824,
+      "grad_norm": 0.9480939507484436,
+      "learning_rate": 2.879951980792317e-05,
+      "loss": 0.51,
+      "step": 10608
+    },
+    {
+      "epoch": 13.57952,
+      "grad_norm": 0.9381304383277893,
+      "learning_rate": 2.879751900760304e-05,
+      "loss": 0.5704,
+      "step": 10609
+    },
+    {
+      "epoch": 13.5808,
+      "grad_norm": 0.9381782412528992,
+      "learning_rate": 2.8795518207282913e-05,
+      "loss": 0.5011,
+      "step": 10610
+    },
+    {
+      "epoch": 13.58208,
+      "grad_norm": 1.024210810661316,
+      "learning_rate": 2.8793517406962788e-05,
+      "loss": 0.5855,
+      "step": 10611
+    },
+    {
+      "epoch": 13.58336,
+      "grad_norm": 0.9958494305610657,
+      "learning_rate": 2.879151660664266e-05,
+      "loss": 0.5279,
+      "step": 10612
+    },
+    {
+      "epoch": 13.58464,
+      "grad_norm": 0.9755289554595947,
+      "learning_rate": 2.8789515806322532e-05,
+      "loss": 0.5981,
+      "step": 10613
+    },
+    {
+      "epoch": 13.58592,
+      "grad_norm": 0.9563091397285461,
+      "learning_rate": 2.8787515006002404e-05,
+      "loss": 0.5707,
+      "step": 10614
+    },
+    {
+      "epoch": 13.5872,
+      "grad_norm": 0.9441842436790466,
+      "learning_rate": 2.8785514205682272e-05,
+      "loss": 0.5695,
+      "step": 10615
+    },
+    {
+      "epoch": 13.58848,
+      "grad_norm": 0.9176508784294128,
+      "learning_rate": 2.8783513405362144e-05,
+      "loss": 0.4954,
+      "step": 10616
+    },
+    {
+      "epoch": 13.58976,
+      "grad_norm": 0.9186686277389526,
+      "learning_rate": 2.8781512605042016e-05,
+      "loss": 0.5372,
+      "step": 10617
+    },
+    {
+      "epoch": 13.59104,
+      "grad_norm": 0.9438495635986328,
+      "learning_rate": 2.877951180472189e-05,
+      "loss": 0.5552,
+      "step": 10618
+    },
+    {
+      "epoch": 13.59232,
+      "grad_norm": 0.9654157161712646,
+      "learning_rate": 2.8777511004401763e-05,
+      "loss": 0.5388,
+      "step": 10619
+    },
+    {
+      "epoch": 13.5936,
+      "grad_norm": 0.9715597629547119,
+      "learning_rate": 2.8775510204081635e-05,
+      "loss": 0.5435,
+      "step": 10620
+    },
+    {
+      "epoch": 13.59488,
+      "grad_norm": 0.9751953482627869,
+      "learning_rate": 2.8773509403761507e-05,
+      "loss": 0.5689,
+      "step": 10621
+    },
+    {
+      "epoch": 13.59616,
+      "grad_norm": 0.9504429697990417,
+      "learning_rate": 2.877150860344138e-05,
+      "loss": 0.5404,
+      "step": 10622
+    },
+    {
+      "epoch": 13.59744,
+      "grad_norm": 0.9462687373161316,
+      "learning_rate": 2.8769507803121247e-05,
+      "loss": 0.527,
+      "step": 10623
+    },
+    {
+      "epoch": 13.59872,
+      "grad_norm": 0.9122000336647034,
+      "learning_rate": 2.876750700280112e-05,
+      "loss": 0.5466,
+      "step": 10624
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 0.9636383056640625,
+      "learning_rate": 2.8765506202480997e-05,
+      "loss": 0.5357,
+      "step": 10625
+    },
+    {
+      "epoch": 13.60128,
+      "grad_norm": 0.9176141023635864,
+      "learning_rate": 2.8763505402160866e-05,
+      "loss": 0.5333,
+      "step": 10626
+    },
+    {
+      "epoch": 13.60256,
+      "grad_norm": 0.935096025466919,
+      "learning_rate": 2.8761504601840738e-05,
+      "loss": 0.5373,
+      "step": 10627
+    },
+    {
+      "epoch": 13.60384,
+      "grad_norm": 0.9804729223251343,
+      "learning_rate": 2.875950380152061e-05,
+      "loss": 0.5202,
+      "step": 10628
+    },
+    {
+      "epoch": 13.60512,
+      "grad_norm": 0.8921434879302979,
+      "learning_rate": 2.875750300120048e-05,
+      "loss": 0.5015,
+      "step": 10629
+    },
+    {
+      "epoch": 13.6064,
+      "grad_norm": 0.9438501000404358,
+      "learning_rate": 2.8755502200880353e-05,
+      "loss": 0.5353,
+      "step": 10630
+    },
+    {
+      "epoch": 13.60768,
+      "grad_norm": 1.0135152339935303,
+      "learning_rate": 2.8753501400560222e-05,
+      "loss": 0.5581,
+      "step": 10631
+    },
+    {
+      "epoch": 13.60896,
+      "grad_norm": 0.9754303693771362,
+      "learning_rate": 2.87515006002401e-05,
+      "loss": 0.5702,
+      "step": 10632
+    },
+    {
+      "epoch": 13.61024,
+      "grad_norm": 0.9811170101165771,
+      "learning_rate": 2.8749499799919972e-05,
+      "loss": 0.5316,
+      "step": 10633
+    },
+    {
+      "epoch": 13.61152,
+      "grad_norm": 0.9455510377883911,
+      "learning_rate": 2.874749899959984e-05,
+      "loss": 0.5505,
+      "step": 10634
+    },
+    {
+      "epoch": 13.6128,
+      "grad_norm": 0.9648592472076416,
+      "learning_rate": 2.8745498199279713e-05,
+      "loss": 0.5324,
+      "step": 10635
+    },
+    {
+      "epoch": 13.61408,
+      "grad_norm": 0.9389111995697021,
+      "learning_rate": 2.8743497398959584e-05,
+      "loss": 0.5199,
+      "step": 10636
+    },
+    {
+      "epoch": 13.61536,
+      "grad_norm": 0.8730208873748779,
+      "learning_rate": 2.8741496598639456e-05,
+      "loss": 0.4906,
+      "step": 10637
+    },
+    {
+      "epoch": 13.61664,
+      "grad_norm": 0.903998613357544,
+      "learning_rate": 2.8739495798319328e-05,
+      "loss": 0.5274,
+      "step": 10638
+    },
+    {
+      "epoch": 13.61792,
+      "grad_norm": 0.9008133411407471,
+      "learning_rate": 2.8737494997999203e-05,
+      "loss": 0.4929,
+      "step": 10639
+    },
+    {
+      "epoch": 13.6192,
+      "grad_norm": 0.9376186728477478,
+      "learning_rate": 2.8735494197679075e-05,
+      "loss": 0.5746,
+      "step": 10640
+    },
+    {
+      "epoch": 13.62048,
+      "grad_norm": 0.8832375407218933,
+      "learning_rate": 2.8733493397358947e-05,
+      "loss": 0.4694,
+      "step": 10641
+    },
+    {
+      "epoch": 13.62176,
+      "grad_norm": 1.0239812135696411,
+      "learning_rate": 2.8731492597038816e-05,
+      "loss": 0.5459,
+      "step": 10642
+    },
+    {
+      "epoch": 13.62304,
+      "grad_norm": 1.0099740028381348,
+      "learning_rate": 2.8729491796718687e-05,
+      "loss": 0.5721,
+      "step": 10643
+    },
+    {
+      "epoch": 13.62432,
+      "grad_norm": 0.9773842096328735,
+      "learning_rate": 2.872749099639856e-05,
+      "loss": 0.5401,
+      "step": 10644
+    },
+    {
+      "epoch": 13.6256,
+      "grad_norm": 0.9006150364875793,
+      "learning_rate": 2.872549019607843e-05,
+      "loss": 0.5157,
+      "step": 10645
+    },
+    {
+      "epoch": 13.62688,
+      "grad_norm": 0.9231019616127014,
+      "learning_rate": 2.8723489395758306e-05,
+      "loss": 0.5334,
+      "step": 10646
+    },
+    {
+      "epoch": 13.62816,
+      "grad_norm": 0.9919240474700928,
+      "learning_rate": 2.8721488595438178e-05,
+      "loss": 0.5759,
+      "step": 10647
+    },
+    {
+      "epoch": 13.62944,
+      "grad_norm": 0.9904776215553284,
+      "learning_rate": 2.871948779511805e-05,
+      "loss": 0.5377,
+      "step": 10648
+    },
+    {
+      "epoch": 13.63072,
+      "grad_norm": 0.9417052268981934,
+      "learning_rate": 2.8717486994797922e-05,
+      "loss": 0.5053,
+      "step": 10649
+    },
+    {
+      "epoch": 13.632,
+      "grad_norm": 0.9234094023704529,
+      "learning_rate": 2.871548619447779e-05,
+      "loss": 0.5466,
+      "step": 10650
+    },
+    {
+      "epoch": 13.63328,
+      "grad_norm": 0.9561535716056824,
+      "learning_rate": 2.8713485394157662e-05,
+      "loss": 0.5591,
+      "step": 10651
+    },
+    {
+      "epoch": 13.63456,
+      "grad_norm": 0.9605762362480164,
+      "learning_rate": 2.8711484593837534e-05,
+      "loss": 0.5438,
+      "step": 10652
+    },
+    {
+      "epoch": 13.63584,
+      "grad_norm": 1.0227789878845215,
+      "learning_rate": 2.870948379351741e-05,
+      "loss": 0.5768,
+      "step": 10653
+    },
+    {
+      "epoch": 13.63712,
+      "grad_norm": 0.9260333776473999,
+      "learning_rate": 2.870748299319728e-05,
+      "loss": 0.4965,
+      "step": 10654
+    },
+    {
+      "epoch": 13.6384,
+      "grad_norm": 1.0059727430343628,
+      "learning_rate": 2.8705482192877153e-05,
+      "loss": 0.5619,
+      "step": 10655
+    },
+    {
+      "epoch": 13.63968,
+      "grad_norm": 0.9554175138473511,
+      "learning_rate": 2.8703481392557025e-05,
+      "loss": 0.5721,
+      "step": 10656
+    },
+    {
+      "epoch": 13.64096,
+      "grad_norm": 0.9822500944137573,
+      "learning_rate": 2.8701480592236897e-05,
+      "loss": 0.5116,
+      "step": 10657
+    },
+    {
+      "epoch": 13.64224,
+      "grad_norm": 0.9823669195175171,
+      "learning_rate": 2.8699479791916765e-05,
+      "loss": 0.5623,
+      "step": 10658
+    },
+    {
+      "epoch": 13.64352,
+      "grad_norm": 0.962337076663971,
+      "learning_rate": 2.8697478991596637e-05,
+      "loss": 0.5153,
+      "step": 10659
+    },
+    {
+      "epoch": 13.6448,
+      "grad_norm": 0.9577946066856384,
+      "learning_rate": 2.8695478191276516e-05,
+      "loss": 0.546,
+      "step": 10660
+    },
+    {
+      "epoch": 13.64608,
+      "grad_norm": 0.9783328771591187,
+      "learning_rate": 2.8693477390956384e-05,
+      "loss": 0.5556,
+      "step": 10661
+    },
+    {
+      "epoch": 13.64736,
+      "grad_norm": 0.9406751990318298,
+      "learning_rate": 2.8691476590636256e-05,
+      "loss": 0.5234,
+      "step": 10662
+    },
+    {
+      "epoch": 13.64864,
+      "grad_norm": 0.9680593013763428,
+      "learning_rate": 2.8689475790316128e-05,
+      "loss": 0.5553,
+      "step": 10663
+    },
+    {
+      "epoch": 13.64992,
+      "grad_norm": 0.9520314335823059,
+      "learning_rate": 2.8687474989996e-05,
+      "loss": 0.5647,
+      "step": 10664
+    },
+    {
+      "epoch": 13.6512,
+      "grad_norm": 0.8977915048599243,
+      "learning_rate": 2.868547418967587e-05,
+      "loss": 0.5178,
+      "step": 10665
+    },
+    {
+      "epoch": 13.65248,
+      "grad_norm": 0.8997889757156372,
+      "learning_rate": 2.868347338935574e-05,
+      "loss": 0.5383,
+      "step": 10666
+    },
+    {
+      "epoch": 13.65376,
+      "grad_norm": 0.9149733781814575,
+      "learning_rate": 2.8681472589035612e-05,
+      "loss": 0.5046,
+      "step": 10667
+    },
+    {
+      "epoch": 13.65504,
+      "grad_norm": 1.018157720565796,
+      "learning_rate": 2.867947178871549e-05,
+      "loss": 0.6097,
+      "step": 10668
+    },
+    {
+      "epoch": 13.656320000000001,
+      "grad_norm": 0.9306768774986267,
+      "learning_rate": 2.867747098839536e-05,
+      "loss": 0.5142,
+      "step": 10669
+    },
+    {
+      "epoch": 13.6576,
+      "grad_norm": 0.9335023760795593,
+      "learning_rate": 2.867547018807523e-05,
+      "loss": 0.5604,
+      "step": 10670
+    },
+    {
+      "epoch": 13.65888,
+      "grad_norm": 0.9208192825317383,
+      "learning_rate": 2.8673469387755103e-05,
+      "loss": 0.5307,
+      "step": 10671
+    },
+    {
+      "epoch": 13.66016,
+      "grad_norm": 0.9682392477989197,
+      "learning_rate": 2.8671468587434975e-05,
+      "loss": 0.5418,
+      "step": 10672
+    },
+    {
+      "epoch": 13.66144,
+      "grad_norm": 0.9635206460952759,
+      "learning_rate": 2.8669467787114846e-05,
+      "loss": 0.5639,
+      "step": 10673
+    },
+    {
+      "epoch": 13.66272,
+      "grad_norm": 0.9996086359024048,
+      "learning_rate": 2.8667466986794715e-05,
+      "loss": 0.5511,
+      "step": 10674
+    },
+    {
+      "epoch": 13.664,
+      "grad_norm": 0.9767386317253113,
+      "learning_rate": 2.8665466186474594e-05,
+      "loss": 0.5375,
+      "step": 10675
+    },
+    {
+      "epoch": 13.66528,
+      "grad_norm": 0.9239612221717834,
+      "learning_rate": 2.8663465386154465e-05,
+      "loss": 0.528,
+      "step": 10676
+    },
+    {
+      "epoch": 13.66656,
+      "grad_norm": 0.9513071179389954,
+      "learning_rate": 2.8661464585834334e-05,
+      "loss": 0.5234,
+      "step": 10677
+    },
+    {
+      "epoch": 13.66784,
+      "grad_norm": 1.0107206106185913,
+      "learning_rate": 2.8659463785514206e-05,
+      "loss": 0.564,
+      "step": 10678
+    },
+    {
+      "epoch": 13.66912,
+      "grad_norm": 0.9832781553268433,
+      "learning_rate": 2.8657462985194078e-05,
+      "loss": 0.5724,
+      "step": 10679
+    },
+    {
+      "epoch": 13.6704,
+      "grad_norm": 0.9765233993530273,
+      "learning_rate": 2.865546218487395e-05,
+      "loss": 0.5414,
+      "step": 10680
+    },
+    {
+      "epoch": 13.67168,
+      "grad_norm": 0.9487534761428833,
+      "learning_rate": 2.865346138455382e-05,
+      "loss": 0.4998,
+      "step": 10681
+    },
+    {
+      "epoch": 13.67296,
+      "grad_norm": 0.9580942988395691,
+      "learning_rate": 2.8651460584233697e-05,
+      "loss": 0.5587,
+      "step": 10682
+    },
+    {
+      "epoch": 13.67424,
+      "grad_norm": 0.9719009399414062,
+      "learning_rate": 2.864945978391357e-05,
+      "loss": 0.5611,
+      "step": 10683
+    },
+    {
+      "epoch": 13.67552,
+      "grad_norm": 0.9432852864265442,
+      "learning_rate": 2.864745898359344e-05,
+      "loss": 0.5628,
+      "step": 10684
+    },
+    {
+      "epoch": 13.6768,
+      "grad_norm": 0.9318643808364868,
+      "learning_rate": 2.864545818327331e-05,
+      "loss": 0.5335,
+      "step": 10685
+    },
+    {
+      "epoch": 13.67808,
+      "grad_norm": 0.9718388915061951,
+      "learning_rate": 2.864345738295318e-05,
+      "loss": 0.5531,
+      "step": 10686
+    },
+    {
+      "epoch": 13.679359999999999,
+      "grad_norm": 0.9416543245315552,
+      "learning_rate": 2.8641456582633052e-05,
+      "loss": 0.5295,
+      "step": 10687
+    },
+    {
+      "epoch": 13.68064,
+      "grad_norm": 0.9783507585525513,
+      "learning_rate": 2.8639455782312924e-05,
+      "loss": 0.5539,
+      "step": 10688
+    },
+    {
+      "epoch": 13.68192,
+      "grad_norm": 0.9382901787757874,
+      "learning_rate": 2.86374549819928e-05,
+      "loss": 0.5215,
+      "step": 10689
+    },
+    {
+      "epoch": 13.6832,
+      "grad_norm": 0.9459569454193115,
+      "learning_rate": 2.863545418167267e-05,
+      "loss": 0.5271,
+      "step": 10690
+    },
+    {
+      "epoch": 13.68448,
+      "grad_norm": 0.9002374410629272,
+      "learning_rate": 2.8633453381352543e-05,
+      "loss": 0.5393,
+      "step": 10691
+    },
+    {
+      "epoch": 13.68576,
+      "grad_norm": 0.9636310338973999,
+      "learning_rate": 2.8631452581032415e-05,
+      "loss": 0.5206,
+      "step": 10692
+    },
+    {
+      "epoch": 13.68704,
+      "grad_norm": 0.9752256274223328,
+      "learning_rate": 2.8629451780712284e-05,
+      "loss": 0.578,
+      "step": 10693
+    },
+    {
+      "epoch": 13.688320000000001,
+      "grad_norm": 0.9666173458099365,
+      "learning_rate": 2.8627450980392155e-05,
+      "loss": 0.5931,
+      "step": 10694
+    },
+    {
+      "epoch": 13.6896,
+      "grad_norm": 0.9382612109184265,
+      "learning_rate": 2.8625450180072027e-05,
+      "loss": 0.5502,
+      "step": 10695
+    },
+    {
+      "epoch": 13.69088,
+      "grad_norm": 0.9369370341300964,
+      "learning_rate": 2.8623449379751906e-05,
+      "loss": 0.5421,
+      "step": 10696
+    },
+    {
+      "epoch": 13.69216,
+      "grad_norm": 0.8949071764945984,
+      "learning_rate": 2.8621448579431774e-05,
+      "loss": 0.5083,
+      "step": 10697
+    },
+    {
+      "epoch": 13.69344,
+      "grad_norm": 0.9892618656158447,
+      "learning_rate": 2.8619447779111646e-05,
+      "loss": 0.5519,
+      "step": 10698
+    },
+    {
+      "epoch": 13.69472,
+      "grad_norm": 0.8985315561294556,
+      "learning_rate": 2.8617446978791518e-05,
+      "loss": 0.4751,
+      "step": 10699
+    },
+    {
+      "epoch": 13.696,
+      "grad_norm": 0.9138407707214355,
+      "learning_rate": 2.861544617847139e-05,
+      "loss": 0.5034,
+      "step": 10700
+    },
+    {
+      "epoch": 13.69728,
+      "grad_norm": 0.8925737142562866,
+      "learning_rate": 2.861344537815126e-05,
+      "loss": 0.5013,
+      "step": 10701
+    },
+    {
+      "epoch": 13.69856,
+      "grad_norm": 0.9540265798568726,
+      "learning_rate": 2.861144457783113e-05,
+      "loss": 0.5579,
+      "step": 10702
+    },
+    {
+      "epoch": 13.69984,
+      "grad_norm": 0.9697530269622803,
+      "learning_rate": 2.860944377751101e-05,
+      "loss": 0.527,
+      "step": 10703
+    },
+    {
+      "epoch": 13.70112,
+      "grad_norm": 0.999180257320404,
+      "learning_rate": 2.860744297719088e-05,
+      "loss": 0.5523,
+      "step": 10704
+    },
+    {
+      "epoch": 13.7024,
+      "grad_norm": 0.9945990443229675,
+      "learning_rate": 2.860544217687075e-05,
+      "loss": 0.5717,
+      "step": 10705
+    },
+    {
+      "epoch": 13.70368,
+      "grad_norm": 0.9230750203132629,
+      "learning_rate": 2.860344137655062e-05,
+      "loss": 0.5279,
+      "step": 10706
+    },
+    {
+      "epoch": 13.70496,
+      "grad_norm": 0.9194453954696655,
+      "learning_rate": 2.8601440576230493e-05,
+      "loss": 0.5262,
+      "step": 10707
+    },
+    {
+      "epoch": 13.70624,
+      "grad_norm": 0.9214555621147156,
+      "learning_rate": 2.8599439775910365e-05,
+      "loss": 0.5166,
+      "step": 10708
+    },
+    {
+      "epoch": 13.70752,
+      "grad_norm": 1.0233759880065918,
+      "learning_rate": 2.8597438975590233e-05,
+      "loss": 0.6114,
+      "step": 10709
+    },
+    {
+      "epoch": 13.7088,
+      "grad_norm": 0.8889565467834473,
+      "learning_rate": 2.8595438175270112e-05,
+      "loss": 0.4844,
+      "step": 10710
+    },
+    {
+      "epoch": 13.71008,
+      "grad_norm": 0.9614165425300598,
+      "learning_rate": 2.8593437374949984e-05,
+      "loss": 0.5454,
+      "step": 10711
+    },
+    {
+      "epoch": 13.711359999999999,
+      "grad_norm": 0.9475215673446655,
+      "learning_rate": 2.8591436574629856e-05,
+      "loss": 0.4936,
+      "step": 10712
+    },
+    {
+      "epoch": 13.71264,
+      "grad_norm": 1.000278353691101,
+      "learning_rate": 2.8589435774309724e-05,
+      "loss": 0.5549,
+      "step": 10713
+    },
+    {
+      "epoch": 13.71392,
+      "grad_norm": 0.918251097202301,
+      "learning_rate": 2.8587434973989596e-05,
+      "loss": 0.5023,
+      "step": 10714
+    },
+    {
+      "epoch": 13.7152,
+      "grad_norm": 0.9441922903060913,
+      "learning_rate": 2.8585434173669468e-05,
+      "loss": 0.5331,
+      "step": 10715
+    },
+    {
+      "epoch": 13.71648,
+      "grad_norm": 0.9622296690940857,
+      "learning_rate": 2.858343337334934e-05,
+      "loss": 0.5316,
+      "step": 10716
+    },
+    {
+      "epoch": 13.71776,
+      "grad_norm": 0.9810208678245544,
+      "learning_rate": 2.8581432573029215e-05,
+      "loss": 0.5379,
+      "step": 10717
+    },
+    {
+      "epoch": 13.71904,
+      "grad_norm": 0.9779439568519592,
+      "learning_rate": 2.8579431772709087e-05,
+      "loss": 0.5698,
+      "step": 10718
+    },
+    {
+      "epoch": 13.72032,
+      "grad_norm": 0.9007058143615723,
+      "learning_rate": 2.857743097238896e-05,
+      "loss": 0.484,
+      "step": 10719
+    },
+    {
+      "epoch": 13.7216,
+      "grad_norm": 0.964832067489624,
+      "learning_rate": 2.857543017206883e-05,
+      "loss": 0.5773,
+      "step": 10720
+    },
+    {
+      "epoch": 13.72288,
+      "grad_norm": 0.9448238015174866,
+      "learning_rate": 2.85734293717487e-05,
+      "loss": 0.5122,
+      "step": 10721
+    },
+    {
+      "epoch": 13.72416,
+      "grad_norm": 0.9428820610046387,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.5472,
+      "step": 10722
+    },
+    {
+      "epoch": 13.72544,
+      "grad_norm": 0.9774311184883118,
+      "learning_rate": 2.8569427771108443e-05,
+      "loss": 0.5224,
+      "step": 10723
+    },
+    {
+      "epoch": 13.72672,
+      "grad_norm": 0.9833608269691467,
+      "learning_rate": 2.8567426970788318e-05,
+      "loss": 0.5713,
+      "step": 10724
+    },
+    {
+      "epoch": 13.728,
+      "grad_norm": 0.906108558177948,
+      "learning_rate": 2.856542617046819e-05,
+      "loss": 0.5289,
+      "step": 10725
+    },
+    {
+      "epoch": 13.72928,
+      "grad_norm": 0.9727580547332764,
+      "learning_rate": 2.856342537014806e-05,
+      "loss": 0.5354,
+      "step": 10726
+    },
+    {
+      "epoch": 13.73056,
+      "grad_norm": 0.9792339205741882,
+      "learning_rate": 2.8561424569827934e-05,
+      "loss": 0.5648,
+      "step": 10727
+    },
+    {
+      "epoch": 13.73184,
+      "grad_norm": 0.962765634059906,
+      "learning_rate": 2.8559423769507805e-05,
+      "loss": 0.5599,
+      "step": 10728
+    },
+    {
+      "epoch": 13.73312,
+      "grad_norm": 0.9317491054534912,
+      "learning_rate": 2.8557422969187674e-05,
+      "loss": 0.5203,
+      "step": 10729
+    },
+    {
+      "epoch": 13.7344,
+      "grad_norm": 0.9817516207695007,
+      "learning_rate": 2.8555422168867546e-05,
+      "loss": 0.5766,
+      "step": 10730
+    },
+    {
+      "epoch": 13.73568,
+      "grad_norm": 0.9059588313102722,
+      "learning_rate": 2.8553421368547424e-05,
+      "loss": 0.5134,
+      "step": 10731
+    },
+    {
+      "epoch": 13.73696,
+      "grad_norm": 0.9380925893783569,
+      "learning_rate": 2.8551420568227293e-05,
+      "loss": 0.5188,
+      "step": 10732
+    },
+    {
+      "epoch": 13.73824,
+      "grad_norm": 0.919386088848114,
+      "learning_rate": 2.8549419767907165e-05,
+      "loss": 0.5315,
+      "step": 10733
+    },
+    {
+      "epoch": 13.73952,
+      "grad_norm": 0.991089403629303,
+      "learning_rate": 2.8547418967587037e-05,
+      "loss": 0.6093,
+      "step": 10734
+    },
+    {
+      "epoch": 13.7408,
+      "grad_norm": 0.9908062219619751,
+      "learning_rate": 2.854541816726691e-05,
+      "loss": 0.5627,
+      "step": 10735
+    },
+    {
+      "epoch": 13.74208,
+      "grad_norm": 0.9567424654960632,
+      "learning_rate": 2.854341736694678e-05,
+      "loss": 0.5442,
+      "step": 10736
+    },
+    {
+      "epoch": 13.74336,
+      "grad_norm": 0.9408556818962097,
+      "learning_rate": 2.854141656662665e-05,
+      "loss": 0.5286,
+      "step": 10737
+    },
+    {
+      "epoch": 13.74464,
+      "grad_norm": 0.9387040734291077,
+      "learning_rate": 2.8539415766306527e-05,
+      "loss": 0.548,
+      "step": 10738
+    },
+    {
+      "epoch": 13.74592,
+      "grad_norm": 0.921125054359436,
+      "learning_rate": 2.85374149659864e-05,
+      "loss": 0.5613,
+      "step": 10739
+    },
+    {
+      "epoch": 13.7472,
+      "grad_norm": 0.9063670635223389,
+      "learning_rate": 2.8535414165666268e-05,
+      "loss": 0.5457,
+      "step": 10740
+    },
+    {
+      "epoch": 13.74848,
+      "grad_norm": 1.002338171005249,
+      "learning_rate": 2.853341336534614e-05,
+      "loss": 0.5935,
+      "step": 10741
+    },
+    {
+      "epoch": 13.74976,
+      "grad_norm": 0.8982512950897217,
+      "learning_rate": 2.853141256502601e-05,
+      "loss": 0.5346,
+      "step": 10742
+    },
+    {
+      "epoch": 13.75104,
+      "grad_norm": 0.8786603808403015,
+      "learning_rate": 2.8529411764705883e-05,
+      "loss": 0.5381,
+      "step": 10743
+    },
+    {
+      "epoch": 13.75232,
+      "grad_norm": 0.9155345559120178,
+      "learning_rate": 2.8527410964385755e-05,
+      "loss": 0.5217,
+      "step": 10744
+    },
+    {
+      "epoch": 13.7536,
+      "grad_norm": 0.9769365787506104,
+      "learning_rate": 2.852541016406563e-05,
+      "loss": 0.5838,
+      "step": 10745
+    },
+    {
+      "epoch": 13.75488,
+      "grad_norm": 0.9437276124954224,
+      "learning_rate": 2.8523409363745502e-05,
+      "loss": 0.5727,
+      "step": 10746
+    },
+    {
+      "epoch": 13.75616,
+      "grad_norm": 0.9436963200569153,
+      "learning_rate": 2.8521408563425374e-05,
+      "loss": 0.5224,
+      "step": 10747
+    },
+    {
+      "epoch": 13.75744,
+      "grad_norm": 0.9029295444488525,
+      "learning_rate": 2.8519407763105243e-05,
+      "loss": 0.5258,
+      "step": 10748
+    },
+    {
+      "epoch": 13.75872,
+      "grad_norm": 0.9773640036582947,
+      "learning_rate": 2.8517406962785114e-05,
+      "loss": 0.5447,
+      "step": 10749
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 0.9137765169143677,
+      "learning_rate": 2.8515406162464986e-05,
+      "loss": 0.543,
+      "step": 10750
+    },
+    {
+      "epoch": 13.76128,
+      "grad_norm": 0.947670042514801,
+      "learning_rate": 2.8513405362144858e-05,
+      "loss": 0.5234,
+      "step": 10751
+    },
+    {
+      "epoch": 13.76256,
+      "grad_norm": 0.9233437776565552,
+      "learning_rate": 2.8511404561824733e-05,
+      "loss": 0.5104,
+      "step": 10752
+    },
+    {
+      "epoch": 13.76384,
+      "grad_norm": 0.9210999011993408,
+      "learning_rate": 2.8509403761504605e-05,
+      "loss": 0.5731,
+      "step": 10753
+    },
+    {
+      "epoch": 13.76512,
+      "grad_norm": 0.9368586540222168,
+      "learning_rate": 2.8507402961184477e-05,
+      "loss": 0.5448,
+      "step": 10754
+    },
+    {
+      "epoch": 13.7664,
+      "grad_norm": 0.933786153793335,
+      "learning_rate": 2.850540216086435e-05,
+      "loss": 0.5324,
+      "step": 10755
+    },
+    {
+      "epoch": 13.76768,
+      "grad_norm": 0.8887897729873657,
+      "learning_rate": 2.8503401360544217e-05,
+      "loss": 0.5083,
+      "step": 10756
+    },
+    {
+      "epoch": 13.76896,
+      "grad_norm": 0.9025307893753052,
+      "learning_rate": 2.850140056022409e-05,
+      "loss": 0.5212,
+      "step": 10757
+    },
+    {
+      "epoch": 13.77024,
+      "grad_norm": 0.936424195766449,
+      "learning_rate": 2.849939975990396e-05,
+      "loss": 0.5719,
+      "step": 10758
+    },
+    {
+      "epoch": 13.77152,
+      "grad_norm": 0.936571478843689,
+      "learning_rate": 2.8497398959583836e-05,
+      "loss": 0.5558,
+      "step": 10759
+    },
+    {
+      "epoch": 13.7728,
+      "grad_norm": 0.9100072979927063,
+      "learning_rate": 2.8495398159263708e-05,
+      "loss": 0.5169,
+      "step": 10760
+    },
+    {
+      "epoch": 13.77408,
+      "grad_norm": 0.8884329199790955,
+      "learning_rate": 2.849339735894358e-05,
+      "loss": 0.5174,
+      "step": 10761
+    },
+    {
+      "epoch": 13.77536,
+      "grad_norm": 0.907352864742279,
+      "learning_rate": 2.8491396558623452e-05,
+      "loss": 0.5166,
+      "step": 10762
+    },
+    {
+      "epoch": 13.77664,
+      "grad_norm": 0.929191529750824,
+      "learning_rate": 2.8489395758303324e-05,
+      "loss": 0.5554,
+      "step": 10763
+    },
+    {
+      "epoch": 13.77792,
+      "grad_norm": 0.8757056593894958,
+      "learning_rate": 2.8487394957983192e-05,
+      "loss": 0.4899,
+      "step": 10764
+    },
+    {
+      "epoch": 13.7792,
+      "grad_norm": 0.9026574492454529,
+      "learning_rate": 2.8485394157663064e-05,
+      "loss": 0.4836,
+      "step": 10765
+    },
+    {
+      "epoch": 13.78048,
+      "grad_norm": 0.9860047101974487,
+      "learning_rate": 2.8483393357342943e-05,
+      "loss": 0.5633,
+      "step": 10766
+    },
+    {
+      "epoch": 13.78176,
+      "grad_norm": 0.8981946110725403,
+      "learning_rate": 2.848139255702281e-05,
+      "loss": 0.5248,
+      "step": 10767
+    },
+    {
+      "epoch": 13.78304,
+      "grad_norm": 0.8830896019935608,
+      "learning_rate": 2.8479391756702683e-05,
+      "loss": 0.5086,
+      "step": 10768
+    },
+    {
+      "epoch": 13.78432,
+      "grad_norm": 0.9160583019256592,
+      "learning_rate": 2.8477390956382555e-05,
+      "loss": 0.5561,
+      "step": 10769
+    },
+    {
+      "epoch": 13.7856,
+      "grad_norm": 0.9723164439201355,
+      "learning_rate": 2.8475390156062427e-05,
+      "loss": 0.6023,
+      "step": 10770
+    },
+    {
+      "epoch": 13.78688,
+      "grad_norm": 0.9493298530578613,
+      "learning_rate": 2.84733893557423e-05,
+      "loss": 0.5237,
+      "step": 10771
+    },
+    {
+      "epoch": 13.78816,
+      "grad_norm": 0.9569862484931946,
+      "learning_rate": 2.8471388555422167e-05,
+      "loss": 0.5436,
+      "step": 10772
+    },
+    {
+      "epoch": 13.78944,
+      "grad_norm": 0.9514901638031006,
+      "learning_rate": 2.8469387755102046e-05,
+      "loss": 0.5559,
+      "step": 10773
+    },
+    {
+      "epoch": 13.79072,
+      "grad_norm": 0.914356529712677,
+      "learning_rate": 2.8467386954781918e-05,
+      "loss": 0.5284,
+      "step": 10774
+    },
+    {
+      "epoch": 13.792,
+      "grad_norm": 0.8957856297492981,
+      "learning_rate": 2.8465386154461786e-05,
+      "loss": 0.51,
+      "step": 10775
+    },
+    {
+      "epoch": 13.79328,
+      "grad_norm": 0.9669928550720215,
+      "learning_rate": 2.8463385354141658e-05,
+      "loss": 0.5284,
+      "step": 10776
+    },
+    {
+      "epoch": 13.79456,
+      "grad_norm": 0.9488115310668945,
+      "learning_rate": 2.846138455382153e-05,
+      "loss": 0.5306,
+      "step": 10777
+    },
+    {
+      "epoch": 13.79584,
+      "grad_norm": 0.9263489246368408,
+      "learning_rate": 2.84593837535014e-05,
+      "loss": 0.4877,
+      "step": 10778
+    },
+    {
+      "epoch": 13.79712,
+      "grad_norm": 0.9073372483253479,
+      "learning_rate": 2.8457382953181273e-05,
+      "loss": 0.5139,
+      "step": 10779
+    },
+    {
+      "epoch": 13.7984,
+      "grad_norm": 0.9395866990089417,
+      "learning_rate": 2.8455382152861142e-05,
+      "loss": 0.5548,
+      "step": 10780
+    },
+    {
+      "epoch": 13.79968,
+      "grad_norm": 0.9865780472755432,
+      "learning_rate": 2.845338135254102e-05,
+      "loss": 0.5339,
+      "step": 10781
+    },
+    {
+      "epoch": 13.80096,
+      "grad_norm": 1.0401817560195923,
+      "learning_rate": 2.8451380552220892e-05,
+      "loss": 0.5783,
+      "step": 10782
+    },
+    {
+      "epoch": 13.80224,
+      "grad_norm": 0.9715459942817688,
+      "learning_rate": 2.844937975190076e-05,
+      "loss": 0.5637,
+      "step": 10783
+    },
+    {
+      "epoch": 13.80352,
+      "grad_norm": 0.9215720891952515,
+      "learning_rate": 2.8447378951580633e-05,
+      "loss": 0.5476,
+      "step": 10784
+    },
+    {
+      "epoch": 13.8048,
+      "grad_norm": 0.9105191826820374,
+      "learning_rate": 2.8445378151260505e-05,
+      "loss": 0.5545,
+      "step": 10785
+    },
+    {
+      "epoch": 13.80608,
+      "grad_norm": 0.9285398125648499,
+      "learning_rate": 2.8443377350940376e-05,
+      "loss": 0.5758,
+      "step": 10786
+    },
+    {
+      "epoch": 13.80736,
+      "grad_norm": 0.9510117173194885,
+      "learning_rate": 2.844137655062025e-05,
+      "loss": 0.5287,
+      "step": 10787
+    },
+    {
+      "epoch": 13.80864,
+      "grad_norm": 0.9962491989135742,
+      "learning_rate": 2.8439375750300124e-05,
+      "loss": 0.6091,
+      "step": 10788
+    },
+    {
+      "epoch": 13.80992,
+      "grad_norm": 0.9605426788330078,
+      "learning_rate": 2.8437374949979995e-05,
+      "loss": 0.5504,
+      "step": 10789
+    },
+    {
+      "epoch": 13.8112,
+      "grad_norm": 0.9639581441879272,
+      "learning_rate": 2.8435374149659867e-05,
+      "loss": 0.5177,
+      "step": 10790
+    },
+    {
+      "epoch": 13.81248,
+      "grad_norm": 0.9726257920265198,
+      "learning_rate": 2.8433373349339736e-05,
+      "loss": 0.5312,
+      "step": 10791
+    },
+    {
+      "epoch": 13.81376,
+      "grad_norm": 0.9021669030189514,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.4934,
+      "step": 10792
+    },
+    {
+      "epoch": 13.81504,
+      "grad_norm": 0.9048178195953369,
+      "learning_rate": 2.842937174869948e-05,
+      "loss": 0.543,
+      "step": 10793
+    },
+    {
+      "epoch": 13.81632,
+      "grad_norm": 0.9584506750106812,
+      "learning_rate": 2.842737094837935e-05,
+      "loss": 0.5404,
+      "step": 10794
+    },
+    {
+      "epoch": 13.8176,
+      "grad_norm": 0.9696805477142334,
+      "learning_rate": 2.8425370148059227e-05,
+      "loss": 0.5502,
+      "step": 10795
+    },
+    {
+      "epoch": 13.81888,
+      "grad_norm": 0.9749921560287476,
+      "learning_rate": 2.84233693477391e-05,
+      "loss": 0.5678,
+      "step": 10796
+    },
+    {
+      "epoch": 13.82016,
+      "grad_norm": 0.959991991519928,
+      "learning_rate": 2.842136854741897e-05,
+      "loss": 0.4897,
+      "step": 10797
+    },
+    {
+      "epoch": 13.821439999999999,
+      "grad_norm": 0.9816640019416809,
+      "learning_rate": 2.8419367747098842e-05,
+      "loss": 0.5713,
+      "step": 10798
+    },
+    {
+      "epoch": 13.82272,
+      "grad_norm": 0.92801833152771,
+      "learning_rate": 2.841736694677871e-05,
+      "loss": 0.4937,
+      "step": 10799
+    },
+    {
+      "epoch": 13.824,
+      "grad_norm": 0.9041526317596436,
+      "learning_rate": 2.8415366146458582e-05,
+      "loss": 0.4989,
+      "step": 10800
+    },
+    {
+      "epoch": 13.82528,
+      "grad_norm": 0.9021725654602051,
+      "learning_rate": 2.8413365346138454e-05,
+      "loss": 0.5101,
+      "step": 10801
+    },
+    {
+      "epoch": 13.82656,
+      "grad_norm": 0.9045724272727966,
+      "learning_rate": 2.841136454581833e-05,
+      "loss": 0.5168,
+      "step": 10802
+    },
+    {
+      "epoch": 13.82784,
+      "grad_norm": 0.9894302487373352,
+      "learning_rate": 2.84093637454982e-05,
+      "loss": 0.5861,
+      "step": 10803
+    },
+    {
+      "epoch": 13.82912,
+      "grad_norm": 0.9526990652084351,
+      "learning_rate": 2.8407362945178073e-05,
+      "loss": 0.5543,
+      "step": 10804
+    },
+    {
+      "epoch": 13.830400000000001,
+      "grad_norm": 0.9203627109527588,
+      "learning_rate": 2.8405362144857945e-05,
+      "loss": 0.5118,
+      "step": 10805
+    },
+    {
+      "epoch": 13.83168,
+      "grad_norm": 0.9406778812408447,
+      "learning_rate": 2.8403361344537817e-05,
+      "loss": 0.5656,
+      "step": 10806
+    },
+    {
+      "epoch": 13.83296,
+      "grad_norm": 0.959984540939331,
+      "learning_rate": 2.8401360544217685e-05,
+      "loss": 0.5362,
+      "step": 10807
+    },
+    {
+      "epoch": 13.83424,
+      "grad_norm": 0.9703862071037292,
+      "learning_rate": 2.8399359743897557e-05,
+      "loss": 0.5502,
+      "step": 10808
+    },
+    {
+      "epoch": 13.83552,
+      "grad_norm": 0.9299468994140625,
+      "learning_rate": 2.8397358943577436e-05,
+      "loss": 0.5166,
+      "step": 10809
+    },
+    {
+      "epoch": 13.8368,
+      "grad_norm": 0.9478683471679688,
+      "learning_rate": 2.8395358143257304e-05,
+      "loss": 0.5688,
+      "step": 10810
+    },
+    {
+      "epoch": 13.83808,
+      "grad_norm": 0.9263501763343811,
+      "learning_rate": 2.8393357342937176e-05,
+      "loss": 0.5294,
+      "step": 10811
+    },
+    {
+      "epoch": 13.83936,
+      "grad_norm": 0.9850642085075378,
+      "learning_rate": 2.8391356542617048e-05,
+      "loss": 0.5819,
+      "step": 10812
+    },
+    {
+      "epoch": 13.84064,
+      "grad_norm": 0.9924013614654541,
+      "learning_rate": 2.838935574229692e-05,
+      "loss": 0.5516,
+      "step": 10813
+    },
+    {
+      "epoch": 13.84192,
+      "grad_norm": 0.9259412288665771,
+      "learning_rate": 2.8387354941976792e-05,
+      "loss": 0.5034,
+      "step": 10814
+    },
+    {
+      "epoch": 13.8432,
+      "grad_norm": 0.9189561605453491,
+      "learning_rate": 2.838535414165666e-05,
+      "loss": 0.5612,
+      "step": 10815
+    },
+    {
+      "epoch": 13.84448,
+      "grad_norm": 0.9896423816680908,
+      "learning_rate": 2.838335334133654e-05,
+      "loss": 0.5521,
+      "step": 10816
+    },
+    {
+      "epoch": 13.84576,
+      "grad_norm": 0.9858946204185486,
+      "learning_rate": 2.838135254101641e-05,
+      "loss": 0.5713,
+      "step": 10817
+    },
+    {
+      "epoch": 13.84704,
+      "grad_norm": 0.9236769080162048,
+      "learning_rate": 2.837935174069628e-05,
+      "loss": 0.5443,
+      "step": 10818
+    },
+    {
+      "epoch": 13.84832,
+      "grad_norm": 0.9041772484779358,
+      "learning_rate": 2.837735094037615e-05,
+      "loss": 0.5474,
+      "step": 10819
+    },
+    {
+      "epoch": 13.8496,
+      "grad_norm": 0.9847750663757324,
+      "learning_rate": 2.8375350140056023e-05,
+      "loss": 0.5703,
+      "step": 10820
+    },
+    {
+      "epoch": 13.85088,
+      "grad_norm": 0.9936345815658569,
+      "learning_rate": 2.8373349339735895e-05,
+      "loss": 0.5808,
+      "step": 10821
+    },
+    {
+      "epoch": 13.85216,
+      "grad_norm": 0.9457690715789795,
+      "learning_rate": 2.8371348539415767e-05,
+      "loss": 0.542,
+      "step": 10822
+    },
+    {
+      "epoch": 13.853439999999999,
+      "grad_norm": 0.9375315308570862,
+      "learning_rate": 2.8369347739095642e-05,
+      "loss": 0.5443,
+      "step": 10823
+    },
+    {
+      "epoch": 13.85472,
+      "grad_norm": 0.9101528525352478,
+      "learning_rate": 2.8367346938775514e-05,
+      "loss": 0.5034,
+      "step": 10824
+    },
+    {
+      "epoch": 13.856,
+      "grad_norm": 0.938879668712616,
+      "learning_rate": 2.8365346138455386e-05,
+      "loss": 0.5565,
+      "step": 10825
+    },
+    {
+      "epoch": 13.85728,
+      "grad_norm": 0.8987931609153748,
+      "learning_rate": 2.8363345338135254e-05,
+      "loss": 0.5148,
+      "step": 10826
+    },
+    {
+      "epoch": 13.85856,
+      "grad_norm": 0.9683275818824768,
+      "learning_rate": 2.8361344537815126e-05,
+      "loss": 0.5337,
+      "step": 10827
+    },
+    {
+      "epoch": 13.85984,
+      "grad_norm": 0.9831743240356445,
+      "learning_rate": 2.8359343737494998e-05,
+      "loss": 0.5295,
+      "step": 10828
+    },
+    {
+      "epoch": 13.86112,
+      "grad_norm": 0.9588016867637634,
+      "learning_rate": 2.835734293717487e-05,
+      "loss": 0.5394,
+      "step": 10829
+    },
+    {
+      "epoch": 13.862400000000001,
+      "grad_norm": 0.9667535424232483,
+      "learning_rate": 2.8355342136854745e-05,
+      "loss": 0.5661,
+      "step": 10830
+    },
+    {
+      "epoch": 13.86368,
+      "grad_norm": 0.9933251738548279,
+      "learning_rate": 2.8353341336534617e-05,
+      "loss": 0.5859,
+      "step": 10831
+    },
+    {
+      "epoch": 13.86496,
+      "grad_norm": 1.0498590469360352,
+      "learning_rate": 2.835134053621449e-05,
+      "loss": 0.6109,
+      "step": 10832
+    },
+    {
+      "epoch": 13.86624,
+      "grad_norm": 0.929767370223999,
+      "learning_rate": 2.834933973589436e-05,
+      "loss": 0.5512,
+      "step": 10833
+    },
+    {
+      "epoch": 13.86752,
+      "grad_norm": 0.9516801238059998,
+      "learning_rate": 2.834733893557423e-05,
+      "loss": 0.5363,
+      "step": 10834
+    },
+    {
+      "epoch": 13.8688,
+      "grad_norm": 0.9916854500770569,
+      "learning_rate": 2.83453381352541e-05,
+      "loss": 0.5854,
+      "step": 10835
+    },
+    {
+      "epoch": 13.87008,
+      "grad_norm": 0.9755607843399048,
+      "learning_rate": 2.8343337334933973e-05,
+      "loss": 0.5362,
+      "step": 10836
+    },
+    {
+      "epoch": 13.87136,
+      "grad_norm": 0.9905216097831726,
+      "learning_rate": 2.8341336534613848e-05,
+      "loss": 0.5287,
+      "step": 10837
+    },
+    {
+      "epoch": 13.87264,
+      "grad_norm": 0.9148777723312378,
+      "learning_rate": 2.833933573429372e-05,
+      "loss": 0.5432,
+      "step": 10838
+    },
+    {
+      "epoch": 13.87392,
+      "grad_norm": 0.978702962398529,
+      "learning_rate": 2.833733493397359e-05,
+      "loss": 0.5745,
+      "step": 10839
+    },
+    {
+      "epoch": 13.8752,
+      "grad_norm": 0.9320308566093445,
+      "learning_rate": 2.8335334133653464e-05,
+      "loss": 0.5186,
+      "step": 10840
+    },
+    {
+      "epoch": 13.87648,
+      "grad_norm": 0.9999722838401794,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.5527,
+      "step": 10841
+    },
+    {
+      "epoch": 13.87776,
+      "grad_norm": 0.9607200026512146,
+      "learning_rate": 2.8331332533013204e-05,
+      "loss": 0.5737,
+      "step": 10842
+    },
+    {
+      "epoch": 13.87904,
+      "grad_norm": 0.9080913662910461,
+      "learning_rate": 2.8329331732693076e-05,
+      "loss": 0.4989,
+      "step": 10843
+    },
+    {
+      "epoch": 13.88032,
+      "grad_norm": 0.9405791759490967,
+      "learning_rate": 2.8327330932372954e-05,
+      "loss": 0.5243,
+      "step": 10844
+    },
+    {
+      "epoch": 13.8816,
+      "grad_norm": 0.9514108896255493,
+      "learning_rate": 2.8325330132052823e-05,
+      "loss": 0.5607,
+      "step": 10845
+    },
+    {
+      "epoch": 13.88288,
+      "grad_norm": 0.9832378029823303,
+      "learning_rate": 2.8323329331732695e-05,
+      "loss": 0.5726,
+      "step": 10846
+    },
+    {
+      "epoch": 13.88416,
+      "grad_norm": 0.9148914813995361,
+      "learning_rate": 2.8321328531412567e-05,
+      "loss": 0.4917,
+      "step": 10847
+    },
+    {
+      "epoch": 13.88544,
+      "grad_norm": 0.9236845970153809,
+      "learning_rate": 2.831932773109244e-05,
+      "loss": 0.5422,
+      "step": 10848
+    },
+    {
+      "epoch": 13.88672,
+      "grad_norm": 0.8827064037322998,
+      "learning_rate": 2.831732693077231e-05,
+      "loss": 0.563,
+      "step": 10849
+    },
+    {
+      "epoch": 13.888,
+      "grad_norm": 0.9374257922172546,
+      "learning_rate": 2.831532613045218e-05,
+      "loss": 0.5358,
+      "step": 10850
+    },
+    {
+      "epoch": 13.88928,
+      "grad_norm": 1.0206034183502197,
+      "learning_rate": 2.8313325330132057e-05,
+      "loss": 0.5629,
+      "step": 10851
+    },
+    {
+      "epoch": 13.89056,
+      "grad_norm": 0.9112451672554016,
+      "learning_rate": 2.831132452981193e-05,
+      "loss": 0.4899,
+      "step": 10852
+    },
+    {
+      "epoch": 13.89184,
+      "grad_norm": 1.0080876350402832,
+      "learning_rate": 2.8309323729491798e-05,
+      "loss": 0.5934,
+      "step": 10853
+    },
+    {
+      "epoch": 13.89312,
+      "grad_norm": 0.9685933589935303,
+      "learning_rate": 2.830732292917167e-05,
+      "loss": 0.5597,
+      "step": 10854
+    },
+    {
+      "epoch": 13.8944,
+      "grad_norm": 0.9162622094154358,
+      "learning_rate": 2.830532212885154e-05,
+      "loss": 0.5136,
+      "step": 10855
+    },
+    {
+      "epoch": 13.89568,
+      "grad_norm": 0.9283410310745239,
+      "learning_rate": 2.8303321328531413e-05,
+      "loss": 0.5515,
+      "step": 10856
+    },
+    {
+      "epoch": 13.89696,
+      "grad_norm": 0.9511690139770508,
+      "learning_rate": 2.8301320528211285e-05,
+      "loss": 0.5062,
+      "step": 10857
+    },
+    {
+      "epoch": 13.89824,
+      "grad_norm": 0.9601842761039734,
+      "learning_rate": 2.829931972789116e-05,
+      "loss": 0.5394,
+      "step": 10858
+    },
+    {
+      "epoch": 13.89952,
+      "grad_norm": 0.946259617805481,
+      "learning_rate": 2.8297318927571032e-05,
+      "loss": 0.5835,
+      "step": 10859
+    },
+    {
+      "epoch": 13.9008,
+      "grad_norm": 0.9118945598602295,
+      "learning_rate": 2.8295318127250904e-05,
+      "loss": 0.5668,
+      "step": 10860
+    },
+    {
+      "epoch": 13.90208,
+      "grad_norm": 0.982496976852417,
+      "learning_rate": 2.8293317326930773e-05,
+      "loss": 0.5393,
+      "step": 10861
+    },
+    {
+      "epoch": 13.90336,
+      "grad_norm": 0.9529083371162415,
+      "learning_rate": 2.8291316526610644e-05,
+      "loss": 0.5617,
+      "step": 10862
+    },
+    {
+      "epoch": 13.90464,
+      "grad_norm": 0.9181329011917114,
+      "learning_rate": 2.8289315726290516e-05,
+      "loss": 0.5187,
+      "step": 10863
+    },
+    {
+      "epoch": 13.90592,
+      "grad_norm": 0.9448521137237549,
+      "learning_rate": 2.8287314925970388e-05,
+      "loss": 0.5783,
+      "step": 10864
+    },
+    {
+      "epoch": 13.9072,
+      "grad_norm": 0.9216022491455078,
+      "learning_rate": 2.8285314125650263e-05,
+      "loss": 0.5452,
+      "step": 10865
+    },
+    {
+      "epoch": 13.90848,
+      "grad_norm": 0.9557605981826782,
+      "learning_rate": 2.8283313325330135e-05,
+      "loss": 0.5593,
+      "step": 10866
+    },
+    {
+      "epoch": 13.90976,
+      "grad_norm": 0.9449323415756226,
+      "learning_rate": 2.8281312525010007e-05,
+      "loss": 0.5575,
+      "step": 10867
+    },
+    {
+      "epoch": 13.91104,
+      "grad_norm": 0.945347249507904,
+      "learning_rate": 2.827931172468988e-05,
+      "loss": 0.5354,
+      "step": 10868
+    },
+    {
+      "epoch": 13.91232,
+      "grad_norm": 0.9745570421218872,
+      "learning_rate": 2.8277310924369747e-05,
+      "loss": 0.5818,
+      "step": 10869
+    },
+    {
+      "epoch": 13.9136,
+      "grad_norm": 0.9462365508079529,
+      "learning_rate": 2.827531012404962e-05,
+      "loss": 0.5278,
+      "step": 10870
+    },
+    {
+      "epoch": 13.91488,
+      "grad_norm": 0.985939621925354,
+      "learning_rate": 2.827330932372949e-05,
+      "loss": 0.5727,
+      "step": 10871
+    },
+    {
+      "epoch": 13.91616,
+      "grad_norm": 0.8899412751197815,
+      "learning_rate": 2.8271308523409366e-05,
+      "loss": 0.5279,
+      "step": 10872
+    },
+    {
+      "epoch": 13.91744,
+      "grad_norm": 0.9762179851531982,
+      "learning_rate": 2.8269307723089238e-05,
+      "loss": 0.5406,
+      "step": 10873
+    },
+    {
+      "epoch": 13.91872,
+      "grad_norm": 0.9601300358772278,
+      "learning_rate": 2.826730692276911e-05,
+      "loss": 0.5518,
+      "step": 10874
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 1.0480626821517944,
+      "learning_rate": 2.8265306122448982e-05,
+      "loss": 0.5649,
+      "step": 10875
+    },
+    {
+      "epoch": 13.92128,
+      "grad_norm": 0.8842355012893677,
+      "learning_rate": 2.8263305322128854e-05,
+      "loss": 0.4951,
+      "step": 10876
+    },
+    {
+      "epoch": 13.92256,
+      "grad_norm": 0.9062721729278564,
+      "learning_rate": 2.8261304521808722e-05,
+      "loss": 0.5746,
+      "step": 10877
+    },
+    {
+      "epoch": 13.92384,
+      "grad_norm": 0.9198760986328125,
+      "learning_rate": 2.8259303721488594e-05,
+      "loss": 0.5183,
+      "step": 10878
+    },
+    {
+      "epoch": 13.92512,
+      "grad_norm": 0.9441488981246948,
+      "learning_rate": 2.8257302921168473e-05,
+      "loss": 0.537,
+      "step": 10879
+    },
+    {
+      "epoch": 13.9264,
+      "grad_norm": 0.9522190093994141,
+      "learning_rate": 2.825530212084834e-05,
+      "loss": 0.514,
+      "step": 10880
+    },
+    {
+      "epoch": 13.92768,
+      "grad_norm": 0.9240538477897644,
+      "learning_rate": 2.8253301320528213e-05,
+      "loss": 0.5422,
+      "step": 10881
+    },
+    {
+      "epoch": 13.92896,
+      "grad_norm": 0.9243524670600891,
+      "learning_rate": 2.8251300520208085e-05,
+      "loss": 0.5353,
+      "step": 10882
+    },
+    {
+      "epoch": 13.93024,
+      "grad_norm": 0.9364755153656006,
+      "learning_rate": 2.8249299719887957e-05,
+      "loss": 0.5339,
+      "step": 10883
+    },
+    {
+      "epoch": 13.93152,
+      "grad_norm": 0.9147524833679199,
+      "learning_rate": 2.824729891956783e-05,
+      "loss": 0.4999,
+      "step": 10884
+    },
+    {
+      "epoch": 13.9328,
+      "grad_norm": 0.956943690776825,
+      "learning_rate": 2.8245298119247697e-05,
+      "loss": 0.531,
+      "step": 10885
+    },
+    {
+      "epoch": 13.93408,
+      "grad_norm": 0.9214387536048889,
+      "learning_rate": 2.8243297318927576e-05,
+      "loss": 0.5403,
+      "step": 10886
+    },
+    {
+      "epoch": 13.93536,
+      "grad_norm": 0.9463467001914978,
+      "learning_rate": 2.8241296518607448e-05,
+      "loss": 0.5471,
+      "step": 10887
+    },
+    {
+      "epoch": 13.93664,
+      "grad_norm": 0.9077194929122925,
+      "learning_rate": 2.8239295718287316e-05,
+      "loss": 0.5406,
+      "step": 10888
+    },
+    {
+      "epoch": 13.93792,
+      "grad_norm": 0.943578839302063,
+      "learning_rate": 2.8237294917967188e-05,
+      "loss": 0.5271,
+      "step": 10889
+    },
+    {
+      "epoch": 13.9392,
+      "grad_norm": 0.9538133144378662,
+      "learning_rate": 2.823529411764706e-05,
+      "loss": 0.5552,
+      "step": 10890
+    },
+    {
+      "epoch": 13.94048,
+      "grad_norm": 0.9875748157501221,
+      "learning_rate": 2.823329331732693e-05,
+      "loss": 0.562,
+      "step": 10891
+    },
+    {
+      "epoch": 13.94176,
+      "grad_norm": 0.8694930076599121,
+      "learning_rate": 2.8231292517006803e-05,
+      "loss": 0.4823,
+      "step": 10892
+    },
+    {
+      "epoch": 13.94304,
+      "grad_norm": 0.9158459305763245,
+      "learning_rate": 2.8229291716686672e-05,
+      "loss": 0.5413,
+      "step": 10893
+    },
+    {
+      "epoch": 13.94432,
+      "grad_norm": 0.9385559558868408,
+      "learning_rate": 2.822729091636655e-05,
+      "loss": 0.5901,
+      "step": 10894
+    },
+    {
+      "epoch": 13.9456,
+      "grad_norm": 0.9433490633964539,
+      "learning_rate": 2.8225290116046422e-05,
+      "loss": 0.5487,
+      "step": 10895
+    },
+    {
+      "epoch": 13.94688,
+      "grad_norm": 0.9610795378684998,
+      "learning_rate": 2.822328931572629e-05,
+      "loss": 0.5375,
+      "step": 10896
+    },
+    {
+      "epoch": 13.94816,
+      "grad_norm": 0.9676675796508789,
+      "learning_rate": 2.8221288515406163e-05,
+      "loss": 0.5581,
+      "step": 10897
+    },
+    {
+      "epoch": 13.94944,
+      "grad_norm": 0.9557973742485046,
+      "learning_rate": 2.8219287715086035e-05,
+      "loss": 0.5599,
+      "step": 10898
+    },
+    {
+      "epoch": 13.95072,
+      "grad_norm": 0.9611872434616089,
+      "learning_rate": 2.8217286914765906e-05,
+      "loss": 0.5688,
+      "step": 10899
+    },
+    {
+      "epoch": 13.952,
+      "grad_norm": 0.9648937582969666,
+      "learning_rate": 2.8215286114445778e-05,
+      "loss": 0.5356,
+      "step": 10900
+    },
+    {
+      "epoch": 13.95328,
+      "grad_norm": 0.9507505893707275,
+      "learning_rate": 2.8213285314125654e-05,
+      "loss": 0.524,
+      "step": 10901
+    },
+    {
+      "epoch": 13.95456,
+      "grad_norm": 0.9989834427833557,
+      "learning_rate": 2.8211284513805525e-05,
+      "loss": 0.5785,
+      "step": 10902
+    },
+    {
+      "epoch": 13.95584,
+      "grad_norm": 0.9809247851371765,
+      "learning_rate": 2.8209283713485397e-05,
+      "loss": 0.5722,
+      "step": 10903
+    },
+    {
+      "epoch": 13.95712,
+      "grad_norm": 0.9577847123146057,
+      "learning_rate": 2.8207282913165266e-05,
+      "loss": 0.5891,
+      "step": 10904
+    },
+    {
+      "epoch": 13.9584,
+      "grad_norm": 1.0070008039474487,
+      "learning_rate": 2.8205282112845138e-05,
+      "loss": 0.6114,
+      "step": 10905
+    },
+    {
+      "epoch": 13.95968,
+      "grad_norm": 0.9523963928222656,
+      "learning_rate": 2.820328131252501e-05,
+      "loss": 0.545,
+      "step": 10906
+    },
+    {
+      "epoch": 13.96096,
+      "grad_norm": 0.9290968179702759,
+      "learning_rate": 2.820128051220488e-05,
+      "loss": 0.5581,
+      "step": 10907
+    },
+    {
+      "epoch": 13.96224,
+      "grad_norm": 0.8966550827026367,
+      "learning_rate": 2.8199279711884757e-05,
+      "loss": 0.5651,
+      "step": 10908
+    },
+    {
+      "epoch": 13.96352,
+      "grad_norm": 0.9531732797622681,
+      "learning_rate": 2.819727891156463e-05,
+      "loss": 0.534,
+      "step": 10909
+    },
+    {
+      "epoch": 13.9648,
+      "grad_norm": 0.9635023474693298,
+      "learning_rate": 2.81952781112445e-05,
+      "loss": 0.5338,
+      "step": 10910
+    },
+    {
+      "epoch": 13.96608,
+      "grad_norm": 0.950612485408783,
+      "learning_rate": 2.8193277310924372e-05,
+      "loss": 0.5763,
+      "step": 10911
+    },
+    {
+      "epoch": 13.96736,
+      "grad_norm": 0.9613419771194458,
+      "learning_rate": 2.819127651060424e-05,
+      "loss": 0.6066,
+      "step": 10912
+    },
+    {
+      "epoch": 13.96864,
+      "grad_norm": 0.9048581123352051,
+      "learning_rate": 2.8189275710284112e-05,
+      "loss": 0.5309,
+      "step": 10913
+    },
+    {
+      "epoch": 13.96992,
+      "grad_norm": 0.931669294834137,
+      "learning_rate": 2.8187274909963984e-05,
+      "loss": 0.5653,
+      "step": 10914
+    },
+    {
+      "epoch": 13.9712,
+      "grad_norm": 0.9555082321166992,
+      "learning_rate": 2.818527410964386e-05,
+      "loss": 0.5659,
+      "step": 10915
+    },
+    {
+      "epoch": 13.972480000000001,
+      "grad_norm": 0.9597335457801819,
+      "learning_rate": 2.818327330932373e-05,
+      "loss": 0.5385,
+      "step": 10916
+    },
+    {
+      "epoch": 13.97376,
+      "grad_norm": 1.0083085298538208,
+      "learning_rate": 2.8181272509003603e-05,
+      "loss": 0.5929,
+      "step": 10917
+    },
+    {
+      "epoch": 13.97504,
+      "grad_norm": 1.0068410634994507,
+      "learning_rate": 2.8179271708683475e-05,
+      "loss": 0.5864,
+      "step": 10918
+    },
+    {
+      "epoch": 13.97632,
+      "grad_norm": 0.9447670578956604,
+      "learning_rate": 2.8177270908363347e-05,
+      "loss": 0.5239,
+      "step": 10919
+    },
+    {
+      "epoch": 13.9776,
+      "grad_norm": 0.9446542263031006,
+      "learning_rate": 2.8175270108043215e-05,
+      "loss": 0.5369,
+      "step": 10920
+    },
+    {
+      "epoch": 13.97888,
+      "grad_norm": 0.9933118224143982,
+      "learning_rate": 2.8173269307723087e-05,
+      "loss": 0.5385,
+      "step": 10921
+    },
+    {
+      "epoch": 13.98016,
+      "grad_norm": 0.9696182012557983,
+      "learning_rate": 2.8171268507402966e-05,
+      "loss": 0.5226,
+      "step": 10922
+    },
+    {
+      "epoch": 13.98144,
+      "grad_norm": 0.9453620314598083,
+      "learning_rate": 2.8169267707082834e-05,
+      "loss": 0.5276,
+      "step": 10923
+    },
+    {
+      "epoch": 13.98272,
+      "grad_norm": 0.9572956562042236,
+      "learning_rate": 2.8167266906762706e-05,
+      "loss": 0.5629,
+      "step": 10924
+    },
+    {
+      "epoch": 13.984,
+      "grad_norm": 0.9128978252410889,
+      "learning_rate": 2.8165266106442578e-05,
+      "loss": 0.5435,
+      "step": 10925
+    },
+    {
+      "epoch": 13.98528,
+      "grad_norm": 0.926874577999115,
+      "learning_rate": 2.816326530612245e-05,
+      "loss": 0.5317,
+      "step": 10926
+    },
+    {
+      "epoch": 13.98656,
+      "grad_norm": 1.0055192708969116,
+      "learning_rate": 2.8161264505802322e-05,
+      "loss": 0.6023,
+      "step": 10927
+    },
+    {
+      "epoch": 13.98784,
+      "grad_norm": 0.9463496804237366,
+      "learning_rate": 2.815926370548219e-05,
+      "loss": 0.5565,
+      "step": 10928
+    },
+    {
+      "epoch": 13.98912,
+      "grad_norm": 0.9550392031669617,
+      "learning_rate": 2.815726290516207e-05,
+      "loss": 0.5601,
+      "step": 10929
+    },
+    {
+      "epoch": 13.9904,
+      "grad_norm": 0.9578831791877747,
+      "learning_rate": 2.815526210484194e-05,
+      "loss": 0.5703,
+      "step": 10930
+    },
+    {
+      "epoch": 13.99168,
+      "grad_norm": 0.9312989711761475,
+      "learning_rate": 2.815326130452181e-05,
+      "loss": 0.5328,
+      "step": 10931
+    },
+    {
+      "epoch": 13.99296,
+      "grad_norm": 1.000938057899475,
+      "learning_rate": 2.815126050420168e-05,
+      "loss": 0.5901,
+      "step": 10932
+    },
+    {
+      "epoch": 13.99424,
+      "grad_norm": 0.9439278244972229,
+      "learning_rate": 2.8149259703881553e-05,
+      "loss": 0.5381,
+      "step": 10933
+    },
+    {
+      "epoch": 13.995519999999999,
+      "grad_norm": 0.9305405616760254,
+      "learning_rate": 2.8147258903561425e-05,
+      "loss": 0.5736,
+      "step": 10934
+    },
+    {
+      "epoch": 13.9968,
+      "grad_norm": 0.8652977347373962,
+      "learning_rate": 2.8145258103241297e-05,
+      "loss": 0.5142,
+      "step": 10935
+    },
+    {
+      "epoch": 13.99808,
+      "grad_norm": 0.9089574217796326,
+      "learning_rate": 2.8143257302921172e-05,
+      "loss": 0.506,
+      "step": 10936
+    },
+    {
+      "epoch": 13.99936,
+      "grad_norm": 0.9164632558822632,
+      "learning_rate": 2.8141256502601044e-05,
+      "loss": 0.5394,
+      "step": 10937
+    },
+    {
+      "epoch": 14.00064,
+      "grad_norm": 1.9779025316238403,
+      "learning_rate": 2.8139255702280916e-05,
+      "loss": 0.969,
+      "step": 10938
+    },
+    {
+      "epoch": 14.00192,
+      "grad_norm": 0.9149560928344727,
+      "learning_rate": 2.8137254901960784e-05,
+      "loss": 0.5095,
+      "step": 10939
+    },
+    {
+      "epoch": 14.0032,
+      "grad_norm": 0.9277960658073425,
+      "learning_rate": 2.8135254101640656e-05,
+      "loss": 0.5741,
+      "step": 10940
+    },
+    {
+      "epoch": 14.00448,
+      "grad_norm": 0.9176803827285767,
+      "learning_rate": 2.8133253301320528e-05,
+      "loss": 0.4634,
+      "step": 10941
+    },
+    {
+      "epoch": 14.00576,
+      "grad_norm": 0.926750123500824,
+      "learning_rate": 2.81312525010004e-05,
+      "loss": 0.5669,
+      "step": 10942
+    },
+    {
+      "epoch": 14.00704,
+      "grad_norm": 0.8823717832565308,
+      "learning_rate": 2.8129251700680275e-05,
+      "loss": 0.4973,
+      "step": 10943
+    },
+    {
+      "epoch": 14.00832,
+      "grad_norm": 0.9568257927894592,
+      "learning_rate": 2.8127250900360147e-05,
+      "loss": 0.5307,
+      "step": 10944
+    },
+    {
+      "epoch": 14.0096,
+      "grad_norm": 0.9589383006095886,
+      "learning_rate": 2.812525010004002e-05,
+      "loss": 0.5054,
+      "step": 10945
+    },
+    {
+      "epoch": 14.01088,
+      "grad_norm": 0.9462265968322754,
+      "learning_rate": 2.812324929971989e-05,
+      "loss": 0.5156,
+      "step": 10946
+    },
+    {
+      "epoch": 14.01216,
+      "grad_norm": 0.9722265601158142,
+      "learning_rate": 2.812124849939976e-05,
+      "loss": 0.504,
+      "step": 10947
+    },
+    {
+      "epoch": 14.01344,
+      "grad_norm": 0.9647181034088135,
+      "learning_rate": 2.811924769907963e-05,
+      "loss": 0.5248,
+      "step": 10948
+    },
+    {
+      "epoch": 14.01472,
+      "grad_norm": 0.9778644442558289,
+      "learning_rate": 2.8117246898759503e-05,
+      "loss": 0.505,
+      "step": 10949
+    },
+    {
+      "epoch": 14.016,
+      "grad_norm": 0.9715518355369568,
+      "learning_rate": 2.8115246098439378e-05,
+      "loss": 0.5486,
+      "step": 10950
+    },
+    {
+      "epoch": 14.01728,
+      "grad_norm": 0.9561265707015991,
+      "learning_rate": 2.811324529811925e-05,
+      "loss": 0.5333,
+      "step": 10951
+    },
+    {
+      "epoch": 14.01856,
+      "grad_norm": 0.9769052863121033,
+      "learning_rate": 2.811124449779912e-05,
+      "loss": 0.5561,
+      "step": 10952
+    },
+    {
+      "epoch": 14.01984,
+      "grad_norm": 0.9301590323448181,
+      "learning_rate": 2.8109243697478993e-05,
+      "loss": 0.505,
+      "step": 10953
+    },
+    {
+      "epoch": 14.02112,
+      "grad_norm": 0.9344640374183655,
+      "learning_rate": 2.8107242897158865e-05,
+      "loss": 0.5225,
+      "step": 10954
+    },
+    {
+      "epoch": 14.0224,
+      "grad_norm": 0.9300322532653809,
+      "learning_rate": 2.8105242096838734e-05,
+      "loss": 0.494,
+      "step": 10955
+    },
+    {
+      "epoch": 14.02368,
+      "grad_norm": 0.9003793001174927,
+      "learning_rate": 2.8103241296518606e-05,
+      "loss": 0.4933,
+      "step": 10956
+    },
+    {
+      "epoch": 14.02496,
+      "grad_norm": 0.9156431555747986,
+      "learning_rate": 2.8101240496198484e-05,
+      "loss": 0.5131,
+      "step": 10957
+    },
+    {
+      "epoch": 14.02624,
+      "grad_norm": 0.929981529712677,
+      "learning_rate": 2.8099239695878353e-05,
+      "loss": 0.5237,
+      "step": 10958
+    },
+    {
+      "epoch": 14.02752,
+      "grad_norm": 0.9370753169059753,
+      "learning_rate": 2.8097238895558225e-05,
+      "loss": 0.4931,
+      "step": 10959
+    },
+    {
+      "epoch": 14.0288,
+      "grad_norm": 0.9672894477844238,
+      "learning_rate": 2.8095238095238096e-05,
+      "loss": 0.4916,
+      "step": 10960
+    },
+    {
+      "epoch": 14.03008,
+      "grad_norm": 0.9758133888244629,
+      "learning_rate": 2.809323729491797e-05,
+      "loss": 0.5458,
+      "step": 10961
+    },
+    {
+      "epoch": 14.03136,
+      "grad_norm": 1.002359390258789,
+      "learning_rate": 2.809123649459784e-05,
+      "loss": 0.5633,
+      "step": 10962
+    },
+    {
+      "epoch": 14.03264,
+      "grad_norm": 0.9776928424835205,
+      "learning_rate": 2.808923569427771e-05,
+      "loss": 0.5331,
+      "step": 10963
+    },
+    {
+      "epoch": 14.03392,
+      "grad_norm": 0.9960092306137085,
+      "learning_rate": 2.8087234893957587e-05,
+      "loss": 0.5458,
+      "step": 10964
+    },
+    {
+      "epoch": 14.0352,
+      "grad_norm": 0.9920620918273926,
+      "learning_rate": 2.808523409363746e-05,
+      "loss": 0.5853,
+      "step": 10965
+    },
+    {
+      "epoch": 14.03648,
+      "grad_norm": 0.9721470475196838,
+      "learning_rate": 2.8083233293317328e-05,
+      "loss": 0.532,
+      "step": 10966
+    },
+    {
+      "epoch": 14.03776,
+      "grad_norm": 0.8846380710601807,
+      "learning_rate": 2.80812324929972e-05,
+      "loss": 0.4606,
+      "step": 10967
+    },
+    {
+      "epoch": 14.03904,
+      "grad_norm": 0.934118926525116,
+      "learning_rate": 2.807923169267707e-05,
+      "loss": 0.4849,
+      "step": 10968
+    },
+    {
+      "epoch": 14.04032,
+      "grad_norm": 0.944482147693634,
+      "learning_rate": 2.8077230892356943e-05,
+      "loss": 0.521,
+      "step": 10969
+    },
+    {
+      "epoch": 14.0416,
+      "grad_norm": 0.9114336967468262,
+      "learning_rate": 2.8075230092036815e-05,
+      "loss": 0.5248,
+      "step": 10970
+    },
+    {
+      "epoch": 14.04288,
+      "grad_norm": 0.9594119787216187,
+      "learning_rate": 2.807322929171669e-05,
+      "loss": 0.5069,
+      "step": 10971
+    },
+    {
+      "epoch": 14.04416,
+      "grad_norm": 0.9807631373405457,
+      "learning_rate": 2.8071228491396562e-05,
+      "loss": 0.5334,
+      "step": 10972
+    },
+    {
+      "epoch": 14.04544,
+      "grad_norm": 0.9795427322387695,
+      "learning_rate": 2.8069227691076434e-05,
+      "loss": 0.5194,
+      "step": 10973
+    },
+    {
+      "epoch": 14.04672,
+      "grad_norm": 0.9742780923843384,
+      "learning_rate": 2.8067226890756302e-05,
+      "loss": 0.5439,
+      "step": 10974
+    },
+    {
+      "epoch": 14.048,
+      "grad_norm": 0.9774748682975769,
+      "learning_rate": 2.8065226090436174e-05,
+      "loss": 0.5262,
+      "step": 10975
+    },
+    {
+      "epoch": 14.04928,
+      "grad_norm": 1.0334399938583374,
+      "learning_rate": 2.8063225290116046e-05,
+      "loss": 0.56,
+      "step": 10976
+    },
+    {
+      "epoch": 14.05056,
+      "grad_norm": 1.013529896736145,
+      "learning_rate": 2.8061224489795918e-05,
+      "loss": 0.5665,
+      "step": 10977
+    },
+    {
+      "epoch": 14.05184,
+      "grad_norm": 0.9661319255828857,
+      "learning_rate": 2.8059223689475793e-05,
+      "loss": 0.5385,
+      "step": 10978
+    },
+    {
+      "epoch": 14.05312,
+      "grad_norm": 0.9551675915718079,
+      "learning_rate": 2.8057222889155665e-05,
+      "loss": 0.533,
+      "step": 10979
+    },
+    {
+      "epoch": 14.0544,
+      "grad_norm": 0.9196348786354065,
+      "learning_rate": 2.8055222088835537e-05,
+      "loss": 0.5153,
+      "step": 10980
+    },
+    {
+      "epoch": 14.05568,
+      "grad_norm": 0.9740005135536194,
+      "learning_rate": 2.805322128851541e-05,
+      "loss": 0.5095,
+      "step": 10981
+    },
+    {
+      "epoch": 14.05696,
+      "grad_norm": 0.9844241142272949,
+      "learning_rate": 2.8051220488195277e-05,
+      "loss": 0.5497,
+      "step": 10982
+    },
+    {
+      "epoch": 14.05824,
+      "grad_norm": 0.9591256380081177,
+      "learning_rate": 2.804921968787515e-05,
+      "loss": 0.5108,
+      "step": 10983
+    },
+    {
+      "epoch": 14.05952,
+      "grad_norm": 0.9638150334358215,
+      "learning_rate": 2.804721888755502e-05,
+      "loss": 0.5081,
+      "step": 10984
+    },
+    {
+      "epoch": 14.0608,
+      "grad_norm": 0.9753637909889221,
+      "learning_rate": 2.8045218087234896e-05,
+      "loss": 0.4962,
+      "step": 10985
+    },
+    {
+      "epoch": 14.06208,
+      "grad_norm": 0.9755197167396545,
+      "learning_rate": 2.8043217286914768e-05,
+      "loss": 0.5239,
+      "step": 10986
+    },
+    {
+      "epoch": 14.06336,
+      "grad_norm": 1.000517725944519,
+      "learning_rate": 2.804121648659464e-05,
+      "loss": 0.5282,
+      "step": 10987
+    },
+    {
+      "epoch": 14.06464,
+      "grad_norm": 0.9386008381843567,
+      "learning_rate": 2.8039215686274512e-05,
+      "loss": 0.5275,
+      "step": 10988
+    },
+    {
+      "epoch": 14.06592,
+      "grad_norm": 0.9570558667182922,
+      "learning_rate": 2.8037214885954384e-05,
+      "loss": 0.5064,
+      "step": 10989
+    },
+    {
+      "epoch": 14.0672,
+      "grad_norm": 0.955383837223053,
+      "learning_rate": 2.8035214085634252e-05,
+      "loss": 0.5152,
+      "step": 10990
+    },
+    {
+      "epoch": 14.06848,
+      "grad_norm": 0.9390836954116821,
+      "learning_rate": 2.8033213285314124e-05,
+      "loss": 0.4959,
+      "step": 10991
+    },
+    {
+      "epoch": 14.06976,
+      "grad_norm": 0.9478831887245178,
+      "learning_rate": 2.8031212484994003e-05,
+      "loss": 0.5058,
+      "step": 10992
+    },
+    {
+      "epoch": 14.07104,
+      "grad_norm": 0.9145018458366394,
+      "learning_rate": 2.802921168467387e-05,
+      "loss": 0.4909,
+      "step": 10993
+    },
+    {
+      "epoch": 14.07232,
+      "grad_norm": 0.9584047198295593,
+      "learning_rate": 2.8027210884353743e-05,
+      "loss": 0.5112,
+      "step": 10994
+    },
+    {
+      "epoch": 14.0736,
+      "grad_norm": 0.973054826259613,
+      "learning_rate": 2.8025210084033615e-05,
+      "loss": 0.5226,
+      "step": 10995
+    },
+    {
+      "epoch": 14.07488,
+      "grad_norm": 0.9144132733345032,
+      "learning_rate": 2.8023209283713487e-05,
+      "loss": 0.4899,
+      "step": 10996
+    },
+    {
+      "epoch": 14.07616,
+      "grad_norm": 0.9813609719276428,
+      "learning_rate": 2.802120848339336e-05,
+      "loss": 0.5154,
+      "step": 10997
+    },
+    {
+      "epoch": 14.07744,
+      "grad_norm": 0.9886991381645203,
+      "learning_rate": 2.8019207683073227e-05,
+      "loss": 0.5467,
+      "step": 10998
+    },
+    {
+      "epoch": 14.07872,
+      "grad_norm": 0.9517562985420227,
+      "learning_rate": 2.8017206882753106e-05,
+      "loss": 0.5436,
+      "step": 10999
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 0.9583877325057983,
+      "learning_rate": 2.8015206082432978e-05,
+      "loss": 0.4949,
+      "step": 11000
+    },
+    {
+      "epoch": 14.08128,
+      "grad_norm": 0.988237738609314,
+      "learning_rate": 2.8013205282112846e-05,
+      "loss": 0.5509,
+      "step": 11001
+    },
+    {
+      "epoch": 14.08256,
+      "grad_norm": 0.956883430480957,
+      "learning_rate": 2.8011204481792718e-05,
+      "loss": 0.5071,
+      "step": 11002
+    },
+    {
+      "epoch": 14.08384,
+      "grad_norm": 0.9071076512336731,
+      "learning_rate": 2.800920368147259e-05,
+      "loss": 0.4785,
+      "step": 11003
+    },
+    {
+      "epoch": 14.08512,
+      "grad_norm": 0.9783008098602295,
+      "learning_rate": 2.800720288115246e-05,
+      "loss": 0.5356,
+      "step": 11004
+    },
+    {
+      "epoch": 14.0864,
+      "grad_norm": 0.9609231352806091,
+      "learning_rate": 2.8005202080832333e-05,
+      "loss": 0.5093,
+      "step": 11005
+    },
+    {
+      "epoch": 14.08768,
+      "grad_norm": 0.9345542192459106,
+      "learning_rate": 2.8003201280512202e-05,
+      "loss": 0.5265,
+      "step": 11006
+    },
+    {
+      "epoch": 14.08896,
+      "grad_norm": 0.94151371717453,
+      "learning_rate": 2.800120048019208e-05,
+      "loss": 0.5561,
+      "step": 11007
+    },
+    {
+      "epoch": 14.09024,
+      "grad_norm": 0.9789606928825378,
+      "learning_rate": 2.7999199679871952e-05,
+      "loss": 0.56,
+      "step": 11008
+    },
+    {
+      "epoch": 14.09152,
+      "grad_norm": 1.0095890760421753,
+      "learning_rate": 2.799719887955182e-05,
+      "loss": 0.5189,
+      "step": 11009
+    },
+    {
+      "epoch": 14.0928,
+      "grad_norm": 0.9807612895965576,
+      "learning_rate": 2.7995198079231693e-05,
+      "loss": 0.5041,
+      "step": 11010
+    },
+    {
+      "epoch": 14.09408,
+      "grad_norm": 0.9574481844902039,
+      "learning_rate": 2.7993197278911565e-05,
+      "loss": 0.5085,
+      "step": 11011
+    },
+    {
+      "epoch": 14.09536,
+      "grad_norm": 0.9162589311599731,
+      "learning_rate": 2.7991196478591436e-05,
+      "loss": 0.5223,
+      "step": 11012
+    },
+    {
+      "epoch": 14.09664,
+      "grad_norm": 0.9016433358192444,
+      "learning_rate": 2.7989195678271308e-05,
+      "loss": 0.5148,
+      "step": 11013
+    },
+    {
+      "epoch": 14.09792,
+      "grad_norm": 0.9376025199890137,
+      "learning_rate": 2.7987194877951184e-05,
+      "loss": 0.5246,
+      "step": 11014
+    },
+    {
+      "epoch": 14.0992,
+      "grad_norm": 0.9752246737480164,
+      "learning_rate": 2.7985194077631055e-05,
+      "loss": 0.5176,
+      "step": 11015
+    },
+    {
+      "epoch": 14.10048,
+      "grad_norm": 0.9751932621002197,
+      "learning_rate": 2.7983193277310927e-05,
+      "loss": 0.5399,
+      "step": 11016
+    },
+    {
+      "epoch": 14.10176,
+      "grad_norm": 0.8926282525062561,
+      "learning_rate": 2.7981192476990796e-05,
+      "loss": 0.4618,
+      "step": 11017
+    },
+    {
+      "epoch": 14.10304,
+      "grad_norm": 0.9403905272483826,
+      "learning_rate": 2.7979191676670668e-05,
+      "loss": 0.5345,
+      "step": 11018
+    },
+    {
+      "epoch": 14.10432,
+      "grad_norm": 0.9546553492546082,
+      "learning_rate": 2.797719087635054e-05,
+      "loss": 0.5458,
+      "step": 11019
+    },
+    {
+      "epoch": 14.1056,
+      "grad_norm": 0.9716793298721313,
+      "learning_rate": 2.797519007603041e-05,
+      "loss": 0.5325,
+      "step": 11020
+    },
+    {
+      "epoch": 14.10688,
+      "grad_norm": 0.9732063412666321,
+      "learning_rate": 2.7973189275710287e-05,
+      "loss": 0.5074,
+      "step": 11021
+    },
+    {
+      "epoch": 14.10816,
+      "grad_norm": 0.9816451072692871,
+      "learning_rate": 2.797118847539016e-05,
+      "loss": 0.5202,
+      "step": 11022
+    },
+    {
+      "epoch": 14.10944,
+      "grad_norm": 0.9631067514419556,
+      "learning_rate": 2.796918767507003e-05,
+      "loss": 0.5571,
+      "step": 11023
+    },
+    {
+      "epoch": 14.11072,
+      "grad_norm": 0.9688462018966675,
+      "learning_rate": 2.7967186874749902e-05,
+      "loss": 0.5302,
+      "step": 11024
+    },
+    {
+      "epoch": 14.112,
+      "grad_norm": 1.0027779340744019,
+      "learning_rate": 2.796518607442977e-05,
+      "loss": 0.5303,
+      "step": 11025
+    },
+    {
+      "epoch": 14.11328,
+      "grad_norm": 0.9702111482620239,
+      "learning_rate": 2.7963185274109642e-05,
+      "loss": 0.5138,
+      "step": 11026
+    },
+    {
+      "epoch": 14.11456,
+      "grad_norm": 1.0045280456542969,
+      "learning_rate": 2.7961184473789514e-05,
+      "loss": 0.5441,
+      "step": 11027
+    },
+    {
+      "epoch": 14.11584,
+      "grad_norm": 0.9333431720733643,
+      "learning_rate": 2.7959183673469393e-05,
+      "loss": 0.5327,
+      "step": 11028
+    },
+    {
+      "epoch": 14.11712,
+      "grad_norm": 0.9155805110931396,
+      "learning_rate": 2.795718287314926e-05,
+      "loss": 0.4828,
+      "step": 11029
+    },
+    {
+      "epoch": 14.1184,
+      "grad_norm": 0.9476427435874939,
+      "learning_rate": 2.7955182072829133e-05,
+      "loss": 0.4894,
+      "step": 11030
+    },
+    {
+      "epoch": 14.11968,
+      "grad_norm": 0.9640700817108154,
+      "learning_rate": 2.7953181272509005e-05,
+      "loss": 0.5322,
+      "step": 11031
+    },
+    {
+      "epoch": 14.12096,
+      "grad_norm": 0.9432242512702942,
+      "learning_rate": 2.7951180472188877e-05,
+      "loss": 0.5413,
+      "step": 11032
+    },
+    {
+      "epoch": 14.12224,
+      "grad_norm": 0.9809675216674805,
+      "learning_rate": 2.7949179671868745e-05,
+      "loss": 0.5346,
+      "step": 11033
+    },
+    {
+      "epoch": 14.12352,
+      "grad_norm": 0.9683305621147156,
+      "learning_rate": 2.7947178871548617e-05,
+      "loss": 0.5097,
+      "step": 11034
+    },
+    {
+      "epoch": 14.1248,
+      "grad_norm": 0.9600328803062439,
+      "learning_rate": 2.7945178071228496e-05,
+      "loss": 0.5013,
+      "step": 11035
+    },
+    {
+      "epoch": 14.12608,
+      "grad_norm": 0.9506182670593262,
+      "learning_rate": 2.7943177270908368e-05,
+      "loss": 0.491,
+      "step": 11036
+    },
+    {
+      "epoch": 14.12736,
+      "grad_norm": 0.9323163628578186,
+      "learning_rate": 2.7941176470588236e-05,
+      "loss": 0.5299,
+      "step": 11037
+    },
+    {
+      "epoch": 14.12864,
+      "grad_norm": 0.9958814978599548,
+      "learning_rate": 2.7939175670268108e-05,
+      "loss": 0.5498,
+      "step": 11038
+    },
+    {
+      "epoch": 14.12992,
+      "grad_norm": 0.9677280187606812,
+      "learning_rate": 2.793717486994798e-05,
+      "loss": 0.5417,
+      "step": 11039
+    },
+    {
+      "epoch": 14.1312,
+      "grad_norm": 0.9545553922653198,
+      "learning_rate": 2.7935174069627852e-05,
+      "loss": 0.4917,
+      "step": 11040
+    },
+    {
+      "epoch": 14.13248,
+      "grad_norm": 0.9735832214355469,
+      "learning_rate": 2.793317326930772e-05,
+      "loss": 0.5313,
+      "step": 11041
+    },
+    {
+      "epoch": 14.13376,
+      "grad_norm": 0.9970044493675232,
+      "learning_rate": 2.79311724689876e-05,
+      "loss": 0.5481,
+      "step": 11042
+    },
+    {
+      "epoch": 14.13504,
+      "grad_norm": 0.9880550503730774,
+      "learning_rate": 2.792917166866747e-05,
+      "loss": 0.5117,
+      "step": 11043
+    },
+    {
+      "epoch": 14.13632,
+      "grad_norm": 0.9498565196990967,
+      "learning_rate": 2.7927170868347343e-05,
+      "loss": 0.5318,
+      "step": 11044
+    },
+    {
+      "epoch": 14.1376,
+      "grad_norm": 0.9987189173698425,
+      "learning_rate": 2.792517006802721e-05,
+      "loss": 0.5665,
+      "step": 11045
+    },
+    {
+      "epoch": 14.13888,
+      "grad_norm": 0.9754243493080139,
+      "learning_rate": 2.7923169267707083e-05,
+      "loss": 0.5245,
+      "step": 11046
+    },
+    {
+      "epoch": 14.14016,
+      "grad_norm": 0.9672796130180359,
+      "learning_rate": 2.7921168467386955e-05,
+      "loss": 0.5555,
+      "step": 11047
+    },
+    {
+      "epoch": 14.14144,
+      "grad_norm": 0.9213780760765076,
+      "learning_rate": 2.7919167667066827e-05,
+      "loss": 0.4929,
+      "step": 11048
+    },
+    {
+      "epoch": 14.14272,
+      "grad_norm": 0.893765389919281,
+      "learning_rate": 2.7917166866746702e-05,
+      "loss": 0.4783,
+      "step": 11049
+    },
+    {
+      "epoch": 14.144,
+      "grad_norm": 0.9702581763267517,
+      "learning_rate": 2.7915166066426574e-05,
+      "loss": 0.5511,
+      "step": 11050
+    },
+    {
+      "epoch": 14.14528,
+      "grad_norm": 0.9254672527313232,
+      "learning_rate": 2.7913165266106446e-05,
+      "loss": 0.5087,
+      "step": 11051
+    },
+    {
+      "epoch": 14.14656,
+      "grad_norm": 0.9144052267074585,
+      "learning_rate": 2.7911164465786317e-05,
+      "loss": 0.5502,
+      "step": 11052
+    },
+    {
+      "epoch": 14.14784,
+      "grad_norm": 0.937017560005188,
+      "learning_rate": 2.7909163665466186e-05,
+      "loss": 0.5214,
+      "step": 11053
+    },
+    {
+      "epoch": 14.14912,
+      "grad_norm": 0.9533378481864929,
+      "learning_rate": 2.7907162865146058e-05,
+      "loss": 0.5256,
+      "step": 11054
+    },
+    {
+      "epoch": 14.1504,
+      "grad_norm": 1.019874095916748,
+      "learning_rate": 2.790516206482593e-05,
+      "loss": 0.5798,
+      "step": 11055
+    },
+    {
+      "epoch": 14.15168,
+      "grad_norm": 1.0000838041305542,
+      "learning_rate": 2.7903161264505805e-05,
+      "loss": 0.5462,
+      "step": 11056
+    },
+    {
+      "epoch": 14.15296,
+      "grad_norm": 0.9633542895317078,
+      "learning_rate": 2.7901160464185677e-05,
+      "loss": 0.5429,
+      "step": 11057
+    },
+    {
+      "epoch": 14.15424,
+      "grad_norm": 0.929341733455658,
+      "learning_rate": 2.789915966386555e-05,
+      "loss": 0.5231,
+      "step": 11058
+    },
+    {
+      "epoch": 14.15552,
+      "grad_norm": 0.9698876142501831,
+      "learning_rate": 2.789715886354542e-05,
+      "loss": 0.5094,
+      "step": 11059
+    },
+    {
+      "epoch": 14.1568,
+      "grad_norm": 0.9775946736335754,
+      "learning_rate": 2.7895158063225292e-05,
+      "loss": 0.5122,
+      "step": 11060
+    },
+    {
+      "epoch": 14.15808,
+      "grad_norm": 0.991977870464325,
+      "learning_rate": 2.789315726290516e-05,
+      "loss": 0.5952,
+      "step": 11061
+    },
+    {
+      "epoch": 14.15936,
+      "grad_norm": 0.9593073129653931,
+      "learning_rate": 2.7891156462585033e-05,
+      "loss": 0.4648,
+      "step": 11062
+    },
+    {
+      "epoch": 14.16064,
+      "grad_norm": 0.9783948063850403,
+      "learning_rate": 2.788915566226491e-05,
+      "loss": 0.5765,
+      "step": 11063
+    },
+    {
+      "epoch": 14.16192,
+      "grad_norm": 0.982947826385498,
+      "learning_rate": 2.788715486194478e-05,
+      "loss": 0.5264,
+      "step": 11064
+    },
+    {
+      "epoch": 14.1632,
+      "grad_norm": 0.9915079474449158,
+      "learning_rate": 2.788515406162465e-05,
+      "loss": 0.5371,
+      "step": 11065
+    },
+    {
+      "epoch": 14.16448,
+      "grad_norm": 0.9839328527450562,
+      "learning_rate": 2.7883153261304523e-05,
+      "loss": 0.5434,
+      "step": 11066
+    },
+    {
+      "epoch": 14.16576,
+      "grad_norm": 0.9635565280914307,
+      "learning_rate": 2.7881152460984395e-05,
+      "loss": 0.5017,
+      "step": 11067
+    },
+    {
+      "epoch": 14.16704,
+      "grad_norm": 0.9151342511177063,
+      "learning_rate": 2.7879151660664267e-05,
+      "loss": 0.5008,
+      "step": 11068
+    },
+    {
+      "epoch": 14.16832,
+      "grad_norm": 0.9379957914352417,
+      "learning_rate": 2.7877150860344136e-05,
+      "loss": 0.5265,
+      "step": 11069
+    },
+    {
+      "epoch": 14.1696,
+      "grad_norm": 0.9489511251449585,
+      "learning_rate": 2.7875150060024014e-05,
+      "loss": 0.4828,
+      "step": 11070
+    },
+    {
+      "epoch": 14.17088,
+      "grad_norm": 0.9574431777000427,
+      "learning_rate": 2.7873149259703886e-05,
+      "loss": 0.5325,
+      "step": 11071
+    },
+    {
+      "epoch": 14.17216,
+      "grad_norm": 0.9298610091209412,
+      "learning_rate": 2.7871148459383755e-05,
+      "loss": 0.5055,
+      "step": 11072
+    },
+    {
+      "epoch": 14.17344,
+      "grad_norm": 1.00288724899292,
+      "learning_rate": 2.7869147659063626e-05,
+      "loss": 0.5837,
+      "step": 11073
+    },
+    {
+      "epoch": 14.17472,
+      "grad_norm": 0.9784044623374939,
+      "learning_rate": 2.78671468587435e-05,
+      "loss": 0.5425,
+      "step": 11074
+    },
+    {
+      "epoch": 14.176,
+      "grad_norm": 0.9960471391677856,
+      "learning_rate": 2.786514605842337e-05,
+      "loss": 0.5137,
+      "step": 11075
+    },
+    {
+      "epoch": 14.17728,
+      "grad_norm": 1.01795494556427,
+      "learning_rate": 2.7863145258103242e-05,
+      "loss": 0.5483,
+      "step": 11076
+    },
+    {
+      "epoch": 14.17856,
+      "grad_norm": 1.0112005472183228,
+      "learning_rate": 2.7861144457783117e-05,
+      "loss": 0.5657,
+      "step": 11077
+    },
+    {
+      "epoch": 14.17984,
+      "grad_norm": 0.9345349073410034,
+      "learning_rate": 2.785914365746299e-05,
+      "loss": 0.5126,
+      "step": 11078
+    },
+    {
+      "epoch": 14.18112,
+      "grad_norm": 0.9684646725654602,
+      "learning_rate": 2.785714285714286e-05,
+      "loss": 0.4951,
+      "step": 11079
+    },
+    {
+      "epoch": 14.1824,
+      "grad_norm": 0.9888444542884827,
+      "learning_rate": 2.785514205682273e-05,
+      "loss": 0.5313,
+      "step": 11080
+    },
+    {
+      "epoch": 14.18368,
+      "grad_norm": 0.9643538594245911,
+      "learning_rate": 2.78531412565026e-05,
+      "loss": 0.5066,
+      "step": 11081
+    },
+    {
+      "epoch": 14.18496,
+      "grad_norm": 1.0346754789352417,
+      "learning_rate": 2.7851140456182473e-05,
+      "loss": 0.5683,
+      "step": 11082
+    },
+    {
+      "epoch": 14.18624,
+      "grad_norm": 0.9812901020050049,
+      "learning_rate": 2.7849139655862345e-05,
+      "loss": 0.5395,
+      "step": 11083
+    },
+    {
+      "epoch": 14.18752,
+      "grad_norm": 1.0163651704788208,
+      "learning_rate": 2.784713885554222e-05,
+      "loss": 0.5592,
+      "step": 11084
+    },
+    {
+      "epoch": 14.1888,
+      "grad_norm": 0.9411138892173767,
+      "learning_rate": 2.7845138055222092e-05,
+      "loss": 0.5013,
+      "step": 11085
+    },
+    {
+      "epoch": 14.19008,
+      "grad_norm": 0.9784141182899475,
+      "learning_rate": 2.7843137254901964e-05,
+      "loss": 0.5097,
+      "step": 11086
+    },
+    {
+      "epoch": 14.19136,
+      "grad_norm": 0.9341786503791809,
+      "learning_rate": 2.7841136454581836e-05,
+      "loss": 0.5253,
+      "step": 11087
+    },
+    {
+      "epoch": 14.19264,
+      "grad_norm": 0.9971571564674377,
+      "learning_rate": 2.7839135654261704e-05,
+      "loss": 0.5746,
+      "step": 11088
+    },
+    {
+      "epoch": 14.19392,
+      "grad_norm": 0.9248690009117126,
+      "learning_rate": 2.7837134853941576e-05,
+      "loss": 0.5079,
+      "step": 11089
+    },
+    {
+      "epoch": 14.1952,
+      "grad_norm": 0.9347603917121887,
+      "learning_rate": 2.7835134053621448e-05,
+      "loss": 0.5212,
+      "step": 11090
+    },
+    {
+      "epoch": 14.19648,
+      "grad_norm": 0.9436104893684387,
+      "learning_rate": 2.7833133253301323e-05,
+      "loss": 0.5332,
+      "step": 11091
+    },
+    {
+      "epoch": 14.19776,
+      "grad_norm": 0.9841783046722412,
+      "learning_rate": 2.7831132452981195e-05,
+      "loss": 0.524,
+      "step": 11092
+    },
+    {
+      "epoch": 14.19904,
+      "grad_norm": 0.9630305171012878,
+      "learning_rate": 2.7829131652661067e-05,
+      "loss": 0.5511,
+      "step": 11093
+    },
+    {
+      "epoch": 14.20032,
+      "grad_norm": 0.9855948090553284,
+      "learning_rate": 2.782713085234094e-05,
+      "loss": 0.573,
+      "step": 11094
+    },
+    {
+      "epoch": 14.2016,
+      "grad_norm": 0.9675716161727905,
+      "learning_rate": 2.782513005202081e-05,
+      "loss": 0.5066,
+      "step": 11095
+    },
+    {
+      "epoch": 14.20288,
+      "grad_norm": 1.0122812986373901,
+      "learning_rate": 2.782312925170068e-05,
+      "loss": 0.59,
+      "step": 11096
+    },
+    {
+      "epoch": 14.20416,
+      "grad_norm": 0.9795871376991272,
+      "learning_rate": 2.782112845138055e-05,
+      "loss": 0.5633,
+      "step": 11097
+    },
+    {
+      "epoch": 14.20544,
+      "grad_norm": 0.9450654983520508,
+      "learning_rate": 2.781912765106043e-05,
+      "loss": 0.5224,
+      "step": 11098
+    },
+    {
+      "epoch": 14.20672,
+      "grad_norm": 0.9646390080451965,
+      "learning_rate": 2.7817126850740298e-05,
+      "loss": 0.5386,
+      "step": 11099
+    },
+    {
+      "epoch": 14.208,
+      "grad_norm": 0.9584577679634094,
+      "learning_rate": 2.781512605042017e-05,
+      "loss": 0.493,
+      "step": 11100
+    },
+    {
+      "epoch": 14.20928,
+      "grad_norm": 1.0012218952178955,
+      "learning_rate": 2.7813125250100042e-05,
+      "loss": 0.5086,
+      "step": 11101
+    },
+    {
+      "epoch": 14.21056,
+      "grad_norm": 0.9794235229492188,
+      "learning_rate": 2.7811124449779914e-05,
+      "loss": 0.5184,
+      "step": 11102
+    },
+    {
+      "epoch": 14.21184,
+      "grad_norm": 0.9602649211883545,
+      "learning_rate": 2.7809123649459786e-05,
+      "loss": 0.558,
+      "step": 11103
+    },
+    {
+      "epoch": 14.21312,
+      "grad_norm": 0.90842205286026,
+      "learning_rate": 2.7807122849139654e-05,
+      "loss": 0.4799,
+      "step": 11104
+    },
+    {
+      "epoch": 14.2144,
+      "grad_norm": 0.9867070913314819,
+      "learning_rate": 2.7805122048819533e-05,
+      "loss": 0.5553,
+      "step": 11105
+    },
+    {
+      "epoch": 14.21568,
+      "grad_norm": 0.9673285484313965,
+      "learning_rate": 2.7803121248499405e-05,
+      "loss": 0.5077,
+      "step": 11106
+    },
+    {
+      "epoch": 14.21696,
+      "grad_norm": 1.000418782234192,
+      "learning_rate": 2.7801120448179273e-05,
+      "loss": 0.5389,
+      "step": 11107
+    },
+    {
+      "epoch": 14.21824,
+      "grad_norm": 0.9656979441642761,
+      "learning_rate": 2.7799119647859145e-05,
+      "loss": 0.5318,
+      "step": 11108
+    },
+    {
+      "epoch": 14.21952,
+      "grad_norm": 0.9900228381156921,
+      "learning_rate": 2.7797118847539017e-05,
+      "loss": 0.5487,
+      "step": 11109
+    },
+    {
+      "epoch": 14.2208,
+      "grad_norm": 0.9100193977355957,
+      "learning_rate": 2.779511804721889e-05,
+      "loss": 0.4788,
+      "step": 11110
+    },
+    {
+      "epoch": 14.22208,
+      "grad_norm": 0.9718891978263855,
+      "learning_rate": 2.779311724689876e-05,
+      "loss": 0.5599,
+      "step": 11111
+    },
+    {
+      "epoch": 14.22336,
+      "grad_norm": 1.0352201461791992,
+      "learning_rate": 2.7791116446578636e-05,
+      "loss": 0.5693,
+      "step": 11112
+    },
+    {
+      "epoch": 14.22464,
+      "grad_norm": 0.9154449105262756,
+      "learning_rate": 2.7789115646258508e-05,
+      "loss": 0.5197,
+      "step": 11113
+    },
+    {
+      "epoch": 14.22592,
+      "grad_norm": 0.9313166737556458,
+      "learning_rate": 2.778711484593838e-05,
+      "loss": 0.5425,
+      "step": 11114
+    },
+    {
+      "epoch": 14.2272,
+      "grad_norm": 0.9067119359970093,
+      "learning_rate": 2.7785114045618248e-05,
+      "loss": 0.4738,
+      "step": 11115
+    },
+    {
+      "epoch": 14.22848,
+      "grad_norm": 0.919452428817749,
+      "learning_rate": 2.778311324529812e-05,
+      "loss": 0.4808,
+      "step": 11116
+    },
+    {
+      "epoch": 14.22976,
+      "grad_norm": 0.9764316082000732,
+      "learning_rate": 2.778111244497799e-05,
+      "loss": 0.5508,
+      "step": 11117
+    },
+    {
+      "epoch": 14.23104,
+      "grad_norm": 1.002416968345642,
+      "learning_rate": 2.7779111644657863e-05,
+      "loss": 0.5727,
+      "step": 11118
+    },
+    {
+      "epoch": 14.23232,
+      "grad_norm": 0.9473809599876404,
+      "learning_rate": 2.7777110844337735e-05,
+      "loss": 0.515,
+      "step": 11119
+    },
+    {
+      "epoch": 14.2336,
+      "grad_norm": 0.9346858263015747,
+      "learning_rate": 2.777511004401761e-05,
+      "loss": 0.5267,
+      "step": 11120
+    },
+    {
+      "epoch": 14.23488,
+      "grad_norm": 0.9307919144630432,
+      "learning_rate": 2.7773109243697482e-05,
+      "loss": 0.4998,
+      "step": 11121
+    },
+    {
+      "epoch": 14.23616,
+      "grad_norm": 0.9973614811897278,
+      "learning_rate": 2.7771108443377354e-05,
+      "loss": 0.582,
+      "step": 11122
+    },
+    {
+      "epoch": 14.23744,
+      "grad_norm": 0.994784414768219,
+      "learning_rate": 2.7769107643057223e-05,
+      "loss": 0.5207,
+      "step": 11123
+    },
+    {
+      "epoch": 14.23872,
+      "grad_norm": 0.9563934206962585,
+      "learning_rate": 2.7767106842737095e-05,
+      "loss": 0.4937,
+      "step": 11124
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 0.9174085855484009,
+      "learning_rate": 2.7765106042416966e-05,
+      "loss": 0.5371,
+      "step": 11125
+    },
+    {
+      "epoch": 14.24128,
+      "grad_norm": 0.9807153940200806,
+      "learning_rate": 2.7763105242096838e-05,
+      "loss": 0.5334,
+      "step": 11126
+    },
+    {
+      "epoch": 14.24256,
+      "grad_norm": 0.9536086916923523,
+      "learning_rate": 2.7761104441776714e-05,
+      "loss": 0.5261,
+      "step": 11127
+    },
+    {
+      "epoch": 14.24384,
+      "grad_norm": 0.9382270574569702,
+      "learning_rate": 2.7759103641456585e-05,
+      "loss": 0.5195,
+      "step": 11128
+    },
+    {
+      "epoch": 14.24512,
+      "grad_norm": 0.98002028465271,
+      "learning_rate": 2.7757102841136457e-05,
+      "loss": 0.5247,
+      "step": 11129
+    },
+    {
+      "epoch": 14.2464,
+      "grad_norm": 0.940831184387207,
+      "learning_rate": 2.775510204081633e-05,
+      "loss": 0.4827,
+      "step": 11130
+    },
+    {
+      "epoch": 14.24768,
+      "grad_norm": 0.8894546031951904,
+      "learning_rate": 2.7753101240496198e-05,
+      "loss": 0.5129,
+      "step": 11131
+    },
+    {
+      "epoch": 14.24896,
+      "grad_norm": 0.8587113618850708,
+      "learning_rate": 2.775110044017607e-05,
+      "loss": 0.4721,
+      "step": 11132
+    },
+    {
+      "epoch": 14.25024,
+      "grad_norm": 0.9889330863952637,
+      "learning_rate": 2.774909963985594e-05,
+      "loss": 0.573,
+      "step": 11133
+    },
+    {
+      "epoch": 14.25152,
+      "grad_norm": 0.9522219300270081,
+      "learning_rate": 2.7747098839535817e-05,
+      "loss": 0.4859,
+      "step": 11134
+    },
+    {
+      "epoch": 14.2528,
+      "grad_norm": 1.0236855745315552,
+      "learning_rate": 2.774509803921569e-05,
+      "loss": 0.5632,
+      "step": 11135
+    },
+    {
+      "epoch": 14.25408,
+      "grad_norm": 0.9715608358383179,
+      "learning_rate": 2.774309723889556e-05,
+      "loss": 0.5655,
+      "step": 11136
+    },
+    {
+      "epoch": 14.25536,
+      "grad_norm": 0.9080358147621155,
+      "learning_rate": 2.7741096438575432e-05,
+      "loss": 0.5086,
+      "step": 11137
+    },
+    {
+      "epoch": 14.25664,
+      "grad_norm": 0.9496445059776306,
+      "learning_rate": 2.7739095638255304e-05,
+      "loss": 0.5476,
+      "step": 11138
+    },
+    {
+      "epoch": 14.25792,
+      "grad_norm": 0.9452351331710815,
+      "learning_rate": 2.7737094837935172e-05,
+      "loss": 0.4651,
+      "step": 11139
+    },
+    {
+      "epoch": 14.2592,
+      "grad_norm": 0.9179670810699463,
+      "learning_rate": 2.7735094037615044e-05,
+      "loss": 0.523,
+      "step": 11140
+    },
+    {
+      "epoch": 14.26048,
+      "grad_norm": 1.0410470962524414,
+      "learning_rate": 2.7733093237294923e-05,
+      "loss": 0.5671,
+      "step": 11141
+    },
+    {
+      "epoch": 14.26176,
+      "grad_norm": 1.0138649940490723,
+      "learning_rate": 2.773109243697479e-05,
+      "loss": 0.5479,
+      "step": 11142
+    },
+    {
+      "epoch": 14.26304,
+      "grad_norm": 1.071816325187683,
+      "learning_rate": 2.7729091636654663e-05,
+      "loss": 0.5894,
+      "step": 11143
+    },
+    {
+      "epoch": 14.26432,
+      "grad_norm": 0.9818564653396606,
+      "learning_rate": 2.7727090836334535e-05,
+      "loss": 0.5264,
+      "step": 11144
+    },
+    {
+      "epoch": 14.2656,
+      "grad_norm": 1.001348853111267,
+      "learning_rate": 2.7725090036014407e-05,
+      "loss": 0.4765,
+      "step": 11145
+    },
+    {
+      "epoch": 14.26688,
+      "grad_norm": 1.0460922718048096,
+      "learning_rate": 2.772308923569428e-05,
+      "loss": 0.5337,
+      "step": 11146
+    },
+    {
+      "epoch": 14.26816,
+      "grad_norm": 0.9920286536216736,
+      "learning_rate": 2.7721088435374147e-05,
+      "loss": 0.5336,
+      "step": 11147
+    },
+    {
+      "epoch": 14.26944,
+      "grad_norm": 0.9713577628135681,
+      "learning_rate": 2.7719087635054026e-05,
+      "loss": 0.5586,
+      "step": 11148
+    },
+    {
+      "epoch": 14.27072,
+      "grad_norm": 0.9439414143562317,
+      "learning_rate": 2.7717086834733898e-05,
+      "loss": 0.5438,
+      "step": 11149
+    },
+    {
+      "epoch": 14.272,
+      "grad_norm": 0.956102192401886,
+      "learning_rate": 2.7715086034413766e-05,
+      "loss": 0.5383,
+      "step": 11150
+    },
+    {
+      "epoch": 14.27328,
+      "grad_norm": 0.9565941095352173,
+      "learning_rate": 2.7713085234093638e-05,
+      "loss": 0.5267,
+      "step": 11151
+    },
+    {
+      "epoch": 14.27456,
+      "grad_norm": 0.9341673851013184,
+      "learning_rate": 2.771108443377351e-05,
+      "loss": 0.5157,
+      "step": 11152
+    },
+    {
+      "epoch": 14.27584,
+      "grad_norm": 0.948688268661499,
+      "learning_rate": 2.7709083633453382e-05,
+      "loss": 0.5132,
+      "step": 11153
+    },
+    {
+      "epoch": 14.27712,
+      "grad_norm": 1.0164926052093506,
+      "learning_rate": 2.7707082833133254e-05,
+      "loss": 0.5641,
+      "step": 11154
+    },
+    {
+      "epoch": 14.2784,
+      "grad_norm": 0.9440988302230835,
+      "learning_rate": 2.770508203281313e-05,
+      "loss": 0.5071,
+      "step": 11155
+    },
+    {
+      "epoch": 14.27968,
+      "grad_norm": 0.9467429518699646,
+      "learning_rate": 2.7703081232493e-05,
+      "loss": 0.4859,
+      "step": 11156
+    },
+    {
+      "epoch": 14.28096,
+      "grad_norm": 0.923644483089447,
+      "learning_rate": 2.7701080432172873e-05,
+      "loss": 0.4908,
+      "step": 11157
+    },
+    {
+      "epoch": 14.28224,
+      "grad_norm": 0.924755334854126,
+      "learning_rate": 2.769907963185274e-05,
+      "loss": 0.5031,
+      "step": 11158
+    },
+    {
+      "epoch": 14.28352,
+      "grad_norm": 0.971537709236145,
+      "learning_rate": 2.7697078831532613e-05,
+      "loss": 0.55,
+      "step": 11159
+    },
+    {
+      "epoch": 14.2848,
+      "grad_norm": 0.9279425144195557,
+      "learning_rate": 2.7695078031212485e-05,
+      "loss": 0.4737,
+      "step": 11160
+    },
+    {
+      "epoch": 14.28608,
+      "grad_norm": 1.0148504972457886,
+      "learning_rate": 2.7693077230892357e-05,
+      "loss": 0.562,
+      "step": 11161
+    },
+    {
+      "epoch": 14.28736,
+      "grad_norm": 1.0114786624908447,
+      "learning_rate": 2.7691076430572232e-05,
+      "loss": 0.5845,
+      "step": 11162
+    },
+    {
+      "epoch": 14.288640000000001,
+      "grad_norm": 1.0205031633377075,
+      "learning_rate": 2.7689075630252104e-05,
+      "loss": 0.5958,
+      "step": 11163
+    },
+    {
+      "epoch": 14.28992,
+      "grad_norm": 1.0409082174301147,
+      "learning_rate": 2.7687074829931976e-05,
+      "loss": 0.5586,
+      "step": 11164
+    },
+    {
+      "epoch": 14.2912,
+      "grad_norm": 0.9876707196235657,
+      "learning_rate": 2.7685074029611847e-05,
+      "loss": 0.5519,
+      "step": 11165
+    },
+    {
+      "epoch": 14.29248,
+      "grad_norm": 1.0097538232803345,
+      "learning_rate": 2.7683073229291716e-05,
+      "loss": 0.5301,
+      "step": 11166
+    },
+    {
+      "epoch": 14.29376,
+      "grad_norm": 0.9804957509040833,
+      "learning_rate": 2.7681072428971588e-05,
+      "loss": 0.5528,
+      "step": 11167
+    },
+    {
+      "epoch": 14.29504,
+      "grad_norm": 0.9977876543998718,
+      "learning_rate": 2.767907162865146e-05,
+      "loss": 0.5245,
+      "step": 11168
+    },
+    {
+      "epoch": 14.29632,
+      "grad_norm": 0.9734411239624023,
+      "learning_rate": 2.7677070828331335e-05,
+      "loss": 0.5306,
+      "step": 11169
+    },
+    {
+      "epoch": 14.2976,
+      "grad_norm": 0.9642010927200317,
+      "learning_rate": 2.7675070028011207e-05,
+      "loss": 0.5472,
+      "step": 11170
+    },
+    {
+      "epoch": 14.29888,
+      "grad_norm": 1.0015202760696411,
+      "learning_rate": 2.767306922769108e-05,
+      "loss": 0.5343,
+      "step": 11171
+    },
+    {
+      "epoch": 14.30016,
+      "grad_norm": 0.9820809364318848,
+      "learning_rate": 2.767106842737095e-05,
+      "loss": 0.5606,
+      "step": 11172
+    },
+    {
+      "epoch": 14.30144,
+      "grad_norm": 0.977370023727417,
+      "learning_rate": 2.7669067627050822e-05,
+      "loss": 0.5411,
+      "step": 11173
+    },
+    {
+      "epoch": 14.30272,
+      "grad_norm": 0.9267496466636658,
+      "learning_rate": 2.766706682673069e-05,
+      "loss": 0.4905,
+      "step": 11174
+    },
+    {
+      "epoch": 14.304,
+      "grad_norm": 0.9372310042381287,
+      "learning_rate": 2.7665066026410563e-05,
+      "loss": 0.521,
+      "step": 11175
+    },
+    {
+      "epoch": 14.30528,
+      "grad_norm": 0.9425508975982666,
+      "learning_rate": 2.766306522609044e-05,
+      "loss": 0.5014,
+      "step": 11176
+    },
+    {
+      "epoch": 14.30656,
+      "grad_norm": 0.9979614019393921,
+      "learning_rate": 2.766106442577031e-05,
+      "loss": 0.5129,
+      "step": 11177
+    },
+    {
+      "epoch": 14.30784,
+      "grad_norm": 1.0363802909851074,
+      "learning_rate": 2.765906362545018e-05,
+      "loss": 0.5654,
+      "step": 11178
+    },
+    {
+      "epoch": 14.30912,
+      "grad_norm": 1.0523051023483276,
+      "learning_rate": 2.7657062825130053e-05,
+      "loss": 0.5443,
+      "step": 11179
+    },
+    {
+      "epoch": 14.3104,
+      "grad_norm": 0.9782631993293762,
+      "learning_rate": 2.7655062024809925e-05,
+      "loss": 0.5205,
+      "step": 11180
+    },
+    {
+      "epoch": 14.31168,
+      "grad_norm": 0.9576646685600281,
+      "learning_rate": 2.7653061224489797e-05,
+      "loss": 0.5151,
+      "step": 11181
+    },
+    {
+      "epoch": 14.31296,
+      "grad_norm": 0.9194750189781189,
+      "learning_rate": 2.7651060424169666e-05,
+      "loss": 0.5067,
+      "step": 11182
+    },
+    {
+      "epoch": 14.31424,
+      "grad_norm": 0.9091538786888123,
+      "learning_rate": 2.7649059623849544e-05,
+      "loss": 0.4496,
+      "step": 11183
+    },
+    {
+      "epoch": 14.31552,
+      "grad_norm": 1.0529427528381348,
+      "learning_rate": 2.7647058823529416e-05,
+      "loss": 0.6212,
+      "step": 11184
+    },
+    {
+      "epoch": 14.3168,
+      "grad_norm": 1.0431804656982422,
+      "learning_rate": 2.7645058023209285e-05,
+      "loss": 0.6295,
+      "step": 11185
+    },
+    {
+      "epoch": 14.31808,
+      "grad_norm": 0.9597348570823669,
+      "learning_rate": 2.7643057222889156e-05,
+      "loss": 0.52,
+      "step": 11186
+    },
+    {
+      "epoch": 14.31936,
+      "grad_norm": 1.0246033668518066,
+      "learning_rate": 2.7641056422569028e-05,
+      "loss": 0.5113,
+      "step": 11187
+    },
+    {
+      "epoch": 14.32064,
+      "grad_norm": 0.9777224659919739,
+      "learning_rate": 2.76390556222489e-05,
+      "loss": 0.4957,
+      "step": 11188
+    },
+    {
+      "epoch": 14.32192,
+      "grad_norm": 0.9850444793701172,
+      "learning_rate": 2.7637054821928772e-05,
+      "loss": 0.5251,
+      "step": 11189
+    },
+    {
+      "epoch": 14.3232,
+      "grad_norm": 0.9483627080917358,
+      "learning_rate": 2.7635054021608647e-05,
+      "loss": 0.5139,
+      "step": 11190
+    },
+    {
+      "epoch": 14.32448,
+      "grad_norm": 0.9668890833854675,
+      "learning_rate": 2.763305322128852e-05,
+      "loss": 0.5483,
+      "step": 11191
+    },
+    {
+      "epoch": 14.32576,
+      "grad_norm": 0.9546102285385132,
+      "learning_rate": 2.763105242096839e-05,
+      "loss": 0.5344,
+      "step": 11192
+    },
+    {
+      "epoch": 14.32704,
+      "grad_norm": 0.9604453444480896,
+      "learning_rate": 2.762905162064826e-05,
+      "loss": 0.5439,
+      "step": 11193
+    },
+    {
+      "epoch": 14.32832,
+      "grad_norm": 0.984946072101593,
+      "learning_rate": 2.762705082032813e-05,
+      "loss": 0.5376,
+      "step": 11194
+    },
+    {
+      "epoch": 14.3296,
+      "grad_norm": 1.022682547569275,
+      "learning_rate": 2.7625050020008003e-05,
+      "loss": 0.5918,
+      "step": 11195
+    },
+    {
+      "epoch": 14.33088,
+      "grad_norm": 0.9947453141212463,
+      "learning_rate": 2.7623049219687875e-05,
+      "loss": 0.556,
+      "step": 11196
+    },
+    {
+      "epoch": 14.33216,
+      "grad_norm": 0.9552220106124878,
+      "learning_rate": 2.762104841936775e-05,
+      "loss": 0.5282,
+      "step": 11197
+    },
+    {
+      "epoch": 14.33344,
+      "grad_norm": 0.9819296598434448,
+      "learning_rate": 2.7619047619047622e-05,
+      "loss": 0.474,
+      "step": 11198
+    },
+    {
+      "epoch": 14.33472,
+      "grad_norm": 1.0161857604980469,
+      "learning_rate": 2.7617046818727494e-05,
+      "loss": 0.5428,
+      "step": 11199
+    },
+    {
+      "epoch": 14.336,
+      "grad_norm": 0.950628399848938,
+      "learning_rate": 2.7615046018407366e-05,
+      "loss": 0.5011,
+      "step": 11200
+    },
+    {
+      "epoch": 14.33728,
+      "grad_norm": 0.9224458336830139,
+      "learning_rate": 2.7613045218087234e-05,
+      "loss": 0.5053,
+      "step": 11201
+    },
+    {
+      "epoch": 14.33856,
+      "grad_norm": 0.9512698650360107,
+      "learning_rate": 2.7611044417767106e-05,
+      "loss": 0.5124,
+      "step": 11202
+    },
+    {
+      "epoch": 14.33984,
+      "grad_norm": 0.905447244644165,
+      "learning_rate": 2.7609043617446978e-05,
+      "loss": 0.4627,
+      "step": 11203
+    },
+    {
+      "epoch": 14.34112,
+      "grad_norm": 0.9614179730415344,
+      "learning_rate": 2.7607042817126853e-05,
+      "loss": 0.5108,
+      "step": 11204
+    },
+    {
+      "epoch": 14.3424,
+      "grad_norm": 1.0302612781524658,
+      "learning_rate": 2.7605042016806725e-05,
+      "loss": 0.588,
+      "step": 11205
+    },
+    {
+      "epoch": 14.343679999999999,
+      "grad_norm": 0.9715131521224976,
+      "learning_rate": 2.7603041216486597e-05,
+      "loss": 0.5215,
+      "step": 11206
+    },
+    {
+      "epoch": 14.34496,
+      "grad_norm": 0.9700056314468384,
+      "learning_rate": 2.760104041616647e-05,
+      "loss": 0.511,
+      "step": 11207
+    },
+    {
+      "epoch": 14.34624,
+      "grad_norm": 0.9388474225997925,
+      "learning_rate": 2.759903961584634e-05,
+      "loss": 0.5359,
+      "step": 11208
+    },
+    {
+      "epoch": 14.34752,
+      "grad_norm": 0.9849310517311096,
+      "learning_rate": 2.759703881552621e-05,
+      "loss": 0.5323,
+      "step": 11209
+    },
+    {
+      "epoch": 14.3488,
+      "grad_norm": 0.990974485874176,
+      "learning_rate": 2.759503801520608e-05,
+      "loss": 0.5483,
+      "step": 11210
+    },
+    {
+      "epoch": 14.35008,
+      "grad_norm": 1.0002540349960327,
+      "learning_rate": 2.759303721488596e-05,
+      "loss": 0.5135,
+      "step": 11211
+    },
+    {
+      "epoch": 14.35136,
+      "grad_norm": 0.9734601378440857,
+      "learning_rate": 2.7591036414565828e-05,
+      "loss": 0.5341,
+      "step": 11212
+    },
+    {
+      "epoch": 14.35264,
+      "grad_norm": 0.9525935649871826,
+      "learning_rate": 2.75890356142457e-05,
+      "loss": 0.5107,
+      "step": 11213
+    },
+    {
+      "epoch": 14.35392,
+      "grad_norm": 0.9455302357673645,
+      "learning_rate": 2.7587034813925572e-05,
+      "loss": 0.528,
+      "step": 11214
+    },
+    {
+      "epoch": 14.3552,
+      "grad_norm": 0.9510558843612671,
+      "learning_rate": 2.7585034013605444e-05,
+      "loss": 0.5205,
+      "step": 11215
+    },
+    {
+      "epoch": 14.35648,
+      "grad_norm": 0.9504905939102173,
+      "learning_rate": 2.7583033213285316e-05,
+      "loss": 0.5393,
+      "step": 11216
+    },
+    {
+      "epoch": 14.35776,
+      "grad_norm": 0.952308714389801,
+      "learning_rate": 2.7581032412965184e-05,
+      "loss": 0.5079,
+      "step": 11217
+    },
+    {
+      "epoch": 14.35904,
+      "grad_norm": 1.0244531631469727,
+      "learning_rate": 2.7579031612645063e-05,
+      "loss": 0.5758,
+      "step": 11218
+    },
+    {
+      "epoch": 14.36032,
+      "grad_norm": 0.9500291347503662,
+      "learning_rate": 2.7577030812324934e-05,
+      "loss": 0.5063,
+      "step": 11219
+    },
+    {
+      "epoch": 14.3616,
+      "grad_norm": 0.9819626808166504,
+      "learning_rate": 2.7575030012004803e-05,
+      "loss": 0.5453,
+      "step": 11220
+    },
+    {
+      "epoch": 14.36288,
+      "grad_norm": 0.9494471549987793,
+      "learning_rate": 2.7573029211684675e-05,
+      "loss": 0.525,
+      "step": 11221
+    },
+    {
+      "epoch": 14.36416,
+      "grad_norm": 0.9260940551757812,
+      "learning_rate": 2.7571028411364547e-05,
+      "loss": 0.5371,
+      "step": 11222
+    },
+    {
+      "epoch": 14.36544,
+      "grad_norm": 0.9827948808670044,
+      "learning_rate": 2.756902761104442e-05,
+      "loss": 0.5559,
+      "step": 11223
+    },
+    {
+      "epoch": 14.36672,
+      "grad_norm": 0.9918002486228943,
+      "learning_rate": 2.756702681072429e-05,
+      "loss": 0.5184,
+      "step": 11224
+    },
+    {
+      "epoch": 14.368,
+      "grad_norm": 0.9910105466842651,
+      "learning_rate": 2.756502601040416e-05,
+      "loss": 0.571,
+      "step": 11225
+    },
+    {
+      "epoch": 14.36928,
+      "grad_norm": 1.0316925048828125,
+      "learning_rate": 2.7563025210084037e-05,
+      "loss": 0.5355,
+      "step": 11226
+    },
+    {
+      "epoch": 14.37056,
+      "grad_norm": 0.9141017198562622,
+      "learning_rate": 2.756102440976391e-05,
+      "loss": 0.4943,
+      "step": 11227
+    },
+    {
+      "epoch": 14.37184,
+      "grad_norm": 0.9571624994277954,
+      "learning_rate": 2.7559023609443778e-05,
+      "loss": 0.5288,
+      "step": 11228
+    },
+    {
+      "epoch": 14.37312,
+      "grad_norm": 0.95545893907547,
+      "learning_rate": 2.755702280912365e-05,
+      "loss": 0.5231,
+      "step": 11229
+    },
+    {
+      "epoch": 14.3744,
+      "grad_norm": 0.8849372863769531,
+      "learning_rate": 2.755502200880352e-05,
+      "loss": 0.4974,
+      "step": 11230
+    },
+    {
+      "epoch": 14.37568,
+      "grad_norm": 0.9513952136039734,
+      "learning_rate": 2.7553021208483393e-05,
+      "loss": 0.5517,
+      "step": 11231
+    },
+    {
+      "epoch": 14.37696,
+      "grad_norm": 0.9136958122253418,
+      "learning_rate": 2.7551020408163265e-05,
+      "loss": 0.5258,
+      "step": 11232
+    },
+    {
+      "epoch": 14.37824,
+      "grad_norm": 0.9673413038253784,
+      "learning_rate": 2.754901960784314e-05,
+      "loss": 0.5092,
+      "step": 11233
+    },
+    {
+      "epoch": 14.37952,
+      "grad_norm": 0.9891186952590942,
+      "learning_rate": 2.7547018807523012e-05,
+      "loss": 0.5038,
+      "step": 11234
+    },
+    {
+      "epoch": 14.3808,
+      "grad_norm": 0.9878220558166504,
+      "learning_rate": 2.7545018007202884e-05,
+      "loss": 0.5578,
+      "step": 11235
+    },
+    {
+      "epoch": 14.38208,
+      "grad_norm": 0.9703464508056641,
+      "learning_rate": 2.7543017206882753e-05,
+      "loss": 0.5463,
+      "step": 11236
+    },
+    {
+      "epoch": 14.38336,
+      "grad_norm": 1.003145456314087,
+      "learning_rate": 2.7541016406562625e-05,
+      "loss": 0.5513,
+      "step": 11237
+    },
+    {
+      "epoch": 14.38464,
+      "grad_norm": 0.9494891166687012,
+      "learning_rate": 2.7539015606242496e-05,
+      "loss": 0.4822,
+      "step": 11238
+    },
+    {
+      "epoch": 14.38592,
+      "grad_norm": 0.9667163491249084,
+      "learning_rate": 2.7537014805922368e-05,
+      "loss": 0.5371,
+      "step": 11239
+    },
+    {
+      "epoch": 14.3872,
+      "grad_norm": 0.9678400754928589,
+      "learning_rate": 2.7535014005602243e-05,
+      "loss": 0.5323,
+      "step": 11240
+    },
+    {
+      "epoch": 14.38848,
+      "grad_norm": 1.001518964767456,
+      "learning_rate": 2.7533013205282115e-05,
+      "loss": 0.5314,
+      "step": 11241
+    },
+    {
+      "epoch": 14.38976,
+      "grad_norm": 1.0005130767822266,
+      "learning_rate": 2.7531012404961987e-05,
+      "loss": 0.5229,
+      "step": 11242
+    },
+    {
+      "epoch": 14.39104,
+      "grad_norm": 0.9950253367424011,
+      "learning_rate": 2.752901160464186e-05,
+      "loss": 0.5628,
+      "step": 11243
+    },
+    {
+      "epoch": 14.39232,
+      "grad_norm": 0.8834913969039917,
+      "learning_rate": 2.7527010804321728e-05,
+      "loss": 0.4599,
+      "step": 11244
+    },
+    {
+      "epoch": 14.3936,
+      "grad_norm": 0.9402278661727905,
+      "learning_rate": 2.75250100040016e-05,
+      "loss": 0.5007,
+      "step": 11245
+    },
+    {
+      "epoch": 14.39488,
+      "grad_norm": 1.0537763833999634,
+      "learning_rate": 2.752300920368147e-05,
+      "loss": 0.569,
+      "step": 11246
+    },
+    {
+      "epoch": 14.39616,
+      "grad_norm": 1.0012062788009644,
+      "learning_rate": 2.7521008403361346e-05,
+      "loss": 0.5324,
+      "step": 11247
+    },
+    {
+      "epoch": 14.39744,
+      "grad_norm": 0.9675048589706421,
+      "learning_rate": 2.751900760304122e-05,
+      "loss": 0.5517,
+      "step": 11248
+    },
+    {
+      "epoch": 14.39872,
+      "grad_norm": 0.9507377743721008,
+      "learning_rate": 2.751700680272109e-05,
+      "loss": 0.5041,
+      "step": 11249
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 0.9572357535362244,
+      "learning_rate": 2.7515006002400962e-05,
+      "loss": 0.5373,
+      "step": 11250
+    },
+    {
+      "epoch": 14.40128,
+      "grad_norm": 0.9814470410346985,
+      "learning_rate": 2.7513005202080834e-05,
+      "loss": 0.5297,
+      "step": 11251
+    },
+    {
+      "epoch": 14.40256,
+      "grad_norm": 0.9352920651435852,
+      "learning_rate": 2.7511004401760702e-05,
+      "loss": 0.4801,
+      "step": 11252
+    },
+    {
+      "epoch": 14.40384,
+      "grad_norm": 0.9089069962501526,
+      "learning_rate": 2.7509003601440574e-05,
+      "loss": 0.4697,
+      "step": 11253
+    },
+    {
+      "epoch": 14.40512,
+      "grad_norm": 0.9568341374397278,
+      "learning_rate": 2.7507002801120453e-05,
+      "loss": 0.53,
+      "step": 11254
+    },
+    {
+      "epoch": 14.4064,
+      "grad_norm": 0.9579848647117615,
+      "learning_rate": 2.750500200080032e-05,
+      "loss": 0.4911,
+      "step": 11255
+    },
+    {
+      "epoch": 14.40768,
+      "grad_norm": 0.9732376337051392,
+      "learning_rate": 2.7503001200480193e-05,
+      "loss": 0.5377,
+      "step": 11256
+    },
+    {
+      "epoch": 14.40896,
+      "grad_norm": 0.9632934331893921,
+      "learning_rate": 2.7501000400160065e-05,
+      "loss": 0.5283,
+      "step": 11257
+    },
+    {
+      "epoch": 14.41024,
+      "grad_norm": 0.9508453011512756,
+      "learning_rate": 2.7498999599839937e-05,
+      "loss": 0.5232,
+      "step": 11258
+    },
+    {
+      "epoch": 14.41152,
+      "grad_norm": 0.9188300371170044,
+      "learning_rate": 2.749699879951981e-05,
+      "loss": 0.522,
+      "step": 11259
+    },
+    {
+      "epoch": 14.4128,
+      "grad_norm": 0.9479809999465942,
+      "learning_rate": 2.7494997999199677e-05,
+      "loss": 0.5126,
+      "step": 11260
+    },
+    {
+      "epoch": 14.41408,
+      "grad_norm": 0.9927715063095093,
+      "learning_rate": 2.7492997198879556e-05,
+      "loss": 0.5527,
+      "step": 11261
+    },
+    {
+      "epoch": 14.41536,
+      "grad_norm": 1.0165045261383057,
+      "learning_rate": 2.7490996398559428e-05,
+      "loss": 0.5617,
+      "step": 11262
+    },
+    {
+      "epoch": 14.41664,
+      "grad_norm": 0.9869691133499146,
+      "learning_rate": 2.7488995598239296e-05,
+      "loss": 0.5413,
+      "step": 11263
+    },
+    {
+      "epoch": 14.41792,
+      "grad_norm": 0.9490245580673218,
+      "learning_rate": 2.7486994797919168e-05,
+      "loss": 0.4778,
+      "step": 11264
+    },
+    {
+      "epoch": 14.4192,
+      "grad_norm": 0.967729389667511,
+      "learning_rate": 2.748499399759904e-05,
+      "loss": 0.5177,
+      "step": 11265
+    },
+    {
+      "epoch": 14.42048,
+      "grad_norm": 0.9395437240600586,
+      "learning_rate": 2.7482993197278912e-05,
+      "loss": 0.5426,
+      "step": 11266
+    },
+    {
+      "epoch": 14.42176,
+      "grad_norm": 0.9222921133041382,
+      "learning_rate": 2.7480992396958784e-05,
+      "loss": 0.4713,
+      "step": 11267
+    },
+    {
+      "epoch": 14.42304,
+      "grad_norm": 0.9864060878753662,
+      "learning_rate": 2.747899159663866e-05,
+      "loss": 0.5087,
+      "step": 11268
+    },
+    {
+      "epoch": 14.42432,
+      "grad_norm": 0.9627559781074524,
+      "learning_rate": 2.747699079631853e-05,
+      "loss": 0.4965,
+      "step": 11269
+    },
+    {
+      "epoch": 14.4256,
+      "grad_norm": 0.9115942120552063,
+      "learning_rate": 2.7474989995998403e-05,
+      "loss": 0.4682,
+      "step": 11270
+    },
+    {
+      "epoch": 14.42688,
+      "grad_norm": 0.96644127368927,
+      "learning_rate": 2.747298919567827e-05,
+      "loss": 0.532,
+      "step": 11271
+    },
+    {
+      "epoch": 14.42816,
+      "grad_norm": 1.0033857822418213,
+      "learning_rate": 2.7470988395358143e-05,
+      "loss": 0.5577,
+      "step": 11272
+    },
+    {
+      "epoch": 14.42944,
+      "grad_norm": 1.0521118640899658,
+      "learning_rate": 2.7468987595038015e-05,
+      "loss": 0.5645,
+      "step": 11273
+    },
+    {
+      "epoch": 14.43072,
+      "grad_norm": 0.985821008682251,
+      "learning_rate": 2.7466986794717887e-05,
+      "loss": 0.5584,
+      "step": 11274
+    },
+    {
+      "epoch": 14.432,
+      "grad_norm": 1.0375871658325195,
+      "learning_rate": 2.7464985994397762e-05,
+      "loss": 0.5631,
+      "step": 11275
+    },
+    {
+      "epoch": 14.43328,
+      "grad_norm": 0.935610294342041,
+      "learning_rate": 2.7462985194077634e-05,
+      "loss": 0.4976,
+      "step": 11276
+    },
+    {
+      "epoch": 14.43456,
+      "grad_norm": 1.0138261318206787,
+      "learning_rate": 2.7460984393757506e-05,
+      "loss": 0.5645,
+      "step": 11277
+    },
+    {
+      "epoch": 14.43584,
+      "grad_norm": 1.0506336688995361,
+      "learning_rate": 2.7458983593437377e-05,
+      "loss": 0.5702,
+      "step": 11278
+    },
+    {
+      "epoch": 14.43712,
+      "grad_norm": 0.951737105846405,
+      "learning_rate": 2.7456982793117246e-05,
+      "loss": 0.4803,
+      "step": 11279
+    },
+    {
+      "epoch": 14.4384,
+      "grad_norm": 0.9894746541976929,
+      "learning_rate": 2.7454981992797118e-05,
+      "loss": 0.5372,
+      "step": 11280
+    },
+    {
+      "epoch": 14.43968,
+      "grad_norm": 0.9843027591705322,
+      "learning_rate": 2.745298119247699e-05,
+      "loss": 0.5739,
+      "step": 11281
+    },
+    {
+      "epoch": 14.44096,
+      "grad_norm": 0.980569064617157,
+      "learning_rate": 2.7450980392156865e-05,
+      "loss": 0.4889,
+      "step": 11282
+    },
+    {
+      "epoch": 14.44224,
+      "grad_norm": 0.9253454208374023,
+      "learning_rate": 2.7448979591836737e-05,
+      "loss": 0.5233,
+      "step": 11283
+    },
+    {
+      "epoch": 14.44352,
+      "grad_norm": 0.9626498818397522,
+      "learning_rate": 2.744697879151661e-05,
+      "loss": 0.5467,
+      "step": 11284
+    },
+    {
+      "epoch": 14.4448,
+      "grad_norm": 0.9599905014038086,
+      "learning_rate": 2.744497799119648e-05,
+      "loss": 0.5341,
+      "step": 11285
+    },
+    {
+      "epoch": 14.44608,
+      "grad_norm": 0.9758439064025879,
+      "learning_rate": 2.7442977190876352e-05,
+      "loss": 0.5201,
+      "step": 11286
+    },
+    {
+      "epoch": 14.44736,
+      "grad_norm": 1.003113865852356,
+      "learning_rate": 2.744097639055622e-05,
+      "loss": 0.523,
+      "step": 11287
+    },
+    {
+      "epoch": 14.44864,
+      "grad_norm": 0.9655167460441589,
+      "learning_rate": 2.7438975590236093e-05,
+      "loss": 0.5333,
+      "step": 11288
+    },
+    {
+      "epoch": 14.44992,
+      "grad_norm": 0.9504974484443665,
+      "learning_rate": 2.743697478991597e-05,
+      "loss": 0.5356,
+      "step": 11289
+    },
+    {
+      "epoch": 14.4512,
+      "grad_norm": 0.978461503982544,
+      "learning_rate": 2.743497398959584e-05,
+      "loss": 0.5443,
+      "step": 11290
+    },
+    {
+      "epoch": 14.45248,
+      "grad_norm": 0.9474901556968689,
+      "learning_rate": 2.743297318927571e-05,
+      "loss": 0.489,
+      "step": 11291
+    },
+    {
+      "epoch": 14.45376,
+      "grad_norm": 0.9023638367652893,
+      "learning_rate": 2.7430972388955583e-05,
+      "loss": 0.4626,
+      "step": 11292
+    },
+    {
+      "epoch": 14.45504,
+      "grad_norm": 0.939704418182373,
+      "learning_rate": 2.7428971588635455e-05,
+      "loss": 0.4852,
+      "step": 11293
+    },
+    {
+      "epoch": 14.45632,
+      "grad_norm": 0.9497430920600891,
+      "learning_rate": 2.7426970788315327e-05,
+      "loss": 0.496,
+      "step": 11294
+    },
+    {
+      "epoch": 14.4576,
+      "grad_norm": 0.9100688695907593,
+      "learning_rate": 2.7424969987995196e-05,
+      "loss": 0.5156,
+      "step": 11295
+    },
+    {
+      "epoch": 14.45888,
+      "grad_norm": 0.9579876661300659,
+      "learning_rate": 2.7422969187675074e-05,
+      "loss": 0.493,
+      "step": 11296
+    },
+    {
+      "epoch": 14.46016,
+      "grad_norm": 0.9984821677207947,
+      "learning_rate": 2.7420968387354946e-05,
+      "loss": 0.5359,
+      "step": 11297
+    },
+    {
+      "epoch": 14.46144,
+      "grad_norm": 0.9672602415084839,
+      "learning_rate": 2.7418967587034815e-05,
+      "loss": 0.5613,
+      "step": 11298
+    },
+    {
+      "epoch": 14.462720000000001,
+      "grad_norm": 1.0040628910064697,
+      "learning_rate": 2.7416966786714686e-05,
+      "loss": 0.5751,
+      "step": 11299
+    },
+    {
+      "epoch": 14.464,
+      "grad_norm": 1.0041166543960571,
+      "learning_rate": 2.7414965986394558e-05,
+      "loss": 0.5865,
+      "step": 11300
+    },
+    {
+      "epoch": 14.46528,
+      "grad_norm": 0.950853168964386,
+      "learning_rate": 2.741296518607443e-05,
+      "loss": 0.5002,
+      "step": 11301
+    },
+    {
+      "epoch": 14.46656,
+      "grad_norm": 0.9579245448112488,
+      "learning_rate": 2.7410964385754302e-05,
+      "loss": 0.5227,
+      "step": 11302
+    },
+    {
+      "epoch": 14.46784,
+      "grad_norm": 0.9853748679161072,
+      "learning_rate": 2.7408963585434177e-05,
+      "loss": 0.5181,
+      "step": 11303
+    },
+    {
+      "epoch": 14.46912,
+      "grad_norm": 0.9100056290626526,
+      "learning_rate": 2.740696278511405e-05,
+      "loss": 0.4892,
+      "step": 11304
+    },
+    {
+      "epoch": 14.4704,
+      "grad_norm": 0.94825679063797,
+      "learning_rate": 2.740496198479392e-05,
+      "loss": 0.4969,
+      "step": 11305
+    },
+    {
+      "epoch": 14.47168,
+      "grad_norm": 0.9380550384521484,
+      "learning_rate": 2.740296118447379e-05,
+      "loss": 0.5133,
+      "step": 11306
+    },
+    {
+      "epoch": 14.47296,
+      "grad_norm": 0.9455284476280212,
+      "learning_rate": 2.740096038415366e-05,
+      "loss": 0.5307,
+      "step": 11307
+    },
+    {
+      "epoch": 14.47424,
+      "grad_norm": 0.95405513048172,
+      "learning_rate": 2.7398959583833533e-05,
+      "loss": 0.4977,
+      "step": 11308
+    },
+    {
+      "epoch": 14.47552,
+      "grad_norm": 0.9854465126991272,
+      "learning_rate": 2.7396958783513405e-05,
+      "loss": 0.5289,
+      "step": 11309
+    },
+    {
+      "epoch": 14.4768,
+      "grad_norm": 1.018682599067688,
+      "learning_rate": 2.739495798319328e-05,
+      "loss": 0.5819,
+      "step": 11310
+    },
+    {
+      "epoch": 14.47808,
+      "grad_norm": 0.9085429310798645,
+      "learning_rate": 2.7392957182873152e-05,
+      "loss": 0.4768,
+      "step": 11311
+    },
+    {
+      "epoch": 14.47936,
+      "grad_norm": 1.025566577911377,
+      "learning_rate": 2.7390956382553024e-05,
+      "loss": 0.5631,
+      "step": 11312
+    },
+    {
+      "epoch": 14.48064,
+      "grad_norm": 0.9369559288024902,
+      "learning_rate": 2.7388955582232896e-05,
+      "loss": 0.5121,
+      "step": 11313
+    },
+    {
+      "epoch": 14.48192,
+      "grad_norm": 0.9811607003211975,
+      "learning_rate": 2.7386954781912764e-05,
+      "loss": 0.5434,
+      "step": 11314
+    },
+    {
+      "epoch": 14.4832,
+      "grad_norm": 0.9261855483055115,
+      "learning_rate": 2.7384953981592636e-05,
+      "loss": 0.547,
+      "step": 11315
+    },
+    {
+      "epoch": 14.48448,
+      "grad_norm": 0.9290302395820618,
+      "learning_rate": 2.7382953181272508e-05,
+      "loss": 0.5299,
+      "step": 11316
+    },
+    {
+      "epoch": 14.48576,
+      "grad_norm": 0.9527462124824524,
+      "learning_rate": 2.7380952380952383e-05,
+      "loss": 0.5147,
+      "step": 11317
+    },
+    {
+      "epoch": 14.48704,
+      "grad_norm": 0.9608799815177917,
+      "learning_rate": 2.7378951580632255e-05,
+      "loss": 0.5475,
+      "step": 11318
+    },
+    {
+      "epoch": 14.48832,
+      "grad_norm": 0.9241880774497986,
+      "learning_rate": 2.7376950780312127e-05,
+      "loss": 0.5598,
+      "step": 11319
+    },
+    {
+      "epoch": 14.4896,
+      "grad_norm": 0.948318362236023,
+      "learning_rate": 2.7374949979992e-05,
+      "loss": 0.5819,
+      "step": 11320
+    },
+    {
+      "epoch": 14.49088,
+      "grad_norm": 0.9118000268936157,
+      "learning_rate": 2.737294917967187e-05,
+      "loss": 0.488,
+      "step": 11321
+    },
+    {
+      "epoch": 14.49216,
+      "grad_norm": 0.9568605422973633,
+      "learning_rate": 2.737094837935174e-05,
+      "loss": 0.5289,
+      "step": 11322
+    },
+    {
+      "epoch": 14.49344,
+      "grad_norm": 0.9479710459709167,
+      "learning_rate": 2.736894757903161e-05,
+      "loss": 0.5361,
+      "step": 11323
+    },
+    {
+      "epoch": 14.49472,
+      "grad_norm": 0.9818686246871948,
+      "learning_rate": 2.736694677871149e-05,
+      "loss": 0.5237,
+      "step": 11324
+    },
+    {
+      "epoch": 14.496,
+      "grad_norm": 0.9445642828941345,
+      "learning_rate": 2.7364945978391358e-05,
+      "loss": 0.4977,
+      "step": 11325
+    },
+    {
+      "epoch": 14.49728,
+      "grad_norm": 0.9737011790275574,
+      "learning_rate": 2.736294517807123e-05,
+      "loss": 0.5182,
+      "step": 11326
+    },
+    {
+      "epoch": 14.49856,
+      "grad_norm": 0.9840577244758606,
+      "learning_rate": 2.7360944377751102e-05,
+      "loss": 0.5137,
+      "step": 11327
+    },
+    {
+      "epoch": 14.49984,
+      "grad_norm": 0.9450149536132812,
+      "learning_rate": 2.7358943577430974e-05,
+      "loss": 0.4973,
+      "step": 11328
+    },
+    {
+      "epoch": 14.50112,
+      "grad_norm": 0.9480723738670349,
+      "learning_rate": 2.7356942777110846e-05,
+      "loss": 0.5709,
+      "step": 11329
+    },
+    {
+      "epoch": 14.5024,
+      "grad_norm": 0.9991519451141357,
+      "learning_rate": 2.7354941976790714e-05,
+      "loss": 0.5573,
+      "step": 11330
+    },
+    {
+      "epoch": 14.50368,
+      "grad_norm": 0.9923670291900635,
+      "learning_rate": 2.7352941176470593e-05,
+      "loss": 0.6224,
+      "step": 11331
+    },
+    {
+      "epoch": 14.50496,
+      "grad_norm": 0.9765622615814209,
+      "learning_rate": 2.7350940376150464e-05,
+      "loss": 0.5312,
+      "step": 11332
+    },
+    {
+      "epoch": 14.50624,
+      "grad_norm": 0.978398859500885,
+      "learning_rate": 2.7348939575830333e-05,
+      "loss": 0.5265,
+      "step": 11333
+    },
+    {
+      "epoch": 14.50752,
+      "grad_norm": 1.0118024349212646,
+      "learning_rate": 2.7346938775510205e-05,
+      "loss": 0.5901,
+      "step": 11334
+    },
+    {
+      "epoch": 14.5088,
+      "grad_norm": 0.9619473814964294,
+      "learning_rate": 2.7344937975190077e-05,
+      "loss": 0.5405,
+      "step": 11335
+    },
+    {
+      "epoch": 14.51008,
+      "grad_norm": 0.970476508140564,
+      "learning_rate": 2.734293717486995e-05,
+      "loss": 0.5543,
+      "step": 11336
+    },
+    {
+      "epoch": 14.51136,
+      "grad_norm": 0.9238805174827576,
+      "learning_rate": 2.734093637454982e-05,
+      "loss": 0.4825,
+      "step": 11337
+    },
+    {
+      "epoch": 14.51264,
+      "grad_norm": 1.0360538959503174,
+      "learning_rate": 2.733893557422969e-05,
+      "loss": 0.5385,
+      "step": 11338
+    },
+    {
+      "epoch": 14.51392,
+      "grad_norm": 0.9870235323905945,
+      "learning_rate": 2.7336934773909567e-05,
+      "loss": 0.5012,
+      "step": 11339
+    },
+    {
+      "epoch": 14.5152,
+      "grad_norm": 0.9728274345397949,
+      "learning_rate": 2.733493397358944e-05,
+      "loss": 0.5382,
+      "step": 11340
+    },
+    {
+      "epoch": 14.51648,
+      "grad_norm": 0.989501416683197,
+      "learning_rate": 2.7332933173269308e-05,
+      "loss": 0.5216,
+      "step": 11341
+    },
+    {
+      "epoch": 14.517759999999999,
+      "grad_norm": 0.987844705581665,
+      "learning_rate": 2.733093237294918e-05,
+      "loss": 0.5553,
+      "step": 11342
+    },
+    {
+      "epoch": 14.51904,
+      "grad_norm": 0.9757768511772156,
+      "learning_rate": 2.732893157262905e-05,
+      "loss": 0.5604,
+      "step": 11343
+    },
+    {
+      "epoch": 14.52032,
+      "grad_norm": 0.9981015920639038,
+      "learning_rate": 2.7326930772308923e-05,
+      "loss": 0.5521,
+      "step": 11344
+    },
+    {
+      "epoch": 14.5216,
+      "grad_norm": 0.9795306921005249,
+      "learning_rate": 2.7324929971988795e-05,
+      "loss": 0.5164,
+      "step": 11345
+    },
+    {
+      "epoch": 14.52288,
+      "grad_norm": 0.9881823062896729,
+      "learning_rate": 2.732292917166867e-05,
+      "loss": 0.519,
+      "step": 11346
+    },
+    {
+      "epoch": 14.52416,
+      "grad_norm": 1.0085053443908691,
+      "learning_rate": 2.7320928371348542e-05,
+      "loss": 0.5578,
+      "step": 11347
+    },
+    {
+      "epoch": 14.52544,
+      "grad_norm": 1.0021884441375732,
+      "learning_rate": 2.7318927571028414e-05,
+      "loss": 0.5362,
+      "step": 11348
+    },
+    {
+      "epoch": 14.52672,
+      "grad_norm": 0.9783837795257568,
+      "learning_rate": 2.7316926770708283e-05,
+      "loss": 0.5256,
+      "step": 11349
+    },
+    {
+      "epoch": 14.528,
+      "grad_norm": 1.005036473274231,
+      "learning_rate": 2.7314925970388155e-05,
+      "loss": 0.5495,
+      "step": 11350
+    },
+    {
+      "epoch": 14.52928,
+      "grad_norm": 0.9625844359397888,
+      "learning_rate": 2.7312925170068026e-05,
+      "loss": 0.5355,
+      "step": 11351
+    },
+    {
+      "epoch": 14.53056,
+      "grad_norm": 0.961077094078064,
+      "learning_rate": 2.7310924369747898e-05,
+      "loss": 0.518,
+      "step": 11352
+    },
+    {
+      "epoch": 14.53184,
+      "grad_norm": 0.9395869970321655,
+      "learning_rate": 2.7308923569427773e-05,
+      "loss": 0.5194,
+      "step": 11353
+    },
+    {
+      "epoch": 14.53312,
+      "grad_norm": 0.9758129715919495,
+      "learning_rate": 2.7306922769107645e-05,
+      "loss": 0.5533,
+      "step": 11354
+    },
+    {
+      "epoch": 14.5344,
+      "grad_norm": 0.9722764492034912,
+      "learning_rate": 2.7304921968787517e-05,
+      "loss": 0.536,
+      "step": 11355
+    },
+    {
+      "epoch": 14.53568,
+      "grad_norm": 0.9164559245109558,
+      "learning_rate": 2.730292116846739e-05,
+      "loss": 0.468,
+      "step": 11356
+    },
+    {
+      "epoch": 14.53696,
+      "grad_norm": 1.0066019296646118,
+      "learning_rate": 2.7300920368147258e-05,
+      "loss": 0.5945,
+      "step": 11357
+    },
+    {
+      "epoch": 14.53824,
+      "grad_norm": 0.9370825886726379,
+      "learning_rate": 2.729891956782713e-05,
+      "loss": 0.5005,
+      "step": 11358
+    },
+    {
+      "epoch": 14.53952,
+      "grad_norm": 0.9669891595840454,
+      "learning_rate": 2.7296918767507e-05,
+      "loss": 0.5536,
+      "step": 11359
+    },
+    {
+      "epoch": 14.5408,
+      "grad_norm": 0.9711414575576782,
+      "learning_rate": 2.729491796718688e-05,
+      "loss": 0.5114,
+      "step": 11360
+    },
+    {
+      "epoch": 14.54208,
+      "grad_norm": 0.9631317257881165,
+      "learning_rate": 2.729291716686675e-05,
+      "loss": 0.5221,
+      "step": 11361
+    },
+    {
+      "epoch": 14.54336,
+      "grad_norm": 0.9969499707221985,
+      "learning_rate": 2.729091636654662e-05,
+      "loss": 0.5097,
+      "step": 11362
+    },
+    {
+      "epoch": 14.54464,
+      "grad_norm": 0.9977988004684448,
+      "learning_rate": 2.7288915566226492e-05,
+      "loss": 0.5569,
+      "step": 11363
+    },
+    {
+      "epoch": 14.54592,
+      "grad_norm": 0.90044105052948,
+      "learning_rate": 2.7286914765906364e-05,
+      "loss": 0.4707,
+      "step": 11364
+    },
+    {
+      "epoch": 14.5472,
+      "grad_norm": 0.9298081398010254,
+      "learning_rate": 2.7284913965586232e-05,
+      "loss": 0.5013,
+      "step": 11365
+    },
+    {
+      "epoch": 14.54848,
+      "grad_norm": 0.987301766872406,
+      "learning_rate": 2.7282913165266104e-05,
+      "loss": 0.5666,
+      "step": 11366
+    },
+    {
+      "epoch": 14.54976,
+      "grad_norm": 0.9727098345756531,
+      "learning_rate": 2.7280912364945983e-05,
+      "loss": 0.5379,
+      "step": 11367
+    },
+    {
+      "epoch": 14.55104,
+      "grad_norm": 0.9952585101127625,
+      "learning_rate": 2.7278911564625855e-05,
+      "loss": 0.5741,
+      "step": 11368
+    },
+    {
+      "epoch": 14.55232,
+      "grad_norm": 0.9384405612945557,
+      "learning_rate": 2.7276910764305723e-05,
+      "loss": 0.502,
+      "step": 11369
+    },
+    {
+      "epoch": 14.5536,
+      "grad_norm": 0.9742892980575562,
+      "learning_rate": 2.7274909963985595e-05,
+      "loss": 0.5407,
+      "step": 11370
+    },
+    {
+      "epoch": 14.55488,
+      "grad_norm": 0.9823377132415771,
+      "learning_rate": 2.7272909163665467e-05,
+      "loss": 0.5218,
+      "step": 11371
+    },
+    {
+      "epoch": 14.55616,
+      "grad_norm": 0.9821147322654724,
+      "learning_rate": 2.727090836334534e-05,
+      "loss": 0.5134,
+      "step": 11372
+    },
+    {
+      "epoch": 14.55744,
+      "grad_norm": 1.0253291130065918,
+      "learning_rate": 2.7268907563025207e-05,
+      "loss": 0.5562,
+      "step": 11373
+    },
+    {
+      "epoch": 14.55872,
+      "grad_norm": 0.9838807582855225,
+      "learning_rate": 2.7266906762705086e-05,
+      "loss": 0.5282,
+      "step": 11374
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 0.9515048861503601,
+      "learning_rate": 2.7264905962384958e-05,
+      "loss": 0.5034,
+      "step": 11375
+    },
+    {
+      "epoch": 14.56128,
+      "grad_norm": 0.9546962976455688,
+      "learning_rate": 2.726290516206483e-05,
+      "loss": 0.5325,
+      "step": 11376
+    },
+    {
+      "epoch": 14.56256,
+      "grad_norm": 0.9903567433357239,
+      "learning_rate": 2.7260904361744698e-05,
+      "loss": 0.5126,
+      "step": 11377
+    },
+    {
+      "epoch": 14.56384,
+      "grad_norm": 0.9805188775062561,
+      "learning_rate": 2.725890356142457e-05,
+      "loss": 0.513,
+      "step": 11378
+    },
+    {
+      "epoch": 14.56512,
+      "grad_norm": 1.0438790321350098,
+      "learning_rate": 2.7256902761104442e-05,
+      "loss": 0.5515,
+      "step": 11379
+    },
+    {
+      "epoch": 14.5664,
+      "grad_norm": 0.9887431263923645,
+      "learning_rate": 2.7254901960784314e-05,
+      "loss": 0.526,
+      "step": 11380
+    },
+    {
+      "epoch": 14.56768,
+      "grad_norm": 0.9955745339393616,
+      "learning_rate": 2.725290116046419e-05,
+      "loss": 0.5261,
+      "step": 11381
+    },
+    {
+      "epoch": 14.56896,
+      "grad_norm": 0.9546988606452942,
+      "learning_rate": 2.725090036014406e-05,
+      "loss": 0.5462,
+      "step": 11382
+    },
+    {
+      "epoch": 14.57024,
+      "grad_norm": 0.966931939125061,
+      "learning_rate": 2.7248899559823933e-05,
+      "loss": 0.5326,
+      "step": 11383
+    },
+    {
+      "epoch": 14.57152,
+      "grad_norm": 0.9506143927574158,
+      "learning_rate": 2.7246898759503804e-05,
+      "loss": 0.4998,
+      "step": 11384
+    },
+    {
+      "epoch": 14.5728,
+      "grad_norm": 0.9455040097236633,
+      "learning_rate": 2.7244897959183673e-05,
+      "loss": 0.5466,
+      "step": 11385
+    },
+    {
+      "epoch": 14.57408,
+      "grad_norm": 0.9447420239448547,
+      "learning_rate": 2.7242897158863545e-05,
+      "loss": 0.4939,
+      "step": 11386
+    },
+    {
+      "epoch": 14.57536,
+      "grad_norm": 0.9344480037689209,
+      "learning_rate": 2.7240896358543417e-05,
+      "loss": 0.5223,
+      "step": 11387
+    },
+    {
+      "epoch": 14.57664,
+      "grad_norm": 0.9362328052520752,
+      "learning_rate": 2.7238895558223292e-05,
+      "loss": 0.5015,
+      "step": 11388
+    },
+    {
+      "epoch": 14.57792,
+      "grad_norm": 0.9401606917381287,
+      "learning_rate": 2.7236894757903164e-05,
+      "loss": 0.4988,
+      "step": 11389
+    },
+    {
+      "epoch": 14.5792,
+      "grad_norm": 0.9665933847427368,
+      "learning_rate": 2.7234893957583036e-05,
+      "loss": 0.5126,
+      "step": 11390
+    },
+    {
+      "epoch": 14.58048,
+      "grad_norm": 0.8998576998710632,
+      "learning_rate": 2.7232893157262907e-05,
+      "loss": 0.505,
+      "step": 11391
+    },
+    {
+      "epoch": 14.58176,
+      "grad_norm": 0.9965265393257141,
+      "learning_rate": 2.723089235694278e-05,
+      "loss": 0.569,
+      "step": 11392
+    },
+    {
+      "epoch": 14.58304,
+      "grad_norm": 0.9181508421897888,
+      "learning_rate": 2.7228891556622648e-05,
+      "loss": 0.4929,
+      "step": 11393
+    },
+    {
+      "epoch": 14.58432,
+      "grad_norm": 0.9306789040565491,
+      "learning_rate": 2.722689075630252e-05,
+      "loss": 0.5179,
+      "step": 11394
+    },
+    {
+      "epoch": 14.5856,
+      "grad_norm": 0.998447597026825,
+      "learning_rate": 2.7224889955982398e-05,
+      "loss": 0.5673,
+      "step": 11395
+    },
+    {
+      "epoch": 14.58688,
+      "grad_norm": 0.9387089014053345,
+      "learning_rate": 2.7222889155662267e-05,
+      "loss": 0.5328,
+      "step": 11396
+    },
+    {
+      "epoch": 14.58816,
+      "grad_norm": 0.9925042986869812,
+      "learning_rate": 2.722088835534214e-05,
+      "loss": 0.5403,
+      "step": 11397
+    },
+    {
+      "epoch": 14.58944,
+      "grad_norm": 0.9419589638710022,
+      "learning_rate": 2.721888755502201e-05,
+      "loss": 0.5322,
+      "step": 11398
+    },
+    {
+      "epoch": 14.59072,
+      "grad_norm": 0.9522401094436646,
+      "learning_rate": 2.7216886754701882e-05,
+      "loss": 0.5214,
+      "step": 11399
+    },
+    {
+      "epoch": 14.592,
+      "grad_norm": 0.9598711729049683,
+      "learning_rate": 2.7214885954381754e-05,
+      "loss": 0.5013,
+      "step": 11400
+    },
+    {
+      "epoch": 14.59328,
+      "grad_norm": 0.9690408706665039,
+      "learning_rate": 2.7212885154061623e-05,
+      "loss": 0.5262,
+      "step": 11401
+    },
+    {
+      "epoch": 14.59456,
+      "grad_norm": 0.9637435078620911,
+      "learning_rate": 2.72108843537415e-05,
+      "loss": 0.5114,
+      "step": 11402
+    },
+    {
+      "epoch": 14.59584,
+      "grad_norm": 0.9893304109573364,
+      "learning_rate": 2.7208883553421373e-05,
+      "loss": 0.5414,
+      "step": 11403
+    },
+    {
+      "epoch": 14.59712,
+      "grad_norm": 1.0310964584350586,
+      "learning_rate": 2.720688275310124e-05,
+      "loss": 0.5245,
+      "step": 11404
+    },
+    {
+      "epoch": 14.5984,
+      "grad_norm": 0.9509328007698059,
+      "learning_rate": 2.7204881952781113e-05,
+      "loss": 0.5463,
+      "step": 11405
+    },
+    {
+      "epoch": 14.59968,
+      "grad_norm": 0.9694564938545227,
+      "learning_rate": 2.7202881152460985e-05,
+      "loss": 0.5419,
+      "step": 11406
+    },
+    {
+      "epoch": 14.60096,
+      "grad_norm": 0.970043420791626,
+      "learning_rate": 2.7200880352140857e-05,
+      "loss": 0.5445,
+      "step": 11407
+    },
+    {
+      "epoch": 14.60224,
+      "grad_norm": 0.9511335492134094,
+      "learning_rate": 2.719887955182073e-05,
+      "loss": 0.5206,
+      "step": 11408
+    },
+    {
+      "epoch": 14.60352,
+      "grad_norm": 1.030502200126648,
+      "learning_rate": 2.7196878751500604e-05,
+      "loss": 0.6213,
+      "step": 11409
+    },
+    {
+      "epoch": 14.604800000000001,
+      "grad_norm": 1.0002635717391968,
+      "learning_rate": 2.7194877951180476e-05,
+      "loss": 0.5467,
+      "step": 11410
+    },
+    {
+      "epoch": 14.60608,
+      "grad_norm": 0.9449097514152527,
+      "learning_rate": 2.7192877150860348e-05,
+      "loss": 0.5446,
+      "step": 11411
+    },
+    {
+      "epoch": 14.60736,
+      "grad_norm": 0.9701420664787292,
+      "learning_rate": 2.7190876350540216e-05,
+      "loss": 0.5615,
+      "step": 11412
+    },
+    {
+      "epoch": 14.60864,
+      "grad_norm": 0.9248257279396057,
+      "learning_rate": 2.7188875550220088e-05,
+      "loss": 0.5041,
+      "step": 11413
+    },
+    {
+      "epoch": 14.60992,
+      "grad_norm": 0.9535519480705261,
+      "learning_rate": 2.718687474989996e-05,
+      "loss": 0.5423,
+      "step": 11414
+    },
+    {
+      "epoch": 14.6112,
+      "grad_norm": 0.9663887619972229,
+      "learning_rate": 2.7184873949579832e-05,
+      "loss": 0.5678,
+      "step": 11415
+    },
+    {
+      "epoch": 14.61248,
+      "grad_norm": 1.0246509313583374,
+      "learning_rate": 2.7182873149259707e-05,
+      "loss": 0.6351,
+      "step": 11416
+    },
+    {
+      "epoch": 14.61376,
+      "grad_norm": 0.9978649616241455,
+      "learning_rate": 2.718087234893958e-05,
+      "loss": 0.5342,
+      "step": 11417
+    },
+    {
+      "epoch": 14.61504,
+      "grad_norm": 0.9514932036399841,
+      "learning_rate": 2.717887154861945e-05,
+      "loss": 0.4586,
+      "step": 11418
+    },
+    {
+      "epoch": 14.61632,
+      "grad_norm": 1.0197707414627075,
+      "learning_rate": 2.7176870748299323e-05,
+      "loss": 0.5932,
+      "step": 11419
+    },
+    {
+      "epoch": 14.6176,
+      "grad_norm": 1.017253041267395,
+      "learning_rate": 2.717486994797919e-05,
+      "loss": 0.5063,
+      "step": 11420
+    },
+    {
+      "epoch": 14.61888,
+      "grad_norm": 0.9654613137245178,
+      "learning_rate": 2.7172869147659063e-05,
+      "loss": 0.5249,
+      "step": 11421
+    },
+    {
+      "epoch": 14.62016,
+      "grad_norm": 0.9739159345626831,
+      "learning_rate": 2.7170868347338935e-05,
+      "loss": 0.5241,
+      "step": 11422
+    },
+    {
+      "epoch": 14.62144,
+      "grad_norm": 0.9443714022636414,
+      "learning_rate": 2.716886754701881e-05,
+      "loss": 0.5465,
+      "step": 11423
+    },
+    {
+      "epoch": 14.62272,
+      "grad_norm": 1.0117498636245728,
+      "learning_rate": 2.7166866746698682e-05,
+      "loss": 0.5488,
+      "step": 11424
+    },
+    {
+      "epoch": 14.624,
+      "grad_norm": 0.9801613092422485,
+      "learning_rate": 2.7164865946378554e-05,
+      "loss": 0.5194,
+      "step": 11425
+    },
+    {
+      "epoch": 14.62528,
+      "grad_norm": 0.9613388180732727,
+      "learning_rate": 2.7162865146058426e-05,
+      "loss": 0.5651,
+      "step": 11426
+    },
+    {
+      "epoch": 14.62656,
+      "grad_norm": 0.9645081758499146,
+      "learning_rate": 2.7160864345738298e-05,
+      "loss": 0.5064,
+      "step": 11427
+    },
+    {
+      "epoch": 14.627839999999999,
+      "grad_norm": 0.9990333318710327,
+      "learning_rate": 2.7158863545418166e-05,
+      "loss": 0.5582,
+      "step": 11428
+    },
+    {
+      "epoch": 14.62912,
+      "grad_norm": 0.9323675036430359,
+      "learning_rate": 2.7156862745098038e-05,
+      "loss": 0.5146,
+      "step": 11429
+    },
+    {
+      "epoch": 14.6304,
+      "grad_norm": 0.9680500030517578,
+      "learning_rate": 2.7154861944777917e-05,
+      "loss": 0.4641,
+      "step": 11430
+    },
+    {
+      "epoch": 14.63168,
+      "grad_norm": 0.9934640526771545,
+      "learning_rate": 2.7152861144457785e-05,
+      "loss": 0.5237,
+      "step": 11431
+    },
+    {
+      "epoch": 14.63296,
+      "grad_norm": 0.9925525188446045,
+      "learning_rate": 2.7150860344137657e-05,
+      "loss": 0.526,
+      "step": 11432
+    },
+    {
+      "epoch": 14.63424,
+      "grad_norm": 0.9799636602401733,
+      "learning_rate": 2.714885954381753e-05,
+      "loss": 0.5483,
+      "step": 11433
+    },
+    {
+      "epoch": 14.63552,
+      "grad_norm": 0.945448637008667,
+      "learning_rate": 2.71468587434974e-05,
+      "loss": 0.5187,
+      "step": 11434
+    },
+    {
+      "epoch": 14.636800000000001,
+      "grad_norm": 0.9392439126968384,
+      "learning_rate": 2.7144857943177272e-05,
+      "loss": 0.5117,
+      "step": 11435
+    },
+    {
+      "epoch": 14.63808,
+      "grad_norm": 0.9761615991592407,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 0.5538,
+      "step": 11436
+    },
+    {
+      "epoch": 14.63936,
+      "grad_norm": 0.9610094428062439,
+      "learning_rate": 2.714085634253702e-05,
+      "loss": 0.5368,
+      "step": 11437
+    },
+    {
+      "epoch": 14.64064,
+      "grad_norm": 0.9939576387405396,
+      "learning_rate": 2.713885554221689e-05,
+      "loss": 0.6027,
+      "step": 11438
+    },
+    {
+      "epoch": 14.64192,
+      "grad_norm": 0.9574117064476013,
+      "learning_rate": 2.713685474189676e-05,
+      "loss": 0.5258,
+      "step": 11439
+    },
+    {
+      "epoch": 14.6432,
+      "grad_norm": 0.9553491473197937,
+      "learning_rate": 2.7134853941576632e-05,
+      "loss": 0.5076,
+      "step": 11440
+    },
+    {
+      "epoch": 14.64448,
+      "grad_norm": 0.9836270809173584,
+      "learning_rate": 2.7132853141256504e-05,
+      "loss": 0.5249,
+      "step": 11441
+    },
+    {
+      "epoch": 14.64576,
+      "grad_norm": 0.9523620009422302,
+      "learning_rate": 2.7130852340936375e-05,
+      "loss": 0.5237,
+      "step": 11442
+    },
+    {
+      "epoch": 14.64704,
+      "grad_norm": 0.9464503526687622,
+      "learning_rate": 2.7128851540616247e-05,
+      "loss": 0.5415,
+      "step": 11443
+    },
+    {
+      "epoch": 14.64832,
+      "grad_norm": 0.9364238381385803,
+      "learning_rate": 2.7126850740296123e-05,
+      "loss": 0.5156,
+      "step": 11444
+    },
+    {
+      "epoch": 14.6496,
+      "grad_norm": 0.943153440952301,
+      "learning_rate": 2.7124849939975994e-05,
+      "loss": 0.516,
+      "step": 11445
+    },
+    {
+      "epoch": 14.65088,
+      "grad_norm": 0.9400473833084106,
+      "learning_rate": 2.7122849139655866e-05,
+      "loss": 0.5315,
+      "step": 11446
+    },
+    {
+      "epoch": 14.65216,
+      "grad_norm": 0.9996716380119324,
+      "learning_rate": 2.7120848339335735e-05,
+      "loss": 0.5381,
+      "step": 11447
+    },
+    {
+      "epoch": 14.65344,
+      "grad_norm": 0.9744866490364075,
+      "learning_rate": 2.7118847539015607e-05,
+      "loss": 0.5502,
+      "step": 11448
+    },
+    {
+      "epoch": 14.65472,
+      "grad_norm": 0.9782332181930542,
+      "learning_rate": 2.711684673869548e-05,
+      "loss": 0.5651,
+      "step": 11449
+    },
+    {
+      "epoch": 14.656,
+      "grad_norm": 1.014737844467163,
+      "learning_rate": 2.711484593837535e-05,
+      "loss": 0.5673,
+      "step": 11450
+    },
+    {
+      "epoch": 14.65728,
+      "grad_norm": 0.9900893568992615,
+      "learning_rate": 2.7112845138055222e-05,
+      "loss": 0.5413,
+      "step": 11451
+    },
+    {
+      "epoch": 14.65856,
+      "grad_norm": 0.9428865909576416,
+      "learning_rate": 2.7110844337735097e-05,
+      "loss": 0.5564,
+      "step": 11452
+    },
+    {
+      "epoch": 14.659839999999999,
+      "grad_norm": 0.973512589931488,
+      "learning_rate": 2.710884353741497e-05,
+      "loss": 0.5656,
+      "step": 11453
+    },
+    {
+      "epoch": 14.66112,
+      "grad_norm": 0.9764522910118103,
+      "learning_rate": 2.710684273709484e-05,
+      "loss": 0.5316,
+      "step": 11454
+    },
+    {
+      "epoch": 14.6624,
+      "grad_norm": 0.962592363357544,
+      "learning_rate": 2.710484193677471e-05,
+      "loss": 0.5184,
+      "step": 11455
+    },
+    {
+      "epoch": 14.66368,
+      "grad_norm": 0.9462336301803589,
+      "learning_rate": 2.710284113645458e-05,
+      "loss": 0.5146,
+      "step": 11456
+    },
+    {
+      "epoch": 14.66496,
+      "grad_norm": 0.9827730655670166,
+      "learning_rate": 2.7100840336134453e-05,
+      "loss": 0.5192,
+      "step": 11457
+    },
+    {
+      "epoch": 14.66624,
+      "grad_norm": 1.0178593397140503,
+      "learning_rate": 2.7098839535814325e-05,
+      "loss": 0.5204,
+      "step": 11458
+    },
+    {
+      "epoch": 14.66752,
+      "grad_norm": 0.9616353511810303,
+      "learning_rate": 2.70968387354942e-05,
+      "loss": 0.5118,
+      "step": 11459
+    },
+    {
+      "epoch": 14.6688,
+      "grad_norm": 0.9945552349090576,
+      "learning_rate": 2.7094837935174072e-05,
+      "loss": 0.5167,
+      "step": 11460
+    },
+    {
+      "epoch": 14.67008,
+      "grad_norm": 0.967586100101471,
+      "learning_rate": 2.7092837134853944e-05,
+      "loss": 0.5346,
+      "step": 11461
+    },
+    {
+      "epoch": 14.67136,
+      "grad_norm": 0.9664031863212585,
+      "learning_rate": 2.7090836334533816e-05,
+      "loss": 0.5105,
+      "step": 11462
+    },
+    {
+      "epoch": 14.67264,
+      "grad_norm": 0.9893897771835327,
+      "learning_rate": 2.7088835534213684e-05,
+      "loss": 0.5307,
+      "step": 11463
+    },
+    {
+      "epoch": 14.67392,
+      "grad_norm": 1.0116097927093506,
+      "learning_rate": 2.7086834733893556e-05,
+      "loss": 0.4891,
+      "step": 11464
+    },
+    {
+      "epoch": 14.6752,
+      "grad_norm": 0.9823921322822571,
+      "learning_rate": 2.7084833933573428e-05,
+      "loss": 0.542,
+      "step": 11465
+    },
+    {
+      "epoch": 14.67648,
+      "grad_norm": 1.0313193798065186,
+      "learning_rate": 2.7082833133253303e-05,
+      "loss": 0.5909,
+      "step": 11466
+    },
+    {
+      "epoch": 14.67776,
+      "grad_norm": 0.9933533072471619,
+      "learning_rate": 2.7080832332933175e-05,
+      "loss": 0.5606,
+      "step": 11467
+    },
+    {
+      "epoch": 14.67904,
+      "grad_norm": 0.9638269543647766,
+      "learning_rate": 2.7078831532613047e-05,
+      "loss": 0.5142,
+      "step": 11468
+    },
+    {
+      "epoch": 14.68032,
+      "grad_norm": 1.006666660308838,
+      "learning_rate": 2.707683073229292e-05,
+      "loss": 0.5522,
+      "step": 11469
+    },
+    {
+      "epoch": 14.6816,
+      "grad_norm": 0.9415097832679749,
+      "learning_rate": 2.707482993197279e-05,
+      "loss": 0.5229,
+      "step": 11470
+    },
+    {
+      "epoch": 14.68288,
+      "grad_norm": 0.9513869881629944,
+      "learning_rate": 2.707282913165266e-05,
+      "loss": 0.5032,
+      "step": 11471
+    },
+    {
+      "epoch": 14.68416,
+      "grad_norm": 1.0023972988128662,
+      "learning_rate": 2.707082833133253e-05,
+      "loss": 0.5571,
+      "step": 11472
+    },
+    {
+      "epoch": 14.68544,
+      "grad_norm": 0.9924998879432678,
+      "learning_rate": 2.706882753101241e-05,
+      "loss": 0.5413,
+      "step": 11473
+    },
+    {
+      "epoch": 14.68672,
+      "grad_norm": 0.9861060380935669,
+      "learning_rate": 2.7066826730692278e-05,
+      "loss": 0.5158,
+      "step": 11474
+    },
+    {
+      "epoch": 14.688,
+      "grad_norm": 0.964842677116394,
+      "learning_rate": 2.706482593037215e-05,
+      "loss": 0.5428,
+      "step": 11475
+    },
+    {
+      "epoch": 14.68928,
+      "grad_norm": 0.9527415633201599,
+      "learning_rate": 2.7062825130052022e-05,
+      "loss": 0.5302,
+      "step": 11476
+    },
+    {
+      "epoch": 14.69056,
+      "grad_norm": 0.9425960779190063,
+      "learning_rate": 2.7060824329731894e-05,
+      "loss": 0.5503,
+      "step": 11477
+    },
+    {
+      "epoch": 14.69184,
+      "grad_norm": 0.9891713261604309,
+      "learning_rate": 2.7058823529411766e-05,
+      "loss": 0.5782,
+      "step": 11478
+    },
+    {
+      "epoch": 14.69312,
+      "grad_norm": 0.9919348359107971,
+      "learning_rate": 2.7056822729091634e-05,
+      "loss": 0.5411,
+      "step": 11479
+    },
+    {
+      "epoch": 14.6944,
+      "grad_norm": 0.9964098334312439,
+      "learning_rate": 2.7054821928771513e-05,
+      "loss": 0.5517,
+      "step": 11480
+    },
+    {
+      "epoch": 14.69568,
+      "grad_norm": 0.9663441181182861,
+      "learning_rate": 2.7052821128451385e-05,
+      "loss": 0.5354,
+      "step": 11481
+    },
+    {
+      "epoch": 14.69696,
+      "grad_norm": 1.0156707763671875,
+      "learning_rate": 2.7050820328131253e-05,
+      "loss": 0.5209,
+      "step": 11482
+    },
+    {
+      "epoch": 14.69824,
+      "grad_norm": 0.9387539625167847,
+      "learning_rate": 2.7048819527811125e-05,
+      "loss": 0.4976,
+      "step": 11483
+    },
+    {
+      "epoch": 14.69952,
+      "grad_norm": 0.9886208772659302,
+      "learning_rate": 2.7046818727490997e-05,
+      "loss": 0.5635,
+      "step": 11484
+    },
+    {
+      "epoch": 14.7008,
+      "grad_norm": 0.9804397225379944,
+      "learning_rate": 2.704481792717087e-05,
+      "loss": 0.5384,
+      "step": 11485
+    },
+    {
+      "epoch": 14.70208,
+      "grad_norm": 0.944187581539154,
+      "learning_rate": 2.704281712685074e-05,
+      "loss": 0.5429,
+      "step": 11486
+    },
+    {
+      "epoch": 14.70336,
+      "grad_norm": 0.9738987684249878,
+      "learning_rate": 2.7040816326530616e-05,
+      "loss": 0.5619,
+      "step": 11487
+    },
+    {
+      "epoch": 14.70464,
+      "grad_norm": 0.9703164100646973,
+      "learning_rate": 2.7038815526210488e-05,
+      "loss": 0.546,
+      "step": 11488
+    },
+    {
+      "epoch": 14.70592,
+      "grad_norm": 0.9917868375778198,
+      "learning_rate": 2.703681472589036e-05,
+      "loss": 0.558,
+      "step": 11489
+    },
+    {
+      "epoch": 14.7072,
+      "grad_norm": 0.9812465906143188,
+      "learning_rate": 2.7034813925570228e-05,
+      "loss": 0.5587,
+      "step": 11490
+    },
+    {
+      "epoch": 14.70848,
+      "grad_norm": 0.9691387414932251,
+      "learning_rate": 2.70328131252501e-05,
+      "loss": 0.5575,
+      "step": 11491
+    },
+    {
+      "epoch": 14.70976,
+      "grad_norm": 0.912678062915802,
+      "learning_rate": 2.7030812324929972e-05,
+      "loss": 0.4773,
+      "step": 11492
+    },
+    {
+      "epoch": 14.71104,
+      "grad_norm": 0.9488722085952759,
+      "learning_rate": 2.7028811524609844e-05,
+      "loss": 0.5633,
+      "step": 11493
+    },
+    {
+      "epoch": 14.71232,
+      "grad_norm": 0.958460807800293,
+      "learning_rate": 2.702681072428972e-05,
+      "loss": 0.509,
+      "step": 11494
+    },
+    {
+      "epoch": 14.7136,
+      "grad_norm": 0.9496520757675171,
+      "learning_rate": 2.702480992396959e-05,
+      "loss": 0.5236,
+      "step": 11495
+    },
+    {
+      "epoch": 14.71488,
+      "grad_norm": 0.9681755900382996,
+      "learning_rate": 2.7022809123649463e-05,
+      "loss": 0.5415,
+      "step": 11496
+    },
+    {
+      "epoch": 14.71616,
+      "grad_norm": 1.0062313079833984,
+      "learning_rate": 2.7020808323329334e-05,
+      "loss": 0.5432,
+      "step": 11497
+    },
+    {
+      "epoch": 14.71744,
+      "grad_norm": 0.962822675704956,
+      "learning_rate": 2.7018807523009203e-05,
+      "loss": 0.5481,
+      "step": 11498
+    },
+    {
+      "epoch": 14.71872,
+      "grad_norm": 1.0065944194793701,
+      "learning_rate": 2.7016806722689075e-05,
+      "loss": 0.5707,
+      "step": 11499
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 0.9401991367340088,
+      "learning_rate": 2.7014805922368947e-05,
+      "loss": 0.5246,
+      "step": 11500
+    },
+    {
+      "epoch": 14.72128,
+      "grad_norm": 0.8997707366943359,
+      "learning_rate": 2.7012805122048822e-05,
+      "loss": 0.4671,
+      "step": 11501
+    },
+    {
+      "epoch": 14.72256,
+      "grad_norm": 0.9574297070503235,
+      "learning_rate": 2.7010804321728694e-05,
+      "loss": 0.5616,
+      "step": 11502
+    },
+    {
+      "epoch": 14.72384,
+      "grad_norm": 0.9319608807563782,
+      "learning_rate": 2.7008803521408566e-05,
+      "loss": 0.5094,
+      "step": 11503
+    },
+    {
+      "epoch": 14.72512,
+      "grad_norm": 0.9995977282524109,
+      "learning_rate": 2.7006802721088437e-05,
+      "loss": 0.5743,
+      "step": 11504
+    },
+    {
+      "epoch": 14.7264,
+      "grad_norm": 0.9879163503646851,
+      "learning_rate": 2.700480192076831e-05,
+      "loss": 0.5305,
+      "step": 11505
+    },
+    {
+      "epoch": 14.72768,
+      "grad_norm": 0.9732846021652222,
+      "learning_rate": 2.7002801120448178e-05,
+      "loss": 0.5276,
+      "step": 11506
+    },
+    {
+      "epoch": 14.72896,
+      "grad_norm": 0.9981286525726318,
+      "learning_rate": 2.700080032012805e-05,
+      "loss": 0.5467,
+      "step": 11507
+    },
+    {
+      "epoch": 14.73024,
+      "grad_norm": 1.0308117866516113,
+      "learning_rate": 2.6998799519807928e-05,
+      "loss": 0.5996,
+      "step": 11508
+    },
+    {
+      "epoch": 14.73152,
+      "grad_norm": 0.9829885363578796,
+      "learning_rate": 2.6996798719487797e-05,
+      "loss": 0.529,
+      "step": 11509
+    },
+    {
+      "epoch": 14.7328,
+      "grad_norm": 0.9994106292724609,
+      "learning_rate": 2.699479791916767e-05,
+      "loss": 0.5486,
+      "step": 11510
+    },
+    {
+      "epoch": 14.73408,
+      "grad_norm": 0.9333558082580566,
+      "learning_rate": 2.699279711884754e-05,
+      "loss": 0.4884,
+      "step": 11511
+    },
+    {
+      "epoch": 14.73536,
+      "grad_norm": 0.9170252680778503,
+      "learning_rate": 2.6990796318527412e-05,
+      "loss": 0.5259,
+      "step": 11512
+    },
+    {
+      "epoch": 14.73664,
+      "grad_norm": 0.9396823644638062,
+      "learning_rate": 2.6988795518207284e-05,
+      "loss": 0.5302,
+      "step": 11513
+    },
+    {
+      "epoch": 14.73792,
+      "grad_norm": 0.9388831853866577,
+      "learning_rate": 2.6986794717887153e-05,
+      "loss": 0.4803,
+      "step": 11514
+    },
+    {
+      "epoch": 14.7392,
+      "grad_norm": 1.0120586156845093,
+      "learning_rate": 2.698479391756703e-05,
+      "loss": 0.581,
+      "step": 11515
+    },
+    {
+      "epoch": 14.74048,
+      "grad_norm": 1.0425275564193726,
+      "learning_rate": 2.6982793117246903e-05,
+      "loss": 0.5611,
+      "step": 11516
+    },
+    {
+      "epoch": 14.74176,
+      "grad_norm": 1.0189977884292603,
+      "learning_rate": 2.698079231692677e-05,
+      "loss": 0.566,
+      "step": 11517
+    },
+    {
+      "epoch": 14.74304,
+      "grad_norm": 0.9831307530403137,
+      "learning_rate": 2.6978791516606643e-05,
+      "loss": 0.5258,
+      "step": 11518
+    },
+    {
+      "epoch": 14.74432,
+      "grad_norm": 0.9544398188591003,
+      "learning_rate": 2.6976790716286515e-05,
+      "loss": 0.5207,
+      "step": 11519
+    },
+    {
+      "epoch": 14.7456,
+      "grad_norm": 0.9739946722984314,
+      "learning_rate": 2.6974789915966387e-05,
+      "loss": 0.5362,
+      "step": 11520
+    },
+    {
+      "epoch": 14.74688,
+      "grad_norm": 1.0266259908676147,
+      "learning_rate": 2.697278911564626e-05,
+      "loss": 0.5669,
+      "step": 11521
+    },
+    {
+      "epoch": 14.74816,
+      "grad_norm": 0.9911479949951172,
+      "learning_rate": 2.6970788315326134e-05,
+      "loss": 0.529,
+      "step": 11522
+    },
+    {
+      "epoch": 14.74944,
+      "grad_norm": 0.978696882724762,
+      "learning_rate": 2.6968787515006006e-05,
+      "loss": 0.5497,
+      "step": 11523
+    },
+    {
+      "epoch": 14.75072,
+      "grad_norm": 0.9742518663406372,
+      "learning_rate": 2.6966786714685878e-05,
+      "loss": 0.4836,
+      "step": 11524
+    },
+    {
+      "epoch": 14.752,
+      "grad_norm": 0.9718377590179443,
+      "learning_rate": 2.6964785914365746e-05,
+      "loss": 0.513,
+      "step": 11525
+    },
+    {
+      "epoch": 14.75328,
+      "grad_norm": 1.0205025672912598,
+      "learning_rate": 2.6962785114045618e-05,
+      "loss": 0.582,
+      "step": 11526
+    },
+    {
+      "epoch": 14.75456,
+      "grad_norm": 0.9338187575340271,
+      "learning_rate": 2.696078431372549e-05,
+      "loss": 0.4745,
+      "step": 11527
+    },
+    {
+      "epoch": 14.75584,
+      "grad_norm": 0.9623693823814392,
+      "learning_rate": 2.6958783513405362e-05,
+      "loss": 0.5462,
+      "step": 11528
+    },
+    {
+      "epoch": 14.75712,
+      "grad_norm": 0.9271833896636963,
+      "learning_rate": 2.6956782713085237e-05,
+      "loss": 0.4699,
+      "step": 11529
+    },
+    {
+      "epoch": 14.7584,
+      "grad_norm": 1.0047674179077148,
+      "learning_rate": 2.695478191276511e-05,
+      "loss": 0.5474,
+      "step": 11530
+    },
+    {
+      "epoch": 14.75968,
+      "grad_norm": 0.9866589903831482,
+      "learning_rate": 2.695278111244498e-05,
+      "loss": 0.5667,
+      "step": 11531
+    },
+    {
+      "epoch": 14.76096,
+      "grad_norm": 0.9769319891929626,
+      "learning_rate": 2.6950780312124853e-05,
+      "loss": 0.5532,
+      "step": 11532
+    },
+    {
+      "epoch": 14.76224,
+      "grad_norm": 0.9511915445327759,
+      "learning_rate": 2.694877951180472e-05,
+      "loss": 0.522,
+      "step": 11533
+    },
+    {
+      "epoch": 14.76352,
+      "grad_norm": 0.9802474975585938,
+      "learning_rate": 2.6946778711484593e-05,
+      "loss": 0.5575,
+      "step": 11534
+    },
+    {
+      "epoch": 14.7648,
+      "grad_norm": 0.8974894285202026,
+      "learning_rate": 2.6944777911164465e-05,
+      "loss": 0.4573,
+      "step": 11535
+    },
+    {
+      "epoch": 14.76608,
+      "grad_norm": 0.9369537830352783,
+      "learning_rate": 2.694277711084434e-05,
+      "loss": 0.5447,
+      "step": 11536
+    },
+    {
+      "epoch": 14.76736,
+      "grad_norm": 0.9481881260871887,
+      "learning_rate": 2.6940776310524212e-05,
+      "loss": 0.5191,
+      "step": 11537
+    },
+    {
+      "epoch": 14.76864,
+      "grad_norm": 0.9813864231109619,
+      "learning_rate": 2.6938775510204084e-05,
+      "loss": 0.5228,
+      "step": 11538
+    },
+    {
+      "epoch": 14.76992,
+      "grad_norm": 0.9654099345207214,
+      "learning_rate": 2.6936774709883956e-05,
+      "loss": 0.4354,
+      "step": 11539
+    },
+    {
+      "epoch": 14.7712,
+      "grad_norm": 1.015555739402771,
+      "learning_rate": 2.6934773909563828e-05,
+      "loss": 0.5356,
+      "step": 11540
+    },
+    {
+      "epoch": 14.77248,
+      "grad_norm": 0.9228314161300659,
+      "learning_rate": 2.6932773109243696e-05,
+      "loss": 0.497,
+      "step": 11541
+    },
+    {
+      "epoch": 14.77376,
+      "grad_norm": 0.9956296682357788,
+      "learning_rate": 2.6930772308923568e-05,
+      "loss": 0.5217,
+      "step": 11542
+    },
+    {
+      "epoch": 14.77504,
+      "grad_norm": 1.0242211818695068,
+      "learning_rate": 2.6928771508603447e-05,
+      "loss": 0.4918,
+      "step": 11543
+    },
+    {
+      "epoch": 14.77632,
+      "grad_norm": 1.0041379928588867,
+      "learning_rate": 2.6926770708283315e-05,
+      "loss": 0.5375,
+      "step": 11544
+    },
+    {
+      "epoch": 14.7776,
+      "grad_norm": 0.9402958750724792,
+      "learning_rate": 2.6924769907963187e-05,
+      "loss": 0.4983,
+      "step": 11545
+    },
+    {
+      "epoch": 14.778880000000001,
+      "grad_norm": 0.9316099882125854,
+      "learning_rate": 2.692276910764306e-05,
+      "loss": 0.4897,
+      "step": 11546
+    },
+    {
+      "epoch": 14.78016,
+      "grad_norm": 1.0062158107757568,
+      "learning_rate": 2.692076830732293e-05,
+      "loss": 0.6163,
+      "step": 11547
+    },
+    {
+      "epoch": 14.78144,
+      "grad_norm": 0.93266361951828,
+      "learning_rate": 2.6918767507002802e-05,
+      "loss": 0.5614,
+      "step": 11548
+    },
+    {
+      "epoch": 14.78272,
+      "grad_norm": 0.9960542321205139,
+      "learning_rate": 2.691676670668267e-05,
+      "loss": 0.556,
+      "step": 11549
+    },
+    {
+      "epoch": 14.784,
+      "grad_norm": 0.9723289608955383,
+      "learning_rate": 2.691476590636255e-05,
+      "loss": 0.5528,
+      "step": 11550
+    },
+    {
+      "epoch": 14.78528,
+      "grad_norm": 1.003118872642517,
+      "learning_rate": 2.691276510604242e-05,
+      "loss": 0.5594,
+      "step": 11551
+    },
+    {
+      "epoch": 14.78656,
+      "grad_norm": 0.8769537806510925,
+      "learning_rate": 2.691076430572229e-05,
+      "loss": 0.4636,
+      "step": 11552
+    },
+    {
+      "epoch": 14.78784,
+      "grad_norm": 0.929743766784668,
+      "learning_rate": 2.6908763505402162e-05,
+      "loss": 0.497,
+      "step": 11553
+    },
+    {
+      "epoch": 14.78912,
+      "grad_norm": 0.9782609343528748,
+      "learning_rate": 2.6906762705082034e-05,
+      "loss": 0.5343,
+      "step": 11554
+    },
+    {
+      "epoch": 14.7904,
+      "grad_norm": 0.9649235010147095,
+      "learning_rate": 2.6904761904761905e-05,
+      "loss": 0.5329,
+      "step": 11555
+    },
+    {
+      "epoch": 14.79168,
+      "grad_norm": 1.0030887126922607,
+      "learning_rate": 2.6902761104441777e-05,
+      "loss": 0.5787,
+      "step": 11556
+    },
+    {
+      "epoch": 14.79296,
+      "grad_norm": 1.0339874029159546,
+      "learning_rate": 2.6900760304121653e-05,
+      "loss": 0.5822,
+      "step": 11557
+    },
+    {
+      "epoch": 14.79424,
+      "grad_norm": 1.0336726903915405,
+      "learning_rate": 2.6898759503801524e-05,
+      "loss": 0.573,
+      "step": 11558
+    },
+    {
+      "epoch": 14.79552,
+      "grad_norm": 0.9409865140914917,
+      "learning_rate": 2.6896758703481396e-05,
+      "loss": 0.5171,
+      "step": 11559
+    },
+    {
+      "epoch": 14.7968,
+      "grad_norm": 0.9395885467529297,
+      "learning_rate": 2.6894757903161265e-05,
+      "loss": 0.52,
+      "step": 11560
+    },
+    {
+      "epoch": 14.79808,
+      "grad_norm": 1.0146170854568481,
+      "learning_rate": 2.6892757102841137e-05,
+      "loss": 0.538,
+      "step": 11561
+    },
+    {
+      "epoch": 14.79936,
+      "grad_norm": 1.051053524017334,
+      "learning_rate": 2.689075630252101e-05,
+      "loss": 0.6081,
+      "step": 11562
+    },
+    {
+      "epoch": 14.80064,
+      "grad_norm": 0.9499222636222839,
+      "learning_rate": 2.688875550220088e-05,
+      "loss": 0.5111,
+      "step": 11563
+    },
+    {
+      "epoch": 14.801919999999999,
+      "grad_norm": 1.0103505849838257,
+      "learning_rate": 2.6886754701880752e-05,
+      "loss": 0.5686,
+      "step": 11564
+    },
+    {
+      "epoch": 14.8032,
+      "grad_norm": 1.0131018161773682,
+      "learning_rate": 2.6884753901560627e-05,
+      "loss": 0.5578,
+      "step": 11565
+    },
+    {
+      "epoch": 14.80448,
+      "grad_norm": 1.0048017501831055,
+      "learning_rate": 2.68827531012405e-05,
+      "loss": 0.5235,
+      "step": 11566
+    },
+    {
+      "epoch": 14.80576,
+      "grad_norm": 1.0479151010513306,
+      "learning_rate": 2.688075230092037e-05,
+      "loss": 0.5589,
+      "step": 11567
+    },
+    {
+      "epoch": 14.80704,
+      "grad_norm": 1.0059313774108887,
+      "learning_rate": 2.687875150060024e-05,
+      "loss": 0.5576,
+      "step": 11568
+    },
+    {
+      "epoch": 14.80832,
+      "grad_norm": 0.9548071622848511,
+      "learning_rate": 2.687675070028011e-05,
+      "loss": 0.5122,
+      "step": 11569
+    },
+    {
+      "epoch": 14.8096,
+      "grad_norm": 0.9280616641044617,
+      "learning_rate": 2.6874749899959983e-05,
+      "loss": 0.5034,
+      "step": 11570
+    },
+    {
+      "epoch": 14.810880000000001,
+      "grad_norm": 0.9565810561180115,
+      "learning_rate": 2.6872749099639855e-05,
+      "loss": 0.516,
+      "step": 11571
+    },
+    {
+      "epoch": 14.81216,
+      "grad_norm": 0.9888594150543213,
+      "learning_rate": 2.687074829931973e-05,
+      "loss": 0.5403,
+      "step": 11572
+    },
+    {
+      "epoch": 14.81344,
+      "grad_norm": 1.0492602586746216,
+      "learning_rate": 2.6868747498999602e-05,
+      "loss": 0.5664,
+      "step": 11573
+    },
+    {
+      "epoch": 14.81472,
+      "grad_norm": 0.9663681983947754,
+      "learning_rate": 2.6866746698679474e-05,
+      "loss": 0.5895,
+      "step": 11574
+    },
+    {
+      "epoch": 14.816,
+      "grad_norm": 0.9314732551574707,
+      "learning_rate": 2.6864745898359346e-05,
+      "loss": 0.4985,
+      "step": 11575
+    },
+    {
+      "epoch": 14.81728,
+      "grad_norm": 0.9417006373405457,
+      "learning_rate": 2.6862745098039214e-05,
+      "loss": 0.5434,
+      "step": 11576
+    },
+    {
+      "epoch": 14.81856,
+      "grad_norm": 0.968472421169281,
+      "learning_rate": 2.6860744297719086e-05,
+      "loss": 0.5591,
+      "step": 11577
+    },
+    {
+      "epoch": 14.81984,
+      "grad_norm": 0.9163793325424194,
+      "learning_rate": 2.6858743497398958e-05,
+      "loss": 0.5035,
+      "step": 11578
+    },
+    {
+      "epoch": 14.82112,
+      "grad_norm": 0.9645053744316101,
+      "learning_rate": 2.6856742697078833e-05,
+      "loss": 0.5461,
+      "step": 11579
+    },
+    {
+      "epoch": 14.8224,
+      "grad_norm": 0.9942086338996887,
+      "learning_rate": 2.6854741896758705e-05,
+      "loss": 0.5775,
+      "step": 11580
+    },
+    {
+      "epoch": 14.82368,
+      "grad_norm": 0.9701970219612122,
+      "learning_rate": 2.6852741096438577e-05,
+      "loss": 0.5237,
+      "step": 11581
+    },
+    {
+      "epoch": 14.82496,
+      "grad_norm": 0.9851134419441223,
+      "learning_rate": 2.685074029611845e-05,
+      "loss": 0.5184,
+      "step": 11582
+    },
+    {
+      "epoch": 14.82624,
+      "grad_norm": 0.8879162073135376,
+      "learning_rate": 2.684873949579832e-05,
+      "loss": 0.4711,
+      "step": 11583
+    },
+    {
+      "epoch": 14.82752,
+      "grad_norm": 0.9536734819412231,
+      "learning_rate": 2.684673869547819e-05,
+      "loss": 0.5822,
+      "step": 11584
+    },
+    {
+      "epoch": 14.8288,
+      "grad_norm": 0.9417433142662048,
+      "learning_rate": 2.684473789515806e-05,
+      "loss": 0.5429,
+      "step": 11585
+    },
+    {
+      "epoch": 14.83008,
+      "grad_norm": 0.9906011819839478,
+      "learning_rate": 2.684273709483794e-05,
+      "loss": 0.5599,
+      "step": 11586
+    },
+    {
+      "epoch": 14.83136,
+      "grad_norm": 1.0022450685501099,
+      "learning_rate": 2.6840736294517808e-05,
+      "loss": 0.565,
+      "step": 11587
+    },
+    {
+      "epoch": 14.83264,
+      "grad_norm": 0.9821887016296387,
+      "learning_rate": 2.683873549419768e-05,
+      "loss": 0.5372,
+      "step": 11588
+    },
+    {
+      "epoch": 14.833919999999999,
+      "grad_norm": 0.9912195801734924,
+      "learning_rate": 2.6836734693877552e-05,
+      "loss": 0.5902,
+      "step": 11589
+    },
+    {
+      "epoch": 14.8352,
+      "grad_norm": 0.940270721912384,
+      "learning_rate": 2.6834733893557424e-05,
+      "loss": 0.5177,
+      "step": 11590
+    },
+    {
+      "epoch": 14.83648,
+      "grad_norm": 0.9974461793899536,
+      "learning_rate": 2.6832733093237296e-05,
+      "loss": 0.5458,
+      "step": 11591
+    },
+    {
+      "epoch": 14.83776,
+      "grad_norm": 1.0043706893920898,
+      "learning_rate": 2.6830732292917164e-05,
+      "loss": 0.6071,
+      "step": 11592
+    },
+    {
+      "epoch": 14.83904,
+      "grad_norm": 0.982296347618103,
+      "learning_rate": 2.6828731492597043e-05,
+      "loss": 0.5237,
+      "step": 11593
+    },
+    {
+      "epoch": 14.84032,
+      "grad_norm": 1.0043891668319702,
+      "learning_rate": 2.6826730692276915e-05,
+      "loss": 0.5551,
+      "step": 11594
+    },
+    {
+      "epoch": 14.8416,
+      "grad_norm": 1.0041606426239014,
+      "learning_rate": 2.6824729891956783e-05,
+      "loss": 0.5376,
+      "step": 11595
+    },
+    {
+      "epoch": 14.84288,
+      "grad_norm": 0.952582061290741,
+      "learning_rate": 2.6822729091636655e-05,
+      "loss": 0.5425,
+      "step": 11596
+    },
+    {
+      "epoch": 14.84416,
+      "grad_norm": 0.905483603477478,
+      "learning_rate": 2.6820728291316527e-05,
+      "loss": 0.5078,
+      "step": 11597
+    },
+    {
+      "epoch": 14.84544,
+      "grad_norm": 0.9707022309303284,
+      "learning_rate": 2.68187274909964e-05,
+      "loss": 0.5593,
+      "step": 11598
+    },
+    {
+      "epoch": 14.84672,
+      "grad_norm": 0.9420384168624878,
+      "learning_rate": 2.681672669067627e-05,
+      "loss": 0.5346,
+      "step": 11599
+    },
+    {
+      "epoch": 14.848,
+      "grad_norm": 1.0081489086151123,
+      "learning_rate": 2.6814725890356146e-05,
+      "loss": 0.5217,
+      "step": 11600
+    },
+    {
+      "epoch": 14.84928,
+      "grad_norm": 1.0194686651229858,
+      "learning_rate": 2.6812725090036018e-05,
+      "loss": 0.6055,
+      "step": 11601
+    },
+    {
+      "epoch": 14.85056,
+      "grad_norm": 0.9814406037330627,
+      "learning_rate": 2.681072428971589e-05,
+      "loss": 0.5336,
+      "step": 11602
+    },
+    {
+      "epoch": 14.85184,
+      "grad_norm": 0.9600327610969543,
+      "learning_rate": 2.6808723489395758e-05,
+      "loss": 0.5383,
+      "step": 11603
+    },
+    {
+      "epoch": 14.85312,
+      "grad_norm": 0.9688712954521179,
+      "learning_rate": 2.680672268907563e-05,
+      "loss": 0.5513,
+      "step": 11604
+    },
+    {
+      "epoch": 14.8544,
+      "grad_norm": 0.9412886500358582,
+      "learning_rate": 2.6804721888755502e-05,
+      "loss": 0.5537,
+      "step": 11605
+    },
+    {
+      "epoch": 14.85568,
+      "grad_norm": 0.9556344151496887,
+      "learning_rate": 2.6802721088435374e-05,
+      "loss": 0.5508,
+      "step": 11606
+    },
+    {
+      "epoch": 14.85696,
+      "grad_norm": 0.9847762584686279,
+      "learning_rate": 2.680072028811525e-05,
+      "loss": 0.5563,
+      "step": 11607
+    },
+    {
+      "epoch": 14.85824,
+      "grad_norm": 0.9027768969535828,
+      "learning_rate": 2.679871948779512e-05,
+      "loss": 0.5691,
+      "step": 11608
+    },
+    {
+      "epoch": 14.85952,
+      "grad_norm": 0.998230516910553,
+      "learning_rate": 2.6796718687474993e-05,
+      "loss": 0.5422,
+      "step": 11609
+    },
+    {
+      "epoch": 14.8608,
+      "grad_norm": 1.0053776502609253,
+      "learning_rate": 2.6794717887154864e-05,
+      "loss": 0.5396,
+      "step": 11610
+    },
+    {
+      "epoch": 14.86208,
+      "grad_norm": 1.0294034481048584,
+      "learning_rate": 2.6792717086834733e-05,
+      "loss": 0.5769,
+      "step": 11611
+    },
+    {
+      "epoch": 14.86336,
+      "grad_norm": 0.9711460471153259,
+      "learning_rate": 2.6790716286514605e-05,
+      "loss": 0.5277,
+      "step": 11612
+    },
+    {
+      "epoch": 14.86464,
+      "grad_norm": 0.9948602318763733,
+      "learning_rate": 2.6788715486194477e-05,
+      "loss": 0.5497,
+      "step": 11613
+    },
+    {
+      "epoch": 14.86592,
+      "grad_norm": 0.9797696471214294,
+      "learning_rate": 2.6786714685874352e-05,
+      "loss": 0.5707,
+      "step": 11614
+    },
+    {
+      "epoch": 14.8672,
+      "grad_norm": 1.006882905960083,
+      "learning_rate": 2.6784713885554224e-05,
+      "loss": 0.5857,
+      "step": 11615
+    },
+    {
+      "epoch": 14.86848,
+      "grad_norm": 0.9534305930137634,
+      "learning_rate": 2.6782713085234096e-05,
+      "loss": 0.5066,
+      "step": 11616
+    },
+    {
+      "epoch": 14.86976,
+      "grad_norm": 1.002995491027832,
+      "learning_rate": 2.6780712284913967e-05,
+      "loss": 0.5673,
+      "step": 11617
+    },
+    {
+      "epoch": 14.87104,
+      "grad_norm": 0.9653766751289368,
+      "learning_rate": 2.677871148459384e-05,
+      "loss": 0.5393,
+      "step": 11618
+    },
+    {
+      "epoch": 14.87232,
+      "grad_norm": 0.9714248776435852,
+      "learning_rate": 2.6776710684273708e-05,
+      "loss": 0.5344,
+      "step": 11619
+    },
+    {
+      "epoch": 14.8736,
+      "grad_norm": 0.9720675349235535,
+      "learning_rate": 2.677470988395358e-05,
+      "loss": 0.5118,
+      "step": 11620
+    },
+    {
+      "epoch": 14.87488,
+      "grad_norm": 0.9434537887573242,
+      "learning_rate": 2.6772709083633458e-05,
+      "loss": 0.5116,
+      "step": 11621
+    },
+    {
+      "epoch": 14.87616,
+      "grad_norm": 0.9508132338523865,
+      "learning_rate": 2.6770708283313327e-05,
+      "loss": 0.5432,
+      "step": 11622
+    },
+    {
+      "epoch": 14.87744,
+      "grad_norm": 0.9770334362983704,
+      "learning_rate": 2.67687074829932e-05,
+      "loss": 0.5533,
+      "step": 11623
+    },
+    {
+      "epoch": 14.87872,
+      "grad_norm": 0.9566938877105713,
+      "learning_rate": 2.676670668267307e-05,
+      "loss": 0.5081,
+      "step": 11624
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 1.0178238153457642,
+      "learning_rate": 2.6764705882352942e-05,
+      "loss": 0.5921,
+      "step": 11625
+    },
+    {
+      "epoch": 14.88128,
+      "grad_norm": 0.9234164357185364,
+      "learning_rate": 2.6762705082032814e-05,
+      "loss": 0.523,
+      "step": 11626
+    },
+    {
+      "epoch": 14.88256,
+      "grad_norm": 0.9184479713439941,
+      "learning_rate": 2.6760704281712683e-05,
+      "loss": 0.5278,
+      "step": 11627
+    },
+    {
+      "epoch": 14.88384,
+      "grad_norm": 0.9186223745346069,
+      "learning_rate": 2.675870348139256e-05,
+      "loss": 0.4659,
+      "step": 11628
+    },
+    {
+      "epoch": 14.88512,
+      "grad_norm": 0.9434010982513428,
+      "learning_rate": 2.6756702681072433e-05,
+      "loss": 0.517,
+      "step": 11629
+    },
+    {
+      "epoch": 14.8864,
+      "grad_norm": 0.9369627833366394,
+      "learning_rate": 2.67547018807523e-05,
+      "loss": 0.5197,
+      "step": 11630
+    },
+    {
+      "epoch": 14.88768,
+      "grad_norm": 0.9298577308654785,
+      "learning_rate": 2.6752701080432173e-05,
+      "loss": 0.5219,
+      "step": 11631
+    },
+    {
+      "epoch": 14.88896,
+      "grad_norm": 0.9377476572990417,
+      "learning_rate": 2.6750700280112045e-05,
+      "loss": 0.5519,
+      "step": 11632
+    },
+    {
+      "epoch": 14.89024,
+      "grad_norm": 0.919325590133667,
+      "learning_rate": 2.6748699479791917e-05,
+      "loss": 0.5225,
+      "step": 11633
+    },
+    {
+      "epoch": 14.89152,
+      "grad_norm": 0.9516395330429077,
+      "learning_rate": 2.674669867947179e-05,
+      "loss": 0.5508,
+      "step": 11634
+    },
+    {
+      "epoch": 14.8928,
+      "grad_norm": 0.9320101737976074,
+      "learning_rate": 2.6744697879151664e-05,
+      "loss": 0.5241,
+      "step": 11635
+    },
+    {
+      "epoch": 14.89408,
+      "grad_norm": 1.018315315246582,
+      "learning_rate": 2.6742697078831536e-05,
+      "loss": 0.5826,
+      "step": 11636
+    },
+    {
+      "epoch": 14.89536,
+      "grad_norm": 0.973010241985321,
+      "learning_rate": 2.6740696278511408e-05,
+      "loss": 0.5401,
+      "step": 11637
+    },
+    {
+      "epoch": 14.89664,
+      "grad_norm": 0.9724129438400269,
+      "learning_rate": 2.6738695478191276e-05,
+      "loss": 0.5432,
+      "step": 11638
+    },
+    {
+      "epoch": 14.89792,
+      "grad_norm": 0.9453209042549133,
+      "learning_rate": 2.6736694677871148e-05,
+      "loss": 0.5506,
+      "step": 11639
+    },
+    {
+      "epoch": 14.8992,
+      "grad_norm": 1.0045145750045776,
+      "learning_rate": 2.673469387755102e-05,
+      "loss": 0.5437,
+      "step": 11640
+    },
+    {
+      "epoch": 14.90048,
+      "grad_norm": 1.0044034719467163,
+      "learning_rate": 2.6732693077230892e-05,
+      "loss": 0.5566,
+      "step": 11641
+    },
+    {
+      "epoch": 14.90176,
+      "grad_norm": 0.9657561182975769,
+      "learning_rate": 2.6730692276910767e-05,
+      "loss": 0.5102,
+      "step": 11642
+    },
+    {
+      "epoch": 14.90304,
+      "grad_norm": 0.9859712719917297,
+      "learning_rate": 2.672869147659064e-05,
+      "loss": 0.5598,
+      "step": 11643
+    },
+    {
+      "epoch": 14.90432,
+      "grad_norm": 0.9568543434143066,
+      "learning_rate": 2.672669067627051e-05,
+      "loss": 0.5439,
+      "step": 11644
+    },
+    {
+      "epoch": 14.9056,
+      "grad_norm": 1.0132269859313965,
+      "learning_rate": 2.6724689875950383e-05,
+      "loss": 0.5572,
+      "step": 11645
+    },
+    {
+      "epoch": 14.90688,
+      "grad_norm": 0.9919137358665466,
+      "learning_rate": 2.672268907563025e-05,
+      "loss": 0.5823,
+      "step": 11646
+    },
+    {
+      "epoch": 14.90816,
+      "grad_norm": 0.9762164354324341,
+      "learning_rate": 2.6720688275310123e-05,
+      "loss": 0.5324,
+      "step": 11647
+    },
+    {
+      "epoch": 14.90944,
+      "grad_norm": 0.9875562191009521,
+      "learning_rate": 2.6718687474989995e-05,
+      "loss": 0.5269,
+      "step": 11648
+    },
+    {
+      "epoch": 14.91072,
+      "grad_norm": 0.9606843590736389,
+      "learning_rate": 2.671668667466987e-05,
+      "loss": 0.5141,
+      "step": 11649
+    },
+    {
+      "epoch": 14.912,
+      "grad_norm": 0.979857861995697,
+      "learning_rate": 2.6714685874349742e-05,
+      "loss": 0.5595,
+      "step": 11650
+    },
+    {
+      "epoch": 14.91328,
+      "grad_norm": 0.9310894012451172,
+      "learning_rate": 2.6712685074029614e-05,
+      "loss": 0.5369,
+      "step": 11651
+    },
+    {
+      "epoch": 14.91456,
+      "grad_norm": 0.9621683955192566,
+      "learning_rate": 2.6710684273709486e-05,
+      "loss": 0.5379,
+      "step": 11652
+    },
+    {
+      "epoch": 14.91584,
+      "grad_norm": 0.9423104524612427,
+      "learning_rate": 2.6708683473389358e-05,
+      "loss": 0.5239,
+      "step": 11653
+    },
+    {
+      "epoch": 14.91712,
+      "grad_norm": 0.9782088398933411,
+      "learning_rate": 2.6706682673069226e-05,
+      "loss": 0.5157,
+      "step": 11654
+    },
+    {
+      "epoch": 14.9184,
+      "grad_norm": 1.0111697912216187,
+      "learning_rate": 2.6704681872749098e-05,
+      "loss": 0.5385,
+      "step": 11655
+    },
+    {
+      "epoch": 14.91968,
+      "grad_norm": 0.9448067545890808,
+      "learning_rate": 2.6702681072428977e-05,
+      "loss": 0.5221,
+      "step": 11656
+    },
+    {
+      "epoch": 14.920960000000001,
+      "grad_norm": 0.9532007575035095,
+      "learning_rate": 2.6700680272108845e-05,
+      "loss": 0.5573,
+      "step": 11657
+    },
+    {
+      "epoch": 14.92224,
+      "grad_norm": 0.9769750237464905,
+      "learning_rate": 2.6698679471788717e-05,
+      "loss": 0.5437,
+      "step": 11658
+    },
+    {
+      "epoch": 14.92352,
+      "grad_norm": 0.9078409671783447,
+      "learning_rate": 2.669667867146859e-05,
+      "loss": 0.4773,
+      "step": 11659
+    },
+    {
+      "epoch": 14.9248,
+      "grad_norm": 0.9596759676933289,
+      "learning_rate": 2.669467787114846e-05,
+      "loss": 0.5117,
+      "step": 11660
+    },
+    {
+      "epoch": 14.92608,
+      "grad_norm": 0.9927992820739746,
+      "learning_rate": 2.6692677070828332e-05,
+      "loss": 0.5405,
+      "step": 11661
+    },
+    {
+      "epoch": 14.92736,
+      "grad_norm": 0.9593184590339661,
+      "learning_rate": 2.66906762705082e-05,
+      "loss": 0.4832,
+      "step": 11662
+    },
+    {
+      "epoch": 14.92864,
+      "grad_norm": 0.9789624214172363,
+      "learning_rate": 2.668867547018808e-05,
+      "loss": 0.5316,
+      "step": 11663
+    },
+    {
+      "epoch": 14.92992,
+      "grad_norm": 0.9484257698059082,
+      "learning_rate": 2.668667466986795e-05,
+      "loss": 0.5088,
+      "step": 11664
+    },
+    {
+      "epoch": 14.9312,
+      "grad_norm": 0.987421989440918,
+      "learning_rate": 2.668467386954782e-05,
+      "loss": 0.5787,
+      "step": 11665
+    },
+    {
+      "epoch": 14.93248,
+      "grad_norm": 0.973088264465332,
+      "learning_rate": 2.6682673069227692e-05,
+      "loss": 0.5346,
+      "step": 11666
+    },
+    {
+      "epoch": 14.93376,
+      "grad_norm": 0.985235333442688,
+      "learning_rate": 2.6680672268907564e-05,
+      "loss": 0.5391,
+      "step": 11667
+    },
+    {
+      "epoch": 14.93504,
+      "grad_norm": 1.0299049615859985,
+      "learning_rate": 2.6678671468587435e-05,
+      "loss": 0.5586,
+      "step": 11668
+    },
+    {
+      "epoch": 14.93632,
+      "grad_norm": 0.9780097603797913,
+      "learning_rate": 2.6676670668267307e-05,
+      "loss": 0.5194,
+      "step": 11669
+    },
+    {
+      "epoch": 14.9376,
+      "grad_norm": 1.037703037261963,
+      "learning_rate": 2.6674669867947183e-05,
+      "loss": 0.5629,
+      "step": 11670
+    },
+    {
+      "epoch": 14.93888,
+      "grad_norm": 0.9950383305549622,
+      "learning_rate": 2.6672669067627054e-05,
+      "loss": 0.5533,
+      "step": 11671
+    },
+    {
+      "epoch": 14.94016,
+      "grad_norm": 0.9457634091377258,
+      "learning_rate": 2.6670668267306926e-05,
+      "loss": 0.5002,
+      "step": 11672
+    },
+    {
+      "epoch": 14.94144,
+      "grad_norm": 1.0216537714004517,
+      "learning_rate": 2.6668667466986795e-05,
+      "loss": 0.5612,
+      "step": 11673
+    },
+    {
+      "epoch": 14.94272,
+      "grad_norm": 0.9794032573699951,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5796,
+      "step": 11674
+    },
+    {
+      "epoch": 14.943999999999999,
+      "grad_norm": 0.965833306312561,
+      "learning_rate": 2.666466586634654e-05,
+      "loss": 0.5874,
+      "step": 11675
+    },
+    {
+      "epoch": 14.94528,
+      "grad_norm": 0.9718341827392578,
+      "learning_rate": 2.666266506602641e-05,
+      "loss": 0.5002,
+      "step": 11676
+    },
+    {
+      "epoch": 14.94656,
+      "grad_norm": 0.9411438703536987,
+      "learning_rate": 2.6660664265706282e-05,
+      "loss": 0.5241,
+      "step": 11677
+    },
+    {
+      "epoch": 14.94784,
+      "grad_norm": 0.9548771381378174,
+      "learning_rate": 2.6658663465386157e-05,
+      "loss": 0.5152,
+      "step": 11678
+    },
+    {
+      "epoch": 14.94912,
+      "grad_norm": 0.9769624471664429,
+      "learning_rate": 2.665666266506603e-05,
+      "loss": 0.5417,
+      "step": 11679
+    },
+    {
+      "epoch": 14.9504,
+      "grad_norm": 0.9090695977210999,
+      "learning_rate": 2.66546618647459e-05,
+      "loss": 0.5338,
+      "step": 11680
+    },
+    {
+      "epoch": 14.95168,
+      "grad_norm": 0.9659811854362488,
+      "learning_rate": 2.665266106442577e-05,
+      "loss": 0.5788,
+      "step": 11681
+    },
+    {
+      "epoch": 14.952960000000001,
+      "grad_norm": 0.94170081615448,
+      "learning_rate": 2.665066026410564e-05,
+      "loss": 0.5491,
+      "step": 11682
+    },
+    {
+      "epoch": 14.95424,
+      "grad_norm": 0.9707318544387817,
+      "learning_rate": 2.6648659463785513e-05,
+      "loss": 0.5736,
+      "step": 11683
+    },
+    {
+      "epoch": 14.95552,
+      "grad_norm": 0.9650655388832092,
+      "learning_rate": 2.6646658663465385e-05,
+      "loss": 0.5422,
+      "step": 11684
+    },
+    {
+      "epoch": 14.9568,
+      "grad_norm": 0.9502249360084534,
+      "learning_rate": 2.664465786314526e-05,
+      "loss": 0.5211,
+      "step": 11685
+    },
+    {
+      "epoch": 14.95808,
+      "grad_norm": 0.9722442030906677,
+      "learning_rate": 2.6642657062825132e-05,
+      "loss": 0.5209,
+      "step": 11686
+    },
+    {
+      "epoch": 14.95936,
+      "grad_norm": 0.9735600352287292,
+      "learning_rate": 2.6640656262505004e-05,
+      "loss": 0.5587,
+      "step": 11687
+    },
+    {
+      "epoch": 14.96064,
+      "grad_norm": 0.9050617218017578,
+      "learning_rate": 2.6638655462184876e-05,
+      "loss": 0.495,
+      "step": 11688
+    },
+    {
+      "epoch": 14.96192,
+      "grad_norm": 1.0128991603851318,
+      "learning_rate": 2.6636654661864744e-05,
+      "loss": 0.5692,
+      "step": 11689
+    },
+    {
+      "epoch": 14.9632,
+      "grad_norm": 0.970399796962738,
+      "learning_rate": 2.6634653861544616e-05,
+      "loss": 0.5175,
+      "step": 11690
+    },
+    {
+      "epoch": 14.96448,
+      "grad_norm": 0.9725888967514038,
+      "learning_rate": 2.6632653061224488e-05,
+      "loss": 0.5323,
+      "step": 11691
+    },
+    {
+      "epoch": 14.96576,
+      "grad_norm": 0.9731080532073975,
+      "learning_rate": 2.6630652260904367e-05,
+      "loss": 0.5548,
+      "step": 11692
+    },
+    {
+      "epoch": 14.96704,
+      "grad_norm": 0.9690375924110413,
+      "learning_rate": 2.6628651460584235e-05,
+      "loss": 0.5147,
+      "step": 11693
+    },
+    {
+      "epoch": 14.96832,
+      "grad_norm": 1.0425267219543457,
+      "learning_rate": 2.6626650660264107e-05,
+      "loss": 0.5284,
+      "step": 11694
+    },
+    {
+      "epoch": 14.9696,
+      "grad_norm": 0.949582576751709,
+      "learning_rate": 2.662464985994398e-05,
+      "loss": 0.4982,
+      "step": 11695
+    },
+    {
+      "epoch": 14.97088,
+      "grad_norm": 0.9666782021522522,
+      "learning_rate": 2.662264905962385e-05,
+      "loss": 0.5504,
+      "step": 11696
+    },
+    {
+      "epoch": 14.97216,
+      "grad_norm": 1.0077440738677979,
+      "learning_rate": 2.662064825930372e-05,
+      "loss": 0.5647,
+      "step": 11697
+    },
+    {
+      "epoch": 14.97344,
+      "grad_norm": 0.930424690246582,
+      "learning_rate": 2.661864745898359e-05,
+      "loss": 0.484,
+      "step": 11698
+    },
+    {
+      "epoch": 14.97472,
+      "grad_norm": 0.9621975421905518,
+      "learning_rate": 2.661664665866347e-05,
+      "loss": 0.5133,
+      "step": 11699
+    },
+    {
+      "epoch": 14.975999999999999,
+      "grad_norm": 0.9930300116539001,
+      "learning_rate": 2.661464585834334e-05,
+      "loss": 0.5704,
+      "step": 11700
+    },
+    {
+      "epoch": 14.97728,
+      "grad_norm": 0.965140163898468,
+      "learning_rate": 2.661264505802321e-05,
+      "loss": 0.5111,
+      "step": 11701
+    },
+    {
+      "epoch": 14.97856,
+      "grad_norm": 0.9955794215202332,
+      "learning_rate": 2.6610644257703082e-05,
+      "loss": 0.5608,
+      "step": 11702
+    },
+    {
+      "epoch": 14.97984,
+      "grad_norm": 0.9986573457717896,
+      "learning_rate": 2.6608643457382954e-05,
+      "loss": 0.5501,
+      "step": 11703
+    },
+    {
+      "epoch": 14.98112,
+      "grad_norm": 0.9306737780570984,
+      "learning_rate": 2.6606642657062826e-05,
+      "loss": 0.5065,
+      "step": 11704
+    },
+    {
+      "epoch": 14.9824,
+      "grad_norm": 0.9776188135147095,
+      "learning_rate": 2.6604641856742694e-05,
+      "loss": 0.5241,
+      "step": 11705
+    },
+    {
+      "epoch": 14.98368,
+      "grad_norm": 0.9727427363395691,
+      "learning_rate": 2.6602641056422573e-05,
+      "loss": 0.515,
+      "step": 11706
+    },
+    {
+      "epoch": 14.98496,
+      "grad_norm": 1.011610507965088,
+      "learning_rate": 2.6600640256102445e-05,
+      "loss": 0.5472,
+      "step": 11707
+    },
+    {
+      "epoch": 14.98624,
+      "grad_norm": 1.0158885717391968,
+      "learning_rate": 2.6598639455782316e-05,
+      "loss": 0.5642,
+      "step": 11708
+    },
+    {
+      "epoch": 14.98752,
+      "grad_norm": 1.021340250968933,
+      "learning_rate": 2.6596638655462185e-05,
+      "loss": 0.5585,
+      "step": 11709
+    },
+    {
+      "epoch": 14.9888,
+      "grad_norm": 0.91712486743927,
+      "learning_rate": 2.6594637855142057e-05,
+      "loss": 0.5085,
+      "step": 11710
+    },
+    {
+      "epoch": 14.99008,
+      "grad_norm": 0.9759134650230408,
+      "learning_rate": 2.659263705482193e-05,
+      "loss": 0.5493,
+      "step": 11711
+    },
+    {
+      "epoch": 14.99136,
+      "grad_norm": 1.0136007070541382,
+      "learning_rate": 2.65906362545018e-05,
+      "loss": 0.5891,
+      "step": 11712
+    },
+    {
+      "epoch": 14.99264,
+      "grad_norm": 0.9995754361152649,
+      "learning_rate": 2.6588635454181676e-05,
+      "loss": 0.4942,
+      "step": 11713
+    },
+    {
+      "epoch": 14.99392,
+      "grad_norm": 0.9608016610145569,
+      "learning_rate": 2.6586634653861548e-05,
+      "loss": 0.5207,
+      "step": 11714
+    },
+    {
+      "epoch": 14.9952,
+      "grad_norm": 0.9471200108528137,
+      "learning_rate": 2.658463385354142e-05,
+      "loss": 0.5334,
+      "step": 11715
+    },
+    {
+      "epoch": 14.99648,
+      "grad_norm": 0.971295952796936,
+      "learning_rate": 2.658263305322129e-05,
+      "loss": 0.5604,
+      "step": 11716
+    },
+    {
+      "epoch": 14.99776,
+      "grad_norm": 0.9903369545936584,
+      "learning_rate": 2.658063225290116e-05,
+      "loss": 0.6,
+      "step": 11717
+    },
+    {
+      "epoch": 14.99904,
+      "grad_norm": 0.9256072044372559,
+      "learning_rate": 2.657863145258103e-05,
+      "loss": 0.5046,
+      "step": 11718
+    },
+    {
+      "epoch": 15.00032,
+      "grad_norm": 1.9394627809524536,
+      "learning_rate": 2.6576630652260904e-05,
+      "loss": 0.813,
+      "step": 11719
+    },
+    {
+      "epoch": 15.0016,
+      "grad_norm": 0.9253193140029907,
+      "learning_rate": 2.657462985194078e-05,
+      "loss": 0.5114,
+      "step": 11720
+    },
+    {
+      "epoch": 15.00288,
+      "grad_norm": 0.9425203800201416,
+      "learning_rate": 2.657262905162065e-05,
+      "loss": 0.4764,
+      "step": 11721
+    },
+    {
+      "epoch": 15.00416,
+      "grad_norm": 0.9434118866920471,
+      "learning_rate": 2.6570628251300522e-05,
+      "loss": 0.4907,
+      "step": 11722
+    },
+    {
+      "epoch": 15.00544,
+      "grad_norm": 0.9589188694953918,
+      "learning_rate": 2.6568627450980394e-05,
+      "loss": 0.5433,
+      "step": 11723
+    },
+    {
+      "epoch": 15.00672,
+      "grad_norm": 0.9243043661117554,
+      "learning_rate": 2.6566626650660266e-05,
+      "loss": 0.5402,
+      "step": 11724
+    },
+    {
+      "epoch": 15.008,
+      "grad_norm": 0.9611799120903015,
+      "learning_rate": 2.6564625850340135e-05,
+      "loss": 0.5056,
+      "step": 11725
+    },
+    {
+      "epoch": 15.00928,
+      "grad_norm": 0.8841933608055115,
+      "learning_rate": 2.6562625050020007e-05,
+      "loss": 0.4763,
+      "step": 11726
+    },
+    {
+      "epoch": 15.01056,
+      "grad_norm": 0.9348115921020508,
+      "learning_rate": 2.6560624249699885e-05,
+      "loss": 0.4711,
+      "step": 11727
+    },
+    {
+      "epoch": 15.01184,
+      "grad_norm": 0.9851812720298767,
+      "learning_rate": 2.6558623449379754e-05,
+      "loss": 0.5166,
+      "step": 11728
+    },
+    {
+      "epoch": 15.01312,
+      "grad_norm": 0.967349648475647,
+      "learning_rate": 2.6556622649059625e-05,
+      "loss": 0.4667,
+      "step": 11729
+    },
+    {
+      "epoch": 15.0144,
+      "grad_norm": 1.0413135290145874,
+      "learning_rate": 2.6554621848739497e-05,
+      "loss": 0.5798,
+      "step": 11730
+    },
+    {
+      "epoch": 15.01568,
+      "grad_norm": 0.9787873029708862,
+      "learning_rate": 2.655262104841937e-05,
+      "loss": 0.5299,
+      "step": 11731
+    },
+    {
+      "epoch": 15.01696,
+      "grad_norm": 0.9922900199890137,
+      "learning_rate": 2.655062024809924e-05,
+      "loss": 0.5363,
+      "step": 11732
+    },
+    {
+      "epoch": 15.01824,
+      "grad_norm": 0.9843411445617676,
+      "learning_rate": 2.654861944777911e-05,
+      "loss": 0.5272,
+      "step": 11733
+    },
+    {
+      "epoch": 15.01952,
+      "grad_norm": 0.9954893589019775,
+      "learning_rate": 2.6546618647458988e-05,
+      "loss": 0.5566,
+      "step": 11734
+    },
+    {
+      "epoch": 15.0208,
+      "grad_norm": 0.9984802603721619,
+      "learning_rate": 2.654461784713886e-05,
+      "loss": 0.561,
+      "step": 11735
+    },
+    {
+      "epoch": 15.02208,
+      "grad_norm": 0.992762565612793,
+      "learning_rate": 2.654261704681873e-05,
+      "loss": 0.4935,
+      "step": 11736
+    },
+    {
+      "epoch": 15.02336,
+      "grad_norm": 0.9480631351470947,
+      "learning_rate": 2.65406162464986e-05,
+      "loss": 0.493,
+      "step": 11737
+    },
+    {
+      "epoch": 15.02464,
+      "grad_norm": 1.0148262977600098,
+      "learning_rate": 2.6538615446178472e-05,
+      "loss": 0.5401,
+      "step": 11738
+    },
+    {
+      "epoch": 15.02592,
+      "grad_norm": 0.9743524193763733,
+      "learning_rate": 2.6536614645858344e-05,
+      "loss": 0.5219,
+      "step": 11739
+    },
+    {
+      "epoch": 15.0272,
+      "grad_norm": 1.019026279449463,
+      "learning_rate": 2.6534613845538216e-05,
+      "loss": 0.5246,
+      "step": 11740
+    },
+    {
+      "epoch": 15.02848,
+      "grad_norm": 0.9823682308197021,
+      "learning_rate": 2.653261304521809e-05,
+      "loss": 0.5357,
+      "step": 11741
+    },
+    {
+      "epoch": 15.02976,
+      "grad_norm": 0.9612826108932495,
+      "learning_rate": 2.6530612244897963e-05,
+      "loss": 0.522,
+      "step": 11742
+    },
+    {
+      "epoch": 15.03104,
+      "grad_norm": 0.9693880677223206,
+      "learning_rate": 2.6528611444577835e-05,
+      "loss": 0.5181,
+      "step": 11743
+    },
+    {
+      "epoch": 15.03232,
+      "grad_norm": 0.9547830820083618,
+      "learning_rate": 2.6526610644257703e-05,
+      "loss": 0.5113,
+      "step": 11744
+    },
+    {
+      "epoch": 15.0336,
+      "grad_norm": 0.9662365317344666,
+      "learning_rate": 2.6524609843937575e-05,
+      "loss": 0.5069,
+      "step": 11745
+    },
+    {
+      "epoch": 15.03488,
+      "grad_norm": 0.9782978296279907,
+      "learning_rate": 2.6522609043617447e-05,
+      "loss": 0.5354,
+      "step": 11746
+    },
+    {
+      "epoch": 15.03616,
+      "grad_norm": 0.9016302824020386,
+      "learning_rate": 2.652060824329732e-05,
+      "loss": 0.4629,
+      "step": 11747
+    },
+    {
+      "epoch": 15.03744,
+      "grad_norm": 0.8719879388809204,
+      "learning_rate": 2.6518607442977194e-05,
+      "loss": 0.4583,
+      "step": 11748
+    },
+    {
+      "epoch": 15.03872,
+      "grad_norm": 1.0020391941070557,
+      "learning_rate": 2.6516606642657066e-05,
+      "loss": 0.5467,
+      "step": 11749
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 0.999260663986206,
+      "learning_rate": 2.6514605842336938e-05,
+      "loss": 0.5304,
+      "step": 11750
+    },
+    {
+      "epoch": 15.04128,
+      "grad_norm": 1.00063157081604,
+      "learning_rate": 2.651260504201681e-05,
+      "loss": 0.5185,
+      "step": 11751
+    },
+    {
+      "epoch": 15.04256,
+      "grad_norm": 1.0226551294326782,
+      "learning_rate": 2.6510604241696678e-05,
+      "loss": 0.5016,
+      "step": 11752
+    },
+    {
+      "epoch": 15.04384,
+      "grad_norm": 0.99168461561203,
+      "learning_rate": 2.650860344137655e-05,
+      "loss": 0.5583,
+      "step": 11753
+    },
+    {
+      "epoch": 15.04512,
+      "grad_norm": 0.9974046349525452,
+      "learning_rate": 2.6506602641056422e-05,
+      "loss": 0.4811,
+      "step": 11754
+    },
+    {
+      "epoch": 15.0464,
+      "grad_norm": 1.019737958908081,
+      "learning_rate": 2.6504601840736297e-05,
+      "loss": 0.5457,
+      "step": 11755
+    },
+    {
+      "epoch": 15.04768,
+      "grad_norm": 0.9845877289772034,
+      "learning_rate": 2.650260104041617e-05,
+      "loss": 0.4997,
+      "step": 11756
+    },
+    {
+      "epoch": 15.04896,
+      "grad_norm": 0.9450182318687439,
+      "learning_rate": 2.650060024009604e-05,
+      "loss": 0.51,
+      "step": 11757
+    },
+    {
+      "epoch": 15.05024,
+      "grad_norm": 1.0113811492919922,
+      "learning_rate": 2.6498599439775913e-05,
+      "loss": 0.5266,
+      "step": 11758
+    },
+    {
+      "epoch": 15.05152,
+      "grad_norm": 0.9556260108947754,
+      "learning_rate": 2.6496598639455785e-05,
+      "loss": 0.517,
+      "step": 11759
+    },
+    {
+      "epoch": 15.0528,
+      "grad_norm": 0.9880708456039429,
+      "learning_rate": 2.6494597839135653e-05,
+      "loss": 0.5527,
+      "step": 11760
+    },
+    {
+      "epoch": 15.05408,
+      "grad_norm": 0.8803279399871826,
+      "learning_rate": 2.6492597038815525e-05,
+      "loss": 0.4476,
+      "step": 11761
+    },
+    {
+      "epoch": 15.05536,
+      "grad_norm": 0.974906861782074,
+      "learning_rate": 2.6490596238495404e-05,
+      "loss": 0.5251,
+      "step": 11762
+    },
+    {
+      "epoch": 15.05664,
+      "grad_norm": 0.9600889086723328,
+      "learning_rate": 2.6488595438175272e-05,
+      "loss": 0.4955,
+      "step": 11763
+    },
+    {
+      "epoch": 15.05792,
+      "grad_norm": 0.9697887301445007,
+      "learning_rate": 2.6486594637855144e-05,
+      "loss": 0.5455,
+      "step": 11764
+    },
+    {
+      "epoch": 15.0592,
+      "grad_norm": 0.9973766207695007,
+      "learning_rate": 2.6484593837535016e-05,
+      "loss": 0.4653,
+      "step": 11765
+    },
+    {
+      "epoch": 15.06048,
+      "grad_norm": 1.0244485139846802,
+      "learning_rate": 2.6482593037214888e-05,
+      "loss": 0.5189,
+      "step": 11766
+    },
+    {
+      "epoch": 15.06176,
+      "grad_norm": 0.9942975640296936,
+      "learning_rate": 2.648059223689476e-05,
+      "loss": 0.5736,
+      "step": 11767
+    },
+    {
+      "epoch": 15.06304,
+      "grad_norm": 0.9399450421333313,
+      "learning_rate": 2.6478591436574628e-05,
+      "loss": 0.5184,
+      "step": 11768
+    },
+    {
+      "epoch": 15.06432,
+      "grad_norm": 0.9376527070999146,
+      "learning_rate": 2.6476590636254507e-05,
+      "loss": 0.494,
+      "step": 11769
+    },
+    {
+      "epoch": 15.0656,
+      "grad_norm": 0.9570677280426025,
+      "learning_rate": 2.647458983593438e-05,
+      "loss": 0.5268,
+      "step": 11770
+    },
+    {
+      "epoch": 15.06688,
+      "grad_norm": 0.9823542833328247,
+      "learning_rate": 2.6472589035614247e-05,
+      "loss": 0.5312,
+      "step": 11771
+    },
+    {
+      "epoch": 15.06816,
+      "grad_norm": 1.0721344947814941,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.52,
+      "step": 11772
+    },
+    {
+      "epoch": 15.06944,
+      "grad_norm": 1.008894681930542,
+      "learning_rate": 2.646858743497399e-05,
+      "loss": 0.5164,
+      "step": 11773
+    },
+    {
+      "epoch": 15.07072,
+      "grad_norm": 0.9895505905151367,
+      "learning_rate": 2.6466586634653862e-05,
+      "loss": 0.5588,
+      "step": 11774
+    },
+    {
+      "epoch": 15.072,
+      "grad_norm": 1.0512263774871826,
+      "learning_rate": 2.6464585834333734e-05,
+      "loss": 0.5231,
+      "step": 11775
+    },
+    {
+      "epoch": 15.07328,
+      "grad_norm": 0.891127347946167,
+      "learning_rate": 2.646258503401361e-05,
+      "loss": 0.4783,
+      "step": 11776
+    },
+    {
+      "epoch": 15.07456,
+      "grad_norm": 0.9510053992271423,
+      "learning_rate": 2.646058423369348e-05,
+      "loss": 0.5321,
+      "step": 11777
+    },
+    {
+      "epoch": 15.07584,
+      "grad_norm": 0.954017698764801,
+      "learning_rate": 2.6458583433373353e-05,
+      "loss": 0.5153,
+      "step": 11778
+    },
+    {
+      "epoch": 15.07712,
+      "grad_norm": 1.0456651449203491,
+      "learning_rate": 2.6456582633053222e-05,
+      "loss": 0.5779,
+      "step": 11779
+    },
+    {
+      "epoch": 15.0784,
+      "grad_norm": 0.9539825320243835,
+      "learning_rate": 2.6454581832733094e-05,
+      "loss": 0.5285,
+      "step": 11780
+    },
+    {
+      "epoch": 15.07968,
+      "grad_norm": 0.952126681804657,
+      "learning_rate": 2.6452581032412965e-05,
+      "loss": 0.512,
+      "step": 11781
+    },
+    {
+      "epoch": 15.08096,
+      "grad_norm": 0.9592217803001404,
+      "learning_rate": 2.6450580232092837e-05,
+      "loss": 0.4852,
+      "step": 11782
+    },
+    {
+      "epoch": 15.08224,
+      "grad_norm": 0.9491934180259705,
+      "learning_rate": 2.644857943177271e-05,
+      "loss": 0.4663,
+      "step": 11783
+    },
+    {
+      "epoch": 15.08352,
+      "grad_norm": 0.9953993558883667,
+      "learning_rate": 2.6446578631452584e-05,
+      "loss": 0.4667,
+      "step": 11784
+    },
+    {
+      "epoch": 15.0848,
+      "grad_norm": 1.0062161684036255,
+      "learning_rate": 2.6444577831132456e-05,
+      "loss": 0.4812,
+      "step": 11785
+    },
+    {
+      "epoch": 15.08608,
+      "grad_norm": 0.9338573217391968,
+      "learning_rate": 2.6442577030812328e-05,
+      "loss": 0.5201,
+      "step": 11786
+    },
+    {
+      "epoch": 15.08736,
+      "grad_norm": 0.9554229378700256,
+      "learning_rate": 2.6440576230492197e-05,
+      "loss": 0.5064,
+      "step": 11787
+    },
+    {
+      "epoch": 15.08864,
+      "grad_norm": 0.9917376041412354,
+      "learning_rate": 2.643857543017207e-05,
+      "loss": 0.528,
+      "step": 11788
+    },
+    {
+      "epoch": 15.08992,
+      "grad_norm": 0.9698293209075928,
+      "learning_rate": 2.643657462985194e-05,
+      "loss": 0.4922,
+      "step": 11789
+    },
+    {
+      "epoch": 15.0912,
+      "grad_norm": 1.039360761642456,
+      "learning_rate": 2.6434573829531812e-05,
+      "loss": 0.5244,
+      "step": 11790
+    },
+    {
+      "epoch": 15.09248,
+      "grad_norm": 0.9987143874168396,
+      "learning_rate": 2.6432573029211687e-05,
+      "loss": 0.5006,
+      "step": 11791
+    },
+    {
+      "epoch": 15.09376,
+      "grad_norm": 0.9327332377433777,
+      "learning_rate": 2.643057222889156e-05,
+      "loss": 0.4795,
+      "step": 11792
+    },
+    {
+      "epoch": 15.09504,
+      "grad_norm": 0.9864438772201538,
+      "learning_rate": 2.642857142857143e-05,
+      "loss": 0.4949,
+      "step": 11793
+    },
+    {
+      "epoch": 15.09632,
+      "grad_norm": 0.99493008852005,
+      "learning_rate": 2.6426570628251303e-05,
+      "loss": 0.5297,
+      "step": 11794
+    },
+    {
+      "epoch": 15.0976,
+      "grad_norm": 0.9412096738815308,
+      "learning_rate": 2.642456982793117e-05,
+      "loss": 0.4775,
+      "step": 11795
+    },
+    {
+      "epoch": 15.09888,
+      "grad_norm": 0.9617460370063782,
+      "learning_rate": 2.6422569027611043e-05,
+      "loss": 0.4855,
+      "step": 11796
+    },
+    {
+      "epoch": 15.10016,
+      "grad_norm": 0.982379674911499,
+      "learning_rate": 2.6420568227290915e-05,
+      "loss": 0.5227,
+      "step": 11797
+    },
+    {
+      "epoch": 15.10144,
+      "grad_norm": 1.020799994468689,
+      "learning_rate": 2.641856742697079e-05,
+      "loss": 0.579,
+      "step": 11798
+    },
+    {
+      "epoch": 15.10272,
+      "grad_norm": 1.0289467573165894,
+      "learning_rate": 2.6416566626650662e-05,
+      "loss": 0.5353,
+      "step": 11799
+    },
+    {
+      "epoch": 15.104,
+      "grad_norm": 1.0211548805236816,
+      "learning_rate": 2.6414565826330534e-05,
+      "loss": 0.5503,
+      "step": 11800
+    },
+    {
+      "epoch": 15.10528,
+      "grad_norm": 1.011851191520691,
+      "learning_rate": 2.6412565026010406e-05,
+      "loss": 0.5338,
+      "step": 11801
+    },
+    {
+      "epoch": 15.10656,
+      "grad_norm": 1.0263190269470215,
+      "learning_rate": 2.6410564225690278e-05,
+      "loss": 0.5643,
+      "step": 11802
+    },
+    {
+      "epoch": 15.10784,
+      "grad_norm": 0.9878085851669312,
+      "learning_rate": 2.6408563425370146e-05,
+      "loss": 0.5279,
+      "step": 11803
+    },
+    {
+      "epoch": 15.10912,
+      "grad_norm": 0.9296584129333496,
+      "learning_rate": 2.6406562625050018e-05,
+      "loss": 0.4767,
+      "step": 11804
+    },
+    {
+      "epoch": 15.1104,
+      "grad_norm": 0.9973602294921875,
+      "learning_rate": 2.6404561824729897e-05,
+      "loss": 0.5266,
+      "step": 11805
+    },
+    {
+      "epoch": 15.11168,
+      "grad_norm": 0.9674133658409119,
+      "learning_rate": 2.6402561024409765e-05,
+      "loss": 0.511,
+      "step": 11806
+    },
+    {
+      "epoch": 15.11296,
+      "grad_norm": 1.0054680109024048,
+      "learning_rate": 2.6400560224089637e-05,
+      "loss": 0.5167,
+      "step": 11807
+    },
+    {
+      "epoch": 15.11424,
+      "grad_norm": 0.9890505075454712,
+      "learning_rate": 2.639855942376951e-05,
+      "loss": 0.5279,
+      "step": 11808
+    },
+    {
+      "epoch": 15.11552,
+      "grad_norm": 0.9986791610717773,
+      "learning_rate": 2.639655862344938e-05,
+      "loss": 0.5025,
+      "step": 11809
+    },
+    {
+      "epoch": 15.1168,
+      "grad_norm": 0.9749217629432678,
+      "learning_rate": 2.6394557823129253e-05,
+      "loss": 0.4886,
+      "step": 11810
+    },
+    {
+      "epoch": 15.11808,
+      "grad_norm": 0.9737854599952698,
+      "learning_rate": 2.639255702280912e-05,
+      "loss": 0.4664,
+      "step": 11811
+    },
+    {
+      "epoch": 15.11936,
+      "grad_norm": 0.9293329119682312,
+      "learning_rate": 2.6390556222489e-05,
+      "loss": 0.4985,
+      "step": 11812
+    },
+    {
+      "epoch": 15.12064,
+      "grad_norm": 0.9365719556808472,
+      "learning_rate": 2.638855542216887e-05,
+      "loss": 0.5201,
+      "step": 11813
+    },
+    {
+      "epoch": 15.12192,
+      "grad_norm": 0.9597119092941284,
+      "learning_rate": 2.638655462184874e-05,
+      "loss": 0.5045,
+      "step": 11814
+    },
+    {
+      "epoch": 15.1232,
+      "grad_norm": 1.0304876565933228,
+      "learning_rate": 2.6384553821528612e-05,
+      "loss": 0.5272,
+      "step": 11815
+    },
+    {
+      "epoch": 15.12448,
+      "grad_norm": 0.9947621822357178,
+      "learning_rate": 2.6382553021208484e-05,
+      "loss": 0.5527,
+      "step": 11816
+    },
+    {
+      "epoch": 15.12576,
+      "grad_norm": 0.9298506379127502,
+      "learning_rate": 2.6380552220888356e-05,
+      "loss": 0.4879,
+      "step": 11817
+    },
+    {
+      "epoch": 15.12704,
+      "grad_norm": 1.0342029333114624,
+      "learning_rate": 2.6378551420568228e-05,
+      "loss": 0.5352,
+      "step": 11818
+    },
+    {
+      "epoch": 15.12832,
+      "grad_norm": 0.9854941368103027,
+      "learning_rate": 2.6376550620248103e-05,
+      "loss": 0.5068,
+      "step": 11819
+    },
+    {
+      "epoch": 15.1296,
+      "grad_norm": 1.0126068592071533,
+      "learning_rate": 2.6374549819927975e-05,
+      "loss": 0.5505,
+      "step": 11820
+    },
+    {
+      "epoch": 15.13088,
+      "grad_norm": 1.0106240510940552,
+      "learning_rate": 2.6372549019607846e-05,
+      "loss": 0.5167,
+      "step": 11821
+    },
+    {
+      "epoch": 15.13216,
+      "grad_norm": 1.0682857036590576,
+      "learning_rate": 2.6370548219287715e-05,
+      "loss": 0.5697,
+      "step": 11822
+    },
+    {
+      "epoch": 15.13344,
+      "grad_norm": 1.0406882762908936,
+      "learning_rate": 2.6368547418967587e-05,
+      "loss": 0.5549,
+      "step": 11823
+    },
+    {
+      "epoch": 15.13472,
+      "grad_norm": 0.9883754253387451,
+      "learning_rate": 2.636654661864746e-05,
+      "loss": 0.4826,
+      "step": 11824
+    },
+    {
+      "epoch": 15.136,
+      "grad_norm": 1.0199334621429443,
+      "learning_rate": 2.636454581832733e-05,
+      "loss": 0.5189,
+      "step": 11825
+    },
+    {
+      "epoch": 15.13728,
+      "grad_norm": 1.0423178672790527,
+      "learning_rate": 2.6362545018007206e-05,
+      "loss": 0.5777,
+      "step": 11826
+    },
+    {
+      "epoch": 15.13856,
+      "grad_norm": 1.055101752281189,
+      "learning_rate": 2.6360544217687078e-05,
+      "loss": 0.555,
+      "step": 11827
+    },
+    {
+      "epoch": 15.13984,
+      "grad_norm": 0.9862048625946045,
+      "learning_rate": 2.635854341736695e-05,
+      "loss": 0.5281,
+      "step": 11828
+    },
+    {
+      "epoch": 15.14112,
+      "grad_norm": 0.9423441886901855,
+      "learning_rate": 2.635654261704682e-05,
+      "loss": 0.4727,
+      "step": 11829
+    },
+    {
+      "epoch": 15.1424,
+      "grad_norm": 0.9520114660263062,
+      "learning_rate": 2.635454181672669e-05,
+      "loss": 0.541,
+      "step": 11830
+    },
+    {
+      "epoch": 15.14368,
+      "grad_norm": 0.9528147578239441,
+      "learning_rate": 2.635254101640656e-05,
+      "loss": 0.5085,
+      "step": 11831
+    },
+    {
+      "epoch": 15.14496,
+      "grad_norm": 0.9359756112098694,
+      "learning_rate": 2.6350540216086434e-05,
+      "loss": 0.4601,
+      "step": 11832
+    },
+    {
+      "epoch": 15.14624,
+      "grad_norm": 0.9768989086151123,
+      "learning_rate": 2.634853941576631e-05,
+      "loss": 0.494,
+      "step": 11833
+    },
+    {
+      "epoch": 15.14752,
+      "grad_norm": 0.9752456545829773,
+      "learning_rate": 2.634653861544618e-05,
+      "loss": 0.5484,
+      "step": 11834
+    },
+    {
+      "epoch": 15.1488,
+      "grad_norm": 0.9548261165618896,
+      "learning_rate": 2.6344537815126052e-05,
+      "loss": 0.5219,
+      "step": 11835
+    },
+    {
+      "epoch": 15.150079999999999,
+      "grad_norm": 1.0379620790481567,
+      "learning_rate": 2.6342537014805924e-05,
+      "loss": 0.5714,
+      "step": 11836
+    },
+    {
+      "epoch": 15.15136,
+      "grad_norm": 0.9867843389511108,
+      "learning_rate": 2.6340536214485796e-05,
+      "loss": 0.5445,
+      "step": 11837
+    },
+    {
+      "epoch": 15.15264,
+      "grad_norm": 0.9981434941291809,
+      "learning_rate": 2.6338535414165665e-05,
+      "loss": 0.4993,
+      "step": 11838
+    },
+    {
+      "epoch": 15.15392,
+      "grad_norm": 1.0145291090011597,
+      "learning_rate": 2.6336534613845537e-05,
+      "loss": 0.52,
+      "step": 11839
+    },
+    {
+      "epoch": 15.1552,
+      "grad_norm": 0.9970989227294922,
+      "learning_rate": 2.6334533813525415e-05,
+      "loss": 0.5516,
+      "step": 11840
+    },
+    {
+      "epoch": 15.15648,
+      "grad_norm": 0.9794284701347351,
+      "learning_rate": 2.6332533013205284e-05,
+      "loss": 0.5283,
+      "step": 11841
+    },
+    {
+      "epoch": 15.15776,
+      "grad_norm": 1.0003759860992432,
+      "learning_rate": 2.6330532212885155e-05,
+      "loss": 0.4844,
+      "step": 11842
+    },
+    {
+      "epoch": 15.15904,
+      "grad_norm": 0.949975311756134,
+      "learning_rate": 2.6328531412565027e-05,
+      "loss": 0.4862,
+      "step": 11843
+    },
+    {
+      "epoch": 15.16032,
+      "grad_norm": 0.9530737400054932,
+      "learning_rate": 2.63265306122449e-05,
+      "loss": 0.4736,
+      "step": 11844
+    },
+    {
+      "epoch": 15.1616,
+      "grad_norm": 0.9621790051460266,
+      "learning_rate": 2.632452981192477e-05,
+      "loss": 0.4617,
+      "step": 11845
+    },
+    {
+      "epoch": 15.16288,
+      "grad_norm": 1.0266780853271484,
+      "learning_rate": 2.632252901160464e-05,
+      "loss": 0.5228,
+      "step": 11846
+    },
+    {
+      "epoch": 15.16416,
+      "grad_norm": 0.9723334908485413,
+      "learning_rate": 2.6320528211284518e-05,
+      "loss": 0.4968,
+      "step": 11847
+    },
+    {
+      "epoch": 15.16544,
+      "grad_norm": 0.9815566539764404,
+      "learning_rate": 2.631852741096439e-05,
+      "loss": 0.5549,
+      "step": 11848
+    },
+    {
+      "epoch": 15.16672,
+      "grad_norm": 0.9635928869247437,
+      "learning_rate": 2.631652661064426e-05,
+      "loss": 0.5149,
+      "step": 11849
+    },
+    {
+      "epoch": 15.168,
+      "grad_norm": 0.9444295167922974,
+      "learning_rate": 2.631452581032413e-05,
+      "loss": 0.5338,
+      "step": 11850
+    },
+    {
+      "epoch": 15.16928,
+      "grad_norm": 0.9650683403015137,
+      "learning_rate": 2.6312525010004002e-05,
+      "loss": 0.5011,
+      "step": 11851
+    },
+    {
+      "epoch": 15.17056,
+      "grad_norm": 0.97637939453125,
+      "learning_rate": 2.6310524209683874e-05,
+      "loss": 0.4643,
+      "step": 11852
+    },
+    {
+      "epoch": 15.17184,
+      "grad_norm": 0.9784893989562988,
+      "learning_rate": 2.6308523409363746e-05,
+      "loss": 0.4875,
+      "step": 11853
+    },
+    {
+      "epoch": 15.17312,
+      "grad_norm": 0.9220801591873169,
+      "learning_rate": 2.630652260904362e-05,
+      "loss": 0.4402,
+      "step": 11854
+    },
+    {
+      "epoch": 15.1744,
+      "grad_norm": 1.0712640285491943,
+      "learning_rate": 2.6304521808723493e-05,
+      "loss": 0.5405,
+      "step": 11855
+    },
+    {
+      "epoch": 15.17568,
+      "grad_norm": 0.9892664551734924,
+      "learning_rate": 2.6302521008403365e-05,
+      "loss": 0.5264,
+      "step": 11856
+    },
+    {
+      "epoch": 15.17696,
+      "grad_norm": 0.9567806720733643,
+      "learning_rate": 2.6300520208083233e-05,
+      "loss": 0.4772,
+      "step": 11857
+    },
+    {
+      "epoch": 15.17824,
+      "grad_norm": 0.9925405979156494,
+      "learning_rate": 2.6298519407763105e-05,
+      "loss": 0.5128,
+      "step": 11858
+    },
+    {
+      "epoch": 15.17952,
+      "grad_norm": 0.9644914865493774,
+      "learning_rate": 2.6296518607442977e-05,
+      "loss": 0.4847,
+      "step": 11859
+    },
+    {
+      "epoch": 15.1808,
+      "grad_norm": 0.9735480546951294,
+      "learning_rate": 2.629451780712285e-05,
+      "loss": 0.5335,
+      "step": 11860
+    },
+    {
+      "epoch": 15.18208,
+      "grad_norm": 0.99360591173172,
+      "learning_rate": 2.6292517006802724e-05,
+      "loss": 0.5185,
+      "step": 11861
+    },
+    {
+      "epoch": 15.18336,
+      "grad_norm": 0.9777383208274841,
+      "learning_rate": 2.6290516206482596e-05,
+      "loss": 0.5173,
+      "step": 11862
+    },
+    {
+      "epoch": 15.18464,
+      "grad_norm": 0.997850775718689,
+      "learning_rate": 2.6288515406162468e-05,
+      "loss": 0.5199,
+      "step": 11863
+    },
+    {
+      "epoch": 15.18592,
+      "grad_norm": 0.9816862344741821,
+      "learning_rate": 2.628651460584234e-05,
+      "loss": 0.5277,
+      "step": 11864
+    },
+    {
+      "epoch": 15.1872,
+      "grad_norm": 1.0436856746673584,
+      "learning_rate": 2.6284513805522208e-05,
+      "loss": 0.5121,
+      "step": 11865
+    },
+    {
+      "epoch": 15.18848,
+      "grad_norm": 0.9631795287132263,
+      "learning_rate": 2.628251300520208e-05,
+      "loss": 0.4858,
+      "step": 11866
+    },
+    {
+      "epoch": 15.18976,
+      "grad_norm": 0.8962351083755493,
+      "learning_rate": 2.6280512204881952e-05,
+      "loss": 0.5049,
+      "step": 11867
+    },
+    {
+      "epoch": 15.19104,
+      "grad_norm": 0.99549400806427,
+      "learning_rate": 2.6278511404561827e-05,
+      "loss": 0.5051,
+      "step": 11868
+    },
+    {
+      "epoch": 15.19232,
+      "grad_norm": 0.926495373249054,
+      "learning_rate": 2.62765106042417e-05,
+      "loss": 0.4831,
+      "step": 11869
+    },
+    {
+      "epoch": 15.1936,
+      "grad_norm": 1.0090394020080566,
+      "learning_rate": 2.627450980392157e-05,
+      "loss": 0.5091,
+      "step": 11870
+    },
+    {
+      "epoch": 15.19488,
+      "grad_norm": 1.0354359149932861,
+      "learning_rate": 2.6272509003601443e-05,
+      "loss": 0.558,
+      "step": 11871
+    },
+    {
+      "epoch": 15.19616,
+      "grad_norm": 1.04960036277771,
+      "learning_rate": 2.6270508203281315e-05,
+      "loss": 0.5582,
+      "step": 11872
+    },
+    {
+      "epoch": 15.19744,
+      "grad_norm": 1.0015323162078857,
+      "learning_rate": 2.6268507402961183e-05,
+      "loss": 0.4953,
+      "step": 11873
+    },
+    {
+      "epoch": 15.19872,
+      "grad_norm": 0.971566915512085,
+      "learning_rate": 2.6266506602641055e-05,
+      "loss": 0.5299,
+      "step": 11874
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 0.9685811996459961,
+      "learning_rate": 2.6264505802320934e-05,
+      "loss": 0.5111,
+      "step": 11875
+    },
+    {
+      "epoch": 15.20128,
+      "grad_norm": 0.9589853882789612,
+      "learning_rate": 2.6262505002000802e-05,
+      "loss": 0.4904,
+      "step": 11876
+    },
+    {
+      "epoch": 15.20256,
+      "grad_norm": 1.025733232498169,
+      "learning_rate": 2.6260504201680674e-05,
+      "loss": 0.4765,
+      "step": 11877
+    },
+    {
+      "epoch": 15.20384,
+      "grad_norm": 1.0392835140228271,
+      "learning_rate": 2.6258503401360546e-05,
+      "loss": 0.5459,
+      "step": 11878
+    },
+    {
+      "epoch": 15.20512,
+      "grad_norm": 0.9788011312484741,
+      "learning_rate": 2.6256502601040418e-05,
+      "loss": 0.4549,
+      "step": 11879
+    },
+    {
+      "epoch": 15.2064,
+      "grad_norm": 0.9786105751991272,
+      "learning_rate": 2.625450180072029e-05,
+      "loss": 0.5015,
+      "step": 11880
+    },
+    {
+      "epoch": 15.20768,
+      "grad_norm": 0.9867866635322571,
+      "learning_rate": 2.6252501000400158e-05,
+      "loss": 0.5019,
+      "step": 11881
+    },
+    {
+      "epoch": 15.20896,
+      "grad_norm": 1.0612010955810547,
+      "learning_rate": 2.6250500200080037e-05,
+      "loss": 0.5723,
+      "step": 11882
+    },
+    {
+      "epoch": 15.21024,
+      "grad_norm": 0.9832754731178284,
+      "learning_rate": 2.624849939975991e-05,
+      "loss": 0.5225,
+      "step": 11883
+    },
+    {
+      "epoch": 15.21152,
+      "grad_norm": 1.044143795967102,
+      "learning_rate": 2.6246498599439777e-05,
+      "loss": 0.5303,
+      "step": 11884
+    },
+    {
+      "epoch": 15.2128,
+      "grad_norm": 0.9782372713088989,
+      "learning_rate": 2.624449779911965e-05,
+      "loss": 0.5279,
+      "step": 11885
+    },
+    {
+      "epoch": 15.21408,
+      "grad_norm": 1.020448923110962,
+      "learning_rate": 2.624249699879952e-05,
+      "loss": 0.5357,
+      "step": 11886
+    },
+    {
+      "epoch": 15.21536,
+      "grad_norm": 1.0035127401351929,
+      "learning_rate": 2.6240496198479392e-05,
+      "loss": 0.5383,
+      "step": 11887
+    },
+    {
+      "epoch": 15.21664,
+      "grad_norm": 0.9856539964675903,
+      "learning_rate": 2.6238495398159264e-05,
+      "loss": 0.5267,
+      "step": 11888
+    },
+    {
+      "epoch": 15.21792,
+      "grad_norm": 0.979160726070404,
+      "learning_rate": 2.623649459783914e-05,
+      "loss": 0.4924,
+      "step": 11889
+    },
+    {
+      "epoch": 15.2192,
+      "grad_norm": 0.9597509503364563,
+      "learning_rate": 2.623449379751901e-05,
+      "loss": 0.5109,
+      "step": 11890
+    },
+    {
+      "epoch": 15.22048,
+      "grad_norm": 0.9971436262130737,
+      "learning_rate": 2.6232492997198883e-05,
+      "loss": 0.5233,
+      "step": 11891
+    },
+    {
+      "epoch": 15.22176,
+      "grad_norm": 0.9758315086364746,
+      "learning_rate": 2.623049219687875e-05,
+      "loss": 0.5042,
+      "step": 11892
+    },
+    {
+      "epoch": 15.22304,
+      "grad_norm": 1.0089176893234253,
+      "learning_rate": 2.6228491396558624e-05,
+      "loss": 0.5225,
+      "step": 11893
+    },
+    {
+      "epoch": 15.22432,
+      "grad_norm": 0.9823707342147827,
+      "learning_rate": 2.6226490596238495e-05,
+      "loss": 0.5532,
+      "step": 11894
+    },
+    {
+      "epoch": 15.2256,
+      "grad_norm": 1.033499836921692,
+      "learning_rate": 2.6224489795918367e-05,
+      "loss": 0.5525,
+      "step": 11895
+    },
+    {
+      "epoch": 15.22688,
+      "grad_norm": 0.9601119160652161,
+      "learning_rate": 2.622248899559824e-05,
+      "loss": 0.5249,
+      "step": 11896
+    },
+    {
+      "epoch": 15.22816,
+      "grad_norm": 0.9555276036262512,
+      "learning_rate": 2.6220488195278114e-05,
+      "loss": 0.4677,
+      "step": 11897
+    },
+    {
+      "epoch": 15.22944,
+      "grad_norm": 1.0126163959503174,
+      "learning_rate": 2.6218487394957986e-05,
+      "loss": 0.5543,
+      "step": 11898
+    },
+    {
+      "epoch": 15.23072,
+      "grad_norm": 1.0423436164855957,
+      "learning_rate": 2.6216486594637858e-05,
+      "loss": 0.5366,
+      "step": 11899
+    },
+    {
+      "epoch": 15.232,
+      "grad_norm": 1.0581696033477783,
+      "learning_rate": 2.6214485794317727e-05,
+      "loss": 0.5205,
+      "step": 11900
+    },
+    {
+      "epoch": 15.23328,
+      "grad_norm": 1.026396632194519,
+      "learning_rate": 2.62124849939976e-05,
+      "loss": 0.4984,
+      "step": 11901
+    },
+    {
+      "epoch": 15.23456,
+      "grad_norm": 1.0131559371948242,
+      "learning_rate": 2.621048419367747e-05,
+      "loss": 0.498,
+      "step": 11902
+    },
+    {
+      "epoch": 15.23584,
+      "grad_norm": 1.0236692428588867,
+      "learning_rate": 2.6208483393357342e-05,
+      "loss": 0.5357,
+      "step": 11903
+    },
+    {
+      "epoch": 15.23712,
+      "grad_norm": 0.9787890911102295,
+      "learning_rate": 2.6206482593037217e-05,
+      "loss": 0.4933,
+      "step": 11904
+    },
+    {
+      "epoch": 15.2384,
+      "grad_norm": 0.9795733094215393,
+      "learning_rate": 2.620448179271709e-05,
+      "loss": 0.523,
+      "step": 11905
+    },
+    {
+      "epoch": 15.23968,
+      "grad_norm": 1.0700660943984985,
+      "learning_rate": 2.620248099239696e-05,
+      "loss": 0.5248,
+      "step": 11906
+    },
+    {
+      "epoch": 15.24096,
+      "grad_norm": 0.9715010523796082,
+      "learning_rate": 2.6200480192076833e-05,
+      "loss": 0.4913,
+      "step": 11907
+    },
+    {
+      "epoch": 15.24224,
+      "grad_norm": 0.9805188179016113,
+      "learning_rate": 2.61984793917567e-05,
+      "loss": 0.5119,
+      "step": 11908
+    },
+    {
+      "epoch": 15.24352,
+      "grad_norm": 1.0102686882019043,
+      "learning_rate": 2.6196478591436573e-05,
+      "loss": 0.5117,
+      "step": 11909
+    },
+    {
+      "epoch": 15.2448,
+      "grad_norm": 1.045552372932434,
+      "learning_rate": 2.6194477791116445e-05,
+      "loss": 0.563,
+      "step": 11910
+    },
+    {
+      "epoch": 15.24608,
+      "grad_norm": 0.9204160571098328,
+      "learning_rate": 2.619247699079632e-05,
+      "loss": 0.4714,
+      "step": 11911
+    },
+    {
+      "epoch": 15.24736,
+      "grad_norm": 1.0518666505813599,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.5461,
+      "step": 11912
+    },
+    {
+      "epoch": 15.24864,
+      "grad_norm": 1.0590685606002808,
+      "learning_rate": 2.6188475390156064e-05,
+      "loss": 0.5677,
+      "step": 11913
+    },
+    {
+      "epoch": 15.24992,
+      "grad_norm": 0.9414172172546387,
+      "learning_rate": 2.6186474589835936e-05,
+      "loss": 0.4822,
+      "step": 11914
+    },
+    {
+      "epoch": 15.2512,
+      "grad_norm": 0.9399723410606384,
+      "learning_rate": 2.6184473789515808e-05,
+      "loss": 0.509,
+      "step": 11915
+    },
+    {
+      "epoch": 15.25248,
+      "grad_norm": 0.9907810688018799,
+      "learning_rate": 2.6182472989195676e-05,
+      "loss": 0.5117,
+      "step": 11916
+    },
+    {
+      "epoch": 15.25376,
+      "grad_norm": 1.0035853385925293,
+      "learning_rate": 2.6180472188875548e-05,
+      "loss": 0.5313,
+      "step": 11917
+    },
+    {
+      "epoch": 15.25504,
+      "grad_norm": 0.9608493447303772,
+      "learning_rate": 2.6178471388555427e-05,
+      "loss": 0.4925,
+      "step": 11918
+    },
+    {
+      "epoch": 15.25632,
+      "grad_norm": 0.933178722858429,
+      "learning_rate": 2.6176470588235295e-05,
+      "loss": 0.5003,
+      "step": 11919
+    },
+    {
+      "epoch": 15.2576,
+      "grad_norm": 1.016798734664917,
+      "learning_rate": 2.6174469787915167e-05,
+      "loss": 0.5243,
+      "step": 11920
+    },
+    {
+      "epoch": 15.25888,
+      "grad_norm": 1.0794856548309326,
+      "learning_rate": 2.617246898759504e-05,
+      "loss": 0.5542,
+      "step": 11921
+    },
+    {
+      "epoch": 15.26016,
+      "grad_norm": 0.988269031047821,
+      "learning_rate": 2.617046818727491e-05,
+      "loss": 0.5095,
+      "step": 11922
+    },
+    {
+      "epoch": 15.26144,
+      "grad_norm": 1.0120782852172852,
+      "learning_rate": 2.6168467386954783e-05,
+      "loss": 0.5286,
+      "step": 11923
+    },
+    {
+      "epoch": 15.26272,
+      "grad_norm": 0.9555339813232422,
+      "learning_rate": 2.616646658663465e-05,
+      "loss": 0.4668,
+      "step": 11924
+    },
+    {
+      "epoch": 15.264,
+      "grad_norm": 0.9648301005363464,
+      "learning_rate": 2.616446578631453e-05,
+      "loss": 0.5279,
+      "step": 11925
+    },
+    {
+      "epoch": 15.26528,
+      "grad_norm": 1.041354775428772,
+      "learning_rate": 2.61624649859944e-05,
+      "loss": 0.5173,
+      "step": 11926
+    },
+    {
+      "epoch": 15.26656,
+      "grad_norm": 0.9829549193382263,
+      "learning_rate": 2.616046418567427e-05,
+      "loss": 0.5126,
+      "step": 11927
+    },
+    {
+      "epoch": 15.26784,
+      "grad_norm": 1.0028904676437378,
+      "learning_rate": 2.6158463385354142e-05,
+      "loss": 0.5008,
+      "step": 11928
+    },
+    {
+      "epoch": 15.269120000000001,
+      "grad_norm": 0.9798859357833862,
+      "learning_rate": 2.6156462585034014e-05,
+      "loss": 0.5285,
+      "step": 11929
+    },
+    {
+      "epoch": 15.2704,
+      "grad_norm": 1.0254261493682861,
+      "learning_rate": 2.6154461784713886e-05,
+      "loss": 0.5012,
+      "step": 11930
+    },
+    {
+      "epoch": 15.27168,
+      "grad_norm": 0.9710574746131897,
+      "learning_rate": 2.6152460984393757e-05,
+      "loss": 0.5052,
+      "step": 11931
+    },
+    {
+      "epoch": 15.27296,
+      "grad_norm": 1.0092233419418335,
+      "learning_rate": 2.6150460184073633e-05,
+      "loss": 0.5666,
+      "step": 11932
+    },
+    {
+      "epoch": 15.27424,
+      "grad_norm": 1.0055060386657715,
+      "learning_rate": 2.6148459383753505e-05,
+      "loss": 0.5386,
+      "step": 11933
+    },
+    {
+      "epoch": 15.27552,
+      "grad_norm": 1.033969759941101,
+      "learning_rate": 2.6146458583433376e-05,
+      "loss": 0.5583,
+      "step": 11934
+    },
+    {
+      "epoch": 15.2768,
+      "grad_norm": 0.95125812292099,
+      "learning_rate": 2.6144457783113245e-05,
+      "loss": 0.4865,
+      "step": 11935
+    },
+    {
+      "epoch": 15.27808,
+      "grad_norm": 1.0022237300872803,
+      "learning_rate": 2.6142456982793117e-05,
+      "loss": 0.5342,
+      "step": 11936
+    },
+    {
+      "epoch": 15.27936,
+      "grad_norm": 0.9819777011871338,
+      "learning_rate": 2.614045618247299e-05,
+      "loss": 0.5323,
+      "step": 11937
+    },
+    {
+      "epoch": 15.28064,
+      "grad_norm": 0.982841968536377,
+      "learning_rate": 2.613845538215286e-05,
+      "loss": 0.4979,
+      "step": 11938
+    },
+    {
+      "epoch": 15.28192,
+      "grad_norm": 1.0762674808502197,
+      "learning_rate": 2.6136454581832736e-05,
+      "loss": 0.559,
+      "step": 11939
+    },
+    {
+      "epoch": 15.2832,
+      "grad_norm": 1.0399508476257324,
+      "learning_rate": 2.6134453781512608e-05,
+      "loss": 0.5466,
+      "step": 11940
+    },
+    {
+      "epoch": 15.28448,
+      "grad_norm": 0.9770174026489258,
+      "learning_rate": 2.613245298119248e-05,
+      "loss": 0.5081,
+      "step": 11941
+    },
+    {
+      "epoch": 15.28576,
+      "grad_norm": 1.06191885471344,
+      "learning_rate": 2.613045218087235e-05,
+      "loss": 0.5334,
+      "step": 11942
+    },
+    {
+      "epoch": 15.28704,
+      "grad_norm": 0.9418057203292847,
+      "learning_rate": 2.612845138055222e-05,
+      "loss": 0.4954,
+      "step": 11943
+    },
+    {
+      "epoch": 15.28832,
+      "grad_norm": 0.9617568254470825,
+      "learning_rate": 2.612645058023209e-05,
+      "loss": 0.5103,
+      "step": 11944
+    },
+    {
+      "epoch": 15.2896,
+      "grad_norm": 1.0067156553268433,
+      "learning_rate": 2.6124449779911963e-05,
+      "loss": 0.5035,
+      "step": 11945
+    },
+    {
+      "epoch": 15.29088,
+      "grad_norm": 1.002389907836914,
+      "learning_rate": 2.612244897959184e-05,
+      "loss": 0.5344,
+      "step": 11946
+    },
+    {
+      "epoch": 15.292159999999999,
+      "grad_norm": 1.0987991094589233,
+      "learning_rate": 2.612044817927171e-05,
+      "loss": 0.5497,
+      "step": 11947
+    },
+    {
+      "epoch": 15.29344,
+      "grad_norm": 1.0003621578216553,
+      "learning_rate": 2.6118447378951582e-05,
+      "loss": 0.4983,
+      "step": 11948
+    },
+    {
+      "epoch": 15.29472,
+      "grad_norm": 0.9967955946922302,
+      "learning_rate": 2.6116446578631454e-05,
+      "loss": 0.5338,
+      "step": 11949
+    },
+    {
+      "epoch": 15.296,
+      "grad_norm": 1.0217162370681763,
+      "learning_rate": 2.6114445778311326e-05,
+      "loss": 0.5592,
+      "step": 11950
+    },
+    {
+      "epoch": 15.29728,
+      "grad_norm": 1.0223591327667236,
+      "learning_rate": 2.6112444977991195e-05,
+      "loss": 0.5181,
+      "step": 11951
+    },
+    {
+      "epoch": 15.29856,
+      "grad_norm": 0.9928198456764221,
+      "learning_rate": 2.6110444177671066e-05,
+      "loss": 0.5228,
+      "step": 11952
+    },
+    {
+      "epoch": 15.29984,
+      "grad_norm": 1.0652741193771362,
+      "learning_rate": 2.6108443377350945e-05,
+      "loss": 0.5615,
+      "step": 11953
+    },
+    {
+      "epoch": 15.30112,
+      "grad_norm": 1.0429683923721313,
+      "learning_rate": 2.6106442577030814e-05,
+      "loss": 0.5492,
+      "step": 11954
+    },
+    {
+      "epoch": 15.3024,
+      "grad_norm": 0.9588310718536377,
+      "learning_rate": 2.6104441776710685e-05,
+      "loss": 0.4757,
+      "step": 11955
+    },
+    {
+      "epoch": 15.30368,
+      "grad_norm": 0.9648290872573853,
+      "learning_rate": 2.6102440976390557e-05,
+      "loss": 0.4989,
+      "step": 11956
+    },
+    {
+      "epoch": 15.30496,
+      "grad_norm": 1.0313849449157715,
+      "learning_rate": 2.610044017607043e-05,
+      "loss": 0.5461,
+      "step": 11957
+    },
+    {
+      "epoch": 15.30624,
+      "grad_norm": 1.0368759632110596,
+      "learning_rate": 2.60984393757503e-05,
+      "loss": 0.5372,
+      "step": 11958
+    },
+    {
+      "epoch": 15.30752,
+      "grad_norm": 0.9525302052497864,
+      "learning_rate": 2.609643857543017e-05,
+      "loss": 0.4805,
+      "step": 11959
+    },
+    {
+      "epoch": 15.3088,
+      "grad_norm": 0.9847881197929382,
+      "learning_rate": 2.6094437775110048e-05,
+      "loss": 0.5264,
+      "step": 11960
+    },
+    {
+      "epoch": 15.31008,
+      "grad_norm": 0.9874829053878784,
+      "learning_rate": 2.609243697478992e-05,
+      "loss": 0.5335,
+      "step": 11961
+    },
+    {
+      "epoch": 15.31136,
+      "grad_norm": 1.0466731786727905,
+      "learning_rate": 2.609043617446979e-05,
+      "loss": 0.5521,
+      "step": 11962
+    },
+    {
+      "epoch": 15.31264,
+      "grad_norm": 1.0074479579925537,
+      "learning_rate": 2.608843537414966e-05,
+      "loss": 0.5141,
+      "step": 11963
+    },
+    {
+      "epoch": 15.31392,
+      "grad_norm": 1.0698914527893066,
+      "learning_rate": 2.6086434573829532e-05,
+      "loss": 0.5662,
+      "step": 11964
+    },
+    {
+      "epoch": 15.3152,
+      "grad_norm": 0.9645558595657349,
+      "learning_rate": 2.6084433773509404e-05,
+      "loss": 0.5126,
+      "step": 11965
+    },
+    {
+      "epoch": 15.31648,
+      "grad_norm": 0.9718447923660278,
+      "learning_rate": 2.6082432973189276e-05,
+      "loss": 0.5003,
+      "step": 11966
+    },
+    {
+      "epoch": 15.31776,
+      "grad_norm": 0.9619754552841187,
+      "learning_rate": 2.608043217286915e-05,
+      "loss": 0.52,
+      "step": 11967
+    },
+    {
+      "epoch": 15.31904,
+      "grad_norm": 0.940642237663269,
+      "learning_rate": 2.6078431372549023e-05,
+      "loss": 0.5113,
+      "step": 11968
+    },
+    {
+      "epoch": 15.32032,
+      "grad_norm": 1.0012346506118774,
+      "learning_rate": 2.6076430572228895e-05,
+      "loss": 0.4932,
+      "step": 11969
+    },
+    {
+      "epoch": 15.3216,
+      "grad_norm": 0.9923637509346008,
+      "learning_rate": 2.6074429771908763e-05,
+      "loss": 0.5567,
+      "step": 11970
+    },
+    {
+      "epoch": 15.32288,
+      "grad_norm": 0.9815508723258972,
+      "learning_rate": 2.6072428971588635e-05,
+      "loss": 0.5412,
+      "step": 11971
+    },
+    {
+      "epoch": 15.32416,
+      "grad_norm": 1.0121711492538452,
+      "learning_rate": 2.6070428171268507e-05,
+      "loss": 0.54,
+      "step": 11972
+    },
+    {
+      "epoch": 15.32544,
+      "grad_norm": 1.0295419692993164,
+      "learning_rate": 2.606842737094838e-05,
+      "loss": 0.5261,
+      "step": 11973
+    },
+    {
+      "epoch": 15.32672,
+      "grad_norm": 0.9861733913421631,
+      "learning_rate": 2.6066426570628254e-05,
+      "loss": 0.5237,
+      "step": 11974
+    },
+    {
+      "epoch": 15.328,
+      "grad_norm": 0.9513747692108154,
+      "learning_rate": 2.6064425770308126e-05,
+      "loss": 0.5107,
+      "step": 11975
+    },
+    {
+      "epoch": 15.32928,
+      "grad_norm": 0.9383821487426758,
+      "learning_rate": 2.6062424969987998e-05,
+      "loss": 0.5109,
+      "step": 11976
+    },
+    {
+      "epoch": 15.33056,
+      "grad_norm": 0.9647513628005981,
+      "learning_rate": 2.606042416966787e-05,
+      "loss": 0.5326,
+      "step": 11977
+    },
+    {
+      "epoch": 15.33184,
+      "grad_norm": 1.0194227695465088,
+      "learning_rate": 2.6058423369347738e-05,
+      "loss": 0.5523,
+      "step": 11978
+    },
+    {
+      "epoch": 15.33312,
+      "grad_norm": 0.9552738070487976,
+      "learning_rate": 2.605642256902761e-05,
+      "loss": 0.494,
+      "step": 11979
+    },
+    {
+      "epoch": 15.3344,
+      "grad_norm": 0.9982839822769165,
+      "learning_rate": 2.6054421768707482e-05,
+      "loss": 0.5138,
+      "step": 11980
+    },
+    {
+      "epoch": 15.33568,
+      "grad_norm": 0.926050066947937,
+      "learning_rate": 2.6052420968387357e-05,
+      "loss": 0.4827,
+      "step": 11981
+    },
+    {
+      "epoch": 15.33696,
+      "grad_norm": 0.9754160046577454,
+      "learning_rate": 2.605042016806723e-05,
+      "loss": 0.4864,
+      "step": 11982
+    },
+    {
+      "epoch": 15.33824,
+      "grad_norm": 1.0375244617462158,
+      "learning_rate": 2.60484193677471e-05,
+      "loss": 0.5324,
+      "step": 11983
+    },
+    {
+      "epoch": 15.33952,
+      "grad_norm": 0.998441755771637,
+      "learning_rate": 2.6046418567426973e-05,
+      "loss": 0.5374,
+      "step": 11984
+    },
+    {
+      "epoch": 15.3408,
+      "grad_norm": 0.9597004055976868,
+      "learning_rate": 2.6044417767106845e-05,
+      "loss": 0.5497,
+      "step": 11985
+    },
+    {
+      "epoch": 15.34208,
+      "grad_norm": 0.984380304813385,
+      "learning_rate": 2.6042416966786713e-05,
+      "loss": 0.4974,
+      "step": 11986
+    },
+    {
+      "epoch": 15.34336,
+      "grad_norm": 1.0443543195724487,
+      "learning_rate": 2.6040416166466585e-05,
+      "loss": 0.5094,
+      "step": 11987
+    },
+    {
+      "epoch": 15.34464,
+      "grad_norm": 1.033092737197876,
+      "learning_rate": 2.6038415366146463e-05,
+      "loss": 0.5502,
+      "step": 11988
+    },
+    {
+      "epoch": 15.34592,
+      "grad_norm": 1.0299837589263916,
+      "learning_rate": 2.6036414565826332e-05,
+      "loss": 0.5757,
+      "step": 11989
+    },
+    {
+      "epoch": 15.3472,
+      "grad_norm": 0.9337131381034851,
+      "learning_rate": 2.6034413765506204e-05,
+      "loss": 0.4797,
+      "step": 11990
+    },
+    {
+      "epoch": 15.34848,
+      "grad_norm": 0.9382844567298889,
+      "learning_rate": 2.6032412965186076e-05,
+      "loss": 0.4631,
+      "step": 11991
+    },
+    {
+      "epoch": 15.34976,
+      "grad_norm": 0.9595440626144409,
+      "learning_rate": 2.6030412164865948e-05,
+      "loss": 0.5302,
+      "step": 11992
+    },
+    {
+      "epoch": 15.35104,
+      "grad_norm": 0.9686310887336731,
+      "learning_rate": 2.602841136454582e-05,
+      "loss": 0.5061,
+      "step": 11993
+    },
+    {
+      "epoch": 15.35232,
+      "grad_norm": 0.9132741093635559,
+      "learning_rate": 2.6026410564225688e-05,
+      "loss": 0.4801,
+      "step": 11994
+    },
+    {
+      "epoch": 15.3536,
+      "grad_norm": 0.9716625213623047,
+      "learning_rate": 2.6024409763905566e-05,
+      "loss": 0.5035,
+      "step": 11995
+    },
+    {
+      "epoch": 15.35488,
+      "grad_norm": 1.003369927406311,
+      "learning_rate": 2.602240896358544e-05,
+      "loss": 0.547,
+      "step": 11996
+    },
+    {
+      "epoch": 15.35616,
+      "grad_norm": 1.0026856660842896,
+      "learning_rate": 2.6020408163265307e-05,
+      "loss": 0.547,
+      "step": 11997
+    },
+    {
+      "epoch": 15.35744,
+      "grad_norm": 0.9878264665603638,
+      "learning_rate": 2.601840736294518e-05,
+      "loss": 0.5284,
+      "step": 11998
+    },
+    {
+      "epoch": 15.35872,
+      "grad_norm": 1.0457453727722168,
+      "learning_rate": 2.601640656262505e-05,
+      "loss": 0.5354,
+      "step": 11999
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 1.0366843938827515,
+      "learning_rate": 2.6014405762304922e-05,
+      "loss": 0.5955,
+      "step": 12000
+    },
+    {
+      "epoch": 15.36128,
+      "grad_norm": 0.9786233901977539,
+      "learning_rate": 2.6012404961984794e-05,
+      "loss": 0.521,
+      "step": 12001
+    },
+    {
+      "epoch": 15.36256,
+      "grad_norm": 1.0292565822601318,
+      "learning_rate": 2.601040416166467e-05,
+      "loss": 0.5014,
+      "step": 12002
+    },
+    {
+      "epoch": 15.36384,
+      "grad_norm": 1.0155521631240845,
+      "learning_rate": 2.600840336134454e-05,
+      "loss": 0.5361,
+      "step": 12003
+    },
+    {
+      "epoch": 15.36512,
+      "grad_norm": 0.9816110134124756,
+      "learning_rate": 2.6006402561024413e-05,
+      "loss": 0.5233,
+      "step": 12004
+    },
+    {
+      "epoch": 15.3664,
+      "grad_norm": 0.9563271999359131,
+      "learning_rate": 2.600440176070428e-05,
+      "loss": 0.5141,
+      "step": 12005
+    },
+    {
+      "epoch": 15.36768,
+      "grad_norm": 1.029533863067627,
+      "learning_rate": 2.6002400960384154e-05,
+      "loss": 0.5331,
+      "step": 12006
+    },
+    {
+      "epoch": 15.36896,
+      "grad_norm": 1.0046855211257935,
+      "learning_rate": 2.6000400160064025e-05,
+      "loss": 0.535,
+      "step": 12007
+    },
+    {
+      "epoch": 15.37024,
+      "grad_norm": 0.9804155826568604,
+      "learning_rate": 2.5998399359743897e-05,
+      "loss": 0.513,
+      "step": 12008
+    },
+    {
+      "epoch": 15.37152,
+      "grad_norm": 0.9942843914031982,
+      "learning_rate": 2.599639855942377e-05,
+      "loss": 0.5471,
+      "step": 12009
+    },
+    {
+      "epoch": 15.3728,
+      "grad_norm": 1.0089268684387207,
+      "learning_rate": 2.5994397759103644e-05,
+      "loss": 0.5792,
+      "step": 12010
+    },
+    {
+      "epoch": 15.37408,
+      "grad_norm": 0.9854798913002014,
+      "learning_rate": 2.5992396958783516e-05,
+      "loss": 0.5647,
+      "step": 12011
+    },
+    {
+      "epoch": 15.37536,
+      "grad_norm": 1.0184553861618042,
+      "learning_rate": 2.5990396158463388e-05,
+      "loss": 0.4929,
+      "step": 12012
+    },
+    {
+      "epoch": 15.37664,
+      "grad_norm": 0.9660348296165466,
+      "learning_rate": 2.5988395358143257e-05,
+      "loss": 0.5246,
+      "step": 12013
+    },
+    {
+      "epoch": 15.37792,
+      "grad_norm": 0.9935752153396606,
+      "learning_rate": 2.598639455782313e-05,
+      "loss": 0.5349,
+      "step": 12014
+    },
+    {
+      "epoch": 15.3792,
+      "grad_norm": 0.953775942325592,
+      "learning_rate": 2.5984393757503e-05,
+      "loss": 0.5124,
+      "step": 12015
+    },
+    {
+      "epoch": 15.38048,
+      "grad_norm": 0.9608764052391052,
+      "learning_rate": 2.5982392957182872e-05,
+      "loss": 0.5077,
+      "step": 12016
+    },
+    {
+      "epoch": 15.38176,
+      "grad_norm": 0.9942414164543152,
+      "learning_rate": 2.5980392156862747e-05,
+      "loss": 0.5343,
+      "step": 12017
+    },
+    {
+      "epoch": 15.38304,
+      "grad_norm": 0.9368318915367126,
+      "learning_rate": 2.597839135654262e-05,
+      "loss": 0.5105,
+      "step": 12018
+    },
+    {
+      "epoch": 15.38432,
+      "grad_norm": 1.0121771097183228,
+      "learning_rate": 2.597639055622249e-05,
+      "loss": 0.5283,
+      "step": 12019
+    },
+    {
+      "epoch": 15.3856,
+      "grad_norm": 1.025707721710205,
+      "learning_rate": 2.5974389755902363e-05,
+      "loss": 0.5341,
+      "step": 12020
+    },
+    {
+      "epoch": 15.38688,
+      "grad_norm": 1.002144694328308,
+      "learning_rate": 2.597238895558223e-05,
+      "loss": 0.5056,
+      "step": 12021
+    },
+    {
+      "epoch": 15.38816,
+      "grad_norm": 0.9893521070480347,
+      "learning_rate": 2.5970388155262103e-05,
+      "loss": 0.5587,
+      "step": 12022
+    },
+    {
+      "epoch": 15.38944,
+      "grad_norm": 0.9755342602729797,
+      "learning_rate": 2.5968387354941975e-05,
+      "loss": 0.5413,
+      "step": 12023
+    },
+    {
+      "epoch": 15.39072,
+      "grad_norm": 0.972171425819397,
+      "learning_rate": 2.5966386554621854e-05,
+      "loss": 0.5081,
+      "step": 12024
+    },
+    {
+      "epoch": 15.392,
+      "grad_norm": 1.0347623825073242,
+      "learning_rate": 2.5964385754301722e-05,
+      "loss": 0.5589,
+      "step": 12025
+    },
+    {
+      "epoch": 15.39328,
+      "grad_norm": 0.9687240123748779,
+      "learning_rate": 2.5962384953981594e-05,
+      "loss": 0.5451,
+      "step": 12026
+    },
+    {
+      "epoch": 15.39456,
+      "grad_norm": 0.9607928991317749,
+      "learning_rate": 2.5960384153661466e-05,
+      "loss": 0.5309,
+      "step": 12027
+    },
+    {
+      "epoch": 15.39584,
+      "grad_norm": 0.9592061042785645,
+      "learning_rate": 2.5958383353341338e-05,
+      "loss": 0.5466,
+      "step": 12028
+    },
+    {
+      "epoch": 15.39712,
+      "grad_norm": 1.015899419784546,
+      "learning_rate": 2.5956382553021206e-05,
+      "loss": 0.5245,
+      "step": 12029
+    },
+    {
+      "epoch": 15.3984,
+      "grad_norm": 0.9896045327186584,
+      "learning_rate": 2.5954381752701078e-05,
+      "loss": 0.5316,
+      "step": 12030
+    },
+    {
+      "epoch": 15.39968,
+      "grad_norm": 0.9522262811660767,
+      "learning_rate": 2.5952380952380957e-05,
+      "loss": 0.5192,
+      "step": 12031
+    },
+    {
+      "epoch": 15.40096,
+      "grad_norm": 0.987667441368103,
+      "learning_rate": 2.595038015206083e-05,
+      "loss": 0.5136,
+      "step": 12032
+    },
+    {
+      "epoch": 15.40224,
+      "grad_norm": 1.0220928192138672,
+      "learning_rate": 2.5948379351740697e-05,
+      "loss": 0.57,
+      "step": 12033
+    },
+    {
+      "epoch": 15.40352,
+      "grad_norm": 1.0101734399795532,
+      "learning_rate": 2.594637855142057e-05,
+      "loss": 0.5165,
+      "step": 12034
+    },
+    {
+      "epoch": 15.4048,
+      "grad_norm": 0.9381194114685059,
+      "learning_rate": 2.594437775110044e-05,
+      "loss": 0.4597,
+      "step": 12035
+    },
+    {
+      "epoch": 15.40608,
+      "grad_norm": 1.0812232494354248,
+      "learning_rate": 2.5942376950780313e-05,
+      "loss": 0.5632,
+      "step": 12036
+    },
+    {
+      "epoch": 15.40736,
+      "grad_norm": 1.0491623878479004,
+      "learning_rate": 2.594037615046018e-05,
+      "loss": 0.538,
+      "step": 12037
+    },
+    {
+      "epoch": 15.40864,
+      "grad_norm": 0.9841446876525879,
+      "learning_rate": 2.593837535014006e-05,
+      "loss": 0.5059,
+      "step": 12038
+    },
+    {
+      "epoch": 15.40992,
+      "grad_norm": 0.9881943464279175,
+      "learning_rate": 2.593637454981993e-05,
+      "loss": 0.5154,
+      "step": 12039
+    },
+    {
+      "epoch": 15.411200000000001,
+      "grad_norm": 1.0402194261550903,
+      "learning_rate": 2.5934373749499803e-05,
+      "loss": 0.5756,
+      "step": 12040
+    },
+    {
+      "epoch": 15.41248,
+      "grad_norm": 1.095178246498108,
+      "learning_rate": 2.5932372949179672e-05,
+      "loss": 0.5601,
+      "step": 12041
+    },
+    {
+      "epoch": 15.41376,
+      "grad_norm": 1.0817269086837769,
+      "learning_rate": 2.5930372148859544e-05,
+      "loss": 0.5595,
+      "step": 12042
+    },
+    {
+      "epoch": 15.41504,
+      "grad_norm": 0.9906639456748962,
+      "learning_rate": 2.5928371348539416e-05,
+      "loss": 0.5067,
+      "step": 12043
+    },
+    {
+      "epoch": 15.41632,
+      "grad_norm": 0.9647719860076904,
+      "learning_rate": 2.5926370548219287e-05,
+      "loss": 0.4848,
+      "step": 12044
+    },
+    {
+      "epoch": 15.4176,
+      "grad_norm": 0.9558652639389038,
+      "learning_rate": 2.5924369747899163e-05,
+      "loss": 0.5381,
+      "step": 12045
+    },
+    {
+      "epoch": 15.41888,
+      "grad_norm": 0.9958375692367554,
+      "learning_rate": 2.5922368947579035e-05,
+      "loss": 0.5672,
+      "step": 12046
+    },
+    {
+      "epoch": 15.42016,
+      "grad_norm": 0.980514645576477,
+      "learning_rate": 2.5920368147258906e-05,
+      "loss": 0.5479,
+      "step": 12047
+    },
+    {
+      "epoch": 15.42144,
+      "grad_norm": 1.041264533996582,
+      "learning_rate": 2.5918367346938778e-05,
+      "loss": 0.5433,
+      "step": 12048
+    },
+    {
+      "epoch": 15.42272,
+      "grad_norm": 0.9948632121086121,
+      "learning_rate": 2.5916366546618647e-05,
+      "loss": 0.5445,
+      "step": 12049
+    },
+    {
+      "epoch": 15.424,
+      "grad_norm": 0.9686235189437866,
+      "learning_rate": 2.591436574629852e-05,
+      "loss": 0.5375,
+      "step": 12050
+    },
+    {
+      "epoch": 15.42528,
+      "grad_norm": 0.9626991748809814,
+      "learning_rate": 2.591236494597839e-05,
+      "loss": 0.5071,
+      "step": 12051
+    },
+    {
+      "epoch": 15.42656,
+      "grad_norm": 0.9826670289039612,
+      "learning_rate": 2.5910364145658266e-05,
+      "loss": 0.52,
+      "step": 12052
+    },
+    {
+      "epoch": 15.42784,
+      "grad_norm": 0.9592919945716858,
+      "learning_rate": 2.5908363345338138e-05,
+      "loss": 0.5125,
+      "step": 12053
+    },
+    {
+      "epoch": 15.42912,
+      "grad_norm": 0.9787027835845947,
+      "learning_rate": 2.590636254501801e-05,
+      "loss": 0.5241,
+      "step": 12054
+    },
+    {
+      "epoch": 15.4304,
+      "grad_norm": 0.9847190380096436,
+      "learning_rate": 2.590436174469788e-05,
+      "loss": 0.4971,
+      "step": 12055
+    },
+    {
+      "epoch": 15.43168,
+      "grad_norm": 0.9840123653411865,
+      "learning_rate": 2.5902360944377753e-05,
+      "loss": 0.55,
+      "step": 12056
+    },
+    {
+      "epoch": 15.43296,
+      "grad_norm": 0.9877153635025024,
+      "learning_rate": 2.590036014405762e-05,
+      "loss": 0.4923,
+      "step": 12057
+    },
+    {
+      "epoch": 15.43424,
+      "grad_norm": 0.9981076717376709,
+      "learning_rate": 2.5898359343737493e-05,
+      "loss": 0.5311,
+      "step": 12058
+    },
+    {
+      "epoch": 15.43552,
+      "grad_norm": 1.1250180006027222,
+      "learning_rate": 2.5896358543417372e-05,
+      "loss": 0.5702,
+      "step": 12059
+    },
+    {
+      "epoch": 15.4368,
+      "grad_norm": 1.031806230545044,
+      "learning_rate": 2.589435774309724e-05,
+      "loss": 0.568,
+      "step": 12060
+    },
+    {
+      "epoch": 15.43808,
+      "grad_norm": 0.9630105495452881,
+      "learning_rate": 2.5892356942777112e-05,
+      "loss": 0.4828,
+      "step": 12061
+    },
+    {
+      "epoch": 15.43936,
+      "grad_norm": 1.0042251348495483,
+      "learning_rate": 2.5890356142456984e-05,
+      "loss": 0.4779,
+      "step": 12062
+    },
+    {
+      "epoch": 15.44064,
+      "grad_norm": 0.9796602725982666,
+      "learning_rate": 2.5888355342136856e-05,
+      "loss": 0.4996,
+      "step": 12063
+    },
+    {
+      "epoch": 15.44192,
+      "grad_norm": 1.0524052381515503,
+      "learning_rate": 2.5886354541816728e-05,
+      "loss": 0.5759,
+      "step": 12064
+    },
+    {
+      "epoch": 15.4432,
+      "grad_norm": 1.0064433813095093,
+      "learning_rate": 2.5884353741496596e-05,
+      "loss": 0.4933,
+      "step": 12065
+    },
+    {
+      "epoch": 15.44448,
+      "grad_norm": 0.9477246403694153,
+      "learning_rate": 2.5882352941176475e-05,
+      "loss": 0.5054,
+      "step": 12066
+    },
+    {
+      "epoch": 15.44576,
+      "grad_norm": 0.9989529848098755,
+      "learning_rate": 2.5880352140856347e-05,
+      "loss": 0.5503,
+      "step": 12067
+    },
+    {
+      "epoch": 15.44704,
+      "grad_norm": 1.0528007745742798,
+      "learning_rate": 2.5878351340536215e-05,
+      "loss": 0.5832,
+      "step": 12068
+    },
+    {
+      "epoch": 15.44832,
+      "grad_norm": 0.9893178939819336,
+      "learning_rate": 2.5876350540216087e-05,
+      "loss": 0.4902,
+      "step": 12069
+    },
+    {
+      "epoch": 15.4496,
+      "grad_norm": 0.9771265387535095,
+      "learning_rate": 2.587434973989596e-05,
+      "loss": 0.531,
+      "step": 12070
+    },
+    {
+      "epoch": 15.45088,
+      "grad_norm": 1.0311421155929565,
+      "learning_rate": 2.587234893957583e-05,
+      "loss": 0.6119,
+      "step": 12071
+    },
+    {
+      "epoch": 15.45216,
+      "grad_norm": 0.9262276291847229,
+      "learning_rate": 2.5870348139255703e-05,
+      "loss": 0.5396,
+      "step": 12072
+    },
+    {
+      "epoch": 15.45344,
+      "grad_norm": 0.9662413001060486,
+      "learning_rate": 2.5868347338935578e-05,
+      "loss": 0.5006,
+      "step": 12073
+    },
+    {
+      "epoch": 15.45472,
+      "grad_norm": 1.0002639293670654,
+      "learning_rate": 2.586634653861545e-05,
+      "loss": 0.5263,
+      "step": 12074
+    },
+    {
+      "epoch": 15.456,
+      "grad_norm": 1.010391354560852,
+      "learning_rate": 2.5864345738295322e-05,
+      "loss": 0.5296,
+      "step": 12075
+    },
+    {
+      "epoch": 15.45728,
+      "grad_norm": 0.9372914433479309,
+      "learning_rate": 2.586234493797519e-05,
+      "loss": 0.4601,
+      "step": 12076
+    },
+    {
+      "epoch": 15.45856,
+      "grad_norm": 0.9640786647796631,
+      "learning_rate": 2.5860344137655062e-05,
+      "loss": 0.5049,
+      "step": 12077
+    },
+    {
+      "epoch": 15.45984,
+      "grad_norm": 0.9395653009414673,
+      "learning_rate": 2.5858343337334934e-05,
+      "loss": 0.4935,
+      "step": 12078
+    },
+    {
+      "epoch": 15.46112,
+      "grad_norm": 0.9529640078544617,
+      "learning_rate": 2.5856342537014806e-05,
+      "loss": 0.5194,
+      "step": 12079
+    },
+    {
+      "epoch": 15.4624,
+      "grad_norm": 0.9985076189041138,
+      "learning_rate": 2.585434173669468e-05,
+      "loss": 0.5531,
+      "step": 12080
+    },
+    {
+      "epoch": 15.46368,
+      "grad_norm": 0.924523115158081,
+      "learning_rate": 2.5852340936374553e-05,
+      "loss": 0.4675,
+      "step": 12081
+    },
+    {
+      "epoch": 15.46496,
+      "grad_norm": 1.0334564447402954,
+      "learning_rate": 2.5850340136054425e-05,
+      "loss": 0.5463,
+      "step": 12082
+    },
+    {
+      "epoch": 15.466239999999999,
+      "grad_norm": 0.9668273329734802,
+      "learning_rate": 2.5848339335734297e-05,
+      "loss": 0.5158,
+      "step": 12083
+    },
+    {
+      "epoch": 15.46752,
+      "grad_norm": 1.0096427202224731,
+      "learning_rate": 2.5846338535414165e-05,
+      "loss": 0.5181,
+      "step": 12084
+    },
+    {
+      "epoch": 15.4688,
+      "grad_norm": 0.9349405169487,
+      "learning_rate": 2.5844337735094037e-05,
+      "loss": 0.4826,
+      "step": 12085
+    },
+    {
+      "epoch": 15.47008,
+      "grad_norm": 0.9573792219161987,
+      "learning_rate": 2.584233693477391e-05,
+      "loss": 0.4755,
+      "step": 12086
+    },
+    {
+      "epoch": 15.47136,
+      "grad_norm": 1.014906644821167,
+      "learning_rate": 2.5840336134453784e-05,
+      "loss": 0.515,
+      "step": 12087
+    },
+    {
+      "epoch": 15.47264,
+      "grad_norm": 0.9619317054748535,
+      "learning_rate": 2.5838335334133656e-05,
+      "loss": 0.5234,
+      "step": 12088
+    },
+    {
+      "epoch": 15.47392,
+      "grad_norm": 0.9710426330566406,
+      "learning_rate": 2.5836334533813528e-05,
+      "loss": 0.4808,
+      "step": 12089
+    },
+    {
+      "epoch": 15.4752,
+      "grad_norm": 1.0196417570114136,
+      "learning_rate": 2.58343337334934e-05,
+      "loss": 0.5433,
+      "step": 12090
+    },
+    {
+      "epoch": 15.47648,
+      "grad_norm": 0.9249922037124634,
+      "learning_rate": 2.583233293317327e-05,
+      "loss": 0.4858,
+      "step": 12091
+    },
+    {
+      "epoch": 15.47776,
+      "grad_norm": 1.008486032485962,
+      "learning_rate": 2.583033213285314e-05,
+      "loss": 0.5453,
+      "step": 12092
+    },
+    {
+      "epoch": 15.47904,
+      "grad_norm": 0.9991050362586975,
+      "learning_rate": 2.5828331332533012e-05,
+      "loss": 0.5167,
+      "step": 12093
+    },
+    {
+      "epoch": 15.48032,
+      "grad_norm": 0.9885373115539551,
+      "learning_rate": 2.582633053221289e-05,
+      "loss": 0.5235,
+      "step": 12094
+    },
+    {
+      "epoch": 15.4816,
+      "grad_norm": 0.9774200320243835,
+      "learning_rate": 2.582432973189276e-05,
+      "loss": 0.4835,
+      "step": 12095
+    },
+    {
+      "epoch": 15.48288,
+      "grad_norm": 0.9708588123321533,
+      "learning_rate": 2.582232893157263e-05,
+      "loss": 0.5181,
+      "step": 12096
+    },
+    {
+      "epoch": 15.48416,
+      "grad_norm": 1.0187863111495972,
+      "learning_rate": 2.5820328131252503e-05,
+      "loss": 0.5702,
+      "step": 12097
+    },
+    {
+      "epoch": 15.48544,
+      "grad_norm": 0.9514103531837463,
+      "learning_rate": 2.5818327330932375e-05,
+      "loss": 0.4906,
+      "step": 12098
+    },
+    {
+      "epoch": 15.48672,
+      "grad_norm": 0.965501606464386,
+      "learning_rate": 2.5816326530612246e-05,
+      "loss": 0.5514,
+      "step": 12099
+    },
+    {
+      "epoch": 15.488,
+      "grad_norm": 1.0036544799804688,
+      "learning_rate": 2.5814325730292115e-05,
+      "loss": 0.4937,
+      "step": 12100
+    },
+    {
+      "epoch": 15.48928,
+      "grad_norm": 1.047786831855774,
+      "learning_rate": 2.5812324929971993e-05,
+      "loss": 0.535,
+      "step": 12101
+    },
+    {
+      "epoch": 15.49056,
+      "grad_norm": 1.0028244256973267,
+      "learning_rate": 2.5810324129651865e-05,
+      "loss": 0.5845,
+      "step": 12102
+    },
+    {
+      "epoch": 15.49184,
+      "grad_norm": 0.9661297798156738,
+      "learning_rate": 2.5808323329331734e-05,
+      "loss": 0.4615,
+      "step": 12103
+    },
+    {
+      "epoch": 15.49312,
+      "grad_norm": 0.9932764172554016,
+      "learning_rate": 2.5806322529011606e-05,
+      "loss": 0.4964,
+      "step": 12104
+    },
+    {
+      "epoch": 15.4944,
+      "grad_norm": 0.9625186920166016,
+      "learning_rate": 2.5804321728691478e-05,
+      "loss": 0.5339,
+      "step": 12105
+    },
+    {
+      "epoch": 15.49568,
+      "grad_norm": 0.9661415219306946,
+      "learning_rate": 2.580232092837135e-05,
+      "loss": 0.5432,
+      "step": 12106
+    },
+    {
+      "epoch": 15.49696,
+      "grad_norm": 1.0386898517608643,
+      "learning_rate": 2.580032012805122e-05,
+      "loss": 0.5427,
+      "step": 12107
+    },
+    {
+      "epoch": 15.49824,
+      "grad_norm": 0.9474533200263977,
+      "learning_rate": 2.5798319327731096e-05,
+      "loss": 0.5278,
+      "step": 12108
+    },
+    {
+      "epoch": 15.49952,
+      "grad_norm": 1.0017415285110474,
+      "learning_rate": 2.579631852741097e-05,
+      "loss": 0.5476,
+      "step": 12109
+    },
+    {
+      "epoch": 15.5008,
+      "grad_norm": 1.017276406288147,
+      "learning_rate": 2.579431772709084e-05,
+      "loss": 0.5015,
+      "step": 12110
+    },
+    {
+      "epoch": 15.50208,
+      "grad_norm": 1.0097792148590088,
+      "learning_rate": 2.579231692677071e-05,
+      "loss": 0.5492,
+      "step": 12111
+    },
+    {
+      "epoch": 15.50336,
+      "grad_norm": 0.9869034290313721,
+      "learning_rate": 2.579031612645058e-05,
+      "loss": 0.535,
+      "step": 12112
+    },
+    {
+      "epoch": 15.50464,
+      "grad_norm": 1.0194756984710693,
+      "learning_rate": 2.5788315326130452e-05,
+      "loss": 0.5019,
+      "step": 12113
+    },
+    {
+      "epoch": 15.50592,
+      "grad_norm": 1.0005152225494385,
+      "learning_rate": 2.5786314525810324e-05,
+      "loss": 0.525,
+      "step": 12114
+    },
+    {
+      "epoch": 15.5072,
+      "grad_norm": 0.9427708387374878,
+      "learning_rate": 2.57843137254902e-05,
+      "loss": 0.5217,
+      "step": 12115
+    },
+    {
+      "epoch": 15.50848,
+      "grad_norm": 0.9799354672431946,
+      "learning_rate": 2.578231292517007e-05,
+      "loss": 0.4981,
+      "step": 12116
+    },
+    {
+      "epoch": 15.50976,
+      "grad_norm": 0.9807402491569519,
+      "learning_rate": 2.5780312124849943e-05,
+      "loss": 0.5026,
+      "step": 12117
+    },
+    {
+      "epoch": 15.51104,
+      "grad_norm": 0.9583036303520203,
+      "learning_rate": 2.5778311324529815e-05,
+      "loss": 0.5072,
+      "step": 12118
+    },
+    {
+      "epoch": 15.51232,
+      "grad_norm": 1.0171658992767334,
+      "learning_rate": 2.5776310524209684e-05,
+      "loss": 0.5592,
+      "step": 12119
+    },
+    {
+      "epoch": 15.5136,
+      "grad_norm": 0.9614086747169495,
+      "learning_rate": 2.5774309723889555e-05,
+      "loss": 0.4961,
+      "step": 12120
+    },
+    {
+      "epoch": 15.51488,
+      "grad_norm": 0.9986427426338196,
+      "learning_rate": 2.5772308923569427e-05,
+      "loss": 0.4993,
+      "step": 12121
+    },
+    {
+      "epoch": 15.51616,
+      "grad_norm": 0.9052167534828186,
+      "learning_rate": 2.57703081232493e-05,
+      "loss": 0.4783,
+      "step": 12122
+    },
+    {
+      "epoch": 15.51744,
+      "grad_norm": 0.9978095889091492,
+      "learning_rate": 2.5768307322929174e-05,
+      "loss": 0.5733,
+      "step": 12123
+    },
+    {
+      "epoch": 15.51872,
+      "grad_norm": 0.9685370326042175,
+      "learning_rate": 2.5766306522609046e-05,
+      "loss": 0.51,
+      "step": 12124
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 1.008240818977356,
+      "learning_rate": 2.5764305722288918e-05,
+      "loss": 0.5129,
+      "step": 12125
+    },
+    {
+      "epoch": 15.52128,
+      "grad_norm": 0.9818560481071472,
+      "learning_rate": 2.576230492196879e-05,
+      "loss": 0.5501,
+      "step": 12126
+    },
+    {
+      "epoch": 15.52256,
+      "grad_norm": 0.9996564984321594,
+      "learning_rate": 2.576030412164866e-05,
+      "loss": 0.4986,
+      "step": 12127
+    },
+    {
+      "epoch": 15.52384,
+      "grad_norm": 1.0474154949188232,
+      "learning_rate": 2.575830332132853e-05,
+      "loss": 0.5526,
+      "step": 12128
+    },
+    {
+      "epoch": 15.52512,
+      "grad_norm": 1.0156049728393555,
+      "learning_rate": 2.5756302521008402e-05,
+      "loss": 0.5273,
+      "step": 12129
+    },
+    {
+      "epoch": 15.5264,
+      "grad_norm": 0.9961782693862915,
+      "learning_rate": 2.5754301720688277e-05,
+      "loss": 0.491,
+      "step": 12130
+    },
+    {
+      "epoch": 15.52768,
+      "grad_norm": 1.0487329959869385,
+      "learning_rate": 2.575230092036815e-05,
+      "loss": 0.5798,
+      "step": 12131
+    },
+    {
+      "epoch": 15.52896,
+      "grad_norm": 0.9667060971260071,
+      "learning_rate": 2.575030012004802e-05,
+      "loss": 0.5385,
+      "step": 12132
+    },
+    {
+      "epoch": 15.53024,
+      "grad_norm": 0.9694793820381165,
+      "learning_rate": 2.5748299319727893e-05,
+      "loss": 0.5248,
+      "step": 12133
+    },
+    {
+      "epoch": 15.53152,
+      "grad_norm": 0.9100515246391296,
+      "learning_rate": 2.5746298519407765e-05,
+      "loss": 0.4769,
+      "step": 12134
+    },
+    {
+      "epoch": 15.5328,
+      "grad_norm": 0.9730302095413208,
+      "learning_rate": 2.5744297719087633e-05,
+      "loss": 0.5182,
+      "step": 12135
+    },
+    {
+      "epoch": 15.53408,
+      "grad_norm": 0.9823408126831055,
+      "learning_rate": 2.5742296918767505e-05,
+      "loss": 0.5195,
+      "step": 12136
+    },
+    {
+      "epoch": 15.53536,
+      "grad_norm": 1.023830771446228,
+      "learning_rate": 2.5740296118447384e-05,
+      "loss": 0.5501,
+      "step": 12137
+    },
+    {
+      "epoch": 15.53664,
+      "grad_norm": 0.9773147106170654,
+      "learning_rate": 2.5738295318127252e-05,
+      "loss": 0.5068,
+      "step": 12138
+    },
+    {
+      "epoch": 15.53792,
+      "grad_norm": 0.9705550074577332,
+      "learning_rate": 2.5736294517807124e-05,
+      "loss": 0.5073,
+      "step": 12139
+    },
+    {
+      "epoch": 15.5392,
+      "grad_norm": 0.9751460552215576,
+      "learning_rate": 2.5734293717486996e-05,
+      "loss": 0.5036,
+      "step": 12140
+    },
+    {
+      "epoch": 15.54048,
+      "grad_norm": 0.9804943799972534,
+      "learning_rate": 2.5732292917166868e-05,
+      "loss": 0.4845,
+      "step": 12141
+    },
+    {
+      "epoch": 15.54176,
+      "grad_norm": 1.020609974861145,
+      "learning_rate": 2.573029211684674e-05,
+      "loss": 0.5516,
+      "step": 12142
+    },
+    {
+      "epoch": 15.54304,
+      "grad_norm": 0.9483124017715454,
+      "learning_rate": 2.5728291316526608e-05,
+      "loss": 0.4902,
+      "step": 12143
+    },
+    {
+      "epoch": 15.54432,
+      "grad_norm": 1.0113613605499268,
+      "learning_rate": 2.5726290516206487e-05,
+      "loss": 0.5527,
+      "step": 12144
+    },
+    {
+      "epoch": 15.5456,
+      "grad_norm": 1.037793755531311,
+      "learning_rate": 2.572428971588636e-05,
+      "loss": 0.543,
+      "step": 12145
+    },
+    {
+      "epoch": 15.54688,
+      "grad_norm": 1.0010935068130493,
+      "learning_rate": 2.5722288915566227e-05,
+      "loss": 0.5157,
+      "step": 12146
+    },
+    {
+      "epoch": 15.54816,
+      "grad_norm": 0.9994384050369263,
+      "learning_rate": 2.57202881152461e-05,
+      "loss": 0.5223,
+      "step": 12147
+    },
+    {
+      "epoch": 15.54944,
+      "grad_norm": 0.9662322998046875,
+      "learning_rate": 2.571828731492597e-05,
+      "loss": 0.5161,
+      "step": 12148
+    },
+    {
+      "epoch": 15.55072,
+      "grad_norm": 1.0219558477401733,
+      "learning_rate": 2.5716286514605843e-05,
+      "loss": 0.534,
+      "step": 12149
+    },
+    {
+      "epoch": 15.552,
+      "grad_norm": 0.9742558002471924,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.5151,
+      "step": 12150
+    },
+    {
+      "epoch": 15.55328,
+      "grad_norm": 1.0059118270874023,
+      "learning_rate": 2.571228491396559e-05,
+      "loss": 0.5637,
+      "step": 12151
+    },
+    {
+      "epoch": 15.55456,
+      "grad_norm": 0.9485430121421814,
+      "learning_rate": 2.571028411364546e-05,
+      "loss": 0.4919,
+      "step": 12152
+    },
+    {
+      "epoch": 15.55584,
+      "grad_norm": 0.9600156545639038,
+      "learning_rate": 2.5708283313325333e-05,
+      "loss": 0.529,
+      "step": 12153
+    },
+    {
+      "epoch": 15.55712,
+      "grad_norm": 0.9811621308326721,
+      "learning_rate": 2.5706282513005202e-05,
+      "loss": 0.5033,
+      "step": 12154
+    },
+    {
+      "epoch": 15.5584,
+      "grad_norm": 1.0301507711410522,
+      "learning_rate": 2.5704281712685074e-05,
+      "loss": 0.501,
+      "step": 12155
+    },
+    {
+      "epoch": 15.55968,
+      "grad_norm": 0.9439396858215332,
+      "learning_rate": 2.5702280912364946e-05,
+      "loss": 0.5265,
+      "step": 12156
+    },
+    {
+      "epoch": 15.56096,
+      "grad_norm": 1.0617479085922241,
+      "learning_rate": 2.5700280112044817e-05,
+      "loss": 0.5651,
+      "step": 12157
+    },
+    {
+      "epoch": 15.56224,
+      "grad_norm": 0.9912658333778381,
+      "learning_rate": 2.5698279311724693e-05,
+      "loss": 0.5087,
+      "step": 12158
+    },
+    {
+      "epoch": 15.56352,
+      "grad_norm": 0.9808444380760193,
+      "learning_rate": 2.5696278511404565e-05,
+      "loss": 0.5334,
+      "step": 12159
+    },
+    {
+      "epoch": 15.5648,
+      "grad_norm": 0.9171379804611206,
+      "learning_rate": 2.5694277711084436e-05,
+      "loss": 0.5009,
+      "step": 12160
+    },
+    {
+      "epoch": 15.56608,
+      "grad_norm": 0.9609277248382568,
+      "learning_rate": 2.5692276910764308e-05,
+      "loss": 0.5078,
+      "step": 12161
+    },
+    {
+      "epoch": 15.56736,
+      "grad_norm": 0.9825374484062195,
+      "learning_rate": 2.5690276110444177e-05,
+      "loss": 0.5374,
+      "step": 12162
+    },
+    {
+      "epoch": 15.56864,
+      "grad_norm": 0.9431901574134827,
+      "learning_rate": 2.568827531012405e-05,
+      "loss": 0.4915,
+      "step": 12163
+    },
+    {
+      "epoch": 15.56992,
+      "grad_norm": 1.0128616094589233,
+      "learning_rate": 2.568627450980392e-05,
+      "loss": 0.5483,
+      "step": 12164
+    },
+    {
+      "epoch": 15.5712,
+      "grad_norm": 1.0459868907928467,
+      "learning_rate": 2.5684273709483796e-05,
+      "loss": 0.5579,
+      "step": 12165
+    },
+    {
+      "epoch": 15.57248,
+      "grad_norm": 1.015820026397705,
+      "learning_rate": 2.5682272909163668e-05,
+      "loss": 0.5315,
+      "step": 12166
+    },
+    {
+      "epoch": 15.57376,
+      "grad_norm": 0.9853371977806091,
+      "learning_rate": 2.568027210884354e-05,
+      "loss": 0.513,
+      "step": 12167
+    },
+    {
+      "epoch": 15.57504,
+      "grad_norm": 0.9857305288314819,
+      "learning_rate": 2.567827130852341e-05,
+      "loss": 0.5006,
+      "step": 12168
+    },
+    {
+      "epoch": 15.57632,
+      "grad_norm": 1.0142581462860107,
+      "learning_rate": 2.5676270508203283e-05,
+      "loss": 0.5309,
+      "step": 12169
+    },
+    {
+      "epoch": 15.5776,
+      "grad_norm": 0.9901478886604309,
+      "learning_rate": 2.567426970788315e-05,
+      "loss": 0.5302,
+      "step": 12170
+    },
+    {
+      "epoch": 15.57888,
+      "grad_norm": 0.9273139238357544,
+      "learning_rate": 2.5672268907563023e-05,
+      "loss": 0.4937,
+      "step": 12171
+    },
+    {
+      "epoch": 15.58016,
+      "grad_norm": 0.984489917755127,
+      "learning_rate": 2.5670268107242902e-05,
+      "loss": 0.547,
+      "step": 12172
+    },
+    {
+      "epoch": 15.58144,
+      "grad_norm": 1.0063872337341309,
+      "learning_rate": 2.566826730692277e-05,
+      "loss": 0.5502,
+      "step": 12173
+    },
+    {
+      "epoch": 15.58272,
+      "grad_norm": 1.0095993280410767,
+      "learning_rate": 2.5666266506602642e-05,
+      "loss": 0.5082,
+      "step": 12174
+    },
+    {
+      "epoch": 15.584,
+      "grad_norm": 0.9566926956176758,
+      "learning_rate": 2.5664265706282514e-05,
+      "loss": 0.5288,
+      "step": 12175
+    },
+    {
+      "epoch": 15.585280000000001,
+      "grad_norm": 0.9695842862129211,
+      "learning_rate": 2.5662264905962386e-05,
+      "loss": 0.5461,
+      "step": 12176
+    },
+    {
+      "epoch": 15.58656,
+      "grad_norm": 0.9567855000495911,
+      "learning_rate": 2.5660264105642258e-05,
+      "loss": 0.5013,
+      "step": 12177
+    },
+    {
+      "epoch": 15.58784,
+      "grad_norm": 0.983958899974823,
+      "learning_rate": 2.5658263305322126e-05,
+      "loss": 0.5322,
+      "step": 12178
+    },
+    {
+      "epoch": 15.58912,
+      "grad_norm": 1.0091723203659058,
+      "learning_rate": 2.5656262505002005e-05,
+      "loss": 0.5172,
+      "step": 12179
+    },
+    {
+      "epoch": 15.5904,
+      "grad_norm": 1.0233546495437622,
+      "learning_rate": 2.5654261704681877e-05,
+      "loss": 0.5247,
+      "step": 12180
+    },
+    {
+      "epoch": 15.59168,
+      "grad_norm": 0.9641100764274597,
+      "learning_rate": 2.5652260904361745e-05,
+      "loss": 0.5123,
+      "step": 12181
+    },
+    {
+      "epoch": 15.59296,
+      "grad_norm": 0.9538204669952393,
+      "learning_rate": 2.5650260104041617e-05,
+      "loss": 0.5048,
+      "step": 12182
+    },
+    {
+      "epoch": 15.59424,
+      "grad_norm": 1.0550053119659424,
+      "learning_rate": 2.564825930372149e-05,
+      "loss": 0.5824,
+      "step": 12183
+    },
+    {
+      "epoch": 15.59552,
+      "grad_norm": 0.9931796193122864,
+      "learning_rate": 2.564625850340136e-05,
+      "loss": 0.526,
+      "step": 12184
+    },
+    {
+      "epoch": 15.5968,
+      "grad_norm": 1.0209304094314575,
+      "learning_rate": 2.5644257703081233e-05,
+      "loss": 0.5275,
+      "step": 12185
+    },
+    {
+      "epoch": 15.59808,
+      "grad_norm": 0.9978035688400269,
+      "learning_rate": 2.5642256902761108e-05,
+      "loss": 0.5207,
+      "step": 12186
+    },
+    {
+      "epoch": 15.59936,
+      "grad_norm": 1.0060955286026,
+      "learning_rate": 2.564025610244098e-05,
+      "loss": 0.5259,
+      "step": 12187
+    },
+    {
+      "epoch": 15.60064,
+      "grad_norm": 1.0725250244140625,
+      "learning_rate": 2.5638255302120852e-05,
+      "loss": 0.5654,
+      "step": 12188
+    },
+    {
+      "epoch": 15.60192,
+      "grad_norm": 1.0049766302108765,
+      "learning_rate": 2.563625450180072e-05,
+      "loss": 0.5537,
+      "step": 12189
+    },
+    {
+      "epoch": 15.6032,
+      "grad_norm": 0.9810140132904053,
+      "learning_rate": 2.5634253701480592e-05,
+      "loss": 0.4873,
+      "step": 12190
+    },
+    {
+      "epoch": 15.60448,
+      "grad_norm": 0.9922715425491333,
+      "learning_rate": 2.5632252901160464e-05,
+      "loss": 0.5372,
+      "step": 12191
+    },
+    {
+      "epoch": 15.60576,
+      "grad_norm": 0.9921250939369202,
+      "learning_rate": 2.5630252100840336e-05,
+      "loss": 0.523,
+      "step": 12192
+    },
+    {
+      "epoch": 15.60704,
+      "grad_norm": 1.0218485593795776,
+      "learning_rate": 2.562825130052021e-05,
+      "loss": 0.5281,
+      "step": 12193
+    },
+    {
+      "epoch": 15.608319999999999,
+      "grad_norm": 1.0210494995117188,
+      "learning_rate": 2.5626250500200083e-05,
+      "loss": 0.5154,
+      "step": 12194
+    },
+    {
+      "epoch": 15.6096,
+      "grad_norm": 1.0493415594100952,
+      "learning_rate": 2.5624249699879955e-05,
+      "loss": 0.5801,
+      "step": 12195
+    },
+    {
+      "epoch": 15.61088,
+      "grad_norm": 0.95387864112854,
+      "learning_rate": 2.5622248899559827e-05,
+      "loss": 0.5033,
+      "step": 12196
+    },
+    {
+      "epoch": 15.61216,
+      "grad_norm": 0.9641955494880676,
+      "learning_rate": 2.5620248099239695e-05,
+      "loss": 0.521,
+      "step": 12197
+    },
+    {
+      "epoch": 15.61344,
+      "grad_norm": 0.979554295539856,
+      "learning_rate": 2.5618247298919567e-05,
+      "loss": 0.5124,
+      "step": 12198
+    },
+    {
+      "epoch": 15.61472,
+      "grad_norm": 1.063732385635376,
+      "learning_rate": 2.561624649859944e-05,
+      "loss": 0.5712,
+      "step": 12199
+    },
+    {
+      "epoch": 15.616,
+      "grad_norm": 1.0626933574676514,
+      "learning_rate": 2.5614245698279314e-05,
+      "loss": 0.5761,
+      "step": 12200
+    },
+    {
+      "epoch": 15.617280000000001,
+      "grad_norm": 1.0086123943328857,
+      "learning_rate": 2.5612244897959186e-05,
+      "loss": 0.5195,
+      "step": 12201
+    },
+    {
+      "epoch": 15.61856,
+      "grad_norm": 0.990342915058136,
+      "learning_rate": 2.5610244097639058e-05,
+      "loss": 0.5288,
+      "step": 12202
+    },
+    {
+      "epoch": 15.61984,
+      "grad_norm": 0.9492666125297546,
+      "learning_rate": 2.560824329731893e-05,
+      "loss": 0.4753,
+      "step": 12203
+    },
+    {
+      "epoch": 15.62112,
+      "grad_norm": 1.011853575706482,
+      "learning_rate": 2.56062424969988e-05,
+      "loss": 0.5095,
+      "step": 12204
+    },
+    {
+      "epoch": 15.6224,
+      "grad_norm": 1.003688931465149,
+      "learning_rate": 2.560424169667867e-05,
+      "loss": 0.4808,
+      "step": 12205
+    },
+    {
+      "epoch": 15.62368,
+      "grad_norm": 0.9985147714614868,
+      "learning_rate": 2.5602240896358542e-05,
+      "loss": 0.5343,
+      "step": 12206
+    },
+    {
+      "epoch": 15.62496,
+      "grad_norm": 0.9690064787864685,
+      "learning_rate": 2.560024009603842e-05,
+      "loss": 0.4995,
+      "step": 12207
+    },
+    {
+      "epoch": 15.62624,
+      "grad_norm": 0.9825969338417053,
+      "learning_rate": 2.559823929571829e-05,
+      "loss": 0.5231,
+      "step": 12208
+    },
+    {
+      "epoch": 15.62752,
+      "grad_norm": 0.995742917060852,
+      "learning_rate": 2.559623849539816e-05,
+      "loss": 0.5243,
+      "step": 12209
+    },
+    {
+      "epoch": 15.6288,
+      "grad_norm": 1.037103295326233,
+      "learning_rate": 2.5594237695078033e-05,
+      "loss": 0.5265,
+      "step": 12210
+    },
+    {
+      "epoch": 15.63008,
+      "grad_norm": 1.0141079425811768,
+      "learning_rate": 2.5592236894757904e-05,
+      "loss": 0.5545,
+      "step": 12211
+    },
+    {
+      "epoch": 15.63136,
+      "grad_norm": 0.9893406629562378,
+      "learning_rate": 2.5590236094437776e-05,
+      "loss": 0.5033,
+      "step": 12212
+    },
+    {
+      "epoch": 15.63264,
+      "grad_norm": 1.0220900774002075,
+      "learning_rate": 2.5588235294117645e-05,
+      "loss": 0.5212,
+      "step": 12213
+    },
+    {
+      "epoch": 15.63392,
+      "grad_norm": 0.9686633348464966,
+      "learning_rate": 2.5586234493797523e-05,
+      "loss": 0.5521,
+      "step": 12214
+    },
+    {
+      "epoch": 15.6352,
+      "grad_norm": 1.001504898071289,
+      "learning_rate": 2.5584233693477395e-05,
+      "loss": 0.5756,
+      "step": 12215
+    },
+    {
+      "epoch": 15.63648,
+      "grad_norm": 1.0070838928222656,
+      "learning_rate": 2.5582232893157264e-05,
+      "loss": 0.5075,
+      "step": 12216
+    },
+    {
+      "epoch": 15.63776,
+      "grad_norm": 0.9980425834655762,
+      "learning_rate": 2.5580232092837136e-05,
+      "loss": 0.5527,
+      "step": 12217
+    },
+    {
+      "epoch": 15.63904,
+      "grad_norm": 0.9670035243034363,
+      "learning_rate": 2.5578231292517007e-05,
+      "loss": 0.49,
+      "step": 12218
+    },
+    {
+      "epoch": 15.64032,
+      "grad_norm": 0.9716927409172058,
+      "learning_rate": 2.557623049219688e-05,
+      "loss": 0.4885,
+      "step": 12219
+    },
+    {
+      "epoch": 15.6416,
+      "grad_norm": 0.9749883413314819,
+      "learning_rate": 2.557422969187675e-05,
+      "loss": 0.5409,
+      "step": 12220
+    },
+    {
+      "epoch": 15.64288,
+      "grad_norm": 0.9985603094100952,
+      "learning_rate": 2.5572228891556626e-05,
+      "loss": 0.5294,
+      "step": 12221
+    },
+    {
+      "epoch": 15.64416,
+      "grad_norm": 1.0051274299621582,
+      "learning_rate": 2.5570228091236498e-05,
+      "loss": 0.5685,
+      "step": 12222
+    },
+    {
+      "epoch": 15.64544,
+      "grad_norm": 1.0328760147094727,
+      "learning_rate": 2.556822729091637e-05,
+      "loss": 0.6053,
+      "step": 12223
+    },
+    {
+      "epoch": 15.64672,
+      "grad_norm": 1.0340243577957153,
+      "learning_rate": 2.556622649059624e-05,
+      "loss": 0.5212,
+      "step": 12224
+    },
+    {
+      "epoch": 15.648,
+      "grad_norm": 1.0244271755218506,
+      "learning_rate": 2.556422569027611e-05,
+      "loss": 0.5555,
+      "step": 12225
+    },
+    {
+      "epoch": 15.64928,
+      "grad_norm": 0.9867590069770813,
+      "learning_rate": 2.5562224889955982e-05,
+      "loss": 0.5331,
+      "step": 12226
+    },
+    {
+      "epoch": 15.65056,
+      "grad_norm": 1.015299916267395,
+      "learning_rate": 2.5560224089635854e-05,
+      "loss": 0.5381,
+      "step": 12227
+    },
+    {
+      "epoch": 15.65184,
+      "grad_norm": 0.9973687529563904,
+      "learning_rate": 2.555822328931573e-05,
+      "loss": 0.5776,
+      "step": 12228
+    },
+    {
+      "epoch": 15.65312,
+      "grad_norm": 0.9658114314079285,
+      "learning_rate": 2.55562224889956e-05,
+      "loss": 0.4938,
+      "step": 12229
+    },
+    {
+      "epoch": 15.6544,
+      "grad_norm": 0.9462615847587585,
+      "learning_rate": 2.5554221688675473e-05,
+      "loss": 0.4878,
+      "step": 12230
+    },
+    {
+      "epoch": 15.65568,
+      "grad_norm": 0.9629518389701843,
+      "learning_rate": 2.5552220888355345e-05,
+      "loss": 0.544,
+      "step": 12231
+    },
+    {
+      "epoch": 15.65696,
+      "grad_norm": 1.0248364210128784,
+      "learning_rate": 2.5550220088035213e-05,
+      "loss": 0.5085,
+      "step": 12232
+    },
+    {
+      "epoch": 15.65824,
+      "grad_norm": 0.9903928637504578,
+      "learning_rate": 2.5548219287715085e-05,
+      "loss": 0.5624,
+      "step": 12233
+    },
+    {
+      "epoch": 15.65952,
+      "grad_norm": 0.9919432401657104,
+      "learning_rate": 2.5546218487394957e-05,
+      "loss": 0.5008,
+      "step": 12234
+    },
+    {
+      "epoch": 15.6608,
+      "grad_norm": 1.0050952434539795,
+      "learning_rate": 2.554421768707483e-05,
+      "loss": 0.5812,
+      "step": 12235
+    },
+    {
+      "epoch": 15.66208,
+      "grad_norm": 0.9851815104484558,
+      "learning_rate": 2.5542216886754704e-05,
+      "loss": 0.5016,
+      "step": 12236
+    },
+    {
+      "epoch": 15.66336,
+      "grad_norm": 0.9741938710212708,
+      "learning_rate": 2.5540216086434576e-05,
+      "loss": 0.5038,
+      "step": 12237
+    },
+    {
+      "epoch": 15.66464,
+      "grad_norm": 1.0300729274749756,
+      "learning_rate": 2.5538215286114448e-05,
+      "loss": 0.5195,
+      "step": 12238
+    },
+    {
+      "epoch": 15.66592,
+      "grad_norm": 0.9596061110496521,
+      "learning_rate": 2.553621448579432e-05,
+      "loss": 0.5052,
+      "step": 12239
+    },
+    {
+      "epoch": 15.6672,
+      "grad_norm": 0.9867904186248779,
+      "learning_rate": 2.553421368547419e-05,
+      "loss": 0.5353,
+      "step": 12240
+    },
+    {
+      "epoch": 15.66848,
+      "grad_norm": 0.9471587538719177,
+      "learning_rate": 2.553221288515406e-05,
+      "loss": 0.4976,
+      "step": 12241
+    },
+    {
+      "epoch": 15.66976,
+      "grad_norm": 0.9685238599777222,
+      "learning_rate": 2.5530212084833932e-05,
+      "loss": 0.5168,
+      "step": 12242
+    },
+    {
+      "epoch": 15.67104,
+      "grad_norm": 0.9731021523475647,
+      "learning_rate": 2.5528211284513807e-05,
+      "loss": 0.5298,
+      "step": 12243
+    },
+    {
+      "epoch": 15.67232,
+      "grad_norm": 0.9390130043029785,
+      "learning_rate": 2.552621048419368e-05,
+      "loss": 0.517,
+      "step": 12244
+    },
+    {
+      "epoch": 15.6736,
+      "grad_norm": 1.0047056674957275,
+      "learning_rate": 2.552420968387355e-05,
+      "loss": 0.5198,
+      "step": 12245
+    },
+    {
+      "epoch": 15.67488,
+      "grad_norm": 1.0100324153900146,
+      "learning_rate": 2.5522208883553423e-05,
+      "loss": 0.5573,
+      "step": 12246
+    },
+    {
+      "epoch": 15.67616,
+      "grad_norm": 1.0420469045639038,
+      "learning_rate": 2.5520208083233295e-05,
+      "loss": 0.5394,
+      "step": 12247
+    },
+    {
+      "epoch": 15.67744,
+      "grad_norm": 1.0220342874526978,
+      "learning_rate": 2.5518207282913163e-05,
+      "loss": 0.5864,
+      "step": 12248
+    },
+    {
+      "epoch": 15.67872,
+      "grad_norm": 1.0437568426132202,
+      "learning_rate": 2.5516206482593035e-05,
+      "loss": 0.5309,
+      "step": 12249
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 1.0043889284133911,
+      "learning_rate": 2.5514205682272914e-05,
+      "loss": 0.5631,
+      "step": 12250
+    },
+    {
+      "epoch": 15.68128,
+      "grad_norm": 1.024249792098999,
+      "learning_rate": 2.5512204881952782e-05,
+      "loss": 0.5329,
+      "step": 12251
+    },
+    {
+      "epoch": 15.68256,
+      "grad_norm": 0.9508040547370911,
+      "learning_rate": 2.5510204081632654e-05,
+      "loss": 0.5257,
+      "step": 12252
+    },
+    {
+      "epoch": 15.68384,
+      "grad_norm": 1.0146198272705078,
+      "learning_rate": 2.5508203281312526e-05,
+      "loss": 0.5494,
+      "step": 12253
+    },
+    {
+      "epoch": 15.68512,
+      "grad_norm": 0.9662773609161377,
+      "learning_rate": 2.5506202480992398e-05,
+      "loss": 0.4899,
+      "step": 12254
+    },
+    {
+      "epoch": 15.6864,
+      "grad_norm": 0.9972757697105408,
+      "learning_rate": 2.550420168067227e-05,
+      "loss": 0.5409,
+      "step": 12255
+    },
+    {
+      "epoch": 15.68768,
+      "grad_norm": 0.9856722950935364,
+      "learning_rate": 2.5502200880352138e-05,
+      "loss": 0.5048,
+      "step": 12256
+    },
+    {
+      "epoch": 15.68896,
+      "grad_norm": 1.0038642883300781,
+      "learning_rate": 2.5500200080032017e-05,
+      "loss": 0.5522,
+      "step": 12257
+    },
+    {
+      "epoch": 15.69024,
+      "grad_norm": 1.0079247951507568,
+      "learning_rate": 2.549819927971189e-05,
+      "loss": 0.5588,
+      "step": 12258
+    },
+    {
+      "epoch": 15.69152,
+      "grad_norm": 1.0044342279434204,
+      "learning_rate": 2.5496198479391757e-05,
+      "loss": 0.5339,
+      "step": 12259
+    },
+    {
+      "epoch": 15.6928,
+      "grad_norm": 0.9543859362602234,
+      "learning_rate": 2.549419767907163e-05,
+      "loss": 0.5141,
+      "step": 12260
+    },
+    {
+      "epoch": 15.69408,
+      "grad_norm": 1.0886956453323364,
+      "learning_rate": 2.54921968787515e-05,
+      "loss": 0.5861,
+      "step": 12261
+    },
+    {
+      "epoch": 15.69536,
+      "grad_norm": 0.9997818470001221,
+      "learning_rate": 2.5490196078431373e-05,
+      "loss": 0.4868,
+      "step": 12262
+    },
+    {
+      "epoch": 15.69664,
+      "grad_norm": 0.9534517526626587,
+      "learning_rate": 2.5488195278111244e-05,
+      "loss": 0.5402,
+      "step": 12263
+    },
+    {
+      "epoch": 15.69792,
+      "grad_norm": 0.9827191233634949,
+      "learning_rate": 2.548619447779112e-05,
+      "loss": 0.5272,
+      "step": 12264
+    },
+    {
+      "epoch": 15.6992,
+      "grad_norm": 0.9801238775253296,
+      "learning_rate": 2.548419367747099e-05,
+      "loss": 0.5243,
+      "step": 12265
+    },
+    {
+      "epoch": 15.70048,
+      "grad_norm": 0.9996204376220703,
+      "learning_rate": 2.5482192877150863e-05,
+      "loss": 0.535,
+      "step": 12266
+    },
+    {
+      "epoch": 15.70176,
+      "grad_norm": 1.0119619369506836,
+      "learning_rate": 2.5480192076830732e-05,
+      "loss": 0.5261,
+      "step": 12267
+    },
+    {
+      "epoch": 15.70304,
+      "grad_norm": 1.0047118663787842,
+      "learning_rate": 2.5478191276510604e-05,
+      "loss": 0.4841,
+      "step": 12268
+    },
+    {
+      "epoch": 15.70432,
+      "grad_norm": 1.0151153802871704,
+      "learning_rate": 2.5476190476190476e-05,
+      "loss": 0.5364,
+      "step": 12269
+    },
+    {
+      "epoch": 15.7056,
+      "grad_norm": 1.0285605192184448,
+      "learning_rate": 2.5474189675870347e-05,
+      "loss": 0.5459,
+      "step": 12270
+    },
+    {
+      "epoch": 15.70688,
+      "grad_norm": 0.9856500625610352,
+      "learning_rate": 2.5472188875550223e-05,
+      "loss": 0.5358,
+      "step": 12271
+    },
+    {
+      "epoch": 15.70816,
+      "grad_norm": 1.0488793849945068,
+      "learning_rate": 2.5470188075230095e-05,
+      "loss": 0.5405,
+      "step": 12272
+    },
+    {
+      "epoch": 15.70944,
+      "grad_norm": 1.0134892463684082,
+      "learning_rate": 2.5468187274909966e-05,
+      "loss": 0.5732,
+      "step": 12273
+    },
+    {
+      "epoch": 15.71072,
+      "grad_norm": 0.9521108269691467,
+      "learning_rate": 2.5466186474589838e-05,
+      "loss": 0.4745,
+      "step": 12274
+    },
+    {
+      "epoch": 15.712,
+      "grad_norm": 0.9995734095573425,
+      "learning_rate": 2.5464185674269707e-05,
+      "loss": 0.5472,
+      "step": 12275
+    },
+    {
+      "epoch": 15.71328,
+      "grad_norm": 1.0318055152893066,
+      "learning_rate": 2.546218487394958e-05,
+      "loss": 0.5725,
+      "step": 12276
+    },
+    {
+      "epoch": 15.71456,
+      "grad_norm": 1.0022940635681152,
+      "learning_rate": 2.546018407362945e-05,
+      "loss": 0.5544,
+      "step": 12277
+    },
+    {
+      "epoch": 15.71584,
+      "grad_norm": 0.9959270358085632,
+      "learning_rate": 2.5458183273309326e-05,
+      "loss": 0.52,
+      "step": 12278
+    },
+    {
+      "epoch": 15.71712,
+      "grad_norm": 1.0101356506347656,
+      "learning_rate": 2.5456182472989198e-05,
+      "loss": 0.5537,
+      "step": 12279
+    },
+    {
+      "epoch": 15.7184,
+      "grad_norm": 0.9922164678573608,
+      "learning_rate": 2.545418167266907e-05,
+      "loss": 0.5343,
+      "step": 12280
+    },
+    {
+      "epoch": 15.71968,
+      "grad_norm": 1.0045051574707031,
+      "learning_rate": 2.545218087234894e-05,
+      "loss": 0.5776,
+      "step": 12281
+    },
+    {
+      "epoch": 15.72096,
+      "grad_norm": 0.9495313763618469,
+      "learning_rate": 2.5450180072028813e-05,
+      "loss": 0.509,
+      "step": 12282
+    },
+    {
+      "epoch": 15.72224,
+      "grad_norm": 0.9955906867980957,
+      "learning_rate": 2.544817927170868e-05,
+      "loss": 0.5389,
+      "step": 12283
+    },
+    {
+      "epoch": 15.72352,
+      "grad_norm": 1.0342867374420166,
+      "learning_rate": 2.5446178471388553e-05,
+      "loss": 0.5492,
+      "step": 12284
+    },
+    {
+      "epoch": 15.7248,
+      "grad_norm": 0.966540515422821,
+      "learning_rate": 2.5444177671068432e-05,
+      "loss": 0.5449,
+      "step": 12285
+    },
+    {
+      "epoch": 15.72608,
+      "grad_norm": 0.9552671313285828,
+      "learning_rate": 2.54421768707483e-05,
+      "loss": 0.5026,
+      "step": 12286
+    },
+    {
+      "epoch": 15.727360000000001,
+      "grad_norm": 0.9747936129570007,
+      "learning_rate": 2.5440176070428172e-05,
+      "loss": 0.524,
+      "step": 12287
+    },
+    {
+      "epoch": 15.72864,
+      "grad_norm": 1.0628736019134521,
+      "learning_rate": 2.5438175270108044e-05,
+      "loss": 0.5114,
+      "step": 12288
+    },
+    {
+      "epoch": 15.72992,
+      "grad_norm": 1.016642689704895,
+      "learning_rate": 2.5436174469787916e-05,
+      "loss": 0.5479,
+      "step": 12289
+    },
+    {
+      "epoch": 15.7312,
+      "grad_norm": 0.95323246717453,
+      "learning_rate": 2.5434173669467788e-05,
+      "loss": 0.5381,
+      "step": 12290
+    },
+    {
+      "epoch": 15.73248,
+      "grad_norm": 1.0170929431915283,
+      "learning_rate": 2.5432172869147656e-05,
+      "loss": 0.5459,
+      "step": 12291
+    },
+    {
+      "epoch": 15.73376,
+      "grad_norm": 0.990404486656189,
+      "learning_rate": 2.5430172068827535e-05,
+      "loss": 0.5469,
+      "step": 12292
+    },
+    {
+      "epoch": 15.73504,
+      "grad_norm": 1.013763189315796,
+      "learning_rate": 2.5428171268507407e-05,
+      "loss": 0.498,
+      "step": 12293
+    },
+    {
+      "epoch": 15.73632,
+      "grad_norm": 0.9921509623527527,
+      "learning_rate": 2.5426170468187275e-05,
+      "loss": 0.5439,
+      "step": 12294
+    },
+    {
+      "epoch": 15.7376,
+      "grad_norm": 0.9957842230796814,
+      "learning_rate": 2.5424169667867147e-05,
+      "loss": 0.5197,
+      "step": 12295
+    },
+    {
+      "epoch": 15.73888,
+      "grad_norm": 0.9728386998176575,
+      "learning_rate": 2.542216886754702e-05,
+      "loss": 0.4966,
+      "step": 12296
+    },
+    {
+      "epoch": 15.74016,
+      "grad_norm": 1.0001862049102783,
+      "learning_rate": 2.542016806722689e-05,
+      "loss": 0.5375,
+      "step": 12297
+    },
+    {
+      "epoch": 15.74144,
+      "grad_norm": 0.9885843992233276,
+      "learning_rate": 2.5418167266906763e-05,
+      "loss": 0.545,
+      "step": 12298
+    },
+    {
+      "epoch": 15.74272,
+      "grad_norm": 1.0801782608032227,
+      "learning_rate": 2.5416166466586638e-05,
+      "loss": 0.548,
+      "step": 12299
+    },
+    {
+      "epoch": 15.744,
+      "grad_norm": 1.014716625213623,
+      "learning_rate": 2.541416566626651e-05,
+      "loss": 0.5159,
+      "step": 12300
+    },
+    {
+      "epoch": 15.74528,
+      "grad_norm": 1.070065975189209,
+      "learning_rate": 2.5412164865946382e-05,
+      "loss": 0.5531,
+      "step": 12301
+    },
+    {
+      "epoch": 15.74656,
+      "grad_norm": 0.9282522201538086,
+      "learning_rate": 2.541016406562625e-05,
+      "loss": 0.5189,
+      "step": 12302
+    },
+    {
+      "epoch": 15.74784,
+      "grad_norm": 0.9973820447921753,
+      "learning_rate": 2.5408163265306122e-05,
+      "loss": 0.5093,
+      "step": 12303
+    },
+    {
+      "epoch": 15.74912,
+      "grad_norm": 0.9329702258110046,
+      "learning_rate": 2.5406162464985994e-05,
+      "loss": 0.4989,
+      "step": 12304
+    },
+    {
+      "epoch": 15.750399999999999,
+      "grad_norm": 0.9909161925315857,
+      "learning_rate": 2.5404161664665866e-05,
+      "loss": 0.5578,
+      "step": 12305
+    },
+    {
+      "epoch": 15.75168,
+      "grad_norm": 1.0105434656143188,
+      "learning_rate": 2.540216086434574e-05,
+      "loss": 0.5251,
+      "step": 12306
+    },
+    {
+      "epoch": 15.75296,
+      "grad_norm": 1.0197798013687134,
+      "learning_rate": 2.5400160064025613e-05,
+      "loss": 0.5527,
+      "step": 12307
+    },
+    {
+      "epoch": 15.75424,
+      "grad_norm": 1.027341604232788,
+      "learning_rate": 2.5398159263705485e-05,
+      "loss": 0.5716,
+      "step": 12308
+    },
+    {
+      "epoch": 15.75552,
+      "grad_norm": 0.9711216688156128,
+      "learning_rate": 2.5396158463385357e-05,
+      "loss": 0.5455,
+      "step": 12309
+    },
+    {
+      "epoch": 15.7568,
+      "grad_norm": 0.9567652940750122,
+      "learning_rate": 2.5394157663065225e-05,
+      "loss": 0.475,
+      "step": 12310
+    },
+    {
+      "epoch": 15.75808,
+      "grad_norm": 0.9756172895431519,
+      "learning_rate": 2.5392156862745097e-05,
+      "loss": 0.534,
+      "step": 12311
+    },
+    {
+      "epoch": 15.759360000000001,
+      "grad_norm": 1.0259792804718018,
+      "learning_rate": 2.539015606242497e-05,
+      "loss": 0.5237,
+      "step": 12312
+    },
+    {
+      "epoch": 15.76064,
+      "grad_norm": 0.9864662885665894,
+      "learning_rate": 2.5388155262104844e-05,
+      "loss": 0.5443,
+      "step": 12313
+    },
+    {
+      "epoch": 15.76192,
+      "grad_norm": 0.9610289335250854,
+      "learning_rate": 2.5386154461784716e-05,
+      "loss": 0.5058,
+      "step": 12314
+    },
+    {
+      "epoch": 15.7632,
+      "grad_norm": 0.9790290594100952,
+      "learning_rate": 2.5384153661464588e-05,
+      "loss": 0.5117,
+      "step": 12315
+    },
+    {
+      "epoch": 15.76448,
+      "grad_norm": 0.9570515751838684,
+      "learning_rate": 2.538215286114446e-05,
+      "loss": 0.5391,
+      "step": 12316
+    },
+    {
+      "epoch": 15.76576,
+      "grad_norm": 0.979626476764679,
+      "learning_rate": 2.538015206082433e-05,
+      "loss": 0.5167,
+      "step": 12317
+    },
+    {
+      "epoch": 15.76704,
+      "grad_norm": 0.9989627599716187,
+      "learning_rate": 2.53781512605042e-05,
+      "loss": 0.5157,
+      "step": 12318
+    },
+    {
+      "epoch": 15.76832,
+      "grad_norm": 1.0324374437332153,
+      "learning_rate": 2.5376150460184072e-05,
+      "loss": 0.5204,
+      "step": 12319
+    },
+    {
+      "epoch": 15.7696,
+      "grad_norm": 0.9673824906349182,
+      "learning_rate": 2.537414965986395e-05,
+      "loss": 0.5313,
+      "step": 12320
+    },
+    {
+      "epoch": 15.77088,
+      "grad_norm": 1.0042349100112915,
+      "learning_rate": 2.537214885954382e-05,
+      "loss": 0.5402,
+      "step": 12321
+    },
+    {
+      "epoch": 15.77216,
+      "grad_norm": 1.0308949947357178,
+      "learning_rate": 2.537014805922369e-05,
+      "loss": 0.5681,
+      "step": 12322
+    },
+    {
+      "epoch": 15.77344,
+      "grad_norm": 0.9506542086601257,
+      "learning_rate": 2.5368147258903563e-05,
+      "loss": 0.5449,
+      "step": 12323
+    },
+    {
+      "epoch": 15.77472,
+      "grad_norm": 0.9522877335548401,
+      "learning_rate": 2.5366146458583434e-05,
+      "loss": 0.5617,
+      "step": 12324
+    },
+    {
+      "epoch": 15.776,
+      "grad_norm": 0.9548900127410889,
+      "learning_rate": 2.5364145658263306e-05,
+      "loss": 0.5247,
+      "step": 12325
+    },
+    {
+      "epoch": 15.77728,
+      "grad_norm": 0.9789671301841736,
+      "learning_rate": 2.5362144857943175e-05,
+      "loss": 0.5114,
+      "step": 12326
+    },
+    {
+      "epoch": 15.77856,
+      "grad_norm": 0.942644476890564,
+      "learning_rate": 2.5360144057623053e-05,
+      "loss": 0.4917,
+      "step": 12327
+    },
+    {
+      "epoch": 15.77984,
+      "grad_norm": 0.9783756732940674,
+      "learning_rate": 2.5358143257302925e-05,
+      "loss": 0.4895,
+      "step": 12328
+    },
+    {
+      "epoch": 15.78112,
+      "grad_norm": 0.9837256669998169,
+      "learning_rate": 2.5356142456982794e-05,
+      "loss": 0.519,
+      "step": 12329
+    },
+    {
+      "epoch": 15.782399999999999,
+      "grad_norm": 0.9281373023986816,
+      "learning_rate": 2.5354141656662666e-05,
+      "loss": 0.4883,
+      "step": 12330
+    },
+    {
+      "epoch": 15.78368,
+      "grad_norm": 1.0375429391860962,
+      "learning_rate": 2.5352140856342537e-05,
+      "loss": 0.5655,
+      "step": 12331
+    },
+    {
+      "epoch": 15.78496,
+      "grad_norm": 1.0052881240844727,
+      "learning_rate": 2.535014005602241e-05,
+      "loss": 0.5547,
+      "step": 12332
+    },
+    {
+      "epoch": 15.78624,
+      "grad_norm": 1.0248472690582275,
+      "learning_rate": 2.534813925570228e-05,
+      "loss": 0.5778,
+      "step": 12333
+    },
+    {
+      "epoch": 15.78752,
+      "grad_norm": 1.0107492208480835,
+      "learning_rate": 2.5346138455382156e-05,
+      "loss": 0.5068,
+      "step": 12334
+    },
+    {
+      "epoch": 15.7888,
+      "grad_norm": 0.9795581698417664,
+      "learning_rate": 2.5344137655062028e-05,
+      "loss": 0.5389,
+      "step": 12335
+    },
+    {
+      "epoch": 15.79008,
+      "grad_norm": 0.9817120432853699,
+      "learning_rate": 2.53421368547419e-05,
+      "loss": 0.5071,
+      "step": 12336
+    },
+    {
+      "epoch": 15.79136,
+      "grad_norm": 0.9362112879753113,
+      "learning_rate": 2.534013605442177e-05,
+      "loss": 0.5537,
+      "step": 12337
+    },
+    {
+      "epoch": 15.79264,
+      "grad_norm": 0.9294254183769226,
+      "learning_rate": 2.533813525410164e-05,
+      "loss": 0.4974,
+      "step": 12338
+    },
+    {
+      "epoch": 15.79392,
+      "grad_norm": 0.9781113862991333,
+      "learning_rate": 2.5336134453781512e-05,
+      "loss": 0.4655,
+      "step": 12339
+    },
+    {
+      "epoch": 15.7952,
+      "grad_norm": 0.9783672094345093,
+      "learning_rate": 2.5334133653461384e-05,
+      "loss": 0.5533,
+      "step": 12340
+    },
+    {
+      "epoch": 15.79648,
+      "grad_norm": 0.9659748673439026,
+      "learning_rate": 2.5332132853141256e-05,
+      "loss": 0.5486,
+      "step": 12341
+    },
+    {
+      "epoch": 15.79776,
+      "grad_norm": 0.9942366480827332,
+      "learning_rate": 2.533013205282113e-05,
+      "loss": 0.5433,
+      "step": 12342
+    },
+    {
+      "epoch": 15.79904,
+      "grad_norm": 0.9866067171096802,
+      "learning_rate": 2.5328131252501003e-05,
+      "loss": 0.5127,
+      "step": 12343
+    },
+    {
+      "epoch": 15.80032,
+      "grad_norm": 1.0013525485992432,
+      "learning_rate": 2.5326130452180875e-05,
+      "loss": 0.5194,
+      "step": 12344
+    },
+    {
+      "epoch": 15.8016,
+      "grad_norm": 0.9939999580383301,
+      "learning_rate": 2.5324129651860743e-05,
+      "loss": 0.5284,
+      "step": 12345
+    },
+    {
+      "epoch": 15.80288,
+      "grad_norm": 0.9597903490066528,
+      "learning_rate": 2.5322128851540615e-05,
+      "loss": 0.5152,
+      "step": 12346
+    },
+    {
+      "epoch": 15.80416,
+      "grad_norm": 0.9546245336532593,
+      "learning_rate": 2.5320128051220487e-05,
+      "loss": 0.487,
+      "step": 12347
+    },
+    {
+      "epoch": 15.80544,
+      "grad_norm": 1.0091270208358765,
+      "learning_rate": 2.531812725090036e-05,
+      "loss": 0.5678,
+      "step": 12348
+    },
+    {
+      "epoch": 15.80672,
+      "grad_norm": 0.9878095984458923,
+      "learning_rate": 2.5316126450580234e-05,
+      "loss": 0.5059,
+      "step": 12349
+    },
+    {
+      "epoch": 15.808,
+      "grad_norm": 1.0209447145462036,
+      "learning_rate": 2.5314125650260106e-05,
+      "loss": 0.5348,
+      "step": 12350
+    },
+    {
+      "epoch": 15.80928,
+      "grad_norm": 1.0176520347595215,
+      "learning_rate": 2.5312124849939978e-05,
+      "loss": 0.4969,
+      "step": 12351
+    },
+    {
+      "epoch": 15.81056,
+      "grad_norm": 1.0019172430038452,
+      "learning_rate": 2.531012404961985e-05,
+      "loss": 0.5235,
+      "step": 12352
+    },
+    {
+      "epoch": 15.81184,
+      "grad_norm": 1.047505259513855,
+      "learning_rate": 2.530812324929972e-05,
+      "loss": 0.6304,
+      "step": 12353
+    },
+    {
+      "epoch": 15.81312,
+      "grad_norm": 0.9699926376342773,
+      "learning_rate": 2.530612244897959e-05,
+      "loss": 0.5278,
+      "step": 12354
+    },
+    {
+      "epoch": 15.8144,
+      "grad_norm": 0.9239999055862427,
+      "learning_rate": 2.5304121648659462e-05,
+      "loss": 0.5137,
+      "step": 12355
+    },
+    {
+      "epoch": 15.81568,
+      "grad_norm": 0.9999382495880127,
+      "learning_rate": 2.530212084833934e-05,
+      "loss": 0.5398,
+      "step": 12356
+    },
+    {
+      "epoch": 15.81696,
+      "grad_norm": 0.9642220139503479,
+      "learning_rate": 2.530012004801921e-05,
+      "loss": 0.5051,
+      "step": 12357
+    },
+    {
+      "epoch": 15.81824,
+      "grad_norm": 0.9850364327430725,
+      "learning_rate": 2.529811924769908e-05,
+      "loss": 0.5086,
+      "step": 12358
+    },
+    {
+      "epoch": 15.81952,
+      "grad_norm": 0.9347919225692749,
+      "learning_rate": 2.5296118447378953e-05,
+      "loss": 0.5029,
+      "step": 12359
+    },
+    {
+      "epoch": 15.8208,
+      "grad_norm": 1.001004695892334,
+      "learning_rate": 2.5294117647058825e-05,
+      "loss": 0.5255,
+      "step": 12360
+    },
+    {
+      "epoch": 15.82208,
+      "grad_norm": 1.0347155332565308,
+      "learning_rate": 2.5292116846738693e-05,
+      "loss": 0.5276,
+      "step": 12361
+    },
+    {
+      "epoch": 15.82336,
+      "grad_norm": 1.0250365734100342,
+      "learning_rate": 2.5290116046418565e-05,
+      "loss": 0.5609,
+      "step": 12362
+    },
+    {
+      "epoch": 15.82464,
+      "grad_norm": 1.0703151226043701,
+      "learning_rate": 2.5288115246098444e-05,
+      "loss": 0.5144,
+      "step": 12363
+    },
+    {
+      "epoch": 15.82592,
+      "grad_norm": 1.0669794082641602,
+      "learning_rate": 2.5286114445778316e-05,
+      "loss": 0.5801,
+      "step": 12364
+    },
+    {
+      "epoch": 15.8272,
+      "grad_norm": 1.0120995044708252,
+      "learning_rate": 2.5284113645458184e-05,
+      "loss": 0.4951,
+      "step": 12365
+    },
+    {
+      "epoch": 15.82848,
+      "grad_norm": 0.9758439660072327,
+      "learning_rate": 2.5282112845138056e-05,
+      "loss": 0.5294,
+      "step": 12366
+    },
+    {
+      "epoch": 15.82976,
+      "grad_norm": 0.9557347893714905,
+      "learning_rate": 2.5280112044817928e-05,
+      "loss": 0.4897,
+      "step": 12367
+    },
+    {
+      "epoch": 15.83104,
+      "grad_norm": 0.9893117547035217,
+      "learning_rate": 2.52781112444978e-05,
+      "loss": 0.5063,
+      "step": 12368
+    },
+    {
+      "epoch": 15.83232,
+      "grad_norm": 1.008176326751709,
+      "learning_rate": 2.5276110444177668e-05,
+      "loss": 0.5255,
+      "step": 12369
+    },
+    {
+      "epoch": 15.8336,
+      "grad_norm": 0.9538099765777588,
+      "learning_rate": 2.5274109643857547e-05,
+      "loss": 0.5154,
+      "step": 12370
+    },
+    {
+      "epoch": 15.83488,
+      "grad_norm": 0.9922953844070435,
+      "learning_rate": 2.527210884353742e-05,
+      "loss": 0.5088,
+      "step": 12371
+    },
+    {
+      "epoch": 15.83616,
+      "grad_norm": 0.9348974823951721,
+      "learning_rate": 2.527010804321729e-05,
+      "loss": 0.4637,
+      "step": 12372
+    },
+    {
+      "epoch": 15.83744,
+      "grad_norm": 1.0178333520889282,
+      "learning_rate": 2.526810724289716e-05,
+      "loss": 0.532,
+      "step": 12373
+    },
+    {
+      "epoch": 15.83872,
+      "grad_norm": 1.0431467294692993,
+      "learning_rate": 2.526610644257703e-05,
+      "loss": 0.5307,
+      "step": 12374
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 1.0939209461212158,
+      "learning_rate": 2.5264105642256903e-05,
+      "loss": 0.5397,
+      "step": 12375
+    },
+    {
+      "epoch": 15.84128,
+      "grad_norm": 1.0265597105026245,
+      "learning_rate": 2.5262104841936774e-05,
+      "loss": 0.5354,
+      "step": 12376
+    },
+    {
+      "epoch": 15.84256,
+      "grad_norm": 0.9757692813873291,
+      "learning_rate": 2.526010404161665e-05,
+      "loss": 0.4957,
+      "step": 12377
+    },
+    {
+      "epoch": 15.84384,
+      "grad_norm": 0.9945160150527954,
+      "learning_rate": 2.525810324129652e-05,
+      "loss": 0.5204,
+      "step": 12378
+    },
+    {
+      "epoch": 15.84512,
+      "grad_norm": 0.9997515082359314,
+      "learning_rate": 2.5256102440976393e-05,
+      "loss": 0.5403,
+      "step": 12379
+    },
+    {
+      "epoch": 15.8464,
+      "grad_norm": 0.9899279475212097,
+      "learning_rate": 2.5254101640656265e-05,
+      "loss": 0.5597,
+      "step": 12380
+    },
+    {
+      "epoch": 15.84768,
+      "grad_norm": 0.941484272480011,
+      "learning_rate": 2.5252100840336134e-05,
+      "loss": 0.5027,
+      "step": 12381
+    },
+    {
+      "epoch": 15.84896,
+      "grad_norm": 0.984250545501709,
+      "learning_rate": 2.5250100040016006e-05,
+      "loss": 0.5449,
+      "step": 12382
+    },
+    {
+      "epoch": 15.85024,
+      "grad_norm": 0.9902025461196899,
+      "learning_rate": 2.5248099239695877e-05,
+      "loss": 0.5443,
+      "step": 12383
+    },
+    {
+      "epoch": 15.85152,
+      "grad_norm": 0.9709980487823486,
+      "learning_rate": 2.5246098439375753e-05,
+      "loss": 0.5031,
+      "step": 12384
+    },
+    {
+      "epoch": 15.8528,
+      "grad_norm": 0.9857783317565918,
+      "learning_rate": 2.5244097639055625e-05,
+      "loss": 0.5027,
+      "step": 12385
+    },
+    {
+      "epoch": 15.85408,
+      "grad_norm": 1.0132930278778076,
+      "learning_rate": 2.5242096838735496e-05,
+      "loss": 0.5154,
+      "step": 12386
+    },
+    {
+      "epoch": 15.85536,
+      "grad_norm": 0.9710603952407837,
+      "learning_rate": 2.5240096038415368e-05,
+      "loss": 0.5182,
+      "step": 12387
+    },
+    {
+      "epoch": 15.85664,
+      "grad_norm": 0.9620773196220398,
+      "learning_rate": 2.523809523809524e-05,
+      "loss": 0.4874,
+      "step": 12388
+    },
+    {
+      "epoch": 15.85792,
+      "grad_norm": 0.9953191876411438,
+      "learning_rate": 2.523609443777511e-05,
+      "loss": 0.5538,
+      "step": 12389
+    },
+    {
+      "epoch": 15.8592,
+      "grad_norm": 0.956991970539093,
+      "learning_rate": 2.523409363745498e-05,
+      "loss": 0.512,
+      "step": 12390
+    },
+    {
+      "epoch": 15.86048,
+      "grad_norm": 1.0285273790359497,
+      "learning_rate": 2.523209283713486e-05,
+      "loss": 0.5422,
+      "step": 12391
+    },
+    {
+      "epoch": 15.86176,
+      "grad_norm": 1.0568751096725464,
+      "learning_rate": 2.5230092036814728e-05,
+      "loss": 0.5794,
+      "step": 12392
+    },
+    {
+      "epoch": 15.86304,
+      "grad_norm": 1.0147608518600464,
+      "learning_rate": 2.52280912364946e-05,
+      "loss": 0.5809,
+      "step": 12393
+    },
+    {
+      "epoch": 15.86432,
+      "grad_norm": 0.9740559458732605,
+      "learning_rate": 2.522609043617447e-05,
+      "loss": 0.5204,
+      "step": 12394
+    },
+    {
+      "epoch": 15.8656,
+      "grad_norm": 1.0392513275146484,
+      "learning_rate": 2.5224089635854343e-05,
+      "loss": 0.5702,
+      "step": 12395
+    },
+    {
+      "epoch": 15.86688,
+      "grad_norm": 0.9570985436439514,
+      "learning_rate": 2.5222088835534215e-05,
+      "loss": 0.5251,
+      "step": 12396
+    },
+    {
+      "epoch": 15.86816,
+      "grad_norm": 1.0077378749847412,
+      "learning_rate": 2.5220088035214083e-05,
+      "loss": 0.5459,
+      "step": 12397
+    },
+    {
+      "epoch": 15.86944,
+      "grad_norm": 0.9814931154251099,
+      "learning_rate": 2.5218087234893962e-05,
+      "loss": 0.5089,
+      "step": 12398
+    },
+    {
+      "epoch": 15.87072,
+      "grad_norm": 1.0061900615692139,
+      "learning_rate": 2.5216086434573834e-05,
+      "loss": 0.534,
+      "step": 12399
+    },
+    {
+      "epoch": 15.872,
+      "grad_norm": 0.9620992541313171,
+      "learning_rate": 2.5214085634253702e-05,
+      "loss": 0.5163,
+      "step": 12400
+    },
+    {
+      "epoch": 15.87328,
+      "grad_norm": 0.9781507253646851,
+      "learning_rate": 2.5212084833933574e-05,
+      "loss": 0.5391,
+      "step": 12401
+    },
+    {
+      "epoch": 15.87456,
+      "grad_norm": 0.9385673403739929,
+      "learning_rate": 2.5210084033613446e-05,
+      "loss": 0.5046,
+      "step": 12402
+    },
+    {
+      "epoch": 15.87584,
+      "grad_norm": 0.9664863348007202,
+      "learning_rate": 2.5208083233293318e-05,
+      "loss": 0.5104,
+      "step": 12403
+    },
+    {
+      "epoch": 15.87712,
+      "grad_norm": 0.9696258306503296,
+      "learning_rate": 2.520608243297319e-05,
+      "loss": 0.4796,
+      "step": 12404
+    },
+    {
+      "epoch": 15.8784,
+      "grad_norm": 0.9488195776939392,
+      "learning_rate": 2.5204081632653065e-05,
+      "loss": 0.4884,
+      "step": 12405
+    },
+    {
+      "epoch": 15.87968,
+      "grad_norm": 1.0127061605453491,
+      "learning_rate": 2.5202080832332937e-05,
+      "loss": 0.5115,
+      "step": 12406
+    },
+    {
+      "epoch": 15.88096,
+      "grad_norm": 0.9693582057952881,
+      "learning_rate": 2.520008003201281e-05,
+      "loss": 0.5386,
+      "step": 12407
+    },
+    {
+      "epoch": 15.88224,
+      "grad_norm": 0.9904215335845947,
+      "learning_rate": 2.5198079231692677e-05,
+      "loss": 0.556,
+      "step": 12408
+    },
+    {
+      "epoch": 15.88352,
+      "grad_norm": 0.9979194402694702,
+      "learning_rate": 2.519607843137255e-05,
+      "loss": 0.5033,
+      "step": 12409
+    },
+    {
+      "epoch": 15.8848,
+      "grad_norm": 1.007330060005188,
+      "learning_rate": 2.519407763105242e-05,
+      "loss": 0.5607,
+      "step": 12410
+    },
+    {
+      "epoch": 15.88608,
+      "grad_norm": 1.0108498334884644,
+      "learning_rate": 2.5192076830732293e-05,
+      "loss": 0.5434,
+      "step": 12411
+    },
+    {
+      "epoch": 15.88736,
+      "grad_norm": 0.9750909805297852,
+      "learning_rate": 2.5190076030412168e-05,
+      "loss": 0.5233,
+      "step": 12412
+    },
+    {
+      "epoch": 15.88864,
+      "grad_norm": 0.982916533946991,
+      "learning_rate": 2.518807523009204e-05,
+      "loss": 0.5241,
+      "step": 12413
+    },
+    {
+      "epoch": 15.88992,
+      "grad_norm": 0.9335137009620667,
+      "learning_rate": 2.5186074429771912e-05,
+      "loss": 0.4939,
+      "step": 12414
+    },
+    {
+      "epoch": 15.8912,
+      "grad_norm": 0.922966718673706,
+      "learning_rate": 2.5184073629451784e-05,
+      "loss": 0.5243,
+      "step": 12415
+    },
+    {
+      "epoch": 15.89248,
+      "grad_norm": 1.0012139081954956,
+      "learning_rate": 2.5182072829131652e-05,
+      "loss": 0.5089,
+      "step": 12416
+    },
+    {
+      "epoch": 15.89376,
+      "grad_norm": 0.9846066832542419,
+      "learning_rate": 2.5180072028811524e-05,
+      "loss": 0.5127,
+      "step": 12417
+    },
+    {
+      "epoch": 15.89504,
+      "grad_norm": 0.9632936120033264,
+      "learning_rate": 2.5178071228491396e-05,
+      "loss": 0.5822,
+      "step": 12418
+    },
+    {
+      "epoch": 15.89632,
+      "grad_norm": 0.9547779560089111,
+      "learning_rate": 2.517607042817127e-05,
+      "loss": 0.5153,
+      "step": 12419
+    },
+    {
+      "epoch": 15.8976,
+      "grad_norm": 0.9924589991569519,
+      "learning_rate": 2.5174069627851143e-05,
+      "loss": 0.558,
+      "step": 12420
+    },
+    {
+      "epoch": 15.89888,
+      "grad_norm": 0.9812869429588318,
+      "learning_rate": 2.5172068827531015e-05,
+      "loss": 0.5061,
+      "step": 12421
+    },
+    {
+      "epoch": 15.90016,
+      "grad_norm": 1.0087809562683105,
+      "learning_rate": 2.5170068027210887e-05,
+      "loss": 0.5524,
+      "step": 12422
+    },
+    {
+      "epoch": 15.901440000000001,
+      "grad_norm": 0.9655302166938782,
+      "learning_rate": 2.516806722689076e-05,
+      "loss": 0.5159,
+      "step": 12423
+    },
+    {
+      "epoch": 15.90272,
+      "grad_norm": 1.032762050628662,
+      "learning_rate": 2.5166066426570627e-05,
+      "loss": 0.5757,
+      "step": 12424
+    },
+    {
+      "epoch": 15.904,
+      "grad_norm": 0.9734963178634644,
+      "learning_rate": 2.51640656262505e-05,
+      "loss": 0.4966,
+      "step": 12425
+    },
+    {
+      "epoch": 15.90528,
+      "grad_norm": 0.9694871306419373,
+      "learning_rate": 2.5162064825930377e-05,
+      "loss": 0.5307,
+      "step": 12426
+    },
+    {
+      "epoch": 15.90656,
+      "grad_norm": 1.0066063404083252,
+      "learning_rate": 2.5160064025610246e-05,
+      "loss": 0.5717,
+      "step": 12427
+    },
+    {
+      "epoch": 15.90784,
+      "grad_norm": 0.9474746584892273,
+      "learning_rate": 2.5158063225290118e-05,
+      "loss": 0.5199,
+      "step": 12428
+    },
+    {
+      "epoch": 15.90912,
+      "grad_norm": 0.9809539914131165,
+      "learning_rate": 2.515606242496999e-05,
+      "loss": 0.5082,
+      "step": 12429
+    },
+    {
+      "epoch": 15.9104,
+      "grad_norm": 1.0064231157302856,
+      "learning_rate": 2.515406162464986e-05,
+      "loss": 0.4924,
+      "step": 12430
+    },
+    {
+      "epoch": 15.91168,
+      "grad_norm": 0.9995524883270264,
+      "learning_rate": 2.5152060824329733e-05,
+      "loss": 0.5142,
+      "step": 12431
+    },
+    {
+      "epoch": 15.91296,
+      "grad_norm": 0.9583382606506348,
+      "learning_rate": 2.5150060024009602e-05,
+      "loss": 0.4687,
+      "step": 12432
+    },
+    {
+      "epoch": 15.91424,
+      "grad_norm": 0.9433925747871399,
+      "learning_rate": 2.514805922368948e-05,
+      "loss": 0.5122,
+      "step": 12433
+    },
+    {
+      "epoch": 15.91552,
+      "grad_norm": 0.9556870460510254,
+      "learning_rate": 2.5146058423369352e-05,
+      "loss": 0.5444,
+      "step": 12434
+    },
+    {
+      "epoch": 15.9168,
+      "grad_norm": 1.0067849159240723,
+      "learning_rate": 2.514405762304922e-05,
+      "loss": 0.5555,
+      "step": 12435
+    },
+    {
+      "epoch": 15.91808,
+      "grad_norm": 0.9572538733482361,
+      "learning_rate": 2.5142056822729093e-05,
+      "loss": 0.5077,
+      "step": 12436
+    },
+    {
+      "epoch": 15.91936,
+      "grad_norm": 1.0347596406936646,
+      "learning_rate": 2.5140056022408964e-05,
+      "loss": 0.5287,
+      "step": 12437
+    },
+    {
+      "epoch": 15.92064,
+      "grad_norm": 0.9975804090499878,
+      "learning_rate": 2.5138055222088836e-05,
+      "loss": 0.5213,
+      "step": 12438
+    },
+    {
+      "epoch": 15.92192,
+      "grad_norm": 0.9736266136169434,
+      "learning_rate": 2.5136054421768708e-05,
+      "loss": 0.5386,
+      "step": 12439
+    },
+    {
+      "epoch": 15.9232,
+      "grad_norm": 1.077915072441101,
+      "learning_rate": 2.5134053621448583e-05,
+      "loss": 0.5772,
+      "step": 12440
+    },
+    {
+      "epoch": 15.924479999999999,
+      "grad_norm": 1.0086665153503418,
+      "learning_rate": 2.5132052821128455e-05,
+      "loss": 0.5173,
+      "step": 12441
+    },
+    {
+      "epoch": 15.92576,
+      "grad_norm": 1.012130856513977,
+      "learning_rate": 2.5130052020808327e-05,
+      "loss": 0.5162,
+      "step": 12442
+    },
+    {
+      "epoch": 15.92704,
+      "grad_norm": 0.9995039701461792,
+      "learning_rate": 2.5128051220488196e-05,
+      "loss": 0.5068,
+      "step": 12443
+    },
+    {
+      "epoch": 15.92832,
+      "grad_norm": 0.9592408537864685,
+      "learning_rate": 2.5126050420168067e-05,
+      "loss": 0.507,
+      "step": 12444
+    },
+    {
+      "epoch": 15.9296,
+      "grad_norm": 0.9532192945480347,
+      "learning_rate": 2.512404961984794e-05,
+      "loss": 0.5134,
+      "step": 12445
+    },
+    {
+      "epoch": 15.93088,
+      "grad_norm": 1.0366421937942505,
+      "learning_rate": 2.512204881952781e-05,
+      "loss": 0.5393,
+      "step": 12446
+    },
+    {
+      "epoch": 15.93216,
+      "grad_norm": 0.9751626253128052,
+      "learning_rate": 2.5120048019207686e-05,
+      "loss": 0.5326,
+      "step": 12447
+    },
+    {
+      "epoch": 15.933440000000001,
+      "grad_norm": 0.9007095694541931,
+      "learning_rate": 2.5118047218887558e-05,
+      "loss": 0.4805,
+      "step": 12448
+    },
+    {
+      "epoch": 15.93472,
+      "grad_norm": 1.0274094343185425,
+      "learning_rate": 2.511604641856743e-05,
+      "loss": 0.5352,
+      "step": 12449
+    },
+    {
+      "epoch": 15.936,
+      "grad_norm": 0.9921882152557373,
+      "learning_rate": 2.5114045618247302e-05,
+      "loss": 0.5379,
+      "step": 12450
+    },
+    {
+      "epoch": 15.93728,
+      "grad_norm": 1.0404365062713623,
+      "learning_rate": 2.511204481792717e-05,
+      "loss": 0.5624,
+      "step": 12451
+    },
+    {
+      "epoch": 15.93856,
+      "grad_norm": 0.9763307571411133,
+      "learning_rate": 2.5110044017607042e-05,
+      "loss": 0.5373,
+      "step": 12452
+    },
+    {
+      "epoch": 15.93984,
+      "grad_norm": 0.939691424369812,
+      "learning_rate": 2.5108043217286914e-05,
+      "loss": 0.5621,
+      "step": 12453
+    },
+    {
+      "epoch": 15.94112,
+      "grad_norm": 0.9770581126213074,
+      "learning_rate": 2.5106042416966786e-05,
+      "loss": 0.5202,
+      "step": 12454
+    },
+    {
+      "epoch": 15.9424,
+      "grad_norm": 1.0360689163208008,
+      "learning_rate": 2.510404161664666e-05,
+      "loss": 0.5714,
+      "step": 12455
+    },
+    {
+      "epoch": 15.94368,
+      "grad_norm": 0.9993544816970825,
+      "learning_rate": 2.5102040816326533e-05,
+      "loss": 0.5125,
+      "step": 12456
+    },
+    {
+      "epoch": 15.94496,
+      "grad_norm": 1.0267115831375122,
+      "learning_rate": 2.5100040016006405e-05,
+      "loss": 0.5312,
+      "step": 12457
+    },
+    {
+      "epoch": 15.94624,
+      "grad_norm": 0.9762638807296753,
+      "learning_rate": 2.5098039215686277e-05,
+      "loss": 0.529,
+      "step": 12458
+    },
+    {
+      "epoch": 15.94752,
+      "grad_norm": 1.0173263549804688,
+      "learning_rate": 2.5096038415366145e-05,
+      "loss": 0.5196,
+      "step": 12459
+    },
+    {
+      "epoch": 15.9488,
+      "grad_norm": 1.0800774097442627,
+      "learning_rate": 2.5094037615046017e-05,
+      "loss": 0.6021,
+      "step": 12460
+    },
+    {
+      "epoch": 15.95008,
+      "grad_norm": 1.0533719062805176,
+      "learning_rate": 2.509203681472589e-05,
+      "loss": 0.5971,
+      "step": 12461
+    },
+    {
+      "epoch": 15.95136,
+      "grad_norm": 1.0251033306121826,
+      "learning_rate": 2.5090036014405764e-05,
+      "loss": 0.5547,
+      "step": 12462
+    },
+    {
+      "epoch": 15.95264,
+      "grad_norm": 0.9884242415428162,
+      "learning_rate": 2.5088035214085636e-05,
+      "loss": 0.5788,
+      "step": 12463
+    },
+    {
+      "epoch": 15.95392,
+      "grad_norm": 0.9535346627235413,
+      "learning_rate": 2.5086034413765508e-05,
+      "loss": 0.488,
+      "step": 12464
+    },
+    {
+      "epoch": 15.9552,
+      "grad_norm": 0.9252434372901917,
+      "learning_rate": 2.508403361344538e-05,
+      "loss": 0.4886,
+      "step": 12465
+    },
+    {
+      "epoch": 15.956479999999999,
+      "grad_norm": 1.00466787815094,
+      "learning_rate": 2.508203281312525e-05,
+      "loss": 0.5489,
+      "step": 12466
+    },
+    {
+      "epoch": 15.95776,
+      "grad_norm": 0.9779563546180725,
+      "learning_rate": 2.508003201280512e-05,
+      "loss": 0.5357,
+      "step": 12467
+    },
+    {
+      "epoch": 15.95904,
+      "grad_norm": 0.9743767380714417,
+      "learning_rate": 2.5078031212484992e-05,
+      "loss": 0.4939,
+      "step": 12468
+    },
+    {
+      "epoch": 15.96032,
+      "grad_norm": 0.9605897665023804,
+      "learning_rate": 2.507603041216487e-05,
+      "loss": 0.4879,
+      "step": 12469
+    },
+    {
+      "epoch": 15.9616,
+      "grad_norm": 0.9653512835502625,
+      "learning_rate": 2.507402961184474e-05,
+      "loss": 0.51,
+      "step": 12470
+    },
+    {
+      "epoch": 15.96288,
+      "grad_norm": 0.9319489002227783,
+      "learning_rate": 2.507202881152461e-05,
+      "loss": 0.4609,
+      "step": 12471
+    },
+    {
+      "epoch": 15.96416,
+      "grad_norm": 0.994040310382843,
+      "learning_rate": 2.5070028011204483e-05,
+      "loss": 0.5493,
+      "step": 12472
+    },
+    {
+      "epoch": 15.96544,
+      "grad_norm": 1.0005146265029907,
+      "learning_rate": 2.5068027210884355e-05,
+      "loss": 0.5584,
+      "step": 12473
+    },
+    {
+      "epoch": 15.96672,
+      "grad_norm": 0.9767276048660278,
+      "learning_rate": 2.5066026410564227e-05,
+      "loss": 0.4839,
+      "step": 12474
+    },
+    {
+      "epoch": 15.968,
+      "grad_norm": 0.9705262780189514,
+      "learning_rate": 2.5064025610244095e-05,
+      "loss": 0.5139,
+      "step": 12475
+    },
+    {
+      "epoch": 15.96928,
+      "grad_norm": 1.0184811353683472,
+      "learning_rate": 2.5062024809923974e-05,
+      "loss": 0.5444,
+      "step": 12476
+    },
+    {
+      "epoch": 15.97056,
+      "grad_norm": 1.0089943408966064,
+      "learning_rate": 2.5060024009603845e-05,
+      "loss": 0.5257,
+      "step": 12477
+    },
+    {
+      "epoch": 15.97184,
+      "grad_norm": 1.0058419704437256,
+      "learning_rate": 2.5058023209283714e-05,
+      "loss": 0.4931,
+      "step": 12478
+    },
+    {
+      "epoch": 15.97312,
+      "grad_norm": 1.0035637617111206,
+      "learning_rate": 2.5056022408963586e-05,
+      "loss": 0.4789,
+      "step": 12479
+    },
+    {
+      "epoch": 15.9744,
+      "grad_norm": 1.051369309425354,
+      "learning_rate": 2.5054021608643458e-05,
+      "loss": 0.5567,
+      "step": 12480
+    },
+    {
+      "epoch": 15.97568,
+      "grad_norm": 1.0520974397659302,
+      "learning_rate": 2.505202080832333e-05,
+      "loss": 0.5678,
+      "step": 12481
+    },
+    {
+      "epoch": 15.97696,
+      "grad_norm": 0.9842495918273926,
+      "learning_rate": 2.50500200080032e-05,
+      "loss": 0.5092,
+      "step": 12482
+    },
+    {
+      "epoch": 15.97824,
+      "grad_norm": 0.9604758620262146,
+      "learning_rate": 2.5048019207683077e-05,
+      "loss": 0.526,
+      "step": 12483
+    },
+    {
+      "epoch": 15.97952,
+      "grad_norm": 0.9821440577507019,
+      "learning_rate": 2.504601840736295e-05,
+      "loss": 0.4723,
+      "step": 12484
+    },
+    {
+      "epoch": 15.9808,
+      "grad_norm": 0.9930018782615662,
+      "learning_rate": 2.504401760704282e-05,
+      "loss": 0.5205,
+      "step": 12485
+    },
+    {
+      "epoch": 15.98208,
+      "grad_norm": 1.0078686475753784,
+      "learning_rate": 2.504201680672269e-05,
+      "loss": 0.5439,
+      "step": 12486
+    },
+    {
+      "epoch": 15.98336,
+      "grad_norm": 1.0228537321090698,
+      "learning_rate": 2.504001600640256e-05,
+      "loss": 0.5895,
+      "step": 12487
+    },
+    {
+      "epoch": 15.98464,
+      "grad_norm": 1.0596520900726318,
+      "learning_rate": 2.5038015206082433e-05,
+      "loss": 0.5966,
+      "step": 12488
+    },
+    {
+      "epoch": 15.98592,
+      "grad_norm": 0.9747412800788879,
+      "learning_rate": 2.5036014405762304e-05,
+      "loss": 0.5209,
+      "step": 12489
+    },
+    {
+      "epoch": 15.9872,
+      "grad_norm": 1.090663194656372,
+      "learning_rate": 2.503401360544218e-05,
+      "loss": 0.545,
+      "step": 12490
+    },
+    {
+      "epoch": 15.98848,
+      "grad_norm": 0.9801415205001831,
+      "learning_rate": 2.503201280512205e-05,
+      "loss": 0.5044,
+      "step": 12491
+    },
+    {
+      "epoch": 15.98976,
+      "grad_norm": 0.9882249236106873,
+      "learning_rate": 2.5030012004801923e-05,
+      "loss": 0.5395,
+      "step": 12492
+    },
+    {
+      "epoch": 15.99104,
+      "grad_norm": 0.9727967381477356,
+      "learning_rate": 2.5028011204481795e-05,
+      "loss": 0.5298,
+      "step": 12493
+    },
+    {
+      "epoch": 15.99232,
+      "grad_norm": 1.004541277885437,
+      "learning_rate": 2.5026010404161664e-05,
+      "loss": 0.5377,
+      "step": 12494
+    },
+    {
+      "epoch": 15.9936,
+      "grad_norm": 0.9552838206291199,
+      "learning_rate": 2.5024009603841536e-05,
+      "loss": 0.5197,
+      "step": 12495
+    },
+    {
+      "epoch": 15.99488,
+      "grad_norm": 0.9584947228431702,
+      "learning_rate": 2.5022008803521407e-05,
+      "loss": 0.5345,
+      "step": 12496
+    },
+    {
+      "epoch": 15.99616,
+      "grad_norm": 0.9853367209434509,
+      "learning_rate": 2.5020008003201283e-05,
+      "loss": 0.5047,
+      "step": 12497
+    },
+    {
+      "epoch": 15.99744,
+      "grad_norm": 0.9928620457649231,
+      "learning_rate": 2.5018007202881154e-05,
+      "loss": 0.5282,
+      "step": 12498
+    },
+    {
+      "epoch": 15.99872,
+      "grad_norm": 1.0169389247894287,
+      "learning_rate": 2.5016006402561026e-05,
+      "loss": 0.5764,
+      "step": 12499
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": Infinity,
+      "learning_rate": 2.5016006402561026e-05,
+      "loss": 0.9264,
+      "step": 12500
+    },
+    {
+      "epoch": 16.00128,
+      "grad_norm": 0.9896093010902405,
+      "learning_rate": 2.5014005602240898e-05,
+      "loss": 0.4865,
+      "step": 12501
+    },
+    {
+      "epoch": 16.00256,
+      "grad_norm": 0.9680584669113159,
+      "learning_rate": 2.501200480192077e-05,
+      "loss": 0.4852,
+      "step": 12502
+    },
+    {
+      "epoch": 16.00384,
+      "grad_norm": 0.9140750765800476,
+      "learning_rate": 2.501000400160064e-05,
+      "loss": 0.48,
+      "step": 12503
+    },
+    {
+      "epoch": 16.00512,
+      "grad_norm": 0.9292126893997192,
+      "learning_rate": 2.500800320128051e-05,
+      "loss": 0.4647,
+      "step": 12504
+    },
+    {
+      "epoch": 16.0064,
+      "grad_norm": 0.9622963070869446,
+      "learning_rate": 2.500600240096039e-05,
+      "loss": 0.4908,
+      "step": 12505
+    },
+    {
+      "epoch": 16.00768,
+      "grad_norm": 0.9972251653671265,
+      "learning_rate": 2.5004001600640257e-05,
+      "loss": 0.5065,
+      "step": 12506
+    },
+    {
+      "epoch": 16.00896,
+      "grad_norm": 1.0023291110992432,
+      "learning_rate": 2.500200080032013e-05,
+      "loss": 0.5344,
+      "step": 12507
+    },
+    {
+      "epoch": 16.01024,
+      "grad_norm": 0.9497984647750854,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4935,
+      "step": 12508
+    },
+    {
+      "epoch": 16.01152,
+      "grad_norm": 0.9904937148094177,
+      "learning_rate": 2.4997999199679873e-05,
+      "loss": 0.4871,
+      "step": 12509
+    },
+    {
+      "epoch": 16.0128,
+      "grad_norm": 0.9689871668815613,
+      "learning_rate": 2.4995998399359745e-05,
+      "loss": 0.471,
+      "step": 12510
+    },
+    {
+      "epoch": 16.01408,
+      "grad_norm": 0.9814728498458862,
+      "learning_rate": 2.4993997599039617e-05,
+      "loss": 0.5067,
+      "step": 12511
+    },
+    {
+      "epoch": 16.01536,
+      "grad_norm": 0.97007817029953,
+      "learning_rate": 2.499199679871949e-05,
+      "loss": 0.507,
+      "step": 12512
+    },
+    {
+      "epoch": 16.01664,
+      "grad_norm": 1.0243626832962036,
+      "learning_rate": 2.498999599839936e-05,
+      "loss": 0.5726,
+      "step": 12513
+    },
+    {
+      "epoch": 16.01792,
+      "grad_norm": 1.0011252164840698,
+      "learning_rate": 2.4987995198079232e-05,
+      "loss": 0.4805,
+      "step": 12514
+    },
+    {
+      "epoch": 16.0192,
+      "grad_norm": 0.9926708936691284,
+      "learning_rate": 2.4985994397759104e-05,
+      "loss": 0.5031,
+      "step": 12515
+    },
+    {
+      "epoch": 16.02048,
+      "grad_norm": 0.9110226035118103,
+      "learning_rate": 2.4983993597438976e-05,
+      "loss": 0.4117,
+      "step": 12516
+    },
+    {
+      "epoch": 16.02176,
+      "grad_norm": 1.0108290910720825,
+      "learning_rate": 2.498199279711885e-05,
+      "loss": 0.5429,
+      "step": 12517
+    },
+    {
+      "epoch": 16.02304,
+      "grad_norm": 0.990385115146637,
+      "learning_rate": 2.497999199679872e-05,
+      "loss": 0.4741,
+      "step": 12518
+    },
+    {
+      "epoch": 16.02432,
+      "grad_norm": 0.982943058013916,
+      "learning_rate": 2.497799119647859e-05,
+      "loss": 0.5495,
+      "step": 12519
+    },
+    {
+      "epoch": 16.0256,
+      "grad_norm": 0.9531762599945068,
+      "learning_rate": 2.4975990396158463e-05,
+      "loss": 0.4758,
+      "step": 12520
+    },
+    {
+      "epoch": 16.02688,
+      "grad_norm": 0.9687629342079163,
+      "learning_rate": 2.497398959583834e-05,
+      "loss": 0.503,
+      "step": 12521
+    },
+    {
+      "epoch": 16.02816,
+      "grad_norm": 1.029827356338501,
+      "learning_rate": 2.4971988795518207e-05,
+      "loss": 0.5105,
+      "step": 12522
+    },
+    {
+      "epoch": 16.02944,
+      "grad_norm": 1.062757968902588,
+      "learning_rate": 2.496998799519808e-05,
+      "loss": 0.5112,
+      "step": 12523
+    },
+    {
+      "epoch": 16.03072,
+      "grad_norm": 1.0137965679168701,
+      "learning_rate": 2.4967987194877954e-05,
+      "loss": 0.5003,
+      "step": 12524
+    },
+    {
+      "epoch": 16.032,
+      "grad_norm": 1.0039608478546143,
+      "learning_rate": 2.4965986394557826e-05,
+      "loss": 0.5081,
+      "step": 12525
+    },
+    {
+      "epoch": 16.03328,
+      "grad_norm": 0.9910953044891357,
+      "learning_rate": 2.4963985594237695e-05,
+      "loss": 0.5416,
+      "step": 12526
+    },
+    {
+      "epoch": 16.03456,
+      "grad_norm": 1.0009675025939941,
+      "learning_rate": 2.4961984793917566e-05,
+      "loss": 0.5186,
+      "step": 12527
+    },
+    {
+      "epoch": 16.03584,
+      "grad_norm": 0.9520965814590454,
+      "learning_rate": 2.4959983993597442e-05,
+      "loss": 0.499,
+      "step": 12528
+    },
+    {
+      "epoch": 16.03712,
+      "grad_norm": 0.9937947392463684,
+      "learning_rate": 2.4957983193277314e-05,
+      "loss": 0.5301,
+      "step": 12529
+    },
+    {
+      "epoch": 16.0384,
+      "grad_norm": 1.0274006128311157,
+      "learning_rate": 2.4955982392957182e-05,
+      "loss": 0.5275,
+      "step": 12530
+    },
+    {
+      "epoch": 16.03968,
+      "grad_norm": 0.9785746932029724,
+      "learning_rate": 2.4953981592637057e-05,
+      "loss": 0.5235,
+      "step": 12531
+    },
+    {
+      "epoch": 16.04096,
+      "grad_norm": 0.9842962026596069,
+      "learning_rate": 2.495198079231693e-05,
+      "loss": 0.4919,
+      "step": 12532
+    },
+    {
+      "epoch": 16.04224,
+      "grad_norm": 0.9303730726242065,
+      "learning_rate": 2.49499799919968e-05,
+      "loss": 0.4506,
+      "step": 12533
+    },
+    {
+      "epoch": 16.04352,
+      "grad_norm": 1.0411185026168823,
+      "learning_rate": 2.494797919167667e-05,
+      "loss": 0.5188,
+      "step": 12534
+    },
+    {
+      "epoch": 16.0448,
+      "grad_norm": 0.9499898552894592,
+      "learning_rate": 2.4945978391356545e-05,
+      "loss": 0.4906,
+      "step": 12535
+    },
+    {
+      "epoch": 16.04608,
+      "grad_norm": 0.9907384514808655,
+      "learning_rate": 2.4943977591036417e-05,
+      "loss": 0.5185,
+      "step": 12536
+    },
+    {
+      "epoch": 16.04736,
+      "grad_norm": 0.9777514338493347,
+      "learning_rate": 2.494197679071629e-05,
+      "loss": 0.5008,
+      "step": 12537
+    },
+    {
+      "epoch": 16.04864,
+      "grad_norm": 1.0125657320022583,
+      "learning_rate": 2.4939975990396157e-05,
+      "loss": 0.4957,
+      "step": 12538
+    },
+    {
+      "epoch": 16.04992,
+      "grad_norm": 1.0564517974853516,
+      "learning_rate": 2.4937975190076032e-05,
+      "loss": 0.5403,
+      "step": 12539
+    },
+    {
+      "epoch": 16.0512,
+      "grad_norm": 1.0502736568450928,
+      "learning_rate": 2.4935974389755904e-05,
+      "loss": 0.5516,
+      "step": 12540
+    },
+    {
+      "epoch": 16.05248,
+      "grad_norm": 0.9993465542793274,
+      "learning_rate": 2.4933973589435776e-05,
+      "loss": 0.5124,
+      "step": 12541
+    },
+    {
+      "epoch": 16.05376,
+      "grad_norm": 1.0142031908035278,
+      "learning_rate": 2.4931972789115648e-05,
+      "loss": 0.4857,
+      "step": 12542
+    },
+    {
+      "epoch": 16.05504,
+      "grad_norm": 1.0206118822097778,
+      "learning_rate": 2.492997198879552e-05,
+      "loss": 0.5747,
+      "step": 12543
+    },
+    {
+      "epoch": 16.05632,
+      "grad_norm": 1.0780686140060425,
+      "learning_rate": 2.492797118847539e-05,
+      "loss": 0.5194,
+      "step": 12544
+    },
+    {
+      "epoch": 16.0576,
+      "grad_norm": 0.9203510284423828,
+      "learning_rate": 2.4925970388155263e-05,
+      "loss": 0.5112,
+      "step": 12545
+    },
+    {
+      "epoch": 16.05888,
+      "grad_norm": 0.9816039800643921,
+      "learning_rate": 2.4923969587835135e-05,
+      "loss": 0.5326,
+      "step": 12546
+    },
+    {
+      "epoch": 16.06016,
+      "grad_norm": 1.0037397146224976,
+      "learning_rate": 2.4921968787515007e-05,
+      "loss": 0.5013,
+      "step": 12547
+    },
+    {
+      "epoch": 16.06144,
+      "grad_norm": 1.0208714008331299,
+      "learning_rate": 2.491996798719488e-05,
+      "loss": 0.4871,
+      "step": 12548
+    },
+    {
+      "epoch": 16.06272,
+      "grad_norm": 1.050378441810608,
+      "learning_rate": 2.491796718687475e-05,
+      "loss": 0.551,
+      "step": 12549
+    },
+    {
+      "epoch": 16.064,
+      "grad_norm": 1.0114853382110596,
+      "learning_rate": 2.4915966386554623e-05,
+      "loss": 0.5621,
+      "step": 12550
+    },
+    {
+      "epoch": 16.06528,
+      "grad_norm": 0.9831593036651611,
+      "learning_rate": 2.4913965586234494e-05,
+      "loss": 0.5084,
+      "step": 12551
+    },
+    {
+      "epoch": 16.06656,
+      "grad_norm": 0.9667739868164062,
+      "learning_rate": 2.4911964785914366e-05,
+      "loss": 0.4955,
+      "step": 12552
+    },
+    {
+      "epoch": 16.06784,
+      "grad_norm": 1.0448590517044067,
+      "learning_rate": 2.4909963985594238e-05,
+      "loss": 0.5371,
+      "step": 12553
+    },
+    {
+      "epoch": 16.06912,
+      "grad_norm": 1.012335181236267,
+      "learning_rate": 2.490796318527411e-05,
+      "loss": 0.505,
+      "step": 12554
+    },
+    {
+      "epoch": 16.0704,
+      "grad_norm": 1.0684611797332764,
+      "learning_rate": 2.4905962384953982e-05,
+      "loss": 0.5381,
+      "step": 12555
+    },
+    {
+      "epoch": 16.07168,
+      "grad_norm": 1.0244140625,
+      "learning_rate": 2.4903961584633857e-05,
+      "loss": 0.4584,
+      "step": 12556
+    },
+    {
+      "epoch": 16.07296,
+      "grad_norm": 1.0357544422149658,
+      "learning_rate": 2.4901960784313726e-05,
+      "loss": 0.5328,
+      "step": 12557
+    },
+    {
+      "epoch": 16.07424,
+      "grad_norm": 1.0519928932189941,
+      "learning_rate": 2.4899959983993597e-05,
+      "loss": 0.4919,
+      "step": 12558
+    },
+    {
+      "epoch": 16.07552,
+      "grad_norm": 1.0284450054168701,
+      "learning_rate": 2.489795918367347e-05,
+      "loss": 0.5221,
+      "step": 12559
+    },
+    {
+      "epoch": 16.0768,
+      "grad_norm": 0.9368478059768677,
+      "learning_rate": 2.4895958383353345e-05,
+      "loss": 0.4828,
+      "step": 12560
+    },
+    {
+      "epoch": 16.07808,
+      "grad_norm": 0.9959625601768494,
+      "learning_rate": 2.4893957583033213e-05,
+      "loss": 0.5332,
+      "step": 12561
+    },
+    {
+      "epoch": 16.07936,
+      "grad_norm": 1.113194465637207,
+      "learning_rate": 2.4891956782713085e-05,
+      "loss": 0.509,
+      "step": 12562
+    },
+    {
+      "epoch": 16.08064,
+      "grad_norm": 0.9688543081283569,
+      "learning_rate": 2.488995598239296e-05,
+      "loss": 0.5131,
+      "step": 12563
+    },
+    {
+      "epoch": 16.08192,
+      "grad_norm": 1.0220595598220825,
+      "learning_rate": 2.4887955182072832e-05,
+      "loss": 0.4878,
+      "step": 12564
+    },
+    {
+      "epoch": 16.0832,
+      "grad_norm": 1.0453155040740967,
+      "learning_rate": 2.48859543817527e-05,
+      "loss": 0.5074,
+      "step": 12565
+    },
+    {
+      "epoch": 16.08448,
+      "grad_norm": 1.0528833866119385,
+      "learning_rate": 2.4883953581432572e-05,
+      "loss": 0.5193,
+      "step": 12566
+    },
+    {
+      "epoch": 16.08576,
+      "grad_norm": 1.0217548608779907,
+      "learning_rate": 2.4881952781112448e-05,
+      "loss": 0.5449,
+      "step": 12567
+    },
+    {
+      "epoch": 16.087040000000002,
+      "grad_norm": 1.0063393115997314,
+      "learning_rate": 2.487995198079232e-05,
+      "loss": 0.4872,
+      "step": 12568
+    },
+    {
+      "epoch": 16.08832,
+      "grad_norm": 1.0022279024124146,
+      "learning_rate": 2.4877951180472188e-05,
+      "loss": 0.5315,
+      "step": 12569
+    },
+    {
+      "epoch": 16.0896,
+      "grad_norm": 0.9380884170532227,
+      "learning_rate": 2.4875950380152063e-05,
+      "loss": 0.4819,
+      "step": 12570
+    },
+    {
+      "epoch": 16.09088,
+      "grad_norm": 1.018511176109314,
+      "learning_rate": 2.4873949579831935e-05,
+      "loss": 0.5077,
+      "step": 12571
+    },
+    {
+      "epoch": 16.09216,
+      "grad_norm": 0.9977579116821289,
+      "learning_rate": 2.4871948779511807e-05,
+      "loss": 0.491,
+      "step": 12572
+    },
+    {
+      "epoch": 16.09344,
+      "grad_norm": 1.025163173675537,
+      "learning_rate": 2.4869947979191675e-05,
+      "loss": 0.4826,
+      "step": 12573
+    },
+    {
+      "epoch": 16.09472,
+      "grad_norm": 1.0350524187088013,
+      "learning_rate": 2.486794717887155e-05,
+      "loss": 0.5445,
+      "step": 12574
+    },
+    {
+      "epoch": 16.096,
+      "grad_norm": 1.023727297782898,
+      "learning_rate": 2.4865946378551422e-05,
+      "loss": 0.5394,
+      "step": 12575
+    },
+    {
+      "epoch": 16.09728,
+      "grad_norm": 0.9656332731246948,
+      "learning_rate": 2.4863945578231294e-05,
+      "loss": 0.4724,
+      "step": 12576
+    },
+    {
+      "epoch": 16.09856,
+      "grad_norm": 1.0451605319976807,
+      "learning_rate": 2.4861944777911166e-05,
+      "loss": 0.5221,
+      "step": 12577
+    },
+    {
+      "epoch": 16.09984,
+      "grad_norm": 1.0490944385528564,
+      "learning_rate": 2.4859943977591038e-05,
+      "loss": 0.5555,
+      "step": 12578
+    },
+    {
+      "epoch": 16.10112,
+      "grad_norm": 1.0549135208129883,
+      "learning_rate": 2.485794317727091e-05,
+      "loss": 0.4902,
+      "step": 12579
+    },
+    {
+      "epoch": 16.1024,
+      "grad_norm": 0.9879729151725769,
+      "learning_rate": 2.485594237695078e-05,
+      "loss": 0.4875,
+      "step": 12580
+    },
+    {
+      "epoch": 16.10368,
+      "grad_norm": 1.011025071144104,
+      "learning_rate": 2.4853941576630654e-05,
+      "loss": 0.4903,
+      "step": 12581
+    },
+    {
+      "epoch": 16.10496,
+      "grad_norm": 0.9577101469039917,
+      "learning_rate": 2.4851940776310525e-05,
+      "loss": 0.5068,
+      "step": 12582
+    },
+    {
+      "epoch": 16.10624,
+      "grad_norm": 0.9981372356414795,
+      "learning_rate": 2.4849939975990397e-05,
+      "loss": 0.5308,
+      "step": 12583
+    },
+    {
+      "epoch": 16.10752,
+      "grad_norm": 0.9399566054344177,
+      "learning_rate": 2.484793917567027e-05,
+      "loss": 0.4912,
+      "step": 12584
+    },
+    {
+      "epoch": 16.1088,
+      "grad_norm": 0.9757966995239258,
+      "learning_rate": 2.484593837535014e-05,
+      "loss": 0.5236,
+      "step": 12585
+    },
+    {
+      "epoch": 16.11008,
+      "grad_norm": 1.0582964420318604,
+      "learning_rate": 2.4843937575030013e-05,
+      "loss": 0.5207,
+      "step": 12586
+    },
+    {
+      "epoch": 16.11136,
+      "grad_norm": 0.9841398596763611,
+      "learning_rate": 2.4841936774709885e-05,
+      "loss": 0.5375,
+      "step": 12587
+    },
+    {
+      "epoch": 16.11264,
+      "grad_norm": 0.9746212959289551,
+      "learning_rate": 2.4839935974389757e-05,
+      "loss": 0.4938,
+      "step": 12588
+    },
+    {
+      "epoch": 16.11392,
+      "grad_norm": 0.9631856679916382,
+      "learning_rate": 2.483793517406963e-05,
+      "loss": 0.4965,
+      "step": 12589
+    },
+    {
+      "epoch": 16.1152,
+      "grad_norm": 0.99147629737854,
+      "learning_rate": 2.48359343737495e-05,
+      "loss": 0.5078,
+      "step": 12590
+    },
+    {
+      "epoch": 16.11648,
+      "grad_norm": 1.013726830482483,
+      "learning_rate": 2.4833933573429375e-05,
+      "loss": 0.5205,
+      "step": 12591
+    },
+    {
+      "epoch": 16.11776,
+      "grad_norm": 1.0126923322677612,
+      "learning_rate": 2.4831932773109244e-05,
+      "loss": 0.47,
+      "step": 12592
+    },
+    {
+      "epoch": 16.11904,
+      "grad_norm": 0.9509507417678833,
+      "learning_rate": 2.4829931972789116e-05,
+      "loss": 0.5231,
+      "step": 12593
+    },
+    {
+      "epoch": 16.12032,
+      "grad_norm": 1.1204897165298462,
+      "learning_rate": 2.4827931172468988e-05,
+      "loss": 0.537,
+      "step": 12594
+    },
+    {
+      "epoch": 16.1216,
+      "grad_norm": 1.0702080726623535,
+      "learning_rate": 2.4825930372148863e-05,
+      "loss": 0.5298,
+      "step": 12595
+    },
+    {
+      "epoch": 16.12288,
+      "grad_norm": 0.982305645942688,
+      "learning_rate": 2.482392957182873e-05,
+      "loss": 0.5061,
+      "step": 12596
+    },
+    {
+      "epoch": 16.12416,
+      "grad_norm": 0.9869879484176636,
+      "learning_rate": 2.4821928771508603e-05,
+      "loss": 0.5259,
+      "step": 12597
+    },
+    {
+      "epoch": 16.12544,
+      "grad_norm": 1.026171088218689,
+      "learning_rate": 2.4819927971188475e-05,
+      "loss": 0.4958,
+      "step": 12598
+    },
+    {
+      "epoch": 16.12672,
+      "grad_norm": 0.9694840312004089,
+      "learning_rate": 2.481792717086835e-05,
+      "loss": 0.5093,
+      "step": 12599
+    },
+    {
+      "epoch": 16.128,
+      "grad_norm": 1.0091004371643066,
+      "learning_rate": 2.481592637054822e-05,
+      "loss": 0.4943,
+      "step": 12600
+    },
+    {
+      "epoch": 16.12928,
+      "grad_norm": 1.0162972211837769,
+      "learning_rate": 2.481392557022809e-05,
+      "loss": 0.4813,
+      "step": 12601
+    },
+    {
+      "epoch": 16.13056,
+      "grad_norm": 1.0108006000518799,
+      "learning_rate": 2.4811924769907966e-05,
+      "loss": 0.5231,
+      "step": 12602
+    },
+    {
+      "epoch": 16.13184,
+      "grad_norm": 1.009252667427063,
+      "learning_rate": 2.4809923969587838e-05,
+      "loss": 0.5657,
+      "step": 12603
+    },
+    {
+      "epoch": 16.13312,
+      "grad_norm": 0.9438086748123169,
+      "learning_rate": 2.4807923169267706e-05,
+      "loss": 0.4868,
+      "step": 12604
+    },
+    {
+      "epoch": 16.1344,
+      "grad_norm": 0.9571229815483093,
+      "learning_rate": 2.4805922368947578e-05,
+      "loss": 0.4877,
+      "step": 12605
+    },
+    {
+      "epoch": 16.13568,
+      "grad_norm": 1.0769860744476318,
+      "learning_rate": 2.4803921568627453e-05,
+      "loss": 0.5616,
+      "step": 12606
+    },
+    {
+      "epoch": 16.13696,
+      "grad_norm": 0.9902837872505188,
+      "learning_rate": 2.4801920768307325e-05,
+      "loss": 0.5381,
+      "step": 12607
+    },
+    {
+      "epoch": 16.13824,
+      "grad_norm": 0.9671089053153992,
+      "learning_rate": 2.4799919967987194e-05,
+      "loss": 0.5305,
+      "step": 12608
+    },
+    {
+      "epoch": 16.13952,
+      "grad_norm": 1.058725118637085,
+      "learning_rate": 2.479791916766707e-05,
+      "loss": 0.4747,
+      "step": 12609
+    },
+    {
+      "epoch": 16.1408,
+      "grad_norm": 1.0563637018203735,
+      "learning_rate": 2.479591836734694e-05,
+      "loss": 0.5424,
+      "step": 12610
+    },
+    {
+      "epoch": 16.14208,
+      "grad_norm": 0.9643771052360535,
+      "learning_rate": 2.4793917567026813e-05,
+      "loss": 0.4893,
+      "step": 12611
+    },
+    {
+      "epoch": 16.14336,
+      "grad_norm": 1.021338939666748,
+      "learning_rate": 2.479191676670668e-05,
+      "loss": 0.5032,
+      "step": 12612
+    },
+    {
+      "epoch": 16.14464,
+      "grad_norm": 0.9763515591621399,
+      "learning_rate": 2.4789915966386556e-05,
+      "loss": 0.5038,
+      "step": 12613
+    },
+    {
+      "epoch": 16.14592,
+      "grad_norm": 1.0414294004440308,
+      "learning_rate": 2.4787915166066428e-05,
+      "loss": 0.4937,
+      "step": 12614
+    },
+    {
+      "epoch": 16.1472,
+      "grad_norm": 1.0470645427703857,
+      "learning_rate": 2.47859143657463e-05,
+      "loss": 0.5073,
+      "step": 12615
+    },
+    {
+      "epoch": 16.14848,
+      "grad_norm": 0.9787927269935608,
+      "learning_rate": 2.4783913565426172e-05,
+      "loss": 0.517,
+      "step": 12616
+    },
+    {
+      "epoch": 16.14976,
+      "grad_norm": 1.0010355710983276,
+      "learning_rate": 2.4781912765106044e-05,
+      "loss": 0.4683,
+      "step": 12617
+    },
+    {
+      "epoch": 16.15104,
+      "grad_norm": 0.978118896484375,
+      "learning_rate": 2.4779911964785916e-05,
+      "loss": 0.4818,
+      "step": 12618
+    },
+    {
+      "epoch": 16.15232,
+      "grad_norm": 0.9748190641403198,
+      "learning_rate": 2.4777911164465787e-05,
+      "loss": 0.5617,
+      "step": 12619
+    },
+    {
+      "epoch": 16.1536,
+      "grad_norm": 0.9519315361976624,
+      "learning_rate": 2.477591036414566e-05,
+      "loss": 0.4726,
+      "step": 12620
+    },
+    {
+      "epoch": 16.15488,
+      "grad_norm": 1.0163748264312744,
+      "learning_rate": 2.477390956382553e-05,
+      "loss": 0.5284,
+      "step": 12621
+    },
+    {
+      "epoch": 16.15616,
+      "grad_norm": 1.035246729850769,
+      "learning_rate": 2.4771908763505403e-05,
+      "loss": 0.5376,
+      "step": 12622
+    },
+    {
+      "epoch": 16.15744,
+      "grad_norm": 1.0474730730056763,
+      "learning_rate": 2.4769907963185275e-05,
+      "loss": 0.5061,
+      "step": 12623
+    },
+    {
+      "epoch": 16.15872,
+      "grad_norm": 0.9941502809524536,
+      "learning_rate": 2.4767907162865147e-05,
+      "loss": 0.5582,
+      "step": 12624
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 1.0640026330947876,
+      "learning_rate": 2.476590636254502e-05,
+      "loss": 0.513,
+      "step": 12625
+    },
+    {
+      "epoch": 16.16128,
+      "grad_norm": 1.030907154083252,
+      "learning_rate": 2.476390556222489e-05,
+      "loss": 0.4757,
+      "step": 12626
+    },
+    {
+      "epoch": 16.16256,
+      "grad_norm": 0.996875524520874,
+      "learning_rate": 2.4761904761904762e-05,
+      "loss": 0.4995,
+      "step": 12627
+    },
+    {
+      "epoch": 16.16384,
+      "grad_norm": 0.9607611894607544,
+      "learning_rate": 2.4759903961584634e-05,
+      "loss": 0.4947,
+      "step": 12628
+    },
+    {
+      "epoch": 16.16512,
+      "grad_norm": 0.9883370995521545,
+      "learning_rate": 2.4757903161264506e-05,
+      "loss": 0.5161,
+      "step": 12629
+    },
+    {
+      "epoch": 16.1664,
+      "grad_norm": 1.0071051120758057,
+      "learning_rate": 2.475590236094438e-05,
+      "loss": 0.4924,
+      "step": 12630
+    },
+    {
+      "epoch": 16.16768,
+      "grad_norm": 1.0023144483566284,
+      "learning_rate": 2.475390156062425e-05,
+      "loss": 0.4846,
+      "step": 12631
+    },
+    {
+      "epoch": 16.16896,
+      "grad_norm": 1.0280662775039673,
+      "learning_rate": 2.475190076030412e-05,
+      "loss": 0.5245,
+      "step": 12632
+    },
+    {
+      "epoch": 16.17024,
+      "grad_norm": 1.0610262155532837,
+      "learning_rate": 2.4749899959983993e-05,
+      "loss": 0.5165,
+      "step": 12633
+    },
+    {
+      "epoch": 16.17152,
+      "grad_norm": 0.9850628972053528,
+      "learning_rate": 2.474789915966387e-05,
+      "loss": 0.4994,
+      "step": 12634
+    },
+    {
+      "epoch": 16.1728,
+      "grad_norm": 1.0291600227355957,
+      "learning_rate": 2.4745898359343737e-05,
+      "loss": 0.5098,
+      "step": 12635
+    },
+    {
+      "epoch": 16.17408,
+      "grad_norm": 0.9818286895751953,
+      "learning_rate": 2.474389755902361e-05,
+      "loss": 0.5086,
+      "step": 12636
+    },
+    {
+      "epoch": 16.17536,
+      "grad_norm": 0.9875121712684631,
+      "learning_rate": 2.4741896758703484e-05,
+      "loss": 0.475,
+      "step": 12637
+    },
+    {
+      "epoch": 16.17664,
+      "grad_norm": 0.9203592538833618,
+      "learning_rate": 2.4739895958383356e-05,
+      "loss": 0.4558,
+      "step": 12638
+    },
+    {
+      "epoch": 16.17792,
+      "grad_norm": 0.9402249455451965,
+      "learning_rate": 2.4737895158063225e-05,
+      "loss": 0.4928,
+      "step": 12639
+    },
+    {
+      "epoch": 16.1792,
+      "grad_norm": 0.9855098724365234,
+      "learning_rate": 2.4735894357743096e-05,
+      "loss": 0.4993,
+      "step": 12640
+    },
+    {
+      "epoch": 16.18048,
+      "grad_norm": 0.9504002332687378,
+      "learning_rate": 2.4733893557422972e-05,
+      "loss": 0.5067,
+      "step": 12641
+    },
+    {
+      "epoch": 16.18176,
+      "grad_norm": 1.0600100755691528,
+      "learning_rate": 2.4731892757102844e-05,
+      "loss": 0.5632,
+      "step": 12642
+    },
+    {
+      "epoch": 16.18304,
+      "grad_norm": 1.0117945671081543,
+      "learning_rate": 2.4729891956782712e-05,
+      "loss": 0.5322,
+      "step": 12643
+    },
+    {
+      "epoch": 16.18432,
+      "grad_norm": 0.9786906242370605,
+      "learning_rate": 2.4727891156462587e-05,
+      "loss": 0.4928,
+      "step": 12644
+    },
+    {
+      "epoch": 16.1856,
+      "grad_norm": 0.9911783337593079,
+      "learning_rate": 2.472589035614246e-05,
+      "loss": 0.4789,
+      "step": 12645
+    },
+    {
+      "epoch": 16.18688,
+      "grad_norm": 1.0481369495391846,
+      "learning_rate": 2.472388955582233e-05,
+      "loss": 0.5374,
+      "step": 12646
+    },
+    {
+      "epoch": 16.18816,
+      "grad_norm": 0.9751612544059753,
+      "learning_rate": 2.47218887555022e-05,
+      "loss": 0.4826,
+      "step": 12647
+    },
+    {
+      "epoch": 16.18944,
+      "grad_norm": 1.0269287824630737,
+      "learning_rate": 2.4719887955182075e-05,
+      "loss": 0.5436,
+      "step": 12648
+    },
+    {
+      "epoch": 16.19072,
+      "grad_norm": 0.966488242149353,
+      "learning_rate": 2.4717887154861947e-05,
+      "loss": 0.4886,
+      "step": 12649
+    },
+    {
+      "epoch": 16.192,
+      "grad_norm": 1.0197010040283203,
+      "learning_rate": 2.471588635454182e-05,
+      "loss": 0.5519,
+      "step": 12650
+    },
+    {
+      "epoch": 16.19328,
+      "grad_norm": 0.9842081665992737,
+      "learning_rate": 2.4713885554221687e-05,
+      "loss": 0.5198,
+      "step": 12651
+    },
+    {
+      "epoch": 16.19456,
+      "grad_norm": 1.0210785865783691,
+      "learning_rate": 2.4711884753901562e-05,
+      "loss": 0.4755,
+      "step": 12652
+    },
+    {
+      "epoch": 16.19584,
+      "grad_norm": 1.0439190864562988,
+      "learning_rate": 2.4709883953581434e-05,
+      "loss": 0.4878,
+      "step": 12653
+    },
+    {
+      "epoch": 16.19712,
+      "grad_norm": 0.9892864227294922,
+      "learning_rate": 2.4707883153261306e-05,
+      "loss": 0.4941,
+      "step": 12654
+    },
+    {
+      "epoch": 16.1984,
+      "grad_norm": 1.1083271503448486,
+      "learning_rate": 2.4705882352941178e-05,
+      "loss": 0.5406,
+      "step": 12655
+    },
+    {
+      "epoch": 16.19968,
+      "grad_norm": 1.0279945135116577,
+      "learning_rate": 2.470388155262105e-05,
+      "loss": 0.5296,
+      "step": 12656
+    },
+    {
+      "epoch": 16.20096,
+      "grad_norm": 0.9678087830543518,
+      "learning_rate": 2.470188075230092e-05,
+      "loss": 0.4879,
+      "step": 12657
+    },
+    {
+      "epoch": 16.20224,
+      "grad_norm": 1.0200779438018799,
+      "learning_rate": 2.4699879951980793e-05,
+      "loss": 0.5342,
+      "step": 12658
+    },
+    {
+      "epoch": 16.20352,
+      "grad_norm": 1.0152758359909058,
+      "learning_rate": 2.4697879151660665e-05,
+      "loss": 0.5535,
+      "step": 12659
+    },
+    {
+      "epoch": 16.2048,
+      "grad_norm": 1.0245957374572754,
+      "learning_rate": 2.4695878351340537e-05,
+      "loss": 0.5309,
+      "step": 12660
+    },
+    {
+      "epoch": 16.20608,
+      "grad_norm": 1.054991364479065,
+      "learning_rate": 2.469387755102041e-05,
+      "loss": 0.5356,
+      "step": 12661
+    },
+    {
+      "epoch": 16.20736,
+      "grad_norm": 1.0215634107589722,
+      "learning_rate": 2.4691876750700284e-05,
+      "loss": 0.5165,
+      "step": 12662
+    },
+    {
+      "epoch": 16.20864,
+      "grad_norm": 0.9687957763671875,
+      "learning_rate": 2.4689875950380153e-05,
+      "loss": 0.4773,
+      "step": 12663
+    },
+    {
+      "epoch": 16.20992,
+      "grad_norm": 1.0658154487609863,
+      "learning_rate": 2.4687875150060024e-05,
+      "loss": 0.582,
+      "step": 12664
+    },
+    {
+      "epoch": 16.2112,
+      "grad_norm": 1.0069797039031982,
+      "learning_rate": 2.4685874349739896e-05,
+      "loss": 0.5316,
+      "step": 12665
+    },
+    {
+      "epoch": 16.21248,
+      "grad_norm": 1.0188541412353516,
+      "learning_rate": 2.468387354941977e-05,
+      "loss": 0.5182,
+      "step": 12666
+    },
+    {
+      "epoch": 16.21376,
+      "grad_norm": 1.0730429887771606,
+      "learning_rate": 2.468187274909964e-05,
+      "loss": 0.5226,
+      "step": 12667
+    },
+    {
+      "epoch": 16.21504,
+      "grad_norm": 1.0113489627838135,
+      "learning_rate": 2.4679871948779512e-05,
+      "loss": 0.4887,
+      "step": 12668
+    },
+    {
+      "epoch": 16.21632,
+      "grad_norm": 0.9426849484443665,
+      "learning_rate": 2.4677871148459387e-05,
+      "loss": 0.4777,
+      "step": 12669
+    },
+    {
+      "epoch": 16.2176,
+      "grad_norm": 0.9736514687538147,
+      "learning_rate": 2.467587034813926e-05,
+      "loss": 0.4941,
+      "step": 12670
+    },
+    {
+      "epoch": 16.21888,
+      "grad_norm": 1.0153673887252808,
+      "learning_rate": 2.4673869547819127e-05,
+      "loss": 0.5524,
+      "step": 12671
+    },
+    {
+      "epoch": 16.22016,
+      "grad_norm": 1.018025279045105,
+      "learning_rate": 2.4671868747499e-05,
+      "loss": 0.4976,
+      "step": 12672
+    },
+    {
+      "epoch": 16.22144,
+      "grad_norm": 0.9701156616210938,
+      "learning_rate": 2.4669867947178875e-05,
+      "loss": 0.4712,
+      "step": 12673
+    },
+    {
+      "epoch": 16.22272,
+      "grad_norm": 1.0330697298049927,
+      "learning_rate": 2.4667867146858746e-05,
+      "loss": 0.4938,
+      "step": 12674
+    },
+    {
+      "epoch": 16.224,
+      "grad_norm": 0.9708216190338135,
+      "learning_rate": 2.4665866346538615e-05,
+      "loss": 0.4665,
+      "step": 12675
+    },
+    {
+      "epoch": 16.22528,
+      "grad_norm": 0.9829319715499878,
+      "learning_rate": 2.466386554621849e-05,
+      "loss": 0.5134,
+      "step": 12676
+    },
+    {
+      "epoch": 16.22656,
+      "grad_norm": 1.0143637657165527,
+      "learning_rate": 2.4661864745898362e-05,
+      "loss": 0.5448,
+      "step": 12677
+    },
+    {
+      "epoch": 16.22784,
+      "grad_norm": 1.0487371683120728,
+      "learning_rate": 2.4659863945578234e-05,
+      "loss": 0.5299,
+      "step": 12678
+    },
+    {
+      "epoch": 16.22912,
+      "grad_norm": 0.9548150300979614,
+      "learning_rate": 2.4657863145258102e-05,
+      "loss": 0.4779,
+      "step": 12679
+    },
+    {
+      "epoch": 16.2304,
+      "grad_norm": 1.0021804571151733,
+      "learning_rate": 2.4655862344937978e-05,
+      "loss": 0.497,
+      "step": 12680
+    },
+    {
+      "epoch": 16.23168,
+      "grad_norm": 0.9995138049125671,
+      "learning_rate": 2.465386154461785e-05,
+      "loss": 0.5175,
+      "step": 12681
+    },
+    {
+      "epoch": 16.23296,
+      "grad_norm": 1.010904312133789,
+      "learning_rate": 2.465186074429772e-05,
+      "loss": 0.5227,
+      "step": 12682
+    },
+    {
+      "epoch": 16.23424,
+      "grad_norm": 0.9441868662834167,
+      "learning_rate": 2.4649859943977593e-05,
+      "loss": 0.5053,
+      "step": 12683
+    },
+    {
+      "epoch": 16.23552,
+      "grad_norm": 1.0118534564971924,
+      "learning_rate": 2.4647859143657465e-05,
+      "loss": 0.5208,
+      "step": 12684
+    },
+    {
+      "epoch": 16.2368,
+      "grad_norm": 1.0776724815368652,
+      "learning_rate": 2.4645858343337337e-05,
+      "loss": 0.5097,
+      "step": 12685
+    },
+    {
+      "epoch": 16.23808,
+      "grad_norm": 1.064697504043579,
+      "learning_rate": 2.464385754301721e-05,
+      "loss": 0.5166,
+      "step": 12686
+    },
+    {
+      "epoch": 16.23936,
+      "grad_norm": 1.0309760570526123,
+      "learning_rate": 2.464185674269708e-05,
+      "loss": 0.5428,
+      "step": 12687
+    },
+    {
+      "epoch": 16.24064,
+      "grad_norm": 0.9784319400787354,
+      "learning_rate": 2.4639855942376952e-05,
+      "loss": 0.4882,
+      "step": 12688
+    },
+    {
+      "epoch": 16.24192,
+      "grad_norm": 1.0319559574127197,
+      "learning_rate": 2.4637855142056824e-05,
+      "loss": 0.527,
+      "step": 12689
+    },
+    {
+      "epoch": 16.2432,
+      "grad_norm": 0.991915762424469,
+      "learning_rate": 2.4635854341736696e-05,
+      "loss": 0.523,
+      "step": 12690
+    },
+    {
+      "epoch": 16.24448,
+      "grad_norm": 1.043216586112976,
+      "learning_rate": 2.4633853541416568e-05,
+      "loss": 0.4977,
+      "step": 12691
+    },
+    {
+      "epoch": 16.24576,
+      "grad_norm": 1.063425064086914,
+      "learning_rate": 2.463185274109644e-05,
+      "loss": 0.5245,
+      "step": 12692
+    },
+    {
+      "epoch": 16.24704,
+      "grad_norm": 1.0236256122589111,
+      "learning_rate": 2.462985194077631e-05,
+      "loss": 0.5112,
+      "step": 12693
+    },
+    {
+      "epoch": 16.24832,
+      "grad_norm": 0.992674708366394,
+      "learning_rate": 2.4627851140456183e-05,
+      "loss": 0.5062,
+      "step": 12694
+    },
+    {
+      "epoch": 16.2496,
+      "grad_norm": 1.0538854598999023,
+      "learning_rate": 2.4625850340136055e-05,
+      "loss": 0.5244,
+      "step": 12695
+    },
+    {
+      "epoch": 16.25088,
+      "grad_norm": 1.021213412284851,
+      "learning_rate": 2.4623849539815927e-05,
+      "loss": 0.5097,
+      "step": 12696
+    },
+    {
+      "epoch": 16.25216,
+      "grad_norm": 1.0427947044372559,
+      "learning_rate": 2.4621848739495802e-05,
+      "loss": 0.5288,
+      "step": 12697
+    },
+    {
+      "epoch": 16.25344,
+      "grad_norm": 1.079925537109375,
+      "learning_rate": 2.461984793917567e-05,
+      "loss": 0.511,
+      "step": 12698
+    },
+    {
+      "epoch": 16.25472,
+      "grad_norm": 1.0293481349945068,
+      "learning_rate": 2.4617847138855543e-05,
+      "loss": 0.534,
+      "step": 12699
+    },
+    {
+      "epoch": 16.256,
+      "grad_norm": 1.0903656482696533,
+      "learning_rate": 2.4615846338535415e-05,
+      "loss": 0.5344,
+      "step": 12700
+    },
+    {
+      "epoch": 16.25728,
+      "grad_norm": 0.9660569429397583,
+      "learning_rate": 2.461384553821529e-05,
+      "loss": 0.507,
+      "step": 12701
+    },
+    {
+      "epoch": 16.25856,
+      "grad_norm": 0.9629020094871521,
+      "learning_rate": 2.461184473789516e-05,
+      "loss": 0.4976,
+      "step": 12702
+    },
+    {
+      "epoch": 16.25984,
+      "grad_norm": 0.9647486209869385,
+      "learning_rate": 2.460984393757503e-05,
+      "loss": 0.504,
+      "step": 12703
+    },
+    {
+      "epoch": 16.26112,
+      "grad_norm": 0.996180534362793,
+      "learning_rate": 2.4607843137254902e-05,
+      "loss": 0.5255,
+      "step": 12704
+    },
+    {
+      "epoch": 16.2624,
+      "grad_norm": 0.9972898960113525,
+      "learning_rate": 2.4605842336934777e-05,
+      "loss": 0.5179,
+      "step": 12705
+    },
+    {
+      "epoch": 16.26368,
+      "grad_norm": 0.97672438621521,
+      "learning_rate": 2.4603841536614646e-05,
+      "loss": 0.4798,
+      "step": 12706
+    },
+    {
+      "epoch": 16.26496,
+      "grad_norm": 1.013742446899414,
+      "learning_rate": 2.4601840736294518e-05,
+      "loss": 0.542,
+      "step": 12707
+    },
+    {
+      "epoch": 16.26624,
+      "grad_norm": 1.0334280729293823,
+      "learning_rate": 2.4599839935974393e-05,
+      "loss": 0.517,
+      "step": 12708
+    },
+    {
+      "epoch": 16.26752,
+      "grad_norm": 0.966270923614502,
+      "learning_rate": 2.4597839135654265e-05,
+      "loss": 0.4617,
+      "step": 12709
+    },
+    {
+      "epoch": 16.2688,
+      "grad_norm": 1.0279269218444824,
+      "learning_rate": 2.4595838335334133e-05,
+      "loss": 0.5156,
+      "step": 12710
+    },
+    {
+      "epoch": 16.27008,
+      "grad_norm": 0.993587851524353,
+      "learning_rate": 2.4593837535014005e-05,
+      "loss": 0.5515,
+      "step": 12711
+    },
+    {
+      "epoch": 16.27136,
+      "grad_norm": 1.055559515953064,
+      "learning_rate": 2.459183673469388e-05,
+      "loss": 0.5437,
+      "step": 12712
+    },
+    {
+      "epoch": 16.27264,
+      "grad_norm": 1.0224788188934326,
+      "learning_rate": 2.4589835934373752e-05,
+      "loss": 0.5237,
+      "step": 12713
+    },
+    {
+      "epoch": 16.27392,
+      "grad_norm": 1.0352107286453247,
+      "learning_rate": 2.458783513405362e-05,
+      "loss": 0.518,
+      "step": 12714
+    },
+    {
+      "epoch": 16.2752,
+      "grad_norm": 1.0856186151504517,
+      "learning_rate": 2.4585834333733496e-05,
+      "loss": 0.5168,
+      "step": 12715
+    },
+    {
+      "epoch": 16.27648,
+      "grad_norm": 1.0590870380401611,
+      "learning_rate": 2.4583833533413368e-05,
+      "loss": 0.517,
+      "step": 12716
+    },
+    {
+      "epoch": 16.27776,
+      "grad_norm": 1.0563161373138428,
+      "learning_rate": 2.458183273309324e-05,
+      "loss": 0.5362,
+      "step": 12717
+    },
+    {
+      "epoch": 16.27904,
+      "grad_norm": 1.0422947406768799,
+      "learning_rate": 2.4579831932773108e-05,
+      "loss": 0.5014,
+      "step": 12718
+    },
+    {
+      "epoch": 16.28032,
+      "grad_norm": 1.0459036827087402,
+      "learning_rate": 2.4577831132452983e-05,
+      "loss": 0.5276,
+      "step": 12719
+    },
+    {
+      "epoch": 16.2816,
+      "grad_norm": 1.057701587677002,
+      "learning_rate": 2.4575830332132855e-05,
+      "loss": 0.4813,
+      "step": 12720
+    },
+    {
+      "epoch": 16.28288,
+      "grad_norm": 1.0466841459274292,
+      "learning_rate": 2.4573829531812727e-05,
+      "loss": 0.5175,
+      "step": 12721
+    },
+    {
+      "epoch": 16.28416,
+      "grad_norm": 0.9757436513900757,
+      "learning_rate": 2.45718287314926e-05,
+      "loss": 0.4863,
+      "step": 12722
+    },
+    {
+      "epoch": 16.28544,
+      "grad_norm": 1.0128544569015503,
+      "learning_rate": 2.456982793117247e-05,
+      "loss": 0.507,
+      "step": 12723
+    },
+    {
+      "epoch": 16.28672,
+      "grad_norm": 0.9682973027229309,
+      "learning_rate": 2.4567827130852343e-05,
+      "loss": 0.5172,
+      "step": 12724
+    },
+    {
+      "epoch": 16.288,
+      "grad_norm": 0.9596603512763977,
+      "learning_rate": 2.4565826330532214e-05,
+      "loss": 0.522,
+      "step": 12725
+    },
+    {
+      "epoch": 16.28928,
+      "grad_norm": 0.9290403723716736,
+      "learning_rate": 2.4563825530212086e-05,
+      "loss": 0.472,
+      "step": 12726
+    },
+    {
+      "epoch": 16.29056,
+      "grad_norm": 1.052056908607483,
+      "learning_rate": 2.4561824729891958e-05,
+      "loss": 0.5415,
+      "step": 12727
+    },
+    {
+      "epoch": 16.29184,
+      "grad_norm": 1.0477123260498047,
+      "learning_rate": 2.455982392957183e-05,
+      "loss": 0.5393,
+      "step": 12728
+    },
+    {
+      "epoch": 16.29312,
+      "grad_norm": 0.9668442010879517,
+      "learning_rate": 2.4557823129251702e-05,
+      "loss": 0.4741,
+      "step": 12729
+    },
+    {
+      "epoch": 16.2944,
+      "grad_norm": 1.0437530279159546,
+      "learning_rate": 2.4555822328931574e-05,
+      "loss": 0.5,
+      "step": 12730
+    },
+    {
+      "epoch": 16.29568,
+      "grad_norm": 0.9850262999534607,
+      "learning_rate": 2.4553821528611446e-05,
+      "loss": 0.5108,
+      "step": 12731
+    },
+    {
+      "epoch": 16.29696,
+      "grad_norm": 0.9622449278831482,
+      "learning_rate": 2.4551820728291317e-05,
+      "loss": 0.506,
+      "step": 12732
+    },
+    {
+      "epoch": 16.29824,
+      "grad_norm": 1.0005285739898682,
+      "learning_rate": 2.454981992797119e-05,
+      "loss": 0.5083,
+      "step": 12733
+    },
+    {
+      "epoch": 16.29952,
+      "grad_norm": 0.9644542336463928,
+      "learning_rate": 2.454781912765106e-05,
+      "loss": 0.4986,
+      "step": 12734
+    },
+    {
+      "epoch": 16.3008,
+      "grad_norm": 1.0753077268600464,
+      "learning_rate": 2.4545818327330933e-05,
+      "loss": 0.5464,
+      "step": 12735
+    },
+    {
+      "epoch": 16.30208,
+      "grad_norm": 1.04744291305542,
+      "learning_rate": 2.4543817527010808e-05,
+      "loss": 0.5277,
+      "step": 12736
+    },
+    {
+      "epoch": 16.30336,
+      "grad_norm": 0.9650123119354248,
+      "learning_rate": 2.4541816726690677e-05,
+      "loss": 0.4648,
+      "step": 12737
+    },
+    {
+      "epoch": 16.30464,
+      "grad_norm": 0.9497844576835632,
+      "learning_rate": 2.453981592637055e-05,
+      "loss": 0.5344,
+      "step": 12738
+    },
+    {
+      "epoch": 16.30592,
+      "grad_norm": 1.0927493572235107,
+      "learning_rate": 2.453781512605042e-05,
+      "loss": 0.5106,
+      "step": 12739
+    },
+    {
+      "epoch": 16.3072,
+      "grad_norm": 1.0164883136749268,
+      "learning_rate": 2.4535814325730296e-05,
+      "loss": 0.5394,
+      "step": 12740
+    },
+    {
+      "epoch": 16.30848,
+      "grad_norm": 0.9408139586448669,
+      "learning_rate": 2.4533813525410164e-05,
+      "loss": 0.4882,
+      "step": 12741
+    },
+    {
+      "epoch": 16.30976,
+      "grad_norm": 0.9942525625228882,
+      "learning_rate": 2.4531812725090036e-05,
+      "loss": 0.5058,
+      "step": 12742
+    },
+    {
+      "epoch": 16.31104,
+      "grad_norm": 1.0041530132293701,
+      "learning_rate": 2.452981192476991e-05,
+      "loss": 0.4444,
+      "step": 12743
+    },
+    {
+      "epoch": 16.31232,
+      "grad_norm": 0.9434515833854675,
+      "learning_rate": 2.4527811124449783e-05,
+      "loss": 0.4969,
+      "step": 12744
+    },
+    {
+      "epoch": 16.3136,
+      "grad_norm": 1.0025871992111206,
+      "learning_rate": 2.452581032412965e-05,
+      "loss": 0.4901,
+      "step": 12745
+    },
+    {
+      "epoch": 16.31488,
+      "grad_norm": 0.9953147768974304,
+      "learning_rate": 2.4523809523809523e-05,
+      "loss": 0.512,
+      "step": 12746
+    },
+    {
+      "epoch": 16.31616,
+      "grad_norm": 1.1160242557525635,
+      "learning_rate": 2.45218087234894e-05,
+      "loss": 0.5995,
+      "step": 12747
+    },
+    {
+      "epoch": 16.31744,
+      "grad_norm": 0.9796026349067688,
+      "learning_rate": 2.451980792316927e-05,
+      "loss": 0.4865,
+      "step": 12748
+    },
+    {
+      "epoch": 16.31872,
+      "grad_norm": 1.0219937562942505,
+      "learning_rate": 2.451780712284914e-05,
+      "loss": 0.4902,
+      "step": 12749
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 1.036454200744629,
+      "learning_rate": 2.4515806322529014e-05,
+      "loss": 0.5686,
+      "step": 12750
+    },
+    {
+      "epoch": 16.32128,
+      "grad_norm": 0.9680700302124023,
+      "learning_rate": 2.4513805522208886e-05,
+      "loss": 0.5085,
+      "step": 12751
+    },
+    {
+      "epoch": 16.32256,
+      "grad_norm": 1.001165509223938,
+      "learning_rate": 2.4511804721888758e-05,
+      "loss": 0.5052,
+      "step": 12752
+    },
+    {
+      "epoch": 16.32384,
+      "grad_norm": 0.9691818356513977,
+      "learning_rate": 2.4509803921568626e-05,
+      "loss": 0.5176,
+      "step": 12753
+    },
+    {
+      "epoch": 16.32512,
+      "grad_norm": 0.9570626616477966,
+      "learning_rate": 2.45078031212485e-05,
+      "loss": 0.4953,
+      "step": 12754
+    },
+    {
+      "epoch": 16.3264,
+      "grad_norm": 1.052276849746704,
+      "learning_rate": 2.4505802320928374e-05,
+      "loss": 0.5756,
+      "step": 12755
+    },
+    {
+      "epoch": 16.32768,
+      "grad_norm": 0.9954919219017029,
+      "learning_rate": 2.4503801520608245e-05,
+      "loss": 0.4592,
+      "step": 12756
+    },
+    {
+      "epoch": 16.32896,
+      "grad_norm": 1.041548490524292,
+      "learning_rate": 2.4501800720288117e-05,
+      "loss": 0.4829,
+      "step": 12757
+    },
+    {
+      "epoch": 16.33024,
+      "grad_norm": 1.0538829565048218,
+      "learning_rate": 2.449979991996799e-05,
+      "loss": 0.5347,
+      "step": 12758
+    },
+    {
+      "epoch": 16.33152,
+      "grad_norm": 1.0308974981307983,
+      "learning_rate": 2.449779911964786e-05,
+      "loss": 0.5045,
+      "step": 12759
+    },
+    {
+      "epoch": 16.3328,
+      "grad_norm": 1.0308443307876587,
+      "learning_rate": 2.4495798319327733e-05,
+      "loss": 0.5121,
+      "step": 12760
+    },
+    {
+      "epoch": 16.33408,
+      "grad_norm": 1.0774739980697632,
+      "learning_rate": 2.4493797519007605e-05,
+      "loss": 0.5345,
+      "step": 12761
+    },
+    {
+      "epoch": 16.33536,
+      "grad_norm": 1.0473657846450806,
+      "learning_rate": 2.4491796718687477e-05,
+      "loss": 0.5251,
+      "step": 12762
+    },
+    {
+      "epoch": 16.33664,
+      "grad_norm": 1.0154544115066528,
+      "learning_rate": 2.448979591836735e-05,
+      "loss": 0.498,
+      "step": 12763
+    },
+    {
+      "epoch": 16.33792,
+      "grad_norm": 0.9398629069328308,
+      "learning_rate": 2.448779511804722e-05,
+      "loss": 0.493,
+      "step": 12764
+    },
+    {
+      "epoch": 16.3392,
+      "grad_norm": 1.0081340074539185,
+      "learning_rate": 2.4485794317727092e-05,
+      "loss": 0.5177,
+      "step": 12765
+    },
+    {
+      "epoch": 16.34048,
+      "grad_norm": 1.0168988704681396,
+      "learning_rate": 2.4483793517406964e-05,
+      "loss": 0.5382,
+      "step": 12766
+    },
+    {
+      "epoch": 16.34176,
+      "grad_norm": 1.0033296346664429,
+      "learning_rate": 2.4481792717086836e-05,
+      "loss": 0.5449,
+      "step": 12767
+    },
+    {
+      "epoch": 16.34304,
+      "grad_norm": 0.9844812154769897,
+      "learning_rate": 2.4479791916766708e-05,
+      "loss": 0.4916,
+      "step": 12768
+    },
+    {
+      "epoch": 16.34432,
+      "grad_norm": 1.042959213256836,
+      "learning_rate": 2.447779111644658e-05,
+      "loss": 0.4871,
+      "step": 12769
+    },
+    {
+      "epoch": 16.3456,
+      "grad_norm": 1.0168309211730957,
+      "learning_rate": 2.447579031612645e-05,
+      "loss": 0.4914,
+      "step": 12770
+    },
+    {
+      "epoch": 16.34688,
+      "grad_norm": 1.0198049545288086,
+      "learning_rate": 2.4473789515806323e-05,
+      "loss": 0.4982,
+      "step": 12771
+    },
+    {
+      "epoch": 16.34816,
+      "grad_norm": 1.0532821416854858,
+      "learning_rate": 2.4471788715486195e-05,
+      "loss": 0.5367,
+      "step": 12772
+    },
+    {
+      "epoch": 16.34944,
+      "grad_norm": 1.0325431823730469,
+      "learning_rate": 2.4469787915166067e-05,
+      "loss": 0.5169,
+      "step": 12773
+    },
+    {
+      "epoch": 16.35072,
+      "grad_norm": 0.9796453714370728,
+      "learning_rate": 2.446778711484594e-05,
+      "loss": 0.4951,
+      "step": 12774
+    },
+    {
+      "epoch": 16.352,
+      "grad_norm": 1.0543535947799683,
+      "learning_rate": 2.4465786314525814e-05,
+      "loss": 0.5832,
+      "step": 12775
+    },
+    {
+      "epoch": 16.35328,
+      "grad_norm": 1.0445586442947388,
+      "learning_rate": 2.4463785514205683e-05,
+      "loss": 0.5339,
+      "step": 12776
+    },
+    {
+      "epoch": 16.35456,
+      "grad_norm": 1.0602067708969116,
+      "learning_rate": 2.4461784713885554e-05,
+      "loss": 0.5414,
+      "step": 12777
+    },
+    {
+      "epoch": 16.35584,
+      "grad_norm": 1.0592375993728638,
+      "learning_rate": 2.4459783913565426e-05,
+      "loss": 0.5309,
+      "step": 12778
+    },
+    {
+      "epoch": 16.35712,
+      "grad_norm": 1.0167428255081177,
+      "learning_rate": 2.44577831132453e-05,
+      "loss": 0.5576,
+      "step": 12779
+    },
+    {
+      "epoch": 16.3584,
+      "grad_norm": 1.0088391304016113,
+      "learning_rate": 2.445578231292517e-05,
+      "loss": 0.4863,
+      "step": 12780
+    },
+    {
+      "epoch": 16.35968,
+      "grad_norm": 1.0060917139053345,
+      "learning_rate": 2.4453781512605042e-05,
+      "loss": 0.4916,
+      "step": 12781
+    },
+    {
+      "epoch": 16.36096,
+      "grad_norm": 0.9650666117668152,
+      "learning_rate": 2.4451780712284917e-05,
+      "loss": 0.483,
+      "step": 12782
+    },
+    {
+      "epoch": 16.36224,
+      "grad_norm": 1.0193016529083252,
+      "learning_rate": 2.444977991196479e-05,
+      "loss": 0.5077,
+      "step": 12783
+    },
+    {
+      "epoch": 16.36352,
+      "grad_norm": 1.00182044506073,
+      "learning_rate": 2.4447779111644657e-05,
+      "loss": 0.4533,
+      "step": 12784
+    },
+    {
+      "epoch": 16.3648,
+      "grad_norm": 1.016948938369751,
+      "learning_rate": 2.444577831132453e-05,
+      "loss": 0.4998,
+      "step": 12785
+    },
+    {
+      "epoch": 16.36608,
+      "grad_norm": 1.0034083127975464,
+      "learning_rate": 2.4443777511004404e-05,
+      "loss": 0.5126,
+      "step": 12786
+    },
+    {
+      "epoch": 16.36736,
+      "grad_norm": 1.0436108112335205,
+      "learning_rate": 2.4441776710684276e-05,
+      "loss": 0.503,
+      "step": 12787
+    },
+    {
+      "epoch": 16.36864,
+      "grad_norm": 1.076514482498169,
+      "learning_rate": 2.4439775910364145e-05,
+      "loss": 0.5608,
+      "step": 12788
+    },
+    {
+      "epoch": 16.36992,
+      "grad_norm": 1.050675868988037,
+      "learning_rate": 2.443777511004402e-05,
+      "loss": 0.5245,
+      "step": 12789
+    },
+    {
+      "epoch": 16.3712,
+      "grad_norm": 0.9627003073692322,
+      "learning_rate": 2.4435774309723892e-05,
+      "loss": 0.4655,
+      "step": 12790
+    },
+    {
+      "epoch": 16.37248,
+      "grad_norm": 1.0105419158935547,
+      "learning_rate": 2.4433773509403764e-05,
+      "loss": 0.5192,
+      "step": 12791
+    },
+    {
+      "epoch": 16.37376,
+      "grad_norm": 1.0531201362609863,
+      "learning_rate": 2.4431772709083632e-05,
+      "loss": 0.5213,
+      "step": 12792
+    },
+    {
+      "epoch": 16.37504,
+      "grad_norm": 1.0447622537612915,
+      "learning_rate": 2.4429771908763507e-05,
+      "loss": 0.4838,
+      "step": 12793
+    },
+    {
+      "epoch": 16.37632,
+      "grad_norm": 1.0392229557037354,
+      "learning_rate": 2.442777110844338e-05,
+      "loss": 0.5205,
+      "step": 12794
+    },
+    {
+      "epoch": 16.3776,
+      "grad_norm": 0.9850015640258789,
+      "learning_rate": 2.442577030812325e-05,
+      "loss": 0.5161,
+      "step": 12795
+    },
+    {
+      "epoch": 16.37888,
+      "grad_norm": 1.0995105504989624,
+      "learning_rate": 2.4423769507803123e-05,
+      "loss": 0.5643,
+      "step": 12796
+    },
+    {
+      "epoch": 16.38016,
+      "grad_norm": 1.0674548149108887,
+      "learning_rate": 2.4421768707482995e-05,
+      "loss": 0.5182,
+      "step": 12797
+    },
+    {
+      "epoch": 16.38144,
+      "grad_norm": 1.1218688488006592,
+      "learning_rate": 2.4419767907162867e-05,
+      "loss": 0.5739,
+      "step": 12798
+    },
+    {
+      "epoch": 16.38272,
+      "grad_norm": 0.980241060256958,
+      "learning_rate": 2.441776710684274e-05,
+      "loss": 0.5029,
+      "step": 12799
+    },
+    {
+      "epoch": 16.384,
+      "grad_norm": 1.0264852046966553,
+      "learning_rate": 2.441576630652261e-05,
+      "loss": 0.5188,
+      "step": 12800
+    },
+    {
+      "epoch": 16.38528,
+      "grad_norm": 1.005523920059204,
+      "learning_rate": 2.4413765506202482e-05,
+      "loss": 0.5254,
+      "step": 12801
+    },
+    {
+      "epoch": 16.38656,
+      "grad_norm": 1.0361162424087524,
+      "learning_rate": 2.4411764705882354e-05,
+      "loss": 0.4976,
+      "step": 12802
+    },
+    {
+      "epoch": 16.38784,
+      "grad_norm": 1.035775899887085,
+      "learning_rate": 2.4409763905562226e-05,
+      "loss": 0.5238,
+      "step": 12803
+    },
+    {
+      "epoch": 16.38912,
+      "grad_norm": 1.037307858467102,
+      "learning_rate": 2.4407763105242098e-05,
+      "loss": 0.5143,
+      "step": 12804
+    },
+    {
+      "epoch": 16.3904,
+      "grad_norm": 0.9811749458312988,
+      "learning_rate": 2.440576230492197e-05,
+      "loss": 0.5074,
+      "step": 12805
+    },
+    {
+      "epoch": 16.39168,
+      "grad_norm": 1.11452317237854,
+      "learning_rate": 2.440376150460184e-05,
+      "loss": 0.4935,
+      "step": 12806
+    },
+    {
+      "epoch": 16.39296,
+      "grad_norm": 0.9943965077400208,
+      "learning_rate": 2.4401760704281713e-05,
+      "loss": 0.5208,
+      "step": 12807
+    },
+    {
+      "epoch": 16.39424,
+      "grad_norm": 1.129791259765625,
+      "learning_rate": 2.4399759903961585e-05,
+      "loss": 0.5443,
+      "step": 12808
+    },
+    {
+      "epoch": 16.39552,
+      "grad_norm": 1.0286704301834106,
+      "learning_rate": 2.4397759103641457e-05,
+      "loss": 0.5206,
+      "step": 12809
+    },
+    {
+      "epoch": 16.3968,
+      "grad_norm": 1.0669382810592651,
+      "learning_rate": 2.4395758303321332e-05,
+      "loss": 0.5445,
+      "step": 12810
+    },
+    {
+      "epoch": 16.39808,
+      "grad_norm": 1.0054514408111572,
+      "learning_rate": 2.43937575030012e-05,
+      "loss": 0.5141,
+      "step": 12811
+    },
+    {
+      "epoch": 16.39936,
+      "grad_norm": 1.0013278722763062,
+      "learning_rate": 2.4391756702681073e-05,
+      "loss": 0.5338,
+      "step": 12812
+    },
+    {
+      "epoch": 16.40064,
+      "grad_norm": 1.0158824920654297,
+      "learning_rate": 2.4389755902360945e-05,
+      "loss": 0.5281,
+      "step": 12813
+    },
+    {
+      "epoch": 16.40192,
+      "grad_norm": 0.9622918367385864,
+      "learning_rate": 2.438775510204082e-05,
+      "loss": 0.4957,
+      "step": 12814
+    },
+    {
+      "epoch": 16.4032,
+      "grad_norm": 1.0246028900146484,
+      "learning_rate": 2.438575430172069e-05,
+      "loss": 0.5386,
+      "step": 12815
+    },
+    {
+      "epoch": 16.40448,
+      "grad_norm": 1.0182970762252808,
+      "learning_rate": 2.438375350140056e-05,
+      "loss": 0.4567,
+      "step": 12816
+    },
+    {
+      "epoch": 16.40576,
+      "grad_norm": 1.0306717157363892,
+      "learning_rate": 2.4381752701080432e-05,
+      "loss": 0.574,
+      "step": 12817
+    },
+    {
+      "epoch": 16.40704,
+      "grad_norm": 1.003909707069397,
+      "learning_rate": 2.4379751900760307e-05,
+      "loss": 0.4996,
+      "step": 12818
+    },
+    {
+      "epoch": 16.40832,
+      "grad_norm": 1.0095973014831543,
+      "learning_rate": 2.4377751100440176e-05,
+      "loss": 0.5096,
+      "step": 12819
+    },
+    {
+      "epoch": 16.4096,
+      "grad_norm": 1.0749574899673462,
+      "learning_rate": 2.4375750300120048e-05,
+      "loss": 0.5673,
+      "step": 12820
+    },
+    {
+      "epoch": 16.41088,
+      "grad_norm": 1.0384900569915771,
+      "learning_rate": 2.4373749499799923e-05,
+      "loss": 0.4899,
+      "step": 12821
+    },
+    {
+      "epoch": 16.41216,
+      "grad_norm": 1.0096982717514038,
+      "learning_rate": 2.4371748699479795e-05,
+      "loss": 0.5459,
+      "step": 12822
+    },
+    {
+      "epoch": 16.41344,
+      "grad_norm": 0.9829618334770203,
+      "learning_rate": 2.4369747899159663e-05,
+      "loss": 0.5178,
+      "step": 12823
+    },
+    {
+      "epoch": 16.41472,
+      "grad_norm": 1.046731948852539,
+      "learning_rate": 2.4367747098839535e-05,
+      "loss": 0.5299,
+      "step": 12824
+    },
+    {
+      "epoch": 16.416,
+      "grad_norm": 1.0445644855499268,
+      "learning_rate": 2.436574629851941e-05,
+      "loss": 0.5256,
+      "step": 12825
+    },
+    {
+      "epoch": 16.41728,
+      "grad_norm": 1.0414682626724243,
+      "learning_rate": 2.4363745498199282e-05,
+      "loss": 0.5141,
+      "step": 12826
+    },
+    {
+      "epoch": 16.41856,
+      "grad_norm": 1.0373891592025757,
+      "learning_rate": 2.436174469787915e-05,
+      "loss": 0.4973,
+      "step": 12827
+    },
+    {
+      "epoch": 16.41984,
+      "grad_norm": 0.9886151552200317,
+      "learning_rate": 2.4359743897559026e-05,
+      "loss": 0.5466,
+      "step": 12828
+    },
+    {
+      "epoch": 16.42112,
+      "grad_norm": 1.0570114850997925,
+      "learning_rate": 2.4357743097238898e-05,
+      "loss": 0.5385,
+      "step": 12829
+    },
+    {
+      "epoch": 16.4224,
+      "grad_norm": 1.0138678550720215,
+      "learning_rate": 2.435574229691877e-05,
+      "loss": 0.5263,
+      "step": 12830
+    },
+    {
+      "epoch": 16.42368,
+      "grad_norm": 1.007300615310669,
+      "learning_rate": 2.4353741496598638e-05,
+      "loss": 0.5041,
+      "step": 12831
+    },
+    {
+      "epoch": 16.42496,
+      "grad_norm": 1.026974081993103,
+      "learning_rate": 2.4351740696278513e-05,
+      "loss": 0.521,
+      "step": 12832
+    },
+    {
+      "epoch": 16.42624,
+      "grad_norm": 0.9623899459838867,
+      "learning_rate": 2.4349739895958385e-05,
+      "loss": 0.4611,
+      "step": 12833
+    },
+    {
+      "epoch": 16.42752,
+      "grad_norm": 1.0570974349975586,
+      "learning_rate": 2.4347739095638257e-05,
+      "loss": 0.5223,
+      "step": 12834
+    },
+    {
+      "epoch": 16.4288,
+      "grad_norm": 0.989608883857727,
+      "learning_rate": 2.434573829531813e-05,
+      "loss": 0.4902,
+      "step": 12835
+    },
+    {
+      "epoch": 16.43008,
+      "grad_norm": 1.0434807538986206,
+      "learning_rate": 2.4343737494998e-05,
+      "loss": 0.5691,
+      "step": 12836
+    },
+    {
+      "epoch": 16.43136,
+      "grad_norm": 1.035138487815857,
+      "learning_rate": 2.4341736694677873e-05,
+      "loss": 0.5493,
+      "step": 12837
+    },
+    {
+      "epoch": 16.43264,
+      "grad_norm": 1.1263927221298218,
+      "learning_rate": 2.4339735894357744e-05,
+      "loss": 0.5908,
+      "step": 12838
+    },
+    {
+      "epoch": 16.43392,
+      "grad_norm": 1.0934780836105347,
+      "learning_rate": 2.4337735094037616e-05,
+      "loss": 0.5754,
+      "step": 12839
+    },
+    {
+      "epoch": 16.4352,
+      "grad_norm": 1.0011106729507446,
+      "learning_rate": 2.4335734293717488e-05,
+      "loss": 0.4959,
+      "step": 12840
+    },
+    {
+      "epoch": 16.43648,
+      "grad_norm": 1.0350295305252075,
+      "learning_rate": 2.433373349339736e-05,
+      "loss": 0.5492,
+      "step": 12841
+    },
+    {
+      "epoch": 16.43776,
+      "grad_norm": 1.044190526008606,
+      "learning_rate": 2.4331732693077232e-05,
+      "loss": 0.5304,
+      "step": 12842
+    },
+    {
+      "epoch": 16.43904,
+      "grad_norm": 0.9419692158699036,
+      "learning_rate": 2.4329731892757104e-05,
+      "loss": 0.4753,
+      "step": 12843
+    },
+    {
+      "epoch": 16.44032,
+      "grad_norm": 0.9518587589263916,
+      "learning_rate": 2.4327731092436976e-05,
+      "loss": 0.4901,
+      "step": 12844
+    },
+    {
+      "epoch": 16.4416,
+      "grad_norm": 0.9863883852958679,
+      "learning_rate": 2.4325730292116847e-05,
+      "loss": 0.5236,
+      "step": 12845
+    },
+    {
+      "epoch": 16.44288,
+      "grad_norm": 1.0617551803588867,
+      "learning_rate": 2.432372949179672e-05,
+      "loss": 0.5631,
+      "step": 12846
+    },
+    {
+      "epoch": 16.44416,
+      "grad_norm": 1.0354115962982178,
+      "learning_rate": 2.432172869147659e-05,
+      "loss": 0.499,
+      "step": 12847
+    },
+    {
+      "epoch": 16.44544,
+      "grad_norm": 1.103115439414978,
+      "learning_rate": 2.4319727891156463e-05,
+      "loss": 0.5406,
+      "step": 12848
+    },
+    {
+      "epoch": 16.44672,
+      "grad_norm": 0.9908240437507629,
+      "learning_rate": 2.4317727090836338e-05,
+      "loss": 0.478,
+      "step": 12849
+    },
+    {
+      "epoch": 16.448,
+      "grad_norm": 1.0611495971679688,
+      "learning_rate": 2.4315726290516207e-05,
+      "loss": 0.5106,
+      "step": 12850
+    },
+    {
+      "epoch": 16.44928,
+      "grad_norm": 1.039263367652893,
+      "learning_rate": 2.431372549019608e-05,
+      "loss": 0.5012,
+      "step": 12851
+    },
+    {
+      "epoch": 16.45056,
+      "grad_norm": 1.0724480152130127,
+      "learning_rate": 2.431172468987595e-05,
+      "loss": 0.5161,
+      "step": 12852
+    },
+    {
+      "epoch": 16.45184,
+      "grad_norm": 1.0709526538848877,
+      "learning_rate": 2.4309723889555826e-05,
+      "loss": 0.4908,
+      "step": 12853
+    },
+    {
+      "epoch": 16.45312,
+      "grad_norm": 1.0412678718566895,
+      "learning_rate": 2.4307723089235694e-05,
+      "loss": 0.5438,
+      "step": 12854
+    },
+    {
+      "epoch": 16.4544,
+      "grad_norm": 0.9475101828575134,
+      "learning_rate": 2.4305722288915566e-05,
+      "loss": 0.4536,
+      "step": 12855
+    },
+    {
+      "epoch": 16.45568,
+      "grad_norm": 0.9541323781013489,
+      "learning_rate": 2.430372148859544e-05,
+      "loss": 0.5116,
+      "step": 12856
+    },
+    {
+      "epoch": 16.45696,
+      "grad_norm": 1.0412318706512451,
+      "learning_rate": 2.4301720688275313e-05,
+      "loss": 0.538,
+      "step": 12857
+    },
+    {
+      "epoch": 16.45824,
+      "grad_norm": 1.0488194227218628,
+      "learning_rate": 2.429971988795518e-05,
+      "loss": 0.5507,
+      "step": 12858
+    },
+    {
+      "epoch": 16.45952,
+      "grad_norm": 0.9851821660995483,
+      "learning_rate": 2.4297719087635053e-05,
+      "loss": 0.5014,
+      "step": 12859
+    },
+    {
+      "epoch": 16.4608,
+      "grad_norm": 0.9385454654693604,
+      "learning_rate": 2.429571828731493e-05,
+      "loss": 0.4872,
+      "step": 12860
+    },
+    {
+      "epoch": 16.46208,
+      "grad_norm": 1.0448015928268433,
+      "learning_rate": 2.42937174869948e-05,
+      "loss": 0.5212,
+      "step": 12861
+    },
+    {
+      "epoch": 16.46336,
+      "grad_norm": 0.9894904494285583,
+      "learning_rate": 2.429171668667467e-05,
+      "loss": 0.5181,
+      "step": 12862
+    },
+    {
+      "epoch": 16.46464,
+      "grad_norm": 1.0554355382919312,
+      "learning_rate": 2.4289715886354544e-05,
+      "loss": 0.557,
+      "step": 12863
+    },
+    {
+      "epoch": 16.46592,
+      "grad_norm": 1.055113434791565,
+      "learning_rate": 2.4287715086034416e-05,
+      "loss": 0.5408,
+      "step": 12864
+    },
+    {
+      "epoch": 16.4672,
+      "grad_norm": 0.9813659191131592,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 0.518,
+      "step": 12865
+    },
+    {
+      "epoch": 16.46848,
+      "grad_norm": 1.0583056211471558,
+      "learning_rate": 2.4283713485394156e-05,
+      "loss": 0.5171,
+      "step": 12866
+    },
+    {
+      "epoch": 16.46976,
+      "grad_norm": 0.9968990087509155,
+      "learning_rate": 2.428171268507403e-05,
+      "loss": 0.5158,
+      "step": 12867
+    },
+    {
+      "epoch": 16.47104,
+      "grad_norm": 0.9917585849761963,
+      "learning_rate": 2.4279711884753904e-05,
+      "loss": 0.5027,
+      "step": 12868
+    },
+    {
+      "epoch": 16.47232,
+      "grad_norm": 0.9894549250602722,
+      "learning_rate": 2.4277711084433775e-05,
+      "loss": 0.4815,
+      "step": 12869
+    },
+    {
+      "epoch": 16.4736,
+      "grad_norm": 1.0116682052612305,
+      "learning_rate": 2.4275710284113647e-05,
+      "loss": 0.5373,
+      "step": 12870
+    },
+    {
+      "epoch": 16.47488,
+      "grad_norm": 0.9584359526634216,
+      "learning_rate": 2.427370948379352e-05,
+      "loss": 0.4933,
+      "step": 12871
+    },
+    {
+      "epoch": 16.47616,
+      "grad_norm": 0.9404587745666504,
+      "learning_rate": 2.427170868347339e-05,
+      "loss": 0.4687,
+      "step": 12872
+    },
+    {
+      "epoch": 16.47744,
+      "grad_norm": 0.9797256588935852,
+      "learning_rate": 2.4269707883153263e-05,
+      "loss": 0.5015,
+      "step": 12873
+    },
+    {
+      "epoch": 16.47872,
+      "grad_norm": 1.0498969554901123,
+      "learning_rate": 2.4267707082833135e-05,
+      "loss": 0.5247,
+      "step": 12874
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 0.9874524474143982,
+      "learning_rate": 2.4265706282513007e-05,
+      "loss": 0.4865,
+      "step": 12875
+    },
+    {
+      "epoch": 16.48128,
+      "grad_norm": 0.9777351021766663,
+      "learning_rate": 2.426370548219288e-05,
+      "loss": 0.4579,
+      "step": 12876
+    },
+    {
+      "epoch": 16.48256,
+      "grad_norm": 1.0781558752059937,
+      "learning_rate": 2.426170468187275e-05,
+      "loss": 0.5586,
+      "step": 12877
+    },
+    {
+      "epoch": 16.48384,
+      "grad_norm": 1.029046893119812,
+      "learning_rate": 2.4259703881552622e-05,
+      "loss": 0.5054,
+      "step": 12878
+    },
+    {
+      "epoch": 16.48512,
+      "grad_norm": 0.9987738728523254,
+      "learning_rate": 2.4257703081232494e-05,
+      "loss": 0.4803,
+      "step": 12879
+    },
+    {
+      "epoch": 16.4864,
+      "grad_norm": 1.0528926849365234,
+      "learning_rate": 2.4255702280912366e-05,
+      "loss": 0.5177,
+      "step": 12880
+    },
+    {
+      "epoch": 16.48768,
+      "grad_norm": 1.0812759399414062,
+      "learning_rate": 2.4253701480592238e-05,
+      "loss": 0.5125,
+      "step": 12881
+    },
+    {
+      "epoch": 16.48896,
+      "grad_norm": 0.9843981266021729,
+      "learning_rate": 2.425170068027211e-05,
+      "loss": 0.4758,
+      "step": 12882
+    },
+    {
+      "epoch": 16.49024,
+      "grad_norm": 0.9296069145202637,
+      "learning_rate": 2.424969987995198e-05,
+      "loss": 0.5027,
+      "step": 12883
+    },
+    {
+      "epoch": 16.49152,
+      "grad_norm": 0.9559636116027832,
+      "learning_rate": 2.4247699079631853e-05,
+      "loss": 0.511,
+      "step": 12884
+    },
+    {
+      "epoch": 16.4928,
+      "grad_norm": 1.0256993770599365,
+      "learning_rate": 2.4245698279311725e-05,
+      "loss": 0.5539,
+      "step": 12885
+    },
+    {
+      "epoch": 16.49408,
+      "grad_norm": 1.0049717426300049,
+      "learning_rate": 2.4243697478991597e-05,
+      "loss": 0.5549,
+      "step": 12886
+    },
+    {
+      "epoch": 16.49536,
+      "grad_norm": 1.005260705947876,
+      "learning_rate": 2.424169667867147e-05,
+      "loss": 0.5164,
+      "step": 12887
+    },
+    {
+      "epoch": 16.49664,
+      "grad_norm": 0.9868449568748474,
+      "learning_rate": 2.4239695878351344e-05,
+      "loss": 0.5297,
+      "step": 12888
+    },
+    {
+      "epoch": 16.49792,
+      "grad_norm": 1.0604848861694336,
+      "learning_rate": 2.4237695078031213e-05,
+      "loss": 0.5506,
+      "step": 12889
+    },
+    {
+      "epoch": 16.4992,
+      "grad_norm": 1.0443841218948364,
+      "learning_rate": 2.4235694277711084e-05,
+      "loss": 0.5574,
+      "step": 12890
+    },
+    {
+      "epoch": 16.50048,
+      "grad_norm": 1.0032176971435547,
+      "learning_rate": 2.4233693477390956e-05,
+      "loss": 0.5244,
+      "step": 12891
+    },
+    {
+      "epoch": 16.50176,
+      "grad_norm": 1.0291366577148438,
+      "learning_rate": 2.423169267707083e-05,
+      "loss": 0.5231,
+      "step": 12892
+    },
+    {
+      "epoch": 16.50304,
+      "grad_norm": 1.0801621675491333,
+      "learning_rate": 2.42296918767507e-05,
+      "loss": 0.5645,
+      "step": 12893
+    },
+    {
+      "epoch": 16.50432,
+      "grad_norm": 1.0463597774505615,
+      "learning_rate": 2.4227691076430572e-05,
+      "loss": 0.4742,
+      "step": 12894
+    },
+    {
+      "epoch": 16.5056,
+      "grad_norm": 1.0148074626922607,
+      "learning_rate": 2.4225690276110447e-05,
+      "loss": 0.5203,
+      "step": 12895
+    },
+    {
+      "epoch": 16.50688,
+      "grad_norm": 0.9906325340270996,
+      "learning_rate": 2.422368947579032e-05,
+      "loss": 0.4931,
+      "step": 12896
+    },
+    {
+      "epoch": 16.50816,
+      "grad_norm": 0.9908265471458435,
+      "learning_rate": 2.4221688675470187e-05,
+      "loss": 0.5016,
+      "step": 12897
+    },
+    {
+      "epoch": 16.50944,
+      "grad_norm": 0.9555728435516357,
+      "learning_rate": 2.421968787515006e-05,
+      "loss": 0.4763,
+      "step": 12898
+    },
+    {
+      "epoch": 16.51072,
+      "grad_norm": 0.9900997281074524,
+      "learning_rate": 2.4217687074829934e-05,
+      "loss": 0.4907,
+      "step": 12899
+    },
+    {
+      "epoch": 16.512,
+      "grad_norm": 0.9736559391021729,
+      "learning_rate": 2.4215686274509806e-05,
+      "loss": 0.4814,
+      "step": 12900
+    },
+    {
+      "epoch": 16.51328,
+      "grad_norm": 1.0533136129379272,
+      "learning_rate": 2.4213685474189675e-05,
+      "loss": 0.5556,
+      "step": 12901
+    },
+    {
+      "epoch": 16.51456,
+      "grad_norm": 0.982058584690094,
+      "learning_rate": 2.421168467386955e-05,
+      "loss": 0.4757,
+      "step": 12902
+    },
+    {
+      "epoch": 16.51584,
+      "grad_norm": 1.0134705305099487,
+      "learning_rate": 2.4209683873549422e-05,
+      "loss": 0.5173,
+      "step": 12903
+    },
+    {
+      "epoch": 16.51712,
+      "grad_norm": 1.0609053373336792,
+      "learning_rate": 2.4207683073229294e-05,
+      "loss": 0.5624,
+      "step": 12904
+    },
+    {
+      "epoch": 16.5184,
+      "grad_norm": 1.0328656435012817,
+      "learning_rate": 2.4205682272909162e-05,
+      "loss": 0.5456,
+      "step": 12905
+    },
+    {
+      "epoch": 16.51968,
+      "grad_norm": 0.997097373008728,
+      "learning_rate": 2.4203681472589037e-05,
+      "loss": 0.4958,
+      "step": 12906
+    },
+    {
+      "epoch": 16.52096,
+      "grad_norm": 1.0784469842910767,
+      "learning_rate": 2.420168067226891e-05,
+      "loss": 0.4847,
+      "step": 12907
+    },
+    {
+      "epoch": 16.52224,
+      "grad_norm": 1.0922346115112305,
+      "learning_rate": 2.419967987194878e-05,
+      "loss": 0.5269,
+      "step": 12908
+    },
+    {
+      "epoch": 16.52352,
+      "grad_norm": 1.0385076999664307,
+      "learning_rate": 2.4197679071628653e-05,
+      "loss": 0.472,
+      "step": 12909
+    },
+    {
+      "epoch": 16.5248,
+      "grad_norm": 0.9775056838989258,
+      "learning_rate": 2.4195678271308525e-05,
+      "loss": 0.5025,
+      "step": 12910
+    },
+    {
+      "epoch": 16.52608,
+      "grad_norm": 1.0673224925994873,
+      "learning_rate": 2.4193677470988397e-05,
+      "loss": 0.5495,
+      "step": 12911
+    },
+    {
+      "epoch": 16.52736,
+      "grad_norm": 1.0127609968185425,
+      "learning_rate": 2.419167667066827e-05,
+      "loss": 0.5031,
+      "step": 12912
+    },
+    {
+      "epoch": 16.52864,
+      "grad_norm": 0.9412071108818054,
+      "learning_rate": 2.418967587034814e-05,
+      "loss": 0.4704,
+      "step": 12913
+    },
+    {
+      "epoch": 16.52992,
+      "grad_norm": 0.9896215200424194,
+      "learning_rate": 2.4187675070028012e-05,
+      "loss": 0.4851,
+      "step": 12914
+    },
+    {
+      "epoch": 16.5312,
+      "grad_norm": 1.0404068231582642,
+      "learning_rate": 2.4185674269707884e-05,
+      "loss": 0.5212,
+      "step": 12915
+    },
+    {
+      "epoch": 16.53248,
+      "grad_norm": 1.0728304386138916,
+      "learning_rate": 2.4183673469387756e-05,
+      "loss": 0.5685,
+      "step": 12916
+    },
+    {
+      "epoch": 16.53376,
+      "grad_norm": 1.0526484251022339,
+      "learning_rate": 2.4181672669067628e-05,
+      "loss": 0.5223,
+      "step": 12917
+    },
+    {
+      "epoch": 16.53504,
+      "grad_norm": 1.0103858709335327,
+      "learning_rate": 2.41796718687475e-05,
+      "loss": 0.5132,
+      "step": 12918
+    },
+    {
+      "epoch": 16.53632,
+      "grad_norm": 1.032500982284546,
+      "learning_rate": 2.417767106842737e-05,
+      "loss": 0.4861,
+      "step": 12919
+    },
+    {
+      "epoch": 16.5376,
+      "grad_norm": 1.0344327688217163,
+      "learning_rate": 2.4175670268107243e-05,
+      "loss": 0.5153,
+      "step": 12920
+    },
+    {
+      "epoch": 16.53888,
+      "grad_norm": 1.008581280708313,
+      "learning_rate": 2.4173669467787115e-05,
+      "loss": 0.5462,
+      "step": 12921
+    },
+    {
+      "epoch": 16.54016,
+      "grad_norm": 1.00281822681427,
+      "learning_rate": 2.4171668667466987e-05,
+      "loss": 0.5099,
+      "step": 12922
+    },
+    {
+      "epoch": 16.54144,
+      "grad_norm": 1.0224007368087769,
+      "learning_rate": 2.4169667867146862e-05,
+      "loss": 0.5216,
+      "step": 12923
+    },
+    {
+      "epoch": 16.54272,
+      "grad_norm": 0.9643240571022034,
+      "learning_rate": 2.416766706682673e-05,
+      "loss": 0.5092,
+      "step": 12924
+    },
+    {
+      "epoch": 16.544,
+      "grad_norm": 1.013013482093811,
+      "learning_rate": 2.4165666266506603e-05,
+      "loss": 0.5128,
+      "step": 12925
+    },
+    {
+      "epoch": 16.545279999999998,
+      "grad_norm": 1.0628739595413208,
+      "learning_rate": 2.4163665466186475e-05,
+      "loss": 0.5326,
+      "step": 12926
+    },
+    {
+      "epoch": 16.54656,
+      "grad_norm": 0.9615119099617004,
+      "learning_rate": 2.416166466586635e-05,
+      "loss": 0.5209,
+      "step": 12927
+    },
+    {
+      "epoch": 16.54784,
+      "grad_norm": 1.0781883001327515,
+      "learning_rate": 2.415966386554622e-05,
+      "loss": 0.5558,
+      "step": 12928
+    },
+    {
+      "epoch": 16.54912,
+      "grad_norm": 0.9772311449050903,
+      "learning_rate": 2.415766306522609e-05,
+      "loss": 0.5129,
+      "step": 12929
+    },
+    {
+      "epoch": 16.5504,
+      "grad_norm": 1.0191627740859985,
+      "learning_rate": 2.4155662264905962e-05,
+      "loss": 0.5298,
+      "step": 12930
+    },
+    {
+      "epoch": 16.55168,
+      "grad_norm": 1.017059087753296,
+      "learning_rate": 2.4153661464585837e-05,
+      "loss": 0.5203,
+      "step": 12931
+    },
+    {
+      "epoch": 16.55296,
+      "grad_norm": 0.9743010997772217,
+      "learning_rate": 2.4151660664265706e-05,
+      "loss": 0.5098,
+      "step": 12932
+    },
+    {
+      "epoch": 16.55424,
+      "grad_norm": 1.0007681846618652,
+      "learning_rate": 2.4149659863945578e-05,
+      "loss": 0.4947,
+      "step": 12933
+    },
+    {
+      "epoch": 16.55552,
+      "grad_norm": 1.0352661609649658,
+      "learning_rate": 2.4147659063625453e-05,
+      "loss": 0.5854,
+      "step": 12934
+    },
+    {
+      "epoch": 16.5568,
+      "grad_norm": 1.0481826066970825,
+      "learning_rate": 2.4145658263305325e-05,
+      "loss": 0.5548,
+      "step": 12935
+    },
+    {
+      "epoch": 16.55808,
+      "grad_norm": 0.9723829030990601,
+      "learning_rate": 2.4143657462985193e-05,
+      "loss": 0.5207,
+      "step": 12936
+    },
+    {
+      "epoch": 16.55936,
+      "grad_norm": 1.0767929553985596,
+      "learning_rate": 2.4141656662665065e-05,
+      "loss": 0.5427,
+      "step": 12937
+    },
+    {
+      "epoch": 16.56064,
+      "grad_norm": 1.0616075992584229,
+      "learning_rate": 2.413965586234494e-05,
+      "loss": 0.5186,
+      "step": 12938
+    },
+    {
+      "epoch": 16.56192,
+      "grad_norm": 1.0053772926330566,
+      "learning_rate": 2.4137655062024812e-05,
+      "loss": 0.5074,
+      "step": 12939
+    },
+    {
+      "epoch": 16.5632,
+      "grad_norm": 0.963870108127594,
+      "learning_rate": 2.413565426170468e-05,
+      "loss": 0.522,
+      "step": 12940
+    },
+    {
+      "epoch": 16.56448,
+      "grad_norm": 1.0813841819763184,
+      "learning_rate": 2.4133653461384556e-05,
+      "loss": 0.523,
+      "step": 12941
+    },
+    {
+      "epoch": 16.56576,
+      "grad_norm": 1.0653431415557861,
+      "learning_rate": 2.4131652661064428e-05,
+      "loss": 0.5574,
+      "step": 12942
+    },
+    {
+      "epoch": 16.56704,
+      "grad_norm": 0.9429889917373657,
+      "learning_rate": 2.41296518607443e-05,
+      "loss": 0.4838,
+      "step": 12943
+    },
+    {
+      "epoch": 16.56832,
+      "grad_norm": 1.001644492149353,
+      "learning_rate": 2.4127651060424168e-05,
+      "loss": 0.5048,
+      "step": 12944
+    },
+    {
+      "epoch": 16.5696,
+      "grad_norm": 1.0215190649032593,
+      "learning_rate": 2.4125650260104043e-05,
+      "loss": 0.5575,
+      "step": 12945
+    },
+    {
+      "epoch": 16.57088,
+      "grad_norm": 1.084291696548462,
+      "learning_rate": 2.4123649459783915e-05,
+      "loss": 0.583,
+      "step": 12946
+    },
+    {
+      "epoch": 16.57216,
+      "grad_norm": 0.9871845245361328,
+      "learning_rate": 2.4121648659463787e-05,
+      "loss": 0.4813,
+      "step": 12947
+    },
+    {
+      "epoch": 16.57344,
+      "grad_norm": 1.0017796754837036,
+      "learning_rate": 2.411964785914366e-05,
+      "loss": 0.5364,
+      "step": 12948
+    },
+    {
+      "epoch": 16.57472,
+      "grad_norm": 0.9939451813697815,
+      "learning_rate": 2.411764705882353e-05,
+      "loss": 0.529,
+      "step": 12949
+    },
+    {
+      "epoch": 16.576,
+      "grad_norm": 1.0428829193115234,
+      "learning_rate": 2.4115646258503403e-05,
+      "loss": 0.5333,
+      "step": 12950
+    },
+    {
+      "epoch": 16.577280000000002,
+      "grad_norm": 1.0062199831008911,
+      "learning_rate": 2.4113645458183274e-05,
+      "loss": 0.4992,
+      "step": 12951
+    },
+    {
+      "epoch": 16.57856,
+      "grad_norm": 1.0425946712493896,
+      "learning_rate": 2.4111644657863146e-05,
+      "loss": 0.5304,
+      "step": 12952
+    },
+    {
+      "epoch": 16.57984,
+      "grad_norm": 1.0212525129318237,
+      "learning_rate": 2.4109643857543018e-05,
+      "loss": 0.52,
+      "step": 12953
+    },
+    {
+      "epoch": 16.58112,
+      "grad_norm": 1.069036841392517,
+      "learning_rate": 2.410764305722289e-05,
+      "loss": 0.5671,
+      "step": 12954
+    },
+    {
+      "epoch": 16.5824,
+      "grad_norm": 1.0660767555236816,
+      "learning_rate": 2.4105642256902762e-05,
+      "loss": 0.5542,
+      "step": 12955
+    },
+    {
+      "epoch": 16.58368,
+      "grad_norm": 1.009103536605835,
+      "learning_rate": 2.4103641456582634e-05,
+      "loss": 0.5098,
+      "step": 12956
+    },
+    {
+      "epoch": 16.58496,
+      "grad_norm": 1.0511337518692017,
+      "learning_rate": 2.4101640656262506e-05,
+      "loss": 0.5654,
+      "step": 12957
+    },
+    {
+      "epoch": 16.58624,
+      "grad_norm": 0.9621989727020264,
+      "learning_rate": 2.4099639855942377e-05,
+      "loss": 0.4795,
+      "step": 12958
+    },
+    {
+      "epoch": 16.58752,
+      "grad_norm": 1.049293041229248,
+      "learning_rate": 2.409763905562225e-05,
+      "loss": 0.5408,
+      "step": 12959
+    },
+    {
+      "epoch": 16.5888,
+      "grad_norm": 1.0931252241134644,
+      "learning_rate": 2.409563825530212e-05,
+      "loss": 0.5048,
+      "step": 12960
+    },
+    {
+      "epoch": 16.59008,
+      "grad_norm": 1.0250253677368164,
+      "learning_rate": 2.4093637454981993e-05,
+      "loss": 0.5235,
+      "step": 12961
+    },
+    {
+      "epoch": 16.59136,
+      "grad_norm": 1.0003156661987305,
+      "learning_rate": 2.4091636654661868e-05,
+      "loss": 0.4948,
+      "step": 12962
+    },
+    {
+      "epoch": 16.59264,
+      "grad_norm": 1.0581244230270386,
+      "learning_rate": 2.4089635854341737e-05,
+      "loss": 0.5393,
+      "step": 12963
+    },
+    {
+      "epoch": 16.59392,
+      "grad_norm": 1.029645323753357,
+      "learning_rate": 2.408763505402161e-05,
+      "loss": 0.5131,
+      "step": 12964
+    },
+    {
+      "epoch": 16.5952,
+      "grad_norm": 0.9942212700843811,
+      "learning_rate": 2.408563425370148e-05,
+      "loss": 0.5036,
+      "step": 12965
+    },
+    {
+      "epoch": 16.59648,
+      "grad_norm": 1.1347296237945557,
+      "learning_rate": 2.4083633453381356e-05,
+      "loss": 0.5557,
+      "step": 12966
+    },
+    {
+      "epoch": 16.59776,
+      "grad_norm": 1.0200735330581665,
+      "learning_rate": 2.4081632653061224e-05,
+      "loss": 0.5125,
+      "step": 12967
+    },
+    {
+      "epoch": 16.59904,
+      "grad_norm": 1.0105204582214355,
+      "learning_rate": 2.4079631852741096e-05,
+      "loss": 0.4875,
+      "step": 12968
+    },
+    {
+      "epoch": 16.60032,
+      "grad_norm": 0.9845042824745178,
+      "learning_rate": 2.407763105242097e-05,
+      "loss": 0.4829,
+      "step": 12969
+    },
+    {
+      "epoch": 16.6016,
+      "grad_norm": 1.0377755165100098,
+      "learning_rate": 2.4075630252100843e-05,
+      "loss": 0.5135,
+      "step": 12970
+    },
+    {
+      "epoch": 16.60288,
+      "grad_norm": 1.0604853630065918,
+      "learning_rate": 2.407362945178071e-05,
+      "loss": 0.5411,
+      "step": 12971
+    },
+    {
+      "epoch": 16.60416,
+      "grad_norm": 0.9907283782958984,
+      "learning_rate": 2.4071628651460583e-05,
+      "loss": 0.4962,
+      "step": 12972
+    },
+    {
+      "epoch": 16.60544,
+      "grad_norm": 1.0216723680496216,
+      "learning_rate": 2.406962785114046e-05,
+      "loss": 0.5194,
+      "step": 12973
+    },
+    {
+      "epoch": 16.60672,
+      "grad_norm": 0.9833100438117981,
+      "learning_rate": 2.406762705082033e-05,
+      "loss": 0.5267,
+      "step": 12974
+    },
+    {
+      "epoch": 16.608,
+      "grad_norm": 0.9777483940124512,
+      "learning_rate": 2.40656262505002e-05,
+      "loss": 0.5334,
+      "step": 12975
+    },
+    {
+      "epoch": 16.60928,
+      "grad_norm": 1.0363882780075073,
+      "learning_rate": 2.4063625450180074e-05,
+      "loss": 0.5079,
+      "step": 12976
+    },
+    {
+      "epoch": 16.61056,
+      "grad_norm": 1.0367538928985596,
+      "learning_rate": 2.4061624649859946e-05,
+      "loss": 0.5033,
+      "step": 12977
+    },
+    {
+      "epoch": 16.61184,
+      "grad_norm": 1.020905613899231,
+      "learning_rate": 2.4059623849539818e-05,
+      "loss": 0.5121,
+      "step": 12978
+    },
+    {
+      "epoch": 16.61312,
+      "grad_norm": 0.9727598428726196,
+      "learning_rate": 2.4057623049219686e-05,
+      "loss": 0.4858,
+      "step": 12979
+    },
+    {
+      "epoch": 16.6144,
+      "grad_norm": 1.0102119445800781,
+      "learning_rate": 2.405562224889956e-05,
+      "loss": 0.5275,
+      "step": 12980
+    },
+    {
+      "epoch": 16.61568,
+      "grad_norm": 1.0345350503921509,
+      "learning_rate": 2.4053621448579433e-05,
+      "loss": 0.5482,
+      "step": 12981
+    },
+    {
+      "epoch": 16.61696,
+      "grad_norm": 1.0067529678344727,
+      "learning_rate": 2.4051620648259305e-05,
+      "loss": 0.5192,
+      "step": 12982
+    },
+    {
+      "epoch": 16.61824,
+      "grad_norm": 0.9604501724243164,
+      "learning_rate": 2.4049619847939174e-05,
+      "loss": 0.4999,
+      "step": 12983
+    },
+    {
+      "epoch": 16.61952,
+      "grad_norm": 0.9932731986045837,
+      "learning_rate": 2.404761904761905e-05,
+      "loss": 0.5266,
+      "step": 12984
+    },
+    {
+      "epoch": 16.6208,
+      "grad_norm": 0.9318062663078308,
+      "learning_rate": 2.404561824729892e-05,
+      "loss": 0.4916,
+      "step": 12985
+    },
+    {
+      "epoch": 16.62208,
+      "grad_norm": 1.0295798778533936,
+      "learning_rate": 2.4043617446978793e-05,
+      "loss": 0.5223,
+      "step": 12986
+    },
+    {
+      "epoch": 16.62336,
+      "grad_norm": 1.082562804222107,
+      "learning_rate": 2.4041616646658665e-05,
+      "loss": 0.5536,
+      "step": 12987
+    },
+    {
+      "epoch": 16.62464,
+      "grad_norm": 0.9973915219306946,
+      "learning_rate": 2.4039615846338536e-05,
+      "loss": 0.4904,
+      "step": 12988
+    },
+    {
+      "epoch": 16.62592,
+      "grad_norm": 0.9800871014595032,
+      "learning_rate": 2.403761504601841e-05,
+      "loss": 0.516,
+      "step": 12989
+    },
+    {
+      "epoch": 16.6272,
+      "grad_norm": 1.0225999355316162,
+      "learning_rate": 2.403561424569828e-05,
+      "loss": 0.5226,
+      "step": 12990
+    },
+    {
+      "epoch": 16.62848,
+      "grad_norm": 0.9972506761550903,
+      "learning_rate": 2.4033613445378152e-05,
+      "loss": 0.5076,
+      "step": 12991
+    },
+    {
+      "epoch": 16.62976,
+      "grad_norm": 1.0018681287765503,
+      "learning_rate": 2.4031612645058024e-05,
+      "loss": 0.5113,
+      "step": 12992
+    },
+    {
+      "epoch": 16.63104,
+      "grad_norm": 1.0438956022262573,
+      "learning_rate": 2.4029611844737896e-05,
+      "loss": 0.5164,
+      "step": 12993
+    },
+    {
+      "epoch": 16.63232,
+      "grad_norm": 1.0386935472488403,
+      "learning_rate": 2.402761104441777e-05,
+      "loss": 0.5395,
+      "step": 12994
+    },
+    {
+      "epoch": 16.6336,
+      "grad_norm": 1.0958852767944336,
+      "learning_rate": 2.402561024409764e-05,
+      "loss": 0.5501,
+      "step": 12995
+    },
+    {
+      "epoch": 16.63488,
+      "grad_norm": 0.9334178566932678,
+      "learning_rate": 2.402360944377751e-05,
+      "loss": 0.4921,
+      "step": 12996
+    },
+    {
+      "epoch": 16.63616,
+      "grad_norm": 1.008756399154663,
+      "learning_rate": 2.4021608643457383e-05,
+      "loss": 0.5099,
+      "step": 12997
+    },
+    {
+      "epoch": 16.63744,
+      "grad_norm": 0.9561886787414551,
+      "learning_rate": 2.401960784313726e-05,
+      "loss": 0.4785,
+      "step": 12998
+    },
+    {
+      "epoch": 16.63872,
+      "grad_norm": 1.0232385396957397,
+      "learning_rate": 2.4017607042817127e-05,
+      "loss": 0.5235,
+      "step": 12999
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 0.9976513981819153,
+      "learning_rate": 2.4015606242497e-05,
+      "loss": 0.5444,
+      "step": 13000
+    },
+    {
+      "epoch": 16.64128,
+      "grad_norm": 1.0245577096939087,
+      "learning_rate": 2.4013605442176874e-05,
+      "loss": 0.5077,
+      "step": 13001
+    },
+    {
+      "epoch": 16.64256,
+      "grad_norm": 1.0046392679214478,
+      "learning_rate": 2.4011604641856746e-05,
+      "loss": 0.5291,
+      "step": 13002
+    },
+    {
+      "epoch": 16.64384,
+      "grad_norm": 1.0134575366973877,
+      "learning_rate": 2.4009603841536614e-05,
+      "loss": 0.5246,
+      "step": 13003
+    },
+    {
+      "epoch": 16.64512,
+      "grad_norm": 1.0273959636688232,
+      "learning_rate": 2.4007603041216486e-05,
+      "loss": 0.5474,
+      "step": 13004
+    },
+    {
+      "epoch": 16.6464,
+      "grad_norm": 1.09654700756073,
+      "learning_rate": 2.400560224089636e-05,
+      "loss": 0.5338,
+      "step": 13005
+    },
+    {
+      "epoch": 16.64768,
+      "grad_norm": 0.9553311467170715,
+      "learning_rate": 2.4003601440576233e-05,
+      "loss": 0.5073,
+      "step": 13006
+    },
+    {
+      "epoch": 16.64896,
+      "grad_norm": 1.0136662721633911,
+      "learning_rate": 2.4001600640256102e-05,
+      "loss": 0.4933,
+      "step": 13007
+    },
+    {
+      "epoch": 16.65024,
+      "grad_norm": 0.9974261522293091,
+      "learning_rate": 2.3999599839935977e-05,
+      "loss": 0.4961,
+      "step": 13008
+    },
+    {
+      "epoch": 16.65152,
+      "grad_norm": 1.0348639488220215,
+      "learning_rate": 2.399759903961585e-05,
+      "loss": 0.5167,
+      "step": 13009
+    },
+    {
+      "epoch": 16.6528,
+      "grad_norm": 1.0358352661132812,
+      "learning_rate": 2.399559823929572e-05,
+      "loss": 0.5012,
+      "step": 13010
+    },
+    {
+      "epoch": 16.65408,
+      "grad_norm": 0.9727697372436523,
+      "learning_rate": 2.399359743897559e-05,
+      "loss": 0.4838,
+      "step": 13011
+    },
+    {
+      "epoch": 16.65536,
+      "grad_norm": 1.0957095623016357,
+      "learning_rate": 2.3991596638655464e-05,
+      "loss": 0.5401,
+      "step": 13012
+    },
+    {
+      "epoch": 16.65664,
+      "grad_norm": 1.0502195358276367,
+      "learning_rate": 2.3989595838335336e-05,
+      "loss": 0.5342,
+      "step": 13013
+    },
+    {
+      "epoch": 16.65792,
+      "grad_norm": 0.9960721135139465,
+      "learning_rate": 2.3987595038015208e-05,
+      "loss": 0.4892,
+      "step": 13014
+    },
+    {
+      "epoch": 16.6592,
+      "grad_norm": 0.9923244714736938,
+      "learning_rate": 2.398559423769508e-05,
+      "loss": 0.5019,
+      "step": 13015
+    },
+    {
+      "epoch": 16.66048,
+      "grad_norm": 1.0378888845443726,
+      "learning_rate": 2.3983593437374952e-05,
+      "loss": 0.4938,
+      "step": 13016
+    },
+    {
+      "epoch": 16.66176,
+      "grad_norm": 0.9792540669441223,
+      "learning_rate": 2.3981592637054824e-05,
+      "loss": 0.5203,
+      "step": 13017
+    },
+    {
+      "epoch": 16.66304,
+      "grad_norm": 1.0740691423416138,
+      "learning_rate": 2.3979591836734696e-05,
+      "loss": 0.4963,
+      "step": 13018
+    },
+    {
+      "epoch": 16.66432,
+      "grad_norm": 1.0150448083877563,
+      "learning_rate": 2.3977591036414567e-05,
+      "loss": 0.5391,
+      "step": 13019
+    },
+    {
+      "epoch": 16.6656,
+      "grad_norm": 0.9979774355888367,
+      "learning_rate": 2.397559023609444e-05,
+      "loss": 0.5039,
+      "step": 13020
+    },
+    {
+      "epoch": 16.66688,
+      "grad_norm": 0.9845618009567261,
+      "learning_rate": 2.397358943577431e-05,
+      "loss": 0.5409,
+      "step": 13021
+    },
+    {
+      "epoch": 16.66816,
+      "grad_norm": 0.9184175133705139,
+      "learning_rate": 2.3971588635454183e-05,
+      "loss": 0.5081,
+      "step": 13022
+    },
+    {
+      "epoch": 16.66944,
+      "grad_norm": 1.0316041707992554,
+      "learning_rate": 2.3969587835134055e-05,
+      "loss": 0.539,
+      "step": 13023
+    },
+    {
+      "epoch": 16.67072,
+      "grad_norm": 1.0392125844955444,
+      "learning_rate": 2.3967587034813927e-05,
+      "loss": 0.5423,
+      "step": 13024
+    },
+    {
+      "epoch": 16.672,
+      "grad_norm": 0.9866761565208435,
+      "learning_rate": 2.39655862344938e-05,
+      "loss": 0.5351,
+      "step": 13025
+    },
+    {
+      "epoch": 16.67328,
+      "grad_norm": 1.0141420364379883,
+      "learning_rate": 2.396358543417367e-05,
+      "loss": 0.5219,
+      "step": 13026
+    },
+    {
+      "epoch": 16.67456,
+      "grad_norm": 1.0628118515014648,
+      "learning_rate": 2.3961584633853542e-05,
+      "loss": 0.5491,
+      "step": 13027
+    },
+    {
+      "epoch": 16.67584,
+      "grad_norm": 1.0219802856445312,
+      "learning_rate": 2.3959583833533414e-05,
+      "loss": 0.5423,
+      "step": 13028
+    },
+    {
+      "epoch": 16.67712,
+      "grad_norm": 1.0285046100616455,
+      "learning_rate": 2.395758303321329e-05,
+      "loss": 0.5219,
+      "step": 13029
+    },
+    {
+      "epoch": 16.6784,
+      "grad_norm": 0.9643558859825134,
+      "learning_rate": 2.3955582232893158e-05,
+      "loss": 0.4952,
+      "step": 13030
+    },
+    {
+      "epoch": 16.67968,
+      "grad_norm": 1.1057844161987305,
+      "learning_rate": 2.395358143257303e-05,
+      "loss": 0.5293,
+      "step": 13031
+    },
+    {
+      "epoch": 16.68096,
+      "grad_norm": 1.0154691934585571,
+      "learning_rate": 2.39515806322529e-05,
+      "loss": 0.5028,
+      "step": 13032
+    },
+    {
+      "epoch": 16.68224,
+      "grad_norm": 0.9593093991279602,
+      "learning_rate": 2.3949579831932777e-05,
+      "loss": 0.4931,
+      "step": 13033
+    },
+    {
+      "epoch": 16.68352,
+      "grad_norm": 1.0532431602478027,
+      "learning_rate": 2.3947579031612645e-05,
+      "loss": 0.5077,
+      "step": 13034
+    },
+    {
+      "epoch": 16.6848,
+      "grad_norm": 0.959784984588623,
+      "learning_rate": 2.3945578231292517e-05,
+      "loss": 0.5014,
+      "step": 13035
+    },
+    {
+      "epoch": 16.68608,
+      "grad_norm": 1.0308352708816528,
+      "learning_rate": 2.3943577430972392e-05,
+      "loss": 0.5016,
+      "step": 13036
+    },
+    {
+      "epoch": 16.687359999999998,
+      "grad_norm": 0.9219370484352112,
+      "learning_rate": 2.3941576630652264e-05,
+      "loss": 0.5001,
+      "step": 13037
+    },
+    {
+      "epoch": 16.68864,
+      "grad_norm": 0.9760138988494873,
+      "learning_rate": 2.3939575830332133e-05,
+      "loss": 0.5212,
+      "step": 13038
+    },
+    {
+      "epoch": 16.68992,
+      "grad_norm": 1.0665395259857178,
+      "learning_rate": 2.3937575030012005e-05,
+      "loss": 0.5621,
+      "step": 13039
+    },
+    {
+      "epoch": 16.6912,
+      "grad_norm": 0.9828092455863953,
+      "learning_rate": 2.393557422969188e-05,
+      "loss": 0.5331,
+      "step": 13040
+    },
+    {
+      "epoch": 16.69248,
+      "grad_norm": 1.0022916793823242,
+      "learning_rate": 2.393357342937175e-05,
+      "loss": 0.5121,
+      "step": 13041
+    },
+    {
+      "epoch": 16.69376,
+      "grad_norm": 0.931938886642456,
+      "learning_rate": 2.393157262905162e-05,
+      "loss": 0.4934,
+      "step": 13042
+    },
+    {
+      "epoch": 16.69504,
+      "grad_norm": 1.0225458145141602,
+      "learning_rate": 2.3929571828731492e-05,
+      "loss": 0.5217,
+      "step": 13043
+    },
+    {
+      "epoch": 16.69632,
+      "grad_norm": 1.0572493076324463,
+      "learning_rate": 2.3927571028411367e-05,
+      "loss": 0.5562,
+      "step": 13044
+    },
+    {
+      "epoch": 16.6976,
+      "grad_norm": 0.9570837616920471,
+      "learning_rate": 2.392557022809124e-05,
+      "loss": 0.4917,
+      "step": 13045
+    },
+    {
+      "epoch": 16.69888,
+      "grad_norm": 0.9432888627052307,
+      "learning_rate": 2.3923569427771108e-05,
+      "loss": 0.4917,
+      "step": 13046
+    },
+    {
+      "epoch": 16.70016,
+      "grad_norm": 0.9940226078033447,
+      "learning_rate": 2.3921568627450983e-05,
+      "loss": 0.5083,
+      "step": 13047
+    },
+    {
+      "epoch": 16.70144,
+      "grad_norm": 0.9966458082199097,
+      "learning_rate": 2.3919567827130855e-05,
+      "loss": 0.5097,
+      "step": 13048
+    },
+    {
+      "epoch": 16.70272,
+      "grad_norm": 1.0165756940841675,
+      "learning_rate": 2.3917567026810727e-05,
+      "loss": 0.5624,
+      "step": 13049
+    },
+    {
+      "epoch": 16.704,
+      "grad_norm": 1.0315126180648804,
+      "learning_rate": 2.3915566226490595e-05,
+      "loss": 0.5547,
+      "step": 13050
+    },
+    {
+      "epoch": 16.70528,
+      "grad_norm": 1.0564510822296143,
+      "learning_rate": 2.391356542617047e-05,
+      "loss": 0.5763,
+      "step": 13051
+    },
+    {
+      "epoch": 16.70656,
+      "grad_norm": 1.0285402536392212,
+      "learning_rate": 2.3911564625850342e-05,
+      "loss": 0.5526,
+      "step": 13052
+    },
+    {
+      "epoch": 16.70784,
+      "grad_norm": 0.9533161520957947,
+      "learning_rate": 2.3909563825530214e-05,
+      "loss": 0.5125,
+      "step": 13053
+    },
+    {
+      "epoch": 16.70912,
+      "grad_norm": 1.0255242586135864,
+      "learning_rate": 2.3907563025210086e-05,
+      "loss": 0.5404,
+      "step": 13054
+    },
+    {
+      "epoch": 16.7104,
+      "grad_norm": 0.9758644104003906,
+      "learning_rate": 2.3905562224889958e-05,
+      "loss": 0.4887,
+      "step": 13055
+    },
+    {
+      "epoch": 16.71168,
+      "grad_norm": 0.9290502667427063,
+      "learning_rate": 2.390356142456983e-05,
+      "loss": 0.4621,
+      "step": 13056
+    },
+    {
+      "epoch": 16.71296,
+      "grad_norm": 1.0182784795761108,
+      "learning_rate": 2.39015606242497e-05,
+      "loss": 0.5575,
+      "step": 13057
+    },
+    {
+      "epoch": 16.71424,
+      "grad_norm": 1.022233009338379,
+      "learning_rate": 2.3899559823929573e-05,
+      "loss": 0.5254,
+      "step": 13058
+    },
+    {
+      "epoch": 16.71552,
+      "grad_norm": 1.0084254741668701,
+      "learning_rate": 2.3897559023609445e-05,
+      "loss": 0.5153,
+      "step": 13059
+    },
+    {
+      "epoch": 16.7168,
+      "grad_norm": 0.9809401631355286,
+      "learning_rate": 2.3895558223289317e-05,
+      "loss": 0.5125,
+      "step": 13060
+    },
+    {
+      "epoch": 16.71808,
+      "grad_norm": 0.9509784579277039,
+      "learning_rate": 2.389355742296919e-05,
+      "loss": 0.4663,
+      "step": 13061
+    },
+    {
+      "epoch": 16.71936,
+      "grad_norm": 1.1078130006790161,
+      "learning_rate": 2.389155662264906e-05,
+      "loss": 0.5595,
+      "step": 13062
+    },
+    {
+      "epoch": 16.72064,
+      "grad_norm": 1.0031615495681763,
+      "learning_rate": 2.3889555822328933e-05,
+      "loss": 0.5288,
+      "step": 13063
+    },
+    {
+      "epoch": 16.72192,
+      "grad_norm": 0.9790761470794678,
+      "learning_rate": 2.3887555022008804e-05,
+      "loss": 0.4976,
+      "step": 13064
+    },
+    {
+      "epoch": 16.7232,
+      "grad_norm": 0.9689204096794128,
+      "learning_rate": 2.3885554221688676e-05,
+      "loss": 0.5193,
+      "step": 13065
+    },
+    {
+      "epoch": 16.72448,
+      "grad_norm": 1.0444097518920898,
+      "learning_rate": 2.3883553421368548e-05,
+      "loss": 0.5472,
+      "step": 13066
+    },
+    {
+      "epoch": 16.72576,
+      "grad_norm": 1.0322445631027222,
+      "learning_rate": 2.388155262104842e-05,
+      "loss": 0.5432,
+      "step": 13067
+    },
+    {
+      "epoch": 16.72704,
+      "grad_norm": 0.9607560038566589,
+      "learning_rate": 2.3879551820728295e-05,
+      "loss": 0.4699,
+      "step": 13068
+    },
+    {
+      "epoch": 16.72832,
+      "grad_norm": 1.0295300483703613,
+      "learning_rate": 2.3877551020408164e-05,
+      "loss": 0.5077,
+      "step": 13069
+    },
+    {
+      "epoch": 16.7296,
+      "grad_norm": 1.0513696670532227,
+      "learning_rate": 2.3875550220088036e-05,
+      "loss": 0.5587,
+      "step": 13070
+    },
+    {
+      "epoch": 16.73088,
+      "grad_norm": 1.0276129245758057,
+      "learning_rate": 2.3873549419767907e-05,
+      "loss": 0.5124,
+      "step": 13071
+    },
+    {
+      "epoch": 16.73216,
+      "grad_norm": 0.9594248533248901,
+      "learning_rate": 2.3871548619447783e-05,
+      "loss": 0.4769,
+      "step": 13072
+    },
+    {
+      "epoch": 16.73344,
+      "grad_norm": 1.0701425075531006,
+      "learning_rate": 2.386954781912765e-05,
+      "loss": 0.5459,
+      "step": 13073
+    },
+    {
+      "epoch": 16.73472,
+      "grad_norm": 1.044277548789978,
+      "learning_rate": 2.3867547018807523e-05,
+      "loss": 0.5692,
+      "step": 13074
+    },
+    {
+      "epoch": 16.736,
+      "grad_norm": 1.0274498462677002,
+      "learning_rate": 2.3865546218487398e-05,
+      "loss": 0.5312,
+      "step": 13075
+    },
+    {
+      "epoch": 16.73728,
+      "grad_norm": 1.0216330289840698,
+      "learning_rate": 2.386354541816727e-05,
+      "loss": 0.5514,
+      "step": 13076
+    },
+    {
+      "epoch": 16.73856,
+      "grad_norm": 1.0108169317245483,
+      "learning_rate": 2.386154461784714e-05,
+      "loss": 0.537,
+      "step": 13077
+    },
+    {
+      "epoch": 16.73984,
+      "grad_norm": 1.015303134918213,
+      "learning_rate": 2.385954381752701e-05,
+      "loss": 0.522,
+      "step": 13078
+    },
+    {
+      "epoch": 16.74112,
+      "grad_norm": 1.0391994714736938,
+      "learning_rate": 2.3857543017206886e-05,
+      "loss": 0.5815,
+      "step": 13079
+    },
+    {
+      "epoch": 16.7424,
+      "grad_norm": 1.0588417053222656,
+      "learning_rate": 2.3855542216886757e-05,
+      "loss": 0.5296,
+      "step": 13080
+    },
+    {
+      "epoch": 16.74368,
+      "grad_norm": 0.9854347109794617,
+      "learning_rate": 2.3853541416566626e-05,
+      "loss": 0.5049,
+      "step": 13081
+    },
+    {
+      "epoch": 16.74496,
+      "grad_norm": 1.0034525394439697,
+      "learning_rate": 2.38515406162465e-05,
+      "loss": 0.5144,
+      "step": 13082
+    },
+    {
+      "epoch": 16.74624,
+      "grad_norm": 1.0394407510757446,
+      "learning_rate": 2.3849539815926373e-05,
+      "loss": 0.4741,
+      "step": 13083
+    },
+    {
+      "epoch": 16.74752,
+      "grad_norm": 0.9602055549621582,
+      "learning_rate": 2.3847539015606245e-05,
+      "loss": 0.5129,
+      "step": 13084
+    },
+    {
+      "epoch": 16.7488,
+      "grad_norm": 1.01785147190094,
+      "learning_rate": 2.3845538215286113e-05,
+      "loss": 0.5215,
+      "step": 13085
+    },
+    {
+      "epoch": 16.75008,
+      "grad_norm": 1.021774411201477,
+      "learning_rate": 2.384353741496599e-05,
+      "loss": 0.5161,
+      "step": 13086
+    },
+    {
+      "epoch": 16.75136,
+      "grad_norm": 1.0218771696090698,
+      "learning_rate": 2.384153661464586e-05,
+      "loss": 0.5066,
+      "step": 13087
+    },
+    {
+      "epoch": 16.75264,
+      "grad_norm": 0.9877253174781799,
+      "learning_rate": 2.3839535814325732e-05,
+      "loss": 0.4995,
+      "step": 13088
+    },
+    {
+      "epoch": 16.75392,
+      "grad_norm": 1.0353214740753174,
+      "learning_rate": 2.3837535014005604e-05,
+      "loss": 0.4538,
+      "step": 13089
+    },
+    {
+      "epoch": 16.7552,
+      "grad_norm": 1.052165150642395,
+      "learning_rate": 2.3835534213685476e-05,
+      "loss": 0.5353,
+      "step": 13090
+    },
+    {
+      "epoch": 16.75648,
+      "grad_norm": 0.9751322269439697,
+      "learning_rate": 2.3833533413365348e-05,
+      "loss": 0.4861,
+      "step": 13091
+    },
+    {
+      "epoch": 16.75776,
+      "grad_norm": 1.0614244937896729,
+      "learning_rate": 2.383153261304522e-05,
+      "loss": 0.562,
+      "step": 13092
+    },
+    {
+      "epoch": 16.75904,
+      "grad_norm": 1.001287817955017,
+      "learning_rate": 2.382953181272509e-05,
+      "loss": 0.5248,
+      "step": 13093
+    },
+    {
+      "epoch": 16.76032,
+      "grad_norm": 1.015852451324463,
+      "learning_rate": 2.3827531012404963e-05,
+      "loss": 0.5339,
+      "step": 13094
+    },
+    {
+      "epoch": 16.7616,
+      "grad_norm": 1.0684832334518433,
+      "learning_rate": 2.3825530212084835e-05,
+      "loss": 0.5295,
+      "step": 13095
+    },
+    {
+      "epoch": 16.76288,
+      "grad_norm": 0.9663676619529724,
+      "learning_rate": 2.3823529411764707e-05,
+      "loss": 0.4559,
+      "step": 13096
+    },
+    {
+      "epoch": 16.76416,
+      "grad_norm": 0.949224591255188,
+      "learning_rate": 2.382152861144458e-05,
+      "loss": 0.5483,
+      "step": 13097
+    },
+    {
+      "epoch": 16.76544,
+      "grad_norm": 1.01194167137146,
+      "learning_rate": 2.381952781112445e-05,
+      "loss": 0.5269,
+      "step": 13098
+    },
+    {
+      "epoch": 16.76672,
+      "grad_norm": 1.0378564596176147,
+      "learning_rate": 2.3817527010804323e-05,
+      "loss": 0.5227,
+      "step": 13099
+    },
+    {
+      "epoch": 16.768,
+      "grad_norm": 1.0125080347061157,
+      "learning_rate": 2.3815526210484195e-05,
+      "loss": 0.5238,
+      "step": 13100
+    },
+    {
+      "epoch": 16.76928,
+      "grad_norm": 1.0728344917297363,
+      "learning_rate": 2.3813525410164066e-05,
+      "loss": 0.5434,
+      "step": 13101
+    },
+    {
+      "epoch": 16.77056,
+      "grad_norm": 0.9800313711166382,
+      "learning_rate": 2.381152460984394e-05,
+      "loss": 0.4794,
+      "step": 13102
+    },
+    {
+      "epoch": 16.77184,
+      "grad_norm": 1.0031391382217407,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.505,
+      "step": 13103
+    },
+    {
+      "epoch": 16.77312,
+      "grad_norm": 0.9527682662010193,
+      "learning_rate": 2.3807523009203682e-05,
+      "loss": 0.5031,
+      "step": 13104
+    },
+    {
+      "epoch": 16.7744,
+      "grad_norm": 1.0169705152511597,
+      "learning_rate": 2.3805522208883554e-05,
+      "loss": 0.4872,
+      "step": 13105
+    },
+    {
+      "epoch": 16.77568,
+      "grad_norm": 0.9913395643234253,
+      "learning_rate": 2.3803521408563426e-05,
+      "loss": 0.5042,
+      "step": 13106
+    },
+    {
+      "epoch": 16.77696,
+      "grad_norm": 1.0133116245269775,
+      "learning_rate": 2.38015206082433e-05,
+      "loss": 0.5202,
+      "step": 13107
+    },
+    {
+      "epoch": 16.77824,
+      "grad_norm": 1.0359431505203247,
+      "learning_rate": 2.379951980792317e-05,
+      "loss": 0.5379,
+      "step": 13108
+    },
+    {
+      "epoch": 16.77952,
+      "grad_norm": 0.9904114603996277,
+      "learning_rate": 2.379751900760304e-05,
+      "loss": 0.486,
+      "step": 13109
+    },
+    {
+      "epoch": 16.7808,
+      "grad_norm": 0.9596001505851746,
+      "learning_rate": 2.3795518207282913e-05,
+      "loss": 0.4627,
+      "step": 13110
+    },
+    {
+      "epoch": 16.78208,
+      "grad_norm": 1.0215861797332764,
+      "learning_rate": 2.379351740696279e-05,
+      "loss": 0.5149,
+      "step": 13111
+    },
+    {
+      "epoch": 16.78336,
+      "grad_norm": 1.0180429220199585,
+      "learning_rate": 2.3791516606642657e-05,
+      "loss": 0.5304,
+      "step": 13112
+    },
+    {
+      "epoch": 16.78464,
+      "grad_norm": 0.942301332950592,
+      "learning_rate": 2.378951580632253e-05,
+      "loss": 0.5205,
+      "step": 13113
+    },
+    {
+      "epoch": 16.78592,
+      "grad_norm": 0.9987159371376038,
+      "learning_rate": 2.3787515006002404e-05,
+      "loss": 0.5267,
+      "step": 13114
+    },
+    {
+      "epoch": 16.7872,
+      "grad_norm": 1.0503463745117188,
+      "learning_rate": 2.3785514205682276e-05,
+      "loss": 0.5245,
+      "step": 13115
+    },
+    {
+      "epoch": 16.78848,
+      "grad_norm": 1.0316410064697266,
+      "learning_rate": 2.3783513405362144e-05,
+      "loss": 0.5262,
+      "step": 13116
+    },
+    {
+      "epoch": 16.78976,
+      "grad_norm": 0.9606108665466309,
+      "learning_rate": 2.3781512605042016e-05,
+      "loss": 0.5006,
+      "step": 13117
+    },
+    {
+      "epoch": 16.79104,
+      "grad_norm": 1.021532654762268,
+      "learning_rate": 2.377951180472189e-05,
+      "loss": 0.5441,
+      "step": 13118
+    },
+    {
+      "epoch": 16.79232,
+      "grad_norm": 0.9799433350563049,
+      "learning_rate": 2.3777511004401763e-05,
+      "loss": 0.5082,
+      "step": 13119
+    },
+    {
+      "epoch": 16.7936,
+      "grad_norm": 0.9904650449752808,
+      "learning_rate": 2.3775510204081632e-05,
+      "loss": 0.55,
+      "step": 13120
+    },
+    {
+      "epoch": 16.79488,
+      "grad_norm": 1.006087303161621,
+      "learning_rate": 2.3773509403761507e-05,
+      "loss": 0.5431,
+      "step": 13121
+    },
+    {
+      "epoch": 16.79616,
+      "grad_norm": 1.0185682773590088,
+      "learning_rate": 2.377150860344138e-05,
+      "loss": 0.5338,
+      "step": 13122
+    },
+    {
+      "epoch": 16.79744,
+      "grad_norm": 1.039462924003601,
+      "learning_rate": 2.376950780312125e-05,
+      "loss": 0.5525,
+      "step": 13123
+    },
+    {
+      "epoch": 16.79872,
+      "grad_norm": 1.048966884613037,
+      "learning_rate": 2.376750700280112e-05,
+      "loss": 0.5458,
+      "step": 13124
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 1.0096321105957031,
+      "learning_rate": 2.3765506202480994e-05,
+      "loss": 0.5175,
+      "step": 13125
+    },
+    {
+      "epoch": 16.80128,
+      "grad_norm": 1.0281754732131958,
+      "learning_rate": 2.3763505402160866e-05,
+      "loss": 0.54,
+      "step": 13126
+    },
+    {
+      "epoch": 16.80256,
+      "grad_norm": 1.0194934606552124,
+      "learning_rate": 2.3761504601840738e-05,
+      "loss": 0.5144,
+      "step": 13127
+    },
+    {
+      "epoch": 16.80384,
+      "grad_norm": 1.0557507276535034,
+      "learning_rate": 2.375950380152061e-05,
+      "loss": 0.537,
+      "step": 13128
+    },
+    {
+      "epoch": 16.80512,
+      "grad_norm": 0.926566481590271,
+      "learning_rate": 2.3757503001200482e-05,
+      "loss": 0.4973,
+      "step": 13129
+    },
+    {
+      "epoch": 16.8064,
+      "grad_norm": 1.016688585281372,
+      "learning_rate": 2.3755502200880354e-05,
+      "loss": 0.524,
+      "step": 13130
+    },
+    {
+      "epoch": 16.80768,
+      "grad_norm": 1.0396153926849365,
+      "learning_rate": 2.3753501400560226e-05,
+      "loss": 0.5364,
+      "step": 13131
+    },
+    {
+      "epoch": 16.80896,
+      "grad_norm": 0.9891914129257202,
+      "learning_rate": 2.3751500600240097e-05,
+      "loss": 0.5359,
+      "step": 13132
+    },
+    {
+      "epoch": 16.81024,
+      "grad_norm": 1.0839710235595703,
+      "learning_rate": 2.374949979991997e-05,
+      "loss": 0.5577,
+      "step": 13133
+    },
+    {
+      "epoch": 16.81152,
+      "grad_norm": 0.9661899209022522,
+      "learning_rate": 2.374749899959984e-05,
+      "loss": 0.4845,
+      "step": 13134
+    },
+    {
+      "epoch": 16.8128,
+      "grad_norm": 0.9867687821388245,
+      "learning_rate": 2.3745498199279713e-05,
+      "loss": 0.5227,
+      "step": 13135
+    },
+    {
+      "epoch": 16.81408,
+      "grad_norm": 1.0412554740905762,
+      "learning_rate": 2.3743497398959585e-05,
+      "loss": 0.5154,
+      "step": 13136
+    },
+    {
+      "epoch": 16.81536,
+      "grad_norm": 0.9606292843818665,
+      "learning_rate": 2.3741496598639457e-05,
+      "loss": 0.4987,
+      "step": 13137
+    },
+    {
+      "epoch": 16.81664,
+      "grad_norm": 0.9731855392456055,
+      "learning_rate": 2.373949579831933e-05,
+      "loss": 0.4813,
+      "step": 13138
+    },
+    {
+      "epoch": 16.81792,
+      "grad_norm": 1.0362776517868042,
+      "learning_rate": 2.37374949979992e-05,
+      "loss": 0.5173,
+      "step": 13139
+    },
+    {
+      "epoch": 16.8192,
+      "grad_norm": 1.0115454196929932,
+      "learning_rate": 2.3735494197679072e-05,
+      "loss": 0.4787,
+      "step": 13140
+    },
+    {
+      "epoch": 16.82048,
+      "grad_norm": 1.0400524139404297,
+      "learning_rate": 2.3733493397358944e-05,
+      "loss": 0.5218,
+      "step": 13141
+    },
+    {
+      "epoch": 16.82176,
+      "grad_norm": 1.0309675931930542,
+      "learning_rate": 2.373149259703882e-05,
+      "loss": 0.5008,
+      "step": 13142
+    },
+    {
+      "epoch": 16.82304,
+      "grad_norm": 0.9864956140518188,
+      "learning_rate": 2.3729491796718688e-05,
+      "loss": 0.4848,
+      "step": 13143
+    },
+    {
+      "epoch": 16.82432,
+      "grad_norm": 0.9826066493988037,
+      "learning_rate": 2.372749099639856e-05,
+      "loss": 0.4918,
+      "step": 13144
+    },
+    {
+      "epoch": 16.8256,
+      "grad_norm": 1.0326327085494995,
+      "learning_rate": 2.372549019607843e-05,
+      "loss": 0.5138,
+      "step": 13145
+    },
+    {
+      "epoch": 16.82688,
+      "grad_norm": 0.9792665839195251,
+      "learning_rate": 2.3723489395758307e-05,
+      "loss": 0.5088,
+      "step": 13146
+    },
+    {
+      "epoch": 16.82816,
+      "grad_norm": 0.9995619654655457,
+      "learning_rate": 2.3721488595438175e-05,
+      "loss": 0.5337,
+      "step": 13147
+    },
+    {
+      "epoch": 16.829439999999998,
+      "grad_norm": 1.030884027481079,
+      "learning_rate": 2.3719487795118047e-05,
+      "loss": 0.541,
+      "step": 13148
+    },
+    {
+      "epoch": 16.83072,
+      "grad_norm": 1.0245966911315918,
+      "learning_rate": 2.3717486994797922e-05,
+      "loss": 0.4978,
+      "step": 13149
+    },
+    {
+      "epoch": 16.832,
+      "grad_norm": 1.035640835762024,
+      "learning_rate": 2.3715486194477794e-05,
+      "loss": 0.5688,
+      "step": 13150
+    },
+    {
+      "epoch": 16.83328,
+      "grad_norm": 1.0522053241729736,
+      "learning_rate": 2.3713485394157663e-05,
+      "loss": 0.5527,
+      "step": 13151
+    },
+    {
+      "epoch": 16.83456,
+      "grad_norm": 1.0623236894607544,
+      "learning_rate": 2.3711484593837535e-05,
+      "loss": 0.5676,
+      "step": 13152
+    },
+    {
+      "epoch": 16.83584,
+      "grad_norm": 0.9654913544654846,
+      "learning_rate": 2.370948379351741e-05,
+      "loss": 0.5085,
+      "step": 13153
+    },
+    {
+      "epoch": 16.83712,
+      "grad_norm": 0.9618136286735535,
+      "learning_rate": 2.370748299319728e-05,
+      "loss": 0.5171,
+      "step": 13154
+    },
+    {
+      "epoch": 16.8384,
+      "grad_norm": 1.009698748588562,
+      "learning_rate": 2.370548219287715e-05,
+      "loss": 0.4982,
+      "step": 13155
+    },
+    {
+      "epoch": 16.83968,
+      "grad_norm": 1.0301358699798584,
+      "learning_rate": 2.3703481392557022e-05,
+      "loss": 0.5463,
+      "step": 13156
+    },
+    {
+      "epoch": 16.84096,
+      "grad_norm": 1.008836269378662,
+      "learning_rate": 2.3701480592236897e-05,
+      "loss": 0.4709,
+      "step": 13157
+    },
+    {
+      "epoch": 16.84224,
+      "grad_norm": 1.0866061449050903,
+      "learning_rate": 2.369947979191677e-05,
+      "loss": 0.5407,
+      "step": 13158
+    },
+    {
+      "epoch": 16.84352,
+      "grad_norm": 1.0506260395050049,
+      "learning_rate": 2.3697478991596638e-05,
+      "loss": 0.5297,
+      "step": 13159
+    },
+    {
+      "epoch": 16.8448,
+      "grad_norm": 1.0603467226028442,
+      "learning_rate": 2.3695478191276513e-05,
+      "loss": 0.5368,
+      "step": 13160
+    },
+    {
+      "epoch": 16.84608,
+      "grad_norm": 0.9975579977035522,
+      "learning_rate": 2.3693477390956385e-05,
+      "loss": 0.5138,
+      "step": 13161
+    },
+    {
+      "epoch": 16.84736,
+      "grad_norm": 1.0625996589660645,
+      "learning_rate": 2.3691476590636257e-05,
+      "loss": 0.5583,
+      "step": 13162
+    },
+    {
+      "epoch": 16.84864,
+      "grad_norm": 1.0248199701309204,
+      "learning_rate": 2.3689475790316125e-05,
+      "loss": 0.5415,
+      "step": 13163
+    },
+    {
+      "epoch": 16.84992,
+      "grad_norm": 1.0083729028701782,
+      "learning_rate": 2.3687474989996e-05,
+      "loss": 0.5248,
+      "step": 13164
+    },
+    {
+      "epoch": 16.8512,
+      "grad_norm": 1.0073386430740356,
+      "learning_rate": 2.3685474189675872e-05,
+      "loss": 0.523,
+      "step": 13165
+    },
+    {
+      "epoch": 16.85248,
+      "grad_norm": 1.023363471031189,
+      "learning_rate": 2.3683473389355744e-05,
+      "loss": 0.5001,
+      "step": 13166
+    },
+    {
+      "epoch": 16.85376,
+      "grad_norm": 0.9759731888771057,
+      "learning_rate": 2.3681472589035616e-05,
+      "loss": 0.5141,
+      "step": 13167
+    },
+    {
+      "epoch": 16.85504,
+      "grad_norm": 1.0707515478134155,
+      "learning_rate": 2.3679471788715488e-05,
+      "loss": 0.5621,
+      "step": 13168
+    },
+    {
+      "epoch": 16.85632,
+      "grad_norm": 1.0116777420043945,
+      "learning_rate": 2.367747098839536e-05,
+      "loss": 0.5197,
+      "step": 13169
+    },
+    {
+      "epoch": 16.8576,
+      "grad_norm": 1.0196561813354492,
+      "learning_rate": 2.367547018807523e-05,
+      "loss": 0.5688,
+      "step": 13170
+    },
+    {
+      "epoch": 16.85888,
+      "grad_norm": 1.0545904636383057,
+      "learning_rate": 2.3673469387755103e-05,
+      "loss": 0.5704,
+      "step": 13171
+    },
+    {
+      "epoch": 16.86016,
+      "grad_norm": 1.0500190258026123,
+      "learning_rate": 2.3671468587434975e-05,
+      "loss": 0.5012,
+      "step": 13172
+    },
+    {
+      "epoch": 16.86144,
+      "grad_norm": 0.9819349646568298,
+      "learning_rate": 2.3669467787114847e-05,
+      "loss": 0.5261,
+      "step": 13173
+    },
+    {
+      "epoch": 16.86272,
+      "grad_norm": 0.993928074836731,
+      "learning_rate": 2.366746698679472e-05,
+      "loss": 0.5433,
+      "step": 13174
+    },
+    {
+      "epoch": 16.864,
+      "grad_norm": 0.9953247308731079,
+      "learning_rate": 2.366546618647459e-05,
+      "loss": 0.5125,
+      "step": 13175
+    },
+    {
+      "epoch": 16.86528,
+      "grad_norm": 1.0047765970230103,
+      "learning_rate": 2.3663465386154463e-05,
+      "loss": 0.5015,
+      "step": 13176
+    },
+    {
+      "epoch": 16.86656,
+      "grad_norm": 1.0829325914382935,
+      "learning_rate": 2.3661464585834334e-05,
+      "loss": 0.5706,
+      "step": 13177
+    },
+    {
+      "epoch": 16.86784,
+      "grad_norm": 1.0480440855026245,
+      "learning_rate": 2.3659463785514206e-05,
+      "loss": 0.5776,
+      "step": 13178
+    },
+    {
+      "epoch": 16.86912,
+      "grad_norm": 1.0479168891906738,
+      "learning_rate": 2.3657462985194078e-05,
+      "loss": 0.5047,
+      "step": 13179
+    },
+    {
+      "epoch": 16.8704,
+      "grad_norm": 0.9871065616607666,
+      "learning_rate": 2.365546218487395e-05,
+      "loss": 0.4929,
+      "step": 13180
+    },
+    {
+      "epoch": 16.87168,
+      "grad_norm": 1.0528372526168823,
+      "learning_rate": 2.3653461384553825e-05,
+      "loss": 0.5289,
+      "step": 13181
+    },
+    {
+      "epoch": 16.87296,
+      "grad_norm": 1.0125209093093872,
+      "learning_rate": 2.3651460584233694e-05,
+      "loss": 0.5539,
+      "step": 13182
+    },
+    {
+      "epoch": 16.87424,
+      "grad_norm": 1.0450409650802612,
+      "learning_rate": 2.3649459783913565e-05,
+      "loss": 0.5374,
+      "step": 13183
+    },
+    {
+      "epoch": 16.87552,
+      "grad_norm": 1.061608910560608,
+      "learning_rate": 2.3647458983593437e-05,
+      "loss": 0.52,
+      "step": 13184
+    },
+    {
+      "epoch": 16.8768,
+      "grad_norm": 1.019355058670044,
+      "learning_rate": 2.3645458183273313e-05,
+      "loss": 0.499,
+      "step": 13185
+    },
+    {
+      "epoch": 16.87808,
+      "grad_norm": 0.9892388582229614,
+      "learning_rate": 2.364345738295318e-05,
+      "loss": 0.5473,
+      "step": 13186
+    },
+    {
+      "epoch": 16.87936,
+      "grad_norm": 1.001673936843872,
+      "learning_rate": 2.3641456582633053e-05,
+      "loss": 0.4966,
+      "step": 13187
+    },
+    {
+      "epoch": 16.88064,
+      "grad_norm": 1.078865647315979,
+      "learning_rate": 2.3639455782312928e-05,
+      "loss": 0.5499,
+      "step": 13188
+    },
+    {
+      "epoch": 16.88192,
+      "grad_norm": 0.9952927827835083,
+      "learning_rate": 2.36374549819928e-05,
+      "loss": 0.5131,
+      "step": 13189
+    },
+    {
+      "epoch": 16.8832,
+      "grad_norm": 0.9789214730262756,
+      "learning_rate": 2.363545418167267e-05,
+      "loss": 0.513,
+      "step": 13190
+    },
+    {
+      "epoch": 16.88448,
+      "grad_norm": 1.0156183242797852,
+      "learning_rate": 2.363345338135254e-05,
+      "loss": 0.5402,
+      "step": 13191
+    },
+    {
+      "epoch": 16.88576,
+      "grad_norm": 1.0390909910202026,
+      "learning_rate": 2.3631452581032416e-05,
+      "loss": 0.571,
+      "step": 13192
+    },
+    {
+      "epoch": 16.88704,
+      "grad_norm": 1.0474461317062378,
+      "learning_rate": 2.3629451780712287e-05,
+      "loss": 0.5456,
+      "step": 13193
+    },
+    {
+      "epoch": 16.88832,
+      "grad_norm": 1.0926357507705688,
+      "learning_rate": 2.3627450980392156e-05,
+      "loss": 0.51,
+      "step": 13194
+    },
+    {
+      "epoch": 16.8896,
+      "grad_norm": 1.0534778833389282,
+      "learning_rate": 2.362545018007203e-05,
+      "loss": 0.5292,
+      "step": 13195
+    },
+    {
+      "epoch": 16.89088,
+      "grad_norm": 1.0174009799957275,
+      "learning_rate": 2.3623449379751903e-05,
+      "loss": 0.5295,
+      "step": 13196
+    },
+    {
+      "epoch": 16.89216,
+      "grad_norm": 1.018369436264038,
+      "learning_rate": 2.3621448579431775e-05,
+      "loss": 0.5362,
+      "step": 13197
+    },
+    {
+      "epoch": 16.89344,
+      "grad_norm": 1.008921504020691,
+      "learning_rate": 2.3619447779111643e-05,
+      "loss": 0.5035,
+      "step": 13198
+    },
+    {
+      "epoch": 16.89472,
+      "grad_norm": 1.040116548538208,
+      "learning_rate": 2.361744697879152e-05,
+      "loss": 0.5335,
+      "step": 13199
+    },
+    {
+      "epoch": 16.896,
+      "grad_norm": 1.0576521158218384,
+      "learning_rate": 2.361544617847139e-05,
+      "loss": 0.5574,
+      "step": 13200
+    },
+    {
+      "epoch": 16.89728,
+      "grad_norm": 1.048539161682129,
+      "learning_rate": 2.3613445378151262e-05,
+      "loss": 0.5106,
+      "step": 13201
+    },
+    {
+      "epoch": 16.89856,
+      "grad_norm": 1.0253418684005737,
+      "learning_rate": 2.3611444577831134e-05,
+      "loss": 0.5217,
+      "step": 13202
+    },
+    {
+      "epoch": 16.89984,
+      "grad_norm": 1.0338140726089478,
+      "learning_rate": 2.3609443777511006e-05,
+      "loss": 0.5126,
+      "step": 13203
+    },
+    {
+      "epoch": 16.90112,
+      "grad_norm": 1.0068703889846802,
+      "learning_rate": 2.3607442977190878e-05,
+      "loss": 0.5127,
+      "step": 13204
+    },
+    {
+      "epoch": 16.9024,
+      "grad_norm": 0.9766025543212891,
+      "learning_rate": 2.360544217687075e-05,
+      "loss": 0.4986,
+      "step": 13205
+    },
+    {
+      "epoch": 16.90368,
+      "grad_norm": 1.0029348134994507,
+      "learning_rate": 2.360344137655062e-05,
+      "loss": 0.509,
+      "step": 13206
+    },
+    {
+      "epoch": 16.90496,
+      "grad_norm": 0.9763104915618896,
+      "learning_rate": 2.3601440576230493e-05,
+      "loss": 0.4673,
+      "step": 13207
+    },
+    {
+      "epoch": 16.90624,
+      "grad_norm": 0.9850437045097351,
+      "learning_rate": 2.3599439775910365e-05,
+      "loss": 0.544,
+      "step": 13208
+    },
+    {
+      "epoch": 16.90752,
+      "grad_norm": 1.063561201095581,
+      "learning_rate": 2.3597438975590237e-05,
+      "loss": 0.5269,
+      "step": 13209
+    },
+    {
+      "epoch": 16.9088,
+      "grad_norm": 1.0321555137634277,
+      "learning_rate": 2.359543817527011e-05,
+      "loss": 0.5038,
+      "step": 13210
+    },
+    {
+      "epoch": 16.91008,
+      "grad_norm": 1.0148087739944458,
+      "learning_rate": 2.359343737494998e-05,
+      "loss": 0.5437,
+      "step": 13211
+    },
+    {
+      "epoch": 16.91136,
+      "grad_norm": 0.9977015852928162,
+      "learning_rate": 2.3591436574629853e-05,
+      "loss": 0.5425,
+      "step": 13212
+    },
+    {
+      "epoch": 16.91264,
+      "grad_norm": 1.0292987823486328,
+      "learning_rate": 2.3589435774309725e-05,
+      "loss": 0.5345,
+      "step": 13213
+    },
+    {
+      "epoch": 16.91392,
+      "grad_norm": 0.9989826679229736,
+      "learning_rate": 2.3587434973989596e-05,
+      "loss": 0.5079,
+      "step": 13214
+    },
+    {
+      "epoch": 16.9152,
+      "grad_norm": 0.987629771232605,
+      "learning_rate": 2.3585434173669468e-05,
+      "loss": 0.5053,
+      "step": 13215
+    },
+    {
+      "epoch": 16.91648,
+      "grad_norm": 1.0139482021331787,
+      "learning_rate": 2.358343337334934e-05,
+      "loss": 0.5206,
+      "step": 13216
+    },
+    {
+      "epoch": 16.91776,
+      "grad_norm": 0.9933033585548401,
+      "learning_rate": 2.3581432573029212e-05,
+      "loss": 0.5358,
+      "step": 13217
+    },
+    {
+      "epoch": 16.91904,
+      "grad_norm": 1.0258675813674927,
+      "learning_rate": 2.3579431772709084e-05,
+      "loss": 0.5204,
+      "step": 13218
+    },
+    {
+      "epoch": 16.92032,
+      "grad_norm": 1.0290369987487793,
+      "learning_rate": 2.3577430972388956e-05,
+      "loss": 0.5262,
+      "step": 13219
+    },
+    {
+      "epoch": 16.9216,
+      "grad_norm": 1.0307328701019287,
+      "learning_rate": 2.357543017206883e-05,
+      "loss": 0.5321,
+      "step": 13220
+    },
+    {
+      "epoch": 16.92288,
+      "grad_norm": 1.0345529317855835,
+      "learning_rate": 2.35734293717487e-05,
+      "loss": 0.5526,
+      "step": 13221
+    },
+    {
+      "epoch": 16.92416,
+      "grad_norm": 1.028792381286621,
+      "learning_rate": 2.357142857142857e-05,
+      "loss": 0.5265,
+      "step": 13222
+    },
+    {
+      "epoch": 16.925440000000002,
+      "grad_norm": 1.0457243919372559,
+      "learning_rate": 2.3569427771108443e-05,
+      "loss": 0.5508,
+      "step": 13223
+    },
+    {
+      "epoch": 16.92672,
+      "grad_norm": 1.011638879776001,
+      "learning_rate": 2.356742697078832e-05,
+      "loss": 0.5209,
+      "step": 13224
+    },
+    {
+      "epoch": 16.928,
+      "grad_norm": 1.0729031562805176,
+      "learning_rate": 2.3565426170468187e-05,
+      "loss": 0.5676,
+      "step": 13225
+    },
+    {
+      "epoch": 16.92928,
+      "grad_norm": 0.988837718963623,
+      "learning_rate": 2.356342537014806e-05,
+      "loss": 0.5399,
+      "step": 13226
+    },
+    {
+      "epoch": 16.93056,
+      "grad_norm": 0.9590950608253479,
+      "learning_rate": 2.3561424569827934e-05,
+      "loss": 0.4938,
+      "step": 13227
+    },
+    {
+      "epoch": 16.93184,
+      "grad_norm": 0.9690542817115784,
+      "learning_rate": 2.3559423769507806e-05,
+      "loss": 0.4957,
+      "step": 13228
+    },
+    {
+      "epoch": 16.93312,
+      "grad_norm": 1.0184234380722046,
+      "learning_rate": 2.3557422969187674e-05,
+      "loss": 0.5152,
+      "step": 13229
+    },
+    {
+      "epoch": 16.9344,
+      "grad_norm": 1.0626813173294067,
+      "learning_rate": 2.3555422168867546e-05,
+      "loss": 0.558,
+      "step": 13230
+    },
+    {
+      "epoch": 16.93568,
+      "grad_norm": 0.9686869978904724,
+      "learning_rate": 2.355342136854742e-05,
+      "loss": 0.5276,
+      "step": 13231
+    },
+    {
+      "epoch": 16.93696,
+      "grad_norm": 0.9972003102302551,
+      "learning_rate": 2.3551420568227293e-05,
+      "loss": 0.5422,
+      "step": 13232
+    },
+    {
+      "epoch": 16.93824,
+      "grad_norm": 1.024174690246582,
+      "learning_rate": 2.3549419767907162e-05,
+      "loss": 0.5273,
+      "step": 13233
+    },
+    {
+      "epoch": 16.93952,
+      "grad_norm": 1.0208815336227417,
+      "learning_rate": 2.3547418967587037e-05,
+      "loss": 0.5221,
+      "step": 13234
+    },
+    {
+      "epoch": 16.9408,
+      "grad_norm": 0.9970613718032837,
+      "learning_rate": 2.354541816726691e-05,
+      "loss": 0.5001,
+      "step": 13235
+    },
+    {
+      "epoch": 16.94208,
+      "grad_norm": 1.0984002351760864,
+      "learning_rate": 2.354341736694678e-05,
+      "loss": 0.5119,
+      "step": 13236
+    },
+    {
+      "epoch": 16.94336,
+      "grad_norm": 1.0116389989852905,
+      "learning_rate": 2.354141656662665e-05,
+      "loss": 0.541,
+      "step": 13237
+    },
+    {
+      "epoch": 16.94464,
+      "grad_norm": 1.0177326202392578,
+      "learning_rate": 2.3539415766306524e-05,
+      "loss": 0.5292,
+      "step": 13238
+    },
+    {
+      "epoch": 16.94592,
+      "grad_norm": 0.9551138877868652,
+      "learning_rate": 2.3537414965986396e-05,
+      "loss": 0.4805,
+      "step": 13239
+    },
+    {
+      "epoch": 16.9472,
+      "grad_norm": 1.0099093914031982,
+      "learning_rate": 2.3535414165666268e-05,
+      "loss": 0.5158,
+      "step": 13240
+    },
+    {
+      "epoch": 16.94848,
+      "grad_norm": 1.0355970859527588,
+      "learning_rate": 2.353341336534614e-05,
+      "loss": 0.517,
+      "step": 13241
+    },
+    {
+      "epoch": 16.94976,
+      "grad_norm": 1.0377235412597656,
+      "learning_rate": 2.3531412565026012e-05,
+      "loss": 0.5256,
+      "step": 13242
+    },
+    {
+      "epoch": 16.95104,
+      "grad_norm": 0.973661482334137,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.4841,
+      "step": 13243
+    },
+    {
+      "epoch": 16.95232,
+      "grad_norm": 0.950306236743927,
+      "learning_rate": 2.3527410964385756e-05,
+      "loss": 0.5096,
+      "step": 13244
+    },
+    {
+      "epoch": 16.9536,
+      "grad_norm": 1.070381999015808,
+      "learning_rate": 2.3525410164065627e-05,
+      "loss": 0.5756,
+      "step": 13245
+    },
+    {
+      "epoch": 16.95488,
+      "grad_norm": 1.0381155014038086,
+      "learning_rate": 2.35234093637455e-05,
+      "loss": 0.5439,
+      "step": 13246
+    },
+    {
+      "epoch": 16.95616,
+      "grad_norm": 1.0218827724456787,
+      "learning_rate": 2.352140856342537e-05,
+      "loss": 0.5185,
+      "step": 13247
+    },
+    {
+      "epoch": 16.95744,
+      "grad_norm": 1.0366841554641724,
+      "learning_rate": 2.3519407763105243e-05,
+      "loss": 0.5542,
+      "step": 13248
+    },
+    {
+      "epoch": 16.95872,
+      "grad_norm": 0.9674044251441956,
+      "learning_rate": 2.3517406962785115e-05,
+      "loss": 0.4658,
+      "step": 13249
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 0.9855802655220032,
+      "learning_rate": 2.3515406162464987e-05,
+      "loss": 0.5267,
+      "step": 13250
+    },
+    {
+      "epoch": 16.96128,
+      "grad_norm": 1.0081998109817505,
+      "learning_rate": 2.351340536214486e-05,
+      "loss": 0.5058,
+      "step": 13251
+    },
+    {
+      "epoch": 16.96256,
+      "grad_norm": 0.9830219149589539,
+      "learning_rate": 2.351140456182473e-05,
+      "loss": 0.4796,
+      "step": 13252
+    },
+    {
+      "epoch": 16.96384,
+      "grad_norm": 0.97882479429245,
+      "learning_rate": 2.3509403761504602e-05,
+      "loss": 0.4897,
+      "step": 13253
+    },
+    {
+      "epoch": 16.96512,
+      "grad_norm": 1.0170930624008179,
+      "learning_rate": 2.3507402961184474e-05,
+      "loss": 0.513,
+      "step": 13254
+    },
+    {
+      "epoch": 16.9664,
+      "grad_norm": 1.030259132385254,
+      "learning_rate": 2.350540216086435e-05,
+      "loss": 0.4973,
+      "step": 13255
+    },
+    {
+      "epoch": 16.96768,
+      "grad_norm": 1.0129339694976807,
+      "learning_rate": 2.3503401360544218e-05,
+      "loss": 0.4936,
+      "step": 13256
+    },
+    {
+      "epoch": 16.96896,
+      "grad_norm": 1.0687109231948853,
+      "learning_rate": 2.350140056022409e-05,
+      "loss": 0.5226,
+      "step": 13257
+    },
+    {
+      "epoch": 16.97024,
+      "grad_norm": 1.0357649326324463,
+      "learning_rate": 2.349939975990396e-05,
+      "loss": 0.5376,
+      "step": 13258
+    },
+    {
+      "epoch": 16.97152,
+      "grad_norm": 1.0332026481628418,
+      "learning_rate": 2.3497398959583837e-05,
+      "loss": 0.4773,
+      "step": 13259
+    },
+    {
+      "epoch": 16.9728,
+      "grad_norm": 0.9495790600776672,
+      "learning_rate": 2.3495398159263705e-05,
+      "loss": 0.4737,
+      "step": 13260
+    },
+    {
+      "epoch": 16.97408,
+      "grad_norm": 1.0137715339660645,
+      "learning_rate": 2.3493397358943577e-05,
+      "loss": 0.5459,
+      "step": 13261
+    },
+    {
+      "epoch": 16.97536,
+      "grad_norm": 0.9835166931152344,
+      "learning_rate": 2.349139655862345e-05,
+      "loss": 0.5019,
+      "step": 13262
+    },
+    {
+      "epoch": 16.97664,
+      "grad_norm": 1.000338077545166,
+      "learning_rate": 2.3489395758303324e-05,
+      "loss": 0.5223,
+      "step": 13263
+    },
+    {
+      "epoch": 16.97792,
+      "grad_norm": 0.9843599796295166,
+      "learning_rate": 2.3487394957983193e-05,
+      "loss": 0.4935,
+      "step": 13264
+    },
+    {
+      "epoch": 16.9792,
+      "grad_norm": 1.0332714319229126,
+      "learning_rate": 2.3485394157663065e-05,
+      "loss": 0.5263,
+      "step": 13265
+    },
+    {
+      "epoch": 16.98048,
+      "grad_norm": 1.008151650428772,
+      "learning_rate": 2.348339335734294e-05,
+      "loss": 0.5458,
+      "step": 13266
+    },
+    {
+      "epoch": 16.98176,
+      "grad_norm": 1.0616765022277832,
+      "learning_rate": 2.348139255702281e-05,
+      "loss": 0.5493,
+      "step": 13267
+    },
+    {
+      "epoch": 16.98304,
+      "grad_norm": 1.007568359375,
+      "learning_rate": 2.347939175670268e-05,
+      "loss": 0.5289,
+      "step": 13268
+    },
+    {
+      "epoch": 16.98432,
+      "grad_norm": 1.0306227207183838,
+      "learning_rate": 2.3477390956382552e-05,
+      "loss": 0.5618,
+      "step": 13269
+    },
+    {
+      "epoch": 16.9856,
+      "grad_norm": 1.0478519201278687,
+      "learning_rate": 2.3475390156062427e-05,
+      "loss": 0.5575,
+      "step": 13270
+    },
+    {
+      "epoch": 16.98688,
+      "grad_norm": 0.9916083216667175,
+      "learning_rate": 2.34733893557423e-05,
+      "loss": 0.5117,
+      "step": 13271
+    },
+    {
+      "epoch": 16.98816,
+      "grad_norm": 1.0166401863098145,
+      "learning_rate": 2.3471388555422168e-05,
+      "loss": 0.5053,
+      "step": 13272
+    },
+    {
+      "epoch": 16.98944,
+      "grad_norm": 0.9944577217102051,
+      "learning_rate": 2.3469387755102043e-05,
+      "loss": 0.5063,
+      "step": 13273
+    },
+    {
+      "epoch": 16.99072,
+      "grad_norm": 1.0171867609024048,
+      "learning_rate": 2.3467386954781915e-05,
+      "loss": 0.5275,
+      "step": 13274
+    },
+    {
+      "epoch": 16.992,
+      "grad_norm": 1.002347469329834,
+      "learning_rate": 2.3465386154461786e-05,
+      "loss": 0.504,
+      "step": 13275
+    },
+    {
+      "epoch": 16.99328,
+      "grad_norm": 0.9947987794876099,
+      "learning_rate": 2.3463385354141655e-05,
+      "loss": 0.4975,
+      "step": 13276
+    },
+    {
+      "epoch": 16.99456,
+      "grad_norm": 1.0772300958633423,
+      "learning_rate": 2.346138455382153e-05,
+      "loss": 0.5841,
+      "step": 13277
+    },
+    {
+      "epoch": 16.99584,
+      "grad_norm": 1.0796716213226318,
+      "learning_rate": 2.3459383753501402e-05,
+      "loss": 0.5478,
+      "step": 13278
+    },
+    {
+      "epoch": 16.99712,
+      "grad_norm": 1.0109952688217163,
+      "learning_rate": 2.3457382953181274e-05,
+      "loss": 0.497,
+      "step": 13279
+    },
+    {
+      "epoch": 16.9984,
+      "grad_norm": 1.00393545627594,
+      "learning_rate": 2.3455382152861146e-05,
+      "loss": 0.4905,
+      "step": 13280
+    },
+    {
+      "epoch": 16.99968,
+      "grad_norm": 1.0317285060882568,
+      "learning_rate": 2.3453381352541018e-05,
+      "loss": 0.5052,
+      "step": 13281
+    },
+    {
+      "epoch": 17.00096,
+      "grad_norm": 2.4376354217529297,
+      "learning_rate": 2.345138055222089e-05,
+      "loss": 0.999,
+      "step": 13282
+    },
+    {
+      "epoch": 17.00224,
+      "grad_norm": 0.9784718155860901,
+      "learning_rate": 2.344937975190076e-05,
+      "loss": 0.4915,
+      "step": 13283
+    },
+    {
+      "epoch": 17.00352,
+      "grad_norm": 1.0082448720932007,
+      "learning_rate": 2.3447378951580633e-05,
+      "loss": 0.4813,
+      "step": 13284
+    },
+    {
+      "epoch": 17.0048,
+      "grad_norm": 1.0144582986831665,
+      "learning_rate": 2.3445378151260505e-05,
+      "loss": 0.5056,
+      "step": 13285
+    },
+    {
+      "epoch": 17.00608,
+      "grad_norm": 0.9649600386619568,
+      "learning_rate": 2.3443377350940377e-05,
+      "loss": 0.4901,
+      "step": 13286
+    },
+    {
+      "epoch": 17.00736,
+      "grad_norm": 0.9816195368766785,
+      "learning_rate": 2.344137655062025e-05,
+      "loss": 0.491,
+      "step": 13287
+    },
+    {
+      "epoch": 17.00864,
+      "grad_norm": 0.9808714985847473,
+      "learning_rate": 2.343937575030012e-05,
+      "loss": 0.4976,
+      "step": 13288
+    },
+    {
+      "epoch": 17.00992,
+      "grad_norm": 0.9497635364532471,
+      "learning_rate": 2.3437374949979992e-05,
+      "loss": 0.4637,
+      "step": 13289
+    },
+    {
+      "epoch": 17.0112,
+      "grad_norm": 1.0557515621185303,
+      "learning_rate": 2.3435374149659864e-05,
+      "loss": 0.5267,
+      "step": 13290
+    },
+    {
+      "epoch": 17.01248,
+      "grad_norm": 1.0407460927963257,
+      "learning_rate": 2.3433373349339736e-05,
+      "loss": 0.5172,
+      "step": 13291
+    },
+    {
+      "epoch": 17.01376,
+      "grad_norm": 1.0044418573379517,
+      "learning_rate": 2.3431372549019608e-05,
+      "loss": 0.5482,
+      "step": 13292
+    },
+    {
+      "epoch": 17.01504,
+      "grad_norm": 1.036288857460022,
+      "learning_rate": 2.342937174869948e-05,
+      "loss": 0.54,
+      "step": 13293
+    },
+    {
+      "epoch": 17.01632,
+      "grad_norm": 1.0211939811706543,
+      "learning_rate": 2.3427370948379355e-05,
+      "loss": 0.4912,
+      "step": 13294
+    },
+    {
+      "epoch": 17.0176,
+      "grad_norm": 0.9974328279495239,
+      "learning_rate": 2.3425370148059224e-05,
+      "loss": 0.4945,
+      "step": 13295
+    },
+    {
+      "epoch": 17.01888,
+      "grad_norm": 1.0395445823669434,
+      "learning_rate": 2.3423369347739095e-05,
+      "loss": 0.5235,
+      "step": 13296
+    },
+    {
+      "epoch": 17.02016,
+      "grad_norm": 1.0685923099517822,
+      "learning_rate": 2.3421368547418967e-05,
+      "loss": 0.5011,
+      "step": 13297
+    },
+    {
+      "epoch": 17.02144,
+      "grad_norm": 0.9720517992973328,
+      "learning_rate": 2.3419367747098843e-05,
+      "loss": 0.4791,
+      "step": 13298
+    },
+    {
+      "epoch": 17.02272,
+      "grad_norm": 1.0298750400543213,
+      "learning_rate": 2.341736694677871e-05,
+      "loss": 0.5083,
+      "step": 13299
+    },
+    {
+      "epoch": 17.024,
+      "grad_norm": 1.0151759386062622,
+      "learning_rate": 2.3415366146458583e-05,
+      "loss": 0.4656,
+      "step": 13300
+    },
+    {
+      "epoch": 17.02528,
+      "grad_norm": 1.1034287214279175,
+      "learning_rate": 2.3413365346138458e-05,
+      "loss": 0.5559,
+      "step": 13301
+    },
+    {
+      "epoch": 17.02656,
+      "grad_norm": 0.9982373714447021,
+      "learning_rate": 2.341136454581833e-05,
+      "loss": 0.4586,
+      "step": 13302
+    },
+    {
+      "epoch": 17.02784,
+      "grad_norm": 1.0372883081436157,
+      "learning_rate": 2.34093637454982e-05,
+      "loss": 0.4949,
+      "step": 13303
+    },
+    {
+      "epoch": 17.02912,
+      "grad_norm": 1.0765658617019653,
+      "learning_rate": 2.340736294517807e-05,
+      "loss": 0.5206,
+      "step": 13304
+    },
+    {
+      "epoch": 17.0304,
+      "grad_norm": 1.0272283554077148,
+      "learning_rate": 2.3405362144857946e-05,
+      "loss": 0.4676,
+      "step": 13305
+    },
+    {
+      "epoch": 17.03168,
+      "grad_norm": 1.091143012046814,
+      "learning_rate": 2.3403361344537817e-05,
+      "loss": 0.5431,
+      "step": 13306
+    },
+    {
+      "epoch": 17.03296,
+      "grad_norm": 0.9813713431358337,
+      "learning_rate": 2.3401360544217686e-05,
+      "loss": 0.4549,
+      "step": 13307
+    },
+    {
+      "epoch": 17.03424,
+      "grad_norm": 0.9769755601882935,
+      "learning_rate": 2.339935974389756e-05,
+      "loss": 0.468,
+      "step": 13308
+    },
+    {
+      "epoch": 17.03552,
+      "grad_norm": 1.0127575397491455,
+      "learning_rate": 2.3397358943577433e-05,
+      "loss": 0.5391,
+      "step": 13309
+    },
+    {
+      "epoch": 17.0368,
+      "grad_norm": 1.0613768100738525,
+      "learning_rate": 2.3395358143257305e-05,
+      "loss": 0.5372,
+      "step": 13310
+    },
+    {
+      "epoch": 17.03808,
+      "grad_norm": 0.9925457239151001,
+      "learning_rate": 2.3393357342937173e-05,
+      "loss": 0.4997,
+      "step": 13311
+    },
+    {
+      "epoch": 17.03936,
+      "grad_norm": 0.9370602965354919,
+      "learning_rate": 2.339135654261705e-05,
+      "loss": 0.474,
+      "step": 13312
+    },
+    {
+      "epoch": 17.04064,
+      "grad_norm": 1.0538114309310913,
+      "learning_rate": 2.338935574229692e-05,
+      "loss": 0.4834,
+      "step": 13313
+    },
+    {
+      "epoch": 17.04192,
+      "grad_norm": 1.0021767616271973,
+      "learning_rate": 2.3387354941976792e-05,
+      "loss": 0.503,
+      "step": 13314
+    },
+    {
+      "epoch": 17.0432,
+      "grad_norm": 1.0035548210144043,
+      "learning_rate": 2.3385354141656664e-05,
+      "loss": 0.4991,
+      "step": 13315
+    },
+    {
+      "epoch": 17.04448,
+      "grad_norm": 1.0278247594833374,
+      "learning_rate": 2.3383353341336536e-05,
+      "loss": 0.4885,
+      "step": 13316
+    },
+    {
+      "epoch": 17.04576,
+      "grad_norm": 1.0547192096710205,
+      "learning_rate": 2.3381352541016408e-05,
+      "loss": 0.5255,
+      "step": 13317
+    },
+    {
+      "epoch": 17.04704,
+      "grad_norm": 1.0186964273452759,
+      "learning_rate": 2.337935174069628e-05,
+      "loss": 0.5052,
+      "step": 13318
+    },
+    {
+      "epoch": 17.04832,
+      "grad_norm": 1.0712857246398926,
+      "learning_rate": 2.337735094037615e-05,
+      "loss": 0.4951,
+      "step": 13319
+    },
+    {
+      "epoch": 17.0496,
+      "grad_norm": 1.0552232265472412,
+      "learning_rate": 2.3375350140056023e-05,
+      "loss": 0.5391,
+      "step": 13320
+    },
+    {
+      "epoch": 17.05088,
+      "grad_norm": 0.9899628758430481,
+      "learning_rate": 2.3373349339735895e-05,
+      "loss": 0.4996,
+      "step": 13321
+    },
+    {
+      "epoch": 17.05216,
+      "grad_norm": 1.085460901260376,
+      "learning_rate": 2.3371348539415767e-05,
+      "loss": 0.5247,
+      "step": 13322
+    },
+    {
+      "epoch": 17.05344,
+      "grad_norm": 1.0540425777435303,
+      "learning_rate": 2.336934773909564e-05,
+      "loss": 0.5785,
+      "step": 13323
+    },
+    {
+      "epoch": 17.05472,
+      "grad_norm": 1.0214723348617554,
+      "learning_rate": 2.336734693877551e-05,
+      "loss": 0.5305,
+      "step": 13324
+    },
+    {
+      "epoch": 17.056,
+      "grad_norm": 1.0059022903442383,
+      "learning_rate": 2.3365346138455383e-05,
+      "loss": 0.487,
+      "step": 13325
+    },
+    {
+      "epoch": 17.05728,
+      "grad_norm": 1.063796877861023,
+      "learning_rate": 2.3363345338135258e-05,
+      "loss": 0.5328,
+      "step": 13326
+    },
+    {
+      "epoch": 17.05856,
+      "grad_norm": 1.0171808004379272,
+      "learning_rate": 2.3361344537815126e-05,
+      "loss": 0.4785,
+      "step": 13327
+    },
+    {
+      "epoch": 17.05984,
+      "grad_norm": 1.0254714488983154,
+      "learning_rate": 2.3359343737494998e-05,
+      "loss": 0.501,
+      "step": 13328
+    },
+    {
+      "epoch": 17.06112,
+      "grad_norm": 1.0111802816390991,
+      "learning_rate": 2.335734293717487e-05,
+      "loss": 0.5116,
+      "step": 13329
+    },
+    {
+      "epoch": 17.0624,
+      "grad_norm": 1.0492706298828125,
+      "learning_rate": 2.3355342136854745e-05,
+      "loss": 0.5129,
+      "step": 13330
+    },
+    {
+      "epoch": 17.06368,
+      "grad_norm": 1.0869669914245605,
+      "learning_rate": 2.3353341336534614e-05,
+      "loss": 0.4696,
+      "step": 13331
+    },
+    {
+      "epoch": 17.06496,
+      "grad_norm": 1.0306788682937622,
+      "learning_rate": 2.3351340536214486e-05,
+      "loss": 0.4977,
+      "step": 13332
+    },
+    {
+      "epoch": 17.06624,
+      "grad_norm": 0.9910968542098999,
+      "learning_rate": 2.334933973589436e-05,
+      "loss": 0.4801,
+      "step": 13333
+    },
+    {
+      "epoch": 17.06752,
+      "grad_norm": 0.9870019555091858,
+      "learning_rate": 2.3347338935574233e-05,
+      "loss": 0.4426,
+      "step": 13334
+    },
+    {
+      "epoch": 17.0688,
+      "grad_norm": 1.0062360763549805,
+      "learning_rate": 2.33453381352541e-05,
+      "loss": 0.5187,
+      "step": 13335
+    },
+    {
+      "epoch": 17.07008,
+      "grad_norm": 1.0343031883239746,
+      "learning_rate": 2.3343337334933973e-05,
+      "loss": 0.5119,
+      "step": 13336
+    },
+    {
+      "epoch": 17.07136,
+      "grad_norm": 1.0217961072921753,
+      "learning_rate": 2.334133653461385e-05,
+      "loss": 0.5541,
+      "step": 13337
+    },
+    {
+      "epoch": 17.07264,
+      "grad_norm": 1.0784296989440918,
+      "learning_rate": 2.333933573429372e-05,
+      "loss": 0.5345,
+      "step": 13338
+    },
+    {
+      "epoch": 17.07392,
+      "grad_norm": 1.029133677482605,
+      "learning_rate": 2.333733493397359e-05,
+      "loss": 0.5253,
+      "step": 13339
+    },
+    {
+      "epoch": 17.0752,
+      "grad_norm": 1.0548806190490723,
+      "learning_rate": 2.3335334133653464e-05,
+      "loss": 0.5053,
+      "step": 13340
+    },
+    {
+      "epoch": 17.07648,
+      "grad_norm": 1.0157487392425537,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.4751,
+      "step": 13341
+    },
+    {
+      "epoch": 17.07776,
+      "grad_norm": 1.0173470973968506,
+      "learning_rate": 2.3331332533013208e-05,
+      "loss": 0.4752,
+      "step": 13342
+    },
+    {
+      "epoch": 17.07904,
+      "grad_norm": 0.9745169281959534,
+      "learning_rate": 2.3329331732693076e-05,
+      "loss": 0.486,
+      "step": 13343
+    },
+    {
+      "epoch": 17.08032,
+      "grad_norm": 0.9814679622650146,
+      "learning_rate": 2.332733093237295e-05,
+      "loss": 0.5124,
+      "step": 13344
+    },
+    {
+      "epoch": 17.0816,
+      "grad_norm": 0.9977136850357056,
+      "learning_rate": 2.3325330132052823e-05,
+      "loss": 0.4984,
+      "step": 13345
+    },
+    {
+      "epoch": 17.08288,
+      "grad_norm": 0.9431197643280029,
+      "learning_rate": 2.3323329331732695e-05,
+      "loss": 0.4411,
+      "step": 13346
+    },
+    {
+      "epoch": 17.08416,
+      "grad_norm": 1.0017660856246948,
+      "learning_rate": 2.3321328531412567e-05,
+      "loss": 0.5211,
+      "step": 13347
+    },
+    {
+      "epoch": 17.08544,
+      "grad_norm": 0.9581298232078552,
+      "learning_rate": 2.331932773109244e-05,
+      "loss": 0.5069,
+      "step": 13348
+    },
+    {
+      "epoch": 17.08672,
+      "grad_norm": 1.0260916948318481,
+      "learning_rate": 2.331732693077231e-05,
+      "loss": 0.5305,
+      "step": 13349
+    },
+    {
+      "epoch": 17.088,
+      "grad_norm": 1.032827615737915,
+      "learning_rate": 2.3315326130452183e-05,
+      "loss": 0.5041,
+      "step": 13350
+    },
+    {
+      "epoch": 17.08928,
+      "grad_norm": 1.0370451211929321,
+      "learning_rate": 2.3313325330132054e-05,
+      "loss": 0.4872,
+      "step": 13351
+    },
+    {
+      "epoch": 17.09056,
+      "grad_norm": 1.0834031105041504,
+      "learning_rate": 2.3311324529811926e-05,
+      "loss": 0.5399,
+      "step": 13352
+    },
+    {
+      "epoch": 17.09184,
+      "grad_norm": 1.0523909330368042,
+      "learning_rate": 2.3309323729491798e-05,
+      "loss": 0.4924,
+      "step": 13353
+    },
+    {
+      "epoch": 17.09312,
+      "grad_norm": 1.0100966691970825,
+      "learning_rate": 2.330732292917167e-05,
+      "loss": 0.5309,
+      "step": 13354
+    },
+    {
+      "epoch": 17.0944,
+      "grad_norm": 1.0047094821929932,
+      "learning_rate": 2.3305322128851542e-05,
+      "loss": 0.4504,
+      "step": 13355
+    },
+    {
+      "epoch": 17.09568,
+      "grad_norm": 1.0484347343444824,
+      "learning_rate": 2.3303321328531414e-05,
+      "loss": 0.5239,
+      "step": 13356
+    },
+    {
+      "epoch": 17.09696,
+      "grad_norm": 1.0204596519470215,
+      "learning_rate": 2.3301320528211286e-05,
+      "loss": 0.5148,
+      "step": 13357
+    },
+    {
+      "epoch": 17.09824,
+      "grad_norm": 0.9914649128913879,
+      "learning_rate": 2.3299319727891157e-05,
+      "loss": 0.46,
+      "step": 13358
+    },
+    {
+      "epoch": 17.09952,
+      "grad_norm": 1.0509700775146484,
+      "learning_rate": 2.329731892757103e-05,
+      "loss": 0.4891,
+      "step": 13359
+    },
+    {
+      "epoch": 17.1008,
+      "grad_norm": 1.0409367084503174,
+      "learning_rate": 2.32953181272509e-05,
+      "loss": 0.5121,
+      "step": 13360
+    },
+    {
+      "epoch": 17.10208,
+      "grad_norm": 1.0810407400131226,
+      "learning_rate": 2.3293317326930776e-05,
+      "loss": 0.4976,
+      "step": 13361
+    },
+    {
+      "epoch": 17.10336,
+      "grad_norm": 1.0771454572677612,
+      "learning_rate": 2.3291316526610645e-05,
+      "loss": 0.4679,
+      "step": 13362
+    },
+    {
+      "epoch": 17.10464,
+      "grad_norm": 1.1189974546432495,
+      "learning_rate": 2.3289315726290517e-05,
+      "loss": 0.572,
+      "step": 13363
+    },
+    {
+      "epoch": 17.10592,
+      "grad_norm": 1.017249584197998,
+      "learning_rate": 2.328731492597039e-05,
+      "loss": 0.5068,
+      "step": 13364
+    },
+    {
+      "epoch": 17.1072,
+      "grad_norm": 0.990572988986969,
+      "learning_rate": 2.3285314125650264e-05,
+      "loss": 0.4766,
+      "step": 13365
+    },
+    {
+      "epoch": 17.10848,
+      "grad_norm": 1.0602976083755493,
+      "learning_rate": 2.3283313325330132e-05,
+      "loss": 0.4712,
+      "step": 13366
+    },
+    {
+      "epoch": 17.10976,
+      "grad_norm": 1.0583609342575073,
+      "learning_rate": 2.3281312525010004e-05,
+      "loss": 0.5533,
+      "step": 13367
+    },
+    {
+      "epoch": 17.11104,
+      "grad_norm": 1.0422335863113403,
+      "learning_rate": 2.327931172468988e-05,
+      "loss": 0.5399,
+      "step": 13368
+    },
+    {
+      "epoch": 17.11232,
+      "grad_norm": 0.9954720139503479,
+      "learning_rate": 2.327731092436975e-05,
+      "loss": 0.4705,
+      "step": 13369
+    },
+    {
+      "epoch": 17.1136,
+      "grad_norm": 1.0625905990600586,
+      "learning_rate": 2.327531012404962e-05,
+      "loss": 0.5102,
+      "step": 13370
+    },
+    {
+      "epoch": 17.11488,
+      "grad_norm": 1.03073251247406,
+      "learning_rate": 2.327330932372949e-05,
+      "loss": 0.5117,
+      "step": 13371
+    },
+    {
+      "epoch": 17.11616,
+      "grad_norm": 0.9936673045158386,
+      "learning_rate": 2.3271308523409367e-05,
+      "loss": 0.5368,
+      "step": 13372
+    },
+    {
+      "epoch": 17.11744,
+      "grad_norm": 1.014151692390442,
+      "learning_rate": 2.326930772308924e-05,
+      "loss": 0.4523,
+      "step": 13373
+    },
+    {
+      "epoch": 17.11872,
+      "grad_norm": 1.0433018207550049,
+      "learning_rate": 2.3267306922769107e-05,
+      "loss": 0.5037,
+      "step": 13374
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 1.0825138092041016,
+      "learning_rate": 2.326530612244898e-05,
+      "loss": 0.5088,
+      "step": 13375
+    },
+    {
+      "epoch": 17.12128,
+      "grad_norm": 1.0675740242004395,
+      "learning_rate": 2.3263305322128854e-05,
+      "loss": 0.5024,
+      "step": 13376
+    },
+    {
+      "epoch": 17.12256,
+      "grad_norm": 1.002501130104065,
+      "learning_rate": 2.3261304521808726e-05,
+      "loss": 0.4748,
+      "step": 13377
+    },
+    {
+      "epoch": 17.12384,
+      "grad_norm": 1.042352557182312,
+      "learning_rate": 2.3259303721488595e-05,
+      "loss": 0.4869,
+      "step": 13378
+    },
+    {
+      "epoch": 17.12512,
+      "grad_norm": 1.0172719955444336,
+      "learning_rate": 2.325730292116847e-05,
+      "loss": 0.4917,
+      "step": 13379
+    },
+    {
+      "epoch": 17.1264,
+      "grad_norm": 0.9971911907196045,
+      "learning_rate": 2.325530212084834e-05,
+      "loss": 0.4934,
+      "step": 13380
+    },
+    {
+      "epoch": 17.12768,
+      "grad_norm": 0.9829162359237671,
+      "learning_rate": 2.3253301320528213e-05,
+      "loss": 0.5106,
+      "step": 13381
+    },
+    {
+      "epoch": 17.12896,
+      "grad_norm": 1.0853031873703003,
+      "learning_rate": 2.3251300520208082e-05,
+      "loss": 0.5248,
+      "step": 13382
+    },
+    {
+      "epoch": 17.13024,
+      "grad_norm": 1.0231006145477295,
+      "learning_rate": 2.3249299719887957e-05,
+      "loss": 0.5012,
+      "step": 13383
+    },
+    {
+      "epoch": 17.13152,
+      "grad_norm": 1.0397844314575195,
+      "learning_rate": 2.324729891956783e-05,
+      "loss": 0.5402,
+      "step": 13384
+    },
+    {
+      "epoch": 17.1328,
+      "grad_norm": 1.0465348958969116,
+      "learning_rate": 2.32452981192477e-05,
+      "loss": 0.5115,
+      "step": 13385
+    },
+    {
+      "epoch": 17.13408,
+      "grad_norm": 1.0421459674835205,
+      "learning_rate": 2.3243297318927573e-05,
+      "loss": 0.4913,
+      "step": 13386
+    },
+    {
+      "epoch": 17.13536,
+      "grad_norm": 0.991023063659668,
+      "learning_rate": 2.3241296518607445e-05,
+      "loss": 0.5016,
+      "step": 13387
+    },
+    {
+      "epoch": 17.13664,
+      "grad_norm": 1.0662450790405273,
+      "learning_rate": 2.3239295718287316e-05,
+      "loss": 0.5424,
+      "step": 13388
+    },
+    {
+      "epoch": 17.13792,
+      "grad_norm": 1.0365347862243652,
+      "learning_rate": 2.323729491796719e-05,
+      "loss": 0.534,
+      "step": 13389
+    },
+    {
+      "epoch": 17.1392,
+      "grad_norm": 1.0068080425262451,
+      "learning_rate": 2.323529411764706e-05,
+      "loss": 0.4582,
+      "step": 13390
+    },
+    {
+      "epoch": 17.14048,
+      "grad_norm": 0.9789687991142273,
+      "learning_rate": 2.3233293317326932e-05,
+      "loss": 0.4683,
+      "step": 13391
+    },
+    {
+      "epoch": 17.14176,
+      "grad_norm": 0.9680244326591492,
+      "learning_rate": 2.3231292517006804e-05,
+      "loss": 0.47,
+      "step": 13392
+    },
+    {
+      "epoch": 17.14304,
+      "grad_norm": 1.0483875274658203,
+      "learning_rate": 2.3229291716686676e-05,
+      "loss": 0.5163,
+      "step": 13393
+    },
+    {
+      "epoch": 17.14432,
+      "grad_norm": 0.998321533203125,
+      "learning_rate": 2.3227290916366548e-05,
+      "loss": 0.5058,
+      "step": 13394
+    },
+    {
+      "epoch": 17.1456,
+      "grad_norm": 1.0452215671539307,
+      "learning_rate": 2.322529011604642e-05,
+      "loss": 0.5377,
+      "step": 13395
+    },
+    {
+      "epoch": 17.14688,
+      "grad_norm": 1.0813066959381104,
+      "learning_rate": 2.322328931572629e-05,
+      "loss": 0.4744,
+      "step": 13396
+    },
+    {
+      "epoch": 17.14816,
+      "grad_norm": 1.0850361585617065,
+      "learning_rate": 2.3221288515406163e-05,
+      "loss": 0.5534,
+      "step": 13397
+    },
+    {
+      "epoch": 17.14944,
+      "grad_norm": 1.034781575202942,
+      "learning_rate": 2.3219287715086035e-05,
+      "loss": 0.5279,
+      "step": 13398
+    },
+    {
+      "epoch": 17.15072,
+      "grad_norm": 1.0639398097991943,
+      "learning_rate": 2.3217286914765907e-05,
+      "loss": 0.5009,
+      "step": 13399
+    },
+    {
+      "epoch": 17.152,
+      "grad_norm": 1.0536409616470337,
+      "learning_rate": 2.3215286114445782e-05,
+      "loss": 0.5132,
+      "step": 13400
+    },
+    {
+      "epoch": 17.15328,
+      "grad_norm": 1.0476963520050049,
+      "learning_rate": 2.321328531412565e-05,
+      "loss": 0.5241,
+      "step": 13401
+    },
+    {
+      "epoch": 17.15456,
+      "grad_norm": 1.0148664712905884,
+      "learning_rate": 2.3211284513805522e-05,
+      "loss": 0.4843,
+      "step": 13402
+    },
+    {
+      "epoch": 17.15584,
+      "grad_norm": 0.9978867769241333,
+      "learning_rate": 2.3209283713485394e-05,
+      "loss": 0.5003,
+      "step": 13403
+    },
+    {
+      "epoch": 17.15712,
+      "grad_norm": 1.0880756378173828,
+      "learning_rate": 2.320728291316527e-05,
+      "loss": 0.5188,
+      "step": 13404
+    },
+    {
+      "epoch": 17.1584,
+      "grad_norm": 1.064671516418457,
+      "learning_rate": 2.3205282112845138e-05,
+      "loss": 0.5898,
+      "step": 13405
+    },
+    {
+      "epoch": 17.15968,
+      "grad_norm": 1.0149790048599243,
+      "learning_rate": 2.320328131252501e-05,
+      "loss": 0.4774,
+      "step": 13406
+    },
+    {
+      "epoch": 17.16096,
+      "grad_norm": 0.992540180683136,
+      "learning_rate": 2.3201280512204885e-05,
+      "loss": 0.4796,
+      "step": 13407
+    },
+    {
+      "epoch": 17.16224,
+      "grad_norm": 1.0193244218826294,
+      "learning_rate": 2.3199279711884757e-05,
+      "loss": 0.503,
+      "step": 13408
+    },
+    {
+      "epoch": 17.16352,
+      "grad_norm": 1.0211812257766724,
+      "learning_rate": 2.3197278911564625e-05,
+      "loss": 0.5344,
+      "step": 13409
+    },
+    {
+      "epoch": 17.1648,
+      "grad_norm": 0.9962586164474487,
+      "learning_rate": 2.3195278111244497e-05,
+      "loss": 0.477,
+      "step": 13410
+    },
+    {
+      "epoch": 17.16608,
+      "grad_norm": 1.0423258543014526,
+      "learning_rate": 2.3193277310924373e-05,
+      "loss": 0.5273,
+      "step": 13411
+    },
+    {
+      "epoch": 17.16736,
+      "grad_norm": 1.0666061639785767,
+      "learning_rate": 2.3191276510604244e-05,
+      "loss": 0.4738,
+      "step": 13412
+    },
+    {
+      "epoch": 17.16864,
+      "grad_norm": 1.0447877645492554,
+      "learning_rate": 2.3189275710284113e-05,
+      "loss": 0.5042,
+      "step": 13413
+    },
+    {
+      "epoch": 17.16992,
+      "grad_norm": 1.041363000869751,
+      "learning_rate": 2.3187274909963988e-05,
+      "loss": 0.5165,
+      "step": 13414
+    },
+    {
+      "epoch": 17.1712,
+      "grad_norm": 0.9531150460243225,
+      "learning_rate": 2.318527410964386e-05,
+      "loss": 0.4656,
+      "step": 13415
+    },
+    {
+      "epoch": 17.17248,
+      "grad_norm": 0.975502610206604,
+      "learning_rate": 2.3183273309323732e-05,
+      "loss": 0.4421,
+      "step": 13416
+    },
+    {
+      "epoch": 17.17376,
+      "grad_norm": 1.0709878206253052,
+      "learning_rate": 2.31812725090036e-05,
+      "loss": 0.5181,
+      "step": 13417
+    },
+    {
+      "epoch": 17.17504,
+      "grad_norm": 1.0228393077850342,
+      "learning_rate": 2.3179271708683476e-05,
+      "loss": 0.488,
+      "step": 13418
+    },
+    {
+      "epoch": 17.17632,
+      "grad_norm": 0.9908788800239563,
+      "learning_rate": 2.3177270908363347e-05,
+      "loss": 0.5135,
+      "step": 13419
+    },
+    {
+      "epoch": 17.1776,
+      "grad_norm": 1.0089093446731567,
+      "learning_rate": 2.317527010804322e-05,
+      "loss": 0.4801,
+      "step": 13420
+    },
+    {
+      "epoch": 17.17888,
+      "grad_norm": 1.0426452159881592,
+      "learning_rate": 2.317326930772309e-05,
+      "loss": 0.5075,
+      "step": 13421
+    },
+    {
+      "epoch": 17.18016,
+      "grad_norm": 1.021355390548706,
+      "learning_rate": 2.3171268507402963e-05,
+      "loss": 0.4907,
+      "step": 13422
+    },
+    {
+      "epoch": 17.18144,
+      "grad_norm": 1.0270839929580688,
+      "learning_rate": 2.3169267707082835e-05,
+      "loss": 0.4559,
+      "step": 13423
+    },
+    {
+      "epoch": 17.18272,
+      "grad_norm": 0.9954252243041992,
+      "learning_rate": 2.3167266906762707e-05,
+      "loss": 0.5026,
+      "step": 13424
+    },
+    {
+      "epoch": 17.184,
+      "grad_norm": 0.9519053101539612,
+      "learning_rate": 2.316526610644258e-05,
+      "loss": 0.465,
+      "step": 13425
+    },
+    {
+      "epoch": 17.18528,
+      "grad_norm": 0.966871976852417,
+      "learning_rate": 2.316326530612245e-05,
+      "loss": 0.4922,
+      "step": 13426
+    },
+    {
+      "epoch": 17.18656,
+      "grad_norm": 1.1276544332504272,
+      "learning_rate": 2.3161264505802322e-05,
+      "loss": 0.5007,
+      "step": 13427
+    },
+    {
+      "epoch": 17.18784,
+      "grad_norm": 1.0328774452209473,
+      "learning_rate": 2.3159263705482194e-05,
+      "loss": 0.4846,
+      "step": 13428
+    },
+    {
+      "epoch": 17.18912,
+      "grad_norm": 1.0253841876983643,
+      "learning_rate": 2.3157262905162066e-05,
+      "loss": 0.48,
+      "step": 13429
+    },
+    {
+      "epoch": 17.1904,
+      "grad_norm": 1.1044909954071045,
+      "learning_rate": 2.3155262104841938e-05,
+      "loss": 0.4854,
+      "step": 13430
+    },
+    {
+      "epoch": 17.19168,
+      "grad_norm": 1.1031279563903809,
+      "learning_rate": 2.315326130452181e-05,
+      "loss": 0.5431,
+      "step": 13431
+    },
+    {
+      "epoch": 17.19296,
+      "grad_norm": 1.0845425128936768,
+      "learning_rate": 2.315126050420168e-05,
+      "loss": 0.5446,
+      "step": 13432
+    },
+    {
+      "epoch": 17.19424,
+      "grad_norm": 1.0719091892242432,
+      "learning_rate": 2.3149259703881553e-05,
+      "loss": 0.525,
+      "step": 13433
+    },
+    {
+      "epoch": 17.19552,
+      "grad_norm": 1.115065097808838,
+      "learning_rate": 2.3147258903561425e-05,
+      "loss": 0.5301,
+      "step": 13434
+    },
+    {
+      "epoch": 17.1968,
+      "grad_norm": 1.0398420095443726,
+      "learning_rate": 2.3145258103241297e-05,
+      "loss": 0.5075,
+      "step": 13435
+    },
+    {
+      "epoch": 17.19808,
+      "grad_norm": 0.9958305358886719,
+      "learning_rate": 2.314325730292117e-05,
+      "loss": 0.4951,
+      "step": 13436
+    },
+    {
+      "epoch": 17.19936,
+      "grad_norm": 1.0259790420532227,
+      "learning_rate": 2.314125650260104e-05,
+      "loss": 0.4844,
+      "step": 13437
+    },
+    {
+      "epoch": 17.20064,
+      "grad_norm": 1.1417039632797241,
+      "learning_rate": 2.3139255702280913e-05,
+      "loss": 0.5455,
+      "step": 13438
+    },
+    {
+      "epoch": 17.20192,
+      "grad_norm": 1.0283827781677246,
+      "learning_rate": 2.3137254901960788e-05,
+      "loss": 0.4931,
+      "step": 13439
+    },
+    {
+      "epoch": 17.2032,
+      "grad_norm": 0.9671483635902405,
+      "learning_rate": 2.3135254101640656e-05,
+      "loss": 0.4683,
+      "step": 13440
+    },
+    {
+      "epoch": 17.20448,
+      "grad_norm": 0.9877151250839233,
+      "learning_rate": 2.3133253301320528e-05,
+      "loss": 0.4812,
+      "step": 13441
+    },
+    {
+      "epoch": 17.20576,
+      "grad_norm": 1.0239356756210327,
+      "learning_rate": 2.31312525010004e-05,
+      "loss": 0.4782,
+      "step": 13442
+    },
+    {
+      "epoch": 17.20704,
+      "grad_norm": 1.1128602027893066,
+      "learning_rate": 2.3129251700680275e-05,
+      "loss": 0.5034,
+      "step": 13443
+    },
+    {
+      "epoch": 17.20832,
+      "grad_norm": 1.0178550481796265,
+      "learning_rate": 2.3127250900360144e-05,
+      "loss": 0.5063,
+      "step": 13444
+    },
+    {
+      "epoch": 17.209600000000002,
+      "grad_norm": 1.0738757848739624,
+      "learning_rate": 2.3125250100040016e-05,
+      "loss": 0.5127,
+      "step": 13445
+    },
+    {
+      "epoch": 17.21088,
+      "grad_norm": 1.0246436595916748,
+      "learning_rate": 2.312324929971989e-05,
+      "loss": 0.5128,
+      "step": 13446
+    },
+    {
+      "epoch": 17.21216,
+      "grad_norm": 1.0808448791503906,
+      "learning_rate": 2.3121248499399763e-05,
+      "loss": 0.5116,
+      "step": 13447
+    },
+    {
+      "epoch": 17.21344,
+      "grad_norm": 1.0280914306640625,
+      "learning_rate": 2.311924769907963e-05,
+      "loss": 0.5175,
+      "step": 13448
+    },
+    {
+      "epoch": 17.21472,
+      "grad_norm": 1.0112377405166626,
+      "learning_rate": 2.3117246898759503e-05,
+      "loss": 0.4666,
+      "step": 13449
+    },
+    {
+      "epoch": 17.216,
+      "grad_norm": 1.0588704347610474,
+      "learning_rate": 2.311524609843938e-05,
+      "loss": 0.4821,
+      "step": 13450
+    },
+    {
+      "epoch": 17.21728,
+      "grad_norm": 1.005056381225586,
+      "learning_rate": 2.311324529811925e-05,
+      "loss": 0.5253,
+      "step": 13451
+    },
+    {
+      "epoch": 17.21856,
+      "grad_norm": 0.979955792427063,
+      "learning_rate": 2.311124449779912e-05,
+      "loss": 0.4832,
+      "step": 13452
+    },
+    {
+      "epoch": 17.21984,
+      "grad_norm": 0.989325225353241,
+      "learning_rate": 2.3109243697478994e-05,
+      "loss": 0.514,
+      "step": 13453
+    },
+    {
+      "epoch": 17.22112,
+      "grad_norm": 1.1037169694900513,
+      "learning_rate": 2.3107242897158866e-05,
+      "loss": 0.5589,
+      "step": 13454
+    },
+    {
+      "epoch": 17.2224,
+      "grad_norm": 1.101965069770813,
+      "learning_rate": 2.3105242096838738e-05,
+      "loss": 0.5082,
+      "step": 13455
+    },
+    {
+      "epoch": 17.22368,
+      "grad_norm": 1.0133678913116455,
+      "learning_rate": 2.3103241296518606e-05,
+      "loss": 0.5253,
+      "step": 13456
+    },
+    {
+      "epoch": 17.22496,
+      "grad_norm": 0.9965882301330566,
+      "learning_rate": 2.310124049619848e-05,
+      "loss": 0.4828,
+      "step": 13457
+    },
+    {
+      "epoch": 17.22624,
+      "grad_norm": 1.0856853723526,
+      "learning_rate": 2.3099239695878353e-05,
+      "loss": 0.533,
+      "step": 13458
+    },
+    {
+      "epoch": 17.22752,
+      "grad_norm": 0.9921742081642151,
+      "learning_rate": 2.3097238895558225e-05,
+      "loss": 0.5221,
+      "step": 13459
+    },
+    {
+      "epoch": 17.2288,
+      "grad_norm": 1.0070276260375977,
+      "learning_rate": 2.3095238095238097e-05,
+      "loss": 0.4822,
+      "step": 13460
+    },
+    {
+      "epoch": 17.23008,
+      "grad_norm": 1.0681109428405762,
+      "learning_rate": 2.309323729491797e-05,
+      "loss": 0.5146,
+      "step": 13461
+    },
+    {
+      "epoch": 17.23136,
+      "grad_norm": 1.0588202476501465,
+      "learning_rate": 2.309123649459784e-05,
+      "loss": 0.484,
+      "step": 13462
+    },
+    {
+      "epoch": 17.23264,
+      "grad_norm": 1.072527527809143,
+      "learning_rate": 2.3089235694277712e-05,
+      "loss": 0.5442,
+      "step": 13463
+    },
+    {
+      "epoch": 17.23392,
+      "grad_norm": 1.0427266359329224,
+      "learning_rate": 2.3087234893957584e-05,
+      "loss": 0.4935,
+      "step": 13464
+    },
+    {
+      "epoch": 17.2352,
+      "grad_norm": 1.0477319955825806,
+      "learning_rate": 2.3085234093637456e-05,
+      "loss": 0.4786,
+      "step": 13465
+    },
+    {
+      "epoch": 17.23648,
+      "grad_norm": 1.1097187995910645,
+      "learning_rate": 2.3083233293317328e-05,
+      "loss": 0.5198,
+      "step": 13466
+    },
+    {
+      "epoch": 17.23776,
+      "grad_norm": 1.0430151224136353,
+      "learning_rate": 2.30812324929972e-05,
+      "loss": 0.5408,
+      "step": 13467
+    },
+    {
+      "epoch": 17.23904,
+      "grad_norm": 1.084143042564392,
+      "learning_rate": 2.3079231692677072e-05,
+      "loss": 0.5259,
+      "step": 13468
+    },
+    {
+      "epoch": 17.24032,
+      "grad_norm": 1.034934163093567,
+      "learning_rate": 2.3077230892356944e-05,
+      "loss": 0.5068,
+      "step": 13469
+    },
+    {
+      "epoch": 17.2416,
+      "grad_norm": 0.9962584972381592,
+      "learning_rate": 2.3075230092036815e-05,
+      "loss": 0.4753,
+      "step": 13470
+    },
+    {
+      "epoch": 17.24288,
+      "grad_norm": 1.0584473609924316,
+      "learning_rate": 2.3073229291716687e-05,
+      "loss": 0.4369,
+      "step": 13471
+    },
+    {
+      "epoch": 17.24416,
+      "grad_norm": 1.0417580604553223,
+      "learning_rate": 2.307122849139656e-05,
+      "loss": 0.547,
+      "step": 13472
+    },
+    {
+      "epoch": 17.24544,
+      "grad_norm": 0.9787613749504089,
+      "learning_rate": 2.306922769107643e-05,
+      "loss": 0.4825,
+      "step": 13473
+    },
+    {
+      "epoch": 17.24672,
+      "grad_norm": 0.97577303647995,
+      "learning_rate": 2.3067226890756306e-05,
+      "loss": 0.4792,
+      "step": 13474
+    },
+    {
+      "epoch": 17.248,
+      "grad_norm": 1.0570039749145508,
+      "learning_rate": 2.3065226090436175e-05,
+      "loss": 0.4757,
+      "step": 13475
+    },
+    {
+      "epoch": 17.24928,
+      "grad_norm": 0.9857975840568542,
+      "learning_rate": 2.3063225290116047e-05,
+      "loss": 0.541,
+      "step": 13476
+    },
+    {
+      "epoch": 17.25056,
+      "grad_norm": 1.0609130859375,
+      "learning_rate": 2.306122448979592e-05,
+      "loss": 0.547,
+      "step": 13477
+    },
+    {
+      "epoch": 17.25184,
+      "grad_norm": 1.0602465867996216,
+      "learning_rate": 2.3059223689475794e-05,
+      "loss": 0.5039,
+      "step": 13478
+    },
+    {
+      "epoch": 17.25312,
+      "grad_norm": 1.033346176147461,
+      "learning_rate": 2.3057222889155662e-05,
+      "loss": 0.4925,
+      "step": 13479
+    },
+    {
+      "epoch": 17.2544,
+      "grad_norm": 1.0470037460327148,
+      "learning_rate": 2.3055222088835534e-05,
+      "loss": 0.5261,
+      "step": 13480
+    },
+    {
+      "epoch": 17.25568,
+      "grad_norm": 1.0750315189361572,
+      "learning_rate": 2.305322128851541e-05,
+      "loss": 0.5026,
+      "step": 13481
+    },
+    {
+      "epoch": 17.25696,
+      "grad_norm": 1.0455011129379272,
+      "learning_rate": 2.305122048819528e-05,
+      "loss": 0.4955,
+      "step": 13482
+    },
+    {
+      "epoch": 17.25824,
+      "grad_norm": 0.9804509282112122,
+      "learning_rate": 2.304921968787515e-05,
+      "loss": 0.4493,
+      "step": 13483
+    },
+    {
+      "epoch": 17.25952,
+      "grad_norm": 1.0425822734832764,
+      "learning_rate": 2.304721888755502e-05,
+      "loss": 0.5265,
+      "step": 13484
+    },
+    {
+      "epoch": 17.2608,
+      "grad_norm": 1.0288803577423096,
+      "learning_rate": 2.3045218087234897e-05,
+      "loss": 0.5335,
+      "step": 13485
+    },
+    {
+      "epoch": 17.26208,
+      "grad_norm": 1.0013490915298462,
+      "learning_rate": 2.304321728691477e-05,
+      "loss": 0.5016,
+      "step": 13486
+    },
+    {
+      "epoch": 17.26336,
+      "grad_norm": 0.9813631176948547,
+      "learning_rate": 2.3041216486594637e-05,
+      "loss": 0.4945,
+      "step": 13487
+    },
+    {
+      "epoch": 17.26464,
+      "grad_norm": 1.004841923713684,
+      "learning_rate": 2.303921568627451e-05,
+      "loss": 0.5357,
+      "step": 13488
+    },
+    {
+      "epoch": 17.26592,
+      "grad_norm": 1.007587194442749,
+      "learning_rate": 2.3037214885954384e-05,
+      "loss": 0.4896,
+      "step": 13489
+    },
+    {
+      "epoch": 17.2672,
+      "grad_norm": 1.0714519023895264,
+      "learning_rate": 2.3035214085634256e-05,
+      "loss": 0.5111,
+      "step": 13490
+    },
+    {
+      "epoch": 17.26848,
+      "grad_norm": 1.0348763465881348,
+      "learning_rate": 2.3033213285314124e-05,
+      "loss": 0.4855,
+      "step": 13491
+    },
+    {
+      "epoch": 17.26976,
+      "grad_norm": 1.0198546648025513,
+      "learning_rate": 2.3031212484994e-05,
+      "loss": 0.5217,
+      "step": 13492
+    },
+    {
+      "epoch": 17.27104,
+      "grad_norm": 1.0204856395721436,
+      "learning_rate": 2.302921168467387e-05,
+      "loss": 0.5029,
+      "step": 13493
+    },
+    {
+      "epoch": 17.27232,
+      "grad_norm": 1.0740776062011719,
+      "learning_rate": 2.3027210884353743e-05,
+      "loss": 0.5458,
+      "step": 13494
+    },
+    {
+      "epoch": 17.2736,
+      "grad_norm": 1.0617952346801758,
+      "learning_rate": 2.3025210084033612e-05,
+      "loss": 0.5018,
+      "step": 13495
+    },
+    {
+      "epoch": 17.27488,
+      "grad_norm": 0.9945394992828369,
+      "learning_rate": 2.3023209283713487e-05,
+      "loss": 0.5174,
+      "step": 13496
+    },
+    {
+      "epoch": 17.27616,
+      "grad_norm": 1.033125400543213,
+      "learning_rate": 2.302120848339336e-05,
+      "loss": 0.5133,
+      "step": 13497
+    },
+    {
+      "epoch": 17.27744,
+      "grad_norm": 1.0384323596954346,
+      "learning_rate": 2.301920768307323e-05,
+      "loss": 0.5283,
+      "step": 13498
+    },
+    {
+      "epoch": 17.27872,
+      "grad_norm": 1.0113089084625244,
+      "learning_rate": 2.3017206882753103e-05,
+      "loss": 0.4772,
+      "step": 13499
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 1.0131186246871948,
+      "learning_rate": 2.3015206082432975e-05,
+      "loss": 0.4923,
+      "step": 13500
+    },
+    {
+      "epoch": 17.28128,
+      "grad_norm": 1.0085313320159912,
+      "learning_rate": 2.3013205282112846e-05,
+      "loss": 0.525,
+      "step": 13501
+    },
+    {
+      "epoch": 17.28256,
+      "grad_norm": 0.9978111386299133,
+      "learning_rate": 2.3011204481792718e-05,
+      "loss": 0.5065,
+      "step": 13502
+    },
+    {
+      "epoch": 17.28384,
+      "grad_norm": 1.0002354383468628,
+      "learning_rate": 2.300920368147259e-05,
+      "loss": 0.4846,
+      "step": 13503
+    },
+    {
+      "epoch": 17.28512,
+      "grad_norm": 1.0522061586380005,
+      "learning_rate": 2.3007202881152462e-05,
+      "loss": 0.5395,
+      "step": 13504
+    },
+    {
+      "epoch": 17.2864,
+      "grad_norm": 1.0922776460647583,
+      "learning_rate": 2.3005202080832334e-05,
+      "loss": 0.5106,
+      "step": 13505
+    },
+    {
+      "epoch": 17.28768,
+      "grad_norm": 1.027981162071228,
+      "learning_rate": 2.3003201280512206e-05,
+      "loss": 0.4703,
+      "step": 13506
+    },
+    {
+      "epoch": 17.28896,
+      "grad_norm": 1.0434186458587646,
+      "learning_rate": 2.3001200480192078e-05,
+      "loss": 0.5184,
+      "step": 13507
+    },
+    {
+      "epoch": 17.29024,
+      "grad_norm": 1.038582682609558,
+      "learning_rate": 2.299919967987195e-05,
+      "loss": 0.5318,
+      "step": 13508
+    },
+    {
+      "epoch": 17.29152,
+      "grad_norm": 1.0460221767425537,
+      "learning_rate": 2.299719887955182e-05,
+      "loss": 0.5173,
+      "step": 13509
+    },
+    {
+      "epoch": 17.2928,
+      "grad_norm": 1.0172295570373535,
+      "learning_rate": 2.2995198079231693e-05,
+      "loss": 0.4987,
+      "step": 13510
+    },
+    {
+      "epoch": 17.29408,
+      "grad_norm": 1.1028789281845093,
+      "learning_rate": 2.2993197278911565e-05,
+      "loss": 0.545,
+      "step": 13511
+    },
+    {
+      "epoch": 17.29536,
+      "grad_norm": 1.0653462409973145,
+      "learning_rate": 2.2991196478591437e-05,
+      "loss": 0.5482,
+      "step": 13512
+    },
+    {
+      "epoch": 17.29664,
+      "grad_norm": 1.0706826448440552,
+      "learning_rate": 2.2989195678271312e-05,
+      "loss": 0.5027,
+      "step": 13513
+    },
+    {
+      "epoch": 17.29792,
+      "grad_norm": 1.0168458223342896,
+      "learning_rate": 2.298719487795118e-05,
+      "loss": 0.5167,
+      "step": 13514
+    },
+    {
+      "epoch": 17.2992,
+      "grad_norm": 0.9857696890830994,
+      "learning_rate": 2.2985194077631052e-05,
+      "loss": 0.5102,
+      "step": 13515
+    },
+    {
+      "epoch": 17.30048,
+      "grad_norm": 1.0181854963302612,
+      "learning_rate": 2.2983193277310924e-05,
+      "loss": 0.4949,
+      "step": 13516
+    },
+    {
+      "epoch": 17.30176,
+      "grad_norm": 1.10274338722229,
+      "learning_rate": 2.29811924769908e-05,
+      "loss": 0.5449,
+      "step": 13517
+    },
+    {
+      "epoch": 17.30304,
+      "grad_norm": 1.0308619737625122,
+      "learning_rate": 2.2979191676670668e-05,
+      "loss": 0.5152,
+      "step": 13518
+    },
+    {
+      "epoch": 17.30432,
+      "grad_norm": 1.0274208784103394,
+      "learning_rate": 2.297719087635054e-05,
+      "loss": 0.5126,
+      "step": 13519
+    },
+    {
+      "epoch": 17.3056,
+      "grad_norm": 1.0590589046478271,
+      "learning_rate": 2.2975190076030415e-05,
+      "loss": 0.516,
+      "step": 13520
+    },
+    {
+      "epoch": 17.30688,
+      "grad_norm": 1.0562139749526978,
+      "learning_rate": 2.2973189275710287e-05,
+      "loss": 0.5627,
+      "step": 13521
+    },
+    {
+      "epoch": 17.30816,
+      "grad_norm": 1.004634141921997,
+      "learning_rate": 2.2971188475390155e-05,
+      "loss": 0.49,
+      "step": 13522
+    },
+    {
+      "epoch": 17.30944,
+      "grad_norm": 1.0177441835403442,
+      "learning_rate": 2.2969187675070027e-05,
+      "loss": 0.4957,
+      "step": 13523
+    },
+    {
+      "epoch": 17.31072,
+      "grad_norm": 1.0136632919311523,
+      "learning_rate": 2.2967186874749903e-05,
+      "loss": 0.507,
+      "step": 13524
+    },
+    {
+      "epoch": 17.312,
+      "grad_norm": 0.9950380325317383,
+      "learning_rate": 2.2965186074429774e-05,
+      "loss": 0.4692,
+      "step": 13525
+    },
+    {
+      "epoch": 17.31328,
+      "grad_norm": 1.0641790628433228,
+      "learning_rate": 2.2963185274109643e-05,
+      "loss": 0.4797,
+      "step": 13526
+    },
+    {
+      "epoch": 17.31456,
+      "grad_norm": 1.120981216430664,
+      "learning_rate": 2.2961184473789518e-05,
+      "loss": 0.5234,
+      "step": 13527
+    },
+    {
+      "epoch": 17.31584,
+      "grad_norm": 1.044729471206665,
+      "learning_rate": 2.295918367346939e-05,
+      "loss": 0.4913,
+      "step": 13528
+    },
+    {
+      "epoch": 17.31712,
+      "grad_norm": 1.029380202293396,
+      "learning_rate": 2.2957182873149262e-05,
+      "loss": 0.5046,
+      "step": 13529
+    },
+    {
+      "epoch": 17.3184,
+      "grad_norm": 1.0592548847198486,
+      "learning_rate": 2.295518207282913e-05,
+      "loss": 0.4982,
+      "step": 13530
+    },
+    {
+      "epoch": 17.31968,
+      "grad_norm": 1.0422892570495605,
+      "learning_rate": 2.2953181272509006e-05,
+      "loss": 0.5258,
+      "step": 13531
+    },
+    {
+      "epoch": 17.32096,
+      "grad_norm": 1.0550442934036255,
+      "learning_rate": 2.2951180472188877e-05,
+      "loss": 0.5656,
+      "step": 13532
+    },
+    {
+      "epoch": 17.32224,
+      "grad_norm": 1.009332299232483,
+      "learning_rate": 2.294917967186875e-05,
+      "loss": 0.473,
+      "step": 13533
+    },
+    {
+      "epoch": 17.32352,
+      "grad_norm": 1.0183335542678833,
+      "learning_rate": 2.294717887154862e-05,
+      "loss": 0.5133,
+      "step": 13534
+    },
+    {
+      "epoch": 17.3248,
+      "grad_norm": 1.013777732849121,
+      "learning_rate": 2.2945178071228493e-05,
+      "loss": 0.5414,
+      "step": 13535
+    },
+    {
+      "epoch": 17.32608,
+      "grad_norm": 1.0590869188308716,
+      "learning_rate": 2.2943177270908365e-05,
+      "loss": 0.5051,
+      "step": 13536
+    },
+    {
+      "epoch": 17.32736,
+      "grad_norm": 1.0427137613296509,
+      "learning_rate": 2.2941176470588237e-05,
+      "loss": 0.5346,
+      "step": 13537
+    },
+    {
+      "epoch": 17.32864,
+      "grad_norm": 0.9942336678504944,
+      "learning_rate": 2.293917567026811e-05,
+      "loss": 0.4948,
+      "step": 13538
+    },
+    {
+      "epoch": 17.32992,
+      "grad_norm": 1.0325044393539429,
+      "learning_rate": 2.293717486994798e-05,
+      "loss": 0.5118,
+      "step": 13539
+    },
+    {
+      "epoch": 17.3312,
+      "grad_norm": 1.0152339935302734,
+      "learning_rate": 2.2935174069627852e-05,
+      "loss": 0.5144,
+      "step": 13540
+    },
+    {
+      "epoch": 17.33248,
+      "grad_norm": 1.0435190200805664,
+      "learning_rate": 2.2933173269307724e-05,
+      "loss": 0.5059,
+      "step": 13541
+    },
+    {
+      "epoch": 17.33376,
+      "grad_norm": 1.0658721923828125,
+      "learning_rate": 2.2931172468987596e-05,
+      "loss": 0.5129,
+      "step": 13542
+    },
+    {
+      "epoch": 17.33504,
+      "grad_norm": 1.0623787641525269,
+      "learning_rate": 2.2929171668667468e-05,
+      "loss": 0.4906,
+      "step": 13543
+    },
+    {
+      "epoch": 17.33632,
+      "grad_norm": 1.0778841972351074,
+      "learning_rate": 2.292717086834734e-05,
+      "loss": 0.5309,
+      "step": 13544
+    },
+    {
+      "epoch": 17.3376,
+      "grad_norm": 1.048268437385559,
+      "learning_rate": 2.292517006802721e-05,
+      "loss": 0.5067,
+      "step": 13545
+    },
+    {
+      "epoch": 17.33888,
+      "grad_norm": 1.0772168636322021,
+      "learning_rate": 2.2923169267707083e-05,
+      "loss": 0.5369,
+      "step": 13546
+    },
+    {
+      "epoch": 17.34016,
+      "grad_norm": 1.0442776679992676,
+      "learning_rate": 2.2921168467386955e-05,
+      "loss": 0.5185,
+      "step": 13547
+    },
+    {
+      "epoch": 17.34144,
+      "grad_norm": 1.0044589042663574,
+      "learning_rate": 2.2919167667066827e-05,
+      "loss": 0.4811,
+      "step": 13548
+    },
+    {
+      "epoch": 17.34272,
+      "grad_norm": 1.067563533782959,
+      "learning_rate": 2.29171668667467e-05,
+      "loss": 0.5228,
+      "step": 13549
+    },
+    {
+      "epoch": 17.344,
+      "grad_norm": 1.026764988899231,
+      "learning_rate": 2.291516606642657e-05,
+      "loss": 0.5043,
+      "step": 13550
+    },
+    {
+      "epoch": 17.34528,
+      "grad_norm": 1.0351366996765137,
+      "learning_rate": 2.2913165266106443e-05,
+      "loss": 0.5151,
+      "step": 13551
+    },
+    {
+      "epoch": 17.34656,
+      "grad_norm": 1.0662847757339478,
+      "learning_rate": 2.2911164465786318e-05,
+      "loss": 0.5239,
+      "step": 13552
+    },
+    {
+      "epoch": 17.34784,
+      "grad_norm": 1.1130237579345703,
+      "learning_rate": 2.2909163665466186e-05,
+      "loss": 0.5437,
+      "step": 13553
+    },
+    {
+      "epoch": 17.34912,
+      "grad_norm": 1.0121694803237915,
+      "learning_rate": 2.2907162865146058e-05,
+      "loss": 0.4844,
+      "step": 13554
+    },
+    {
+      "epoch": 17.3504,
+      "grad_norm": 1.0110938549041748,
+      "learning_rate": 2.290516206482593e-05,
+      "loss": 0.4779,
+      "step": 13555
+    },
+    {
+      "epoch": 17.35168,
+      "grad_norm": 1.0952173471450806,
+      "learning_rate": 2.2903161264505805e-05,
+      "loss": 0.5603,
+      "step": 13556
+    },
+    {
+      "epoch": 17.35296,
+      "grad_norm": 1.018991231918335,
+      "learning_rate": 2.2901160464185674e-05,
+      "loss": 0.5099,
+      "step": 13557
+    },
+    {
+      "epoch": 17.35424,
+      "grad_norm": 1.0593852996826172,
+      "learning_rate": 2.2899159663865546e-05,
+      "loss": 0.5467,
+      "step": 13558
+    },
+    {
+      "epoch": 17.35552,
+      "grad_norm": 1.0670006275177002,
+      "learning_rate": 2.289715886354542e-05,
+      "loss": 0.495,
+      "step": 13559
+    },
+    {
+      "epoch": 17.3568,
+      "grad_norm": 1.0311278104782104,
+      "learning_rate": 2.2895158063225293e-05,
+      "loss": 0.481,
+      "step": 13560
+    },
+    {
+      "epoch": 17.35808,
+      "grad_norm": 1.0428582429885864,
+      "learning_rate": 2.289315726290516e-05,
+      "loss": 0.4989,
+      "step": 13561
+    },
+    {
+      "epoch": 17.35936,
+      "grad_norm": 1.0917315483093262,
+      "learning_rate": 2.2891156462585033e-05,
+      "loss": 0.5338,
+      "step": 13562
+    },
+    {
+      "epoch": 17.36064,
+      "grad_norm": 0.999539852142334,
+      "learning_rate": 2.288915566226491e-05,
+      "loss": 0.5455,
+      "step": 13563
+    },
+    {
+      "epoch": 17.36192,
+      "grad_norm": 1.0183244943618774,
+      "learning_rate": 2.288715486194478e-05,
+      "loss": 0.515,
+      "step": 13564
+    },
+    {
+      "epoch": 17.3632,
+      "grad_norm": 1.0455024242401123,
+      "learning_rate": 2.288515406162465e-05,
+      "loss": 0.5141,
+      "step": 13565
+    },
+    {
+      "epoch": 17.36448,
+      "grad_norm": 1.0188703536987305,
+      "learning_rate": 2.2883153261304524e-05,
+      "loss": 0.4722,
+      "step": 13566
+    },
+    {
+      "epoch": 17.36576,
+      "grad_norm": 1.0280158519744873,
+      "learning_rate": 2.2881152460984396e-05,
+      "loss": 0.5013,
+      "step": 13567
+    },
+    {
+      "epoch": 17.36704,
+      "grad_norm": 1.048551082611084,
+      "learning_rate": 2.2879151660664268e-05,
+      "loss": 0.5232,
+      "step": 13568
+    },
+    {
+      "epoch": 17.36832,
+      "grad_norm": 0.9960765838623047,
+      "learning_rate": 2.2877150860344136e-05,
+      "loss": 0.5111,
+      "step": 13569
+    },
+    {
+      "epoch": 17.3696,
+      "grad_norm": 0.9852240085601807,
+      "learning_rate": 2.287515006002401e-05,
+      "loss": 0.4906,
+      "step": 13570
+    },
+    {
+      "epoch": 17.37088,
+      "grad_norm": 1.0797041654586792,
+      "learning_rate": 2.2873149259703883e-05,
+      "loss": 0.5014,
+      "step": 13571
+    },
+    {
+      "epoch": 17.37216,
+      "grad_norm": 1.0308854579925537,
+      "learning_rate": 2.2871148459383755e-05,
+      "loss": 0.4957,
+      "step": 13572
+    },
+    {
+      "epoch": 17.37344,
+      "grad_norm": 1.0648754835128784,
+      "learning_rate": 2.2869147659063627e-05,
+      "loss": 0.4862,
+      "step": 13573
+    },
+    {
+      "epoch": 17.37472,
+      "grad_norm": 1.0644230842590332,
+      "learning_rate": 2.28671468587435e-05,
+      "loss": 0.5365,
+      "step": 13574
+    },
+    {
+      "epoch": 17.376,
+      "grad_norm": 1.071189522743225,
+      "learning_rate": 2.286514605842337e-05,
+      "loss": 0.5117,
+      "step": 13575
+    },
+    {
+      "epoch": 17.37728,
+      "grad_norm": 1.0346791744232178,
+      "learning_rate": 2.2863145258103242e-05,
+      "loss": 0.532,
+      "step": 13576
+    },
+    {
+      "epoch": 17.37856,
+      "grad_norm": 1.0855563879013062,
+      "learning_rate": 2.2861144457783114e-05,
+      "loss": 0.5115,
+      "step": 13577
+    },
+    {
+      "epoch": 17.37984,
+      "grad_norm": 0.9857675433158875,
+      "learning_rate": 2.2859143657462986e-05,
+      "loss": 0.473,
+      "step": 13578
+    },
+    {
+      "epoch": 17.38112,
+      "grad_norm": 0.993156373500824,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.4923,
+      "step": 13579
+    },
+    {
+      "epoch": 17.3824,
+      "grad_norm": 1.099673867225647,
+      "learning_rate": 2.285514205682273e-05,
+      "loss": 0.5255,
+      "step": 13580
+    },
+    {
+      "epoch": 17.38368,
+      "grad_norm": 1.0772475004196167,
+      "learning_rate": 2.2853141256502602e-05,
+      "loss": 0.5019,
+      "step": 13581
+    },
+    {
+      "epoch": 17.38496,
+      "grad_norm": 1.078121304512024,
+      "learning_rate": 2.2851140456182474e-05,
+      "loss": 0.5288,
+      "step": 13582
+    },
+    {
+      "epoch": 17.38624,
+      "grad_norm": 0.9957267642021179,
+      "learning_rate": 2.2849139655862345e-05,
+      "loss": 0.4716,
+      "step": 13583
+    },
+    {
+      "epoch": 17.38752,
+      "grad_norm": 1.0318092107772827,
+      "learning_rate": 2.2847138855542217e-05,
+      "loss": 0.4994,
+      "step": 13584
+    },
+    {
+      "epoch": 17.3888,
+      "grad_norm": 1.078931212425232,
+      "learning_rate": 2.284513805522209e-05,
+      "loss": 0.538,
+      "step": 13585
+    },
+    {
+      "epoch": 17.39008,
+      "grad_norm": 1.092041254043579,
+      "learning_rate": 2.284313725490196e-05,
+      "loss": 0.5409,
+      "step": 13586
+    },
+    {
+      "epoch": 17.39136,
+      "grad_norm": 1.0563862323760986,
+      "learning_rate": 2.2841136454581836e-05,
+      "loss": 0.4859,
+      "step": 13587
+    },
+    {
+      "epoch": 17.39264,
+      "grad_norm": 1.045304298400879,
+      "learning_rate": 2.2839135654261705e-05,
+      "loss": 0.4956,
+      "step": 13588
+    },
+    {
+      "epoch": 17.39392,
+      "grad_norm": 1.0128166675567627,
+      "learning_rate": 2.2837134853941577e-05,
+      "loss": 0.476,
+      "step": 13589
+    },
+    {
+      "epoch": 17.3952,
+      "grad_norm": 0.9987632036209106,
+      "learning_rate": 2.283513405362145e-05,
+      "loss": 0.4838,
+      "step": 13590
+    },
+    {
+      "epoch": 17.39648,
+      "grad_norm": 1.012587547302246,
+      "learning_rate": 2.2833133253301324e-05,
+      "loss": 0.4904,
+      "step": 13591
+    },
+    {
+      "epoch": 17.39776,
+      "grad_norm": 1.0492074489593506,
+      "learning_rate": 2.2831132452981192e-05,
+      "loss": 0.5372,
+      "step": 13592
+    },
+    {
+      "epoch": 17.39904,
+      "grad_norm": 1.0137112140655518,
+      "learning_rate": 2.2829131652661064e-05,
+      "loss": 0.4681,
+      "step": 13593
+    },
+    {
+      "epoch": 17.40032,
+      "grad_norm": 1.0790600776672363,
+      "learning_rate": 2.282713085234094e-05,
+      "loss": 0.5596,
+      "step": 13594
+    },
+    {
+      "epoch": 17.4016,
+      "grad_norm": 0.9867367744445801,
+      "learning_rate": 2.282513005202081e-05,
+      "loss": 0.4537,
+      "step": 13595
+    },
+    {
+      "epoch": 17.40288,
+      "grad_norm": 1.0197902917861938,
+      "learning_rate": 2.282312925170068e-05,
+      "loss": 0.5013,
+      "step": 13596
+    },
+    {
+      "epoch": 17.40416,
+      "grad_norm": 1.0174394845962524,
+      "learning_rate": 2.282112845138055e-05,
+      "loss": 0.512,
+      "step": 13597
+    },
+    {
+      "epoch": 17.40544,
+      "grad_norm": 1.000441312789917,
+      "learning_rate": 2.2819127651060427e-05,
+      "loss": 0.4996,
+      "step": 13598
+    },
+    {
+      "epoch": 17.40672,
+      "grad_norm": 1.0385953187942505,
+      "learning_rate": 2.28171268507403e-05,
+      "loss": 0.5161,
+      "step": 13599
+    },
+    {
+      "epoch": 17.408,
+      "grad_norm": 1.1014004945755005,
+      "learning_rate": 2.2815126050420167e-05,
+      "loss": 0.4972,
+      "step": 13600
+    },
+    {
+      "epoch": 17.40928,
+      "grad_norm": 1.0908480882644653,
+      "learning_rate": 2.281312525010004e-05,
+      "loss": 0.5294,
+      "step": 13601
+    },
+    {
+      "epoch": 17.41056,
+      "grad_norm": 1.0491387844085693,
+      "learning_rate": 2.2811124449779914e-05,
+      "loss": 0.5295,
+      "step": 13602
+    },
+    {
+      "epoch": 17.41184,
+      "grad_norm": 1.1002565622329712,
+      "learning_rate": 2.2809123649459786e-05,
+      "loss": 0.5095,
+      "step": 13603
+    },
+    {
+      "epoch": 17.41312,
+      "grad_norm": 1.077600359916687,
+      "learning_rate": 2.2807122849139654e-05,
+      "loss": 0.485,
+      "step": 13604
+    },
+    {
+      "epoch": 17.4144,
+      "grad_norm": 1.1393417119979858,
+      "learning_rate": 2.280512204881953e-05,
+      "loss": 0.5553,
+      "step": 13605
+    },
+    {
+      "epoch": 17.41568,
+      "grad_norm": 1.087091088294983,
+      "learning_rate": 2.28031212484994e-05,
+      "loss": 0.5519,
+      "step": 13606
+    },
+    {
+      "epoch": 17.41696,
+      "grad_norm": 1.0322144031524658,
+      "learning_rate": 2.2801120448179273e-05,
+      "loss": 0.5036,
+      "step": 13607
+    },
+    {
+      "epoch": 17.41824,
+      "grad_norm": 1.0630682706832886,
+      "learning_rate": 2.2799119647859142e-05,
+      "loss": 0.5113,
+      "step": 13608
+    },
+    {
+      "epoch": 17.41952,
+      "grad_norm": 1.0402500629425049,
+      "learning_rate": 2.2797118847539017e-05,
+      "loss": 0.4692,
+      "step": 13609
+    },
+    {
+      "epoch": 17.4208,
+      "grad_norm": 1.02988600730896,
+      "learning_rate": 2.279511804721889e-05,
+      "loss": 0.4878,
+      "step": 13610
+    },
+    {
+      "epoch": 17.42208,
+      "grad_norm": 1.0270915031433105,
+      "learning_rate": 2.279311724689876e-05,
+      "loss": 0.4991,
+      "step": 13611
+    },
+    {
+      "epoch": 17.42336,
+      "grad_norm": 0.9723539352416992,
+      "learning_rate": 2.2791116446578633e-05,
+      "loss": 0.4838,
+      "step": 13612
+    },
+    {
+      "epoch": 17.42464,
+      "grad_norm": 1.06536066532135,
+      "learning_rate": 2.2789115646258505e-05,
+      "loss": 0.5087,
+      "step": 13613
+    },
+    {
+      "epoch": 17.42592,
+      "grad_norm": 1.0233323574066162,
+      "learning_rate": 2.2787114845938376e-05,
+      "loss": 0.5379,
+      "step": 13614
+    },
+    {
+      "epoch": 17.4272,
+      "grad_norm": 1.0208466053009033,
+      "learning_rate": 2.2785114045618248e-05,
+      "loss": 0.4942,
+      "step": 13615
+    },
+    {
+      "epoch": 17.42848,
+      "grad_norm": 1.020458698272705,
+      "learning_rate": 2.278311324529812e-05,
+      "loss": 0.5093,
+      "step": 13616
+    },
+    {
+      "epoch": 17.42976,
+      "grad_norm": 1.0593698024749756,
+      "learning_rate": 2.2781112444977992e-05,
+      "loss": 0.5102,
+      "step": 13617
+    },
+    {
+      "epoch": 17.43104,
+      "grad_norm": 1.0400246381759644,
+      "learning_rate": 2.2779111644657864e-05,
+      "loss": 0.5102,
+      "step": 13618
+    },
+    {
+      "epoch": 17.43232,
+      "grad_norm": 1.0877729654312134,
+      "learning_rate": 2.2777110844337736e-05,
+      "loss": 0.5376,
+      "step": 13619
+    },
+    {
+      "epoch": 17.4336,
+      "grad_norm": 1.0149822235107422,
+      "learning_rate": 2.2775110044017608e-05,
+      "loss": 0.5096,
+      "step": 13620
+    },
+    {
+      "epoch": 17.43488,
+      "grad_norm": 1.0287227630615234,
+      "learning_rate": 2.277310924369748e-05,
+      "loss": 0.5049,
+      "step": 13621
+    },
+    {
+      "epoch": 17.43616,
+      "grad_norm": 0.98603755235672,
+      "learning_rate": 2.277110844337735e-05,
+      "loss": 0.4833,
+      "step": 13622
+    },
+    {
+      "epoch": 17.43744,
+      "grad_norm": 1.0274131298065186,
+      "learning_rate": 2.2769107643057223e-05,
+      "loss": 0.4929,
+      "step": 13623
+    },
+    {
+      "epoch": 17.43872,
+      "grad_norm": 1.006752610206604,
+      "learning_rate": 2.2767106842737095e-05,
+      "loss": 0.4812,
+      "step": 13624
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 1.0845232009887695,
+      "learning_rate": 2.2765106042416967e-05,
+      "loss": 0.5392,
+      "step": 13625
+    },
+    {
+      "epoch": 17.44128,
+      "grad_norm": 1.0171184539794922,
+      "learning_rate": 2.2763105242096842e-05,
+      "loss": 0.4821,
+      "step": 13626
+    },
+    {
+      "epoch": 17.44256,
+      "grad_norm": 1.019047498703003,
+      "learning_rate": 2.276110444177671e-05,
+      "loss": 0.5066,
+      "step": 13627
+    },
+    {
+      "epoch": 17.44384,
+      "grad_norm": 0.9561603665351868,
+      "learning_rate": 2.2759103641456582e-05,
+      "loss": 0.4448,
+      "step": 13628
+    },
+    {
+      "epoch": 17.44512,
+      "grad_norm": 1.011527180671692,
+      "learning_rate": 2.2757102841136454e-05,
+      "loss": 0.4852,
+      "step": 13629
+    },
+    {
+      "epoch": 17.4464,
+      "grad_norm": 1.019681453704834,
+      "learning_rate": 2.275510204081633e-05,
+      "loss": 0.4853,
+      "step": 13630
+    },
+    {
+      "epoch": 17.44768,
+      "grad_norm": 1.005802035331726,
+      "learning_rate": 2.2753101240496198e-05,
+      "loss": 0.4922,
+      "step": 13631
+    },
+    {
+      "epoch": 17.44896,
+      "grad_norm": 1.0661100149154663,
+      "learning_rate": 2.275110044017607e-05,
+      "loss": 0.5282,
+      "step": 13632
+    },
+    {
+      "epoch": 17.45024,
+      "grad_norm": 1.1106263399124146,
+      "learning_rate": 2.2749099639855945e-05,
+      "loss": 0.5464,
+      "step": 13633
+    },
+    {
+      "epoch": 17.45152,
+      "grad_norm": 1.0604326725006104,
+      "learning_rate": 2.2747098839535817e-05,
+      "loss": 0.4823,
+      "step": 13634
+    },
+    {
+      "epoch": 17.4528,
+      "grad_norm": 1.089737057685852,
+      "learning_rate": 2.2745098039215685e-05,
+      "loss": 0.5092,
+      "step": 13635
+    },
+    {
+      "epoch": 17.45408,
+      "grad_norm": 1.071022391319275,
+      "learning_rate": 2.2743097238895557e-05,
+      "loss": 0.5489,
+      "step": 13636
+    },
+    {
+      "epoch": 17.45536,
+      "grad_norm": 1.0083519220352173,
+      "learning_rate": 2.2741096438575433e-05,
+      "loss": 0.479,
+      "step": 13637
+    },
+    {
+      "epoch": 17.45664,
+      "grad_norm": 1.0257757902145386,
+      "learning_rate": 2.2739095638255304e-05,
+      "loss": 0.5014,
+      "step": 13638
+    },
+    {
+      "epoch": 17.45792,
+      "grad_norm": 1.0760114192962646,
+      "learning_rate": 2.2737094837935173e-05,
+      "loss": 0.57,
+      "step": 13639
+    },
+    {
+      "epoch": 17.4592,
+      "grad_norm": 1.0728118419647217,
+      "learning_rate": 2.2735094037615048e-05,
+      "loss": 0.5392,
+      "step": 13640
+    },
+    {
+      "epoch": 17.46048,
+      "grad_norm": 1.043785572052002,
+      "learning_rate": 2.273309323729492e-05,
+      "loss": 0.5259,
+      "step": 13641
+    },
+    {
+      "epoch": 17.46176,
+      "grad_norm": 1.0340181589126587,
+      "learning_rate": 2.2731092436974792e-05,
+      "loss": 0.5326,
+      "step": 13642
+    },
+    {
+      "epoch": 17.46304,
+      "grad_norm": 1.094905138015747,
+      "learning_rate": 2.272909163665466e-05,
+      "loss": 0.5099,
+      "step": 13643
+    },
+    {
+      "epoch": 17.46432,
+      "grad_norm": 1.0578844547271729,
+      "learning_rate": 2.2727090836334536e-05,
+      "loss": 0.5155,
+      "step": 13644
+    },
+    {
+      "epoch": 17.4656,
+      "grad_norm": 1.028712272644043,
+      "learning_rate": 2.2725090036014407e-05,
+      "loss": 0.487,
+      "step": 13645
+    },
+    {
+      "epoch": 17.46688,
+      "grad_norm": 1.080743670463562,
+      "learning_rate": 2.272308923569428e-05,
+      "loss": 0.5096,
+      "step": 13646
+    },
+    {
+      "epoch": 17.46816,
+      "grad_norm": 1.0488895177841187,
+      "learning_rate": 2.272108843537415e-05,
+      "loss": 0.477,
+      "step": 13647
+    },
+    {
+      "epoch": 17.46944,
+      "grad_norm": 1.0804858207702637,
+      "learning_rate": 2.2719087635054023e-05,
+      "loss": 0.5193,
+      "step": 13648
+    },
+    {
+      "epoch": 17.47072,
+      "grad_norm": 1.074198603630066,
+      "learning_rate": 2.2717086834733895e-05,
+      "loss": 0.518,
+      "step": 13649
+    },
+    {
+      "epoch": 17.472,
+      "grad_norm": 1.0862224102020264,
+      "learning_rate": 2.2715086034413767e-05,
+      "loss": 0.5612,
+      "step": 13650
+    },
+    {
+      "epoch": 17.47328,
+      "grad_norm": 1.0850809812545776,
+      "learning_rate": 2.271308523409364e-05,
+      "loss": 0.5247,
+      "step": 13651
+    },
+    {
+      "epoch": 17.47456,
+      "grad_norm": 1.0903500318527222,
+      "learning_rate": 2.271108443377351e-05,
+      "loss": 0.5088,
+      "step": 13652
+    },
+    {
+      "epoch": 17.47584,
+      "grad_norm": 1.0023061037063599,
+      "learning_rate": 2.2709083633453382e-05,
+      "loss": 0.5279,
+      "step": 13653
+    },
+    {
+      "epoch": 17.47712,
+      "grad_norm": 0.980223536491394,
+      "learning_rate": 2.2707082833133254e-05,
+      "loss": 0.5039,
+      "step": 13654
+    },
+    {
+      "epoch": 17.4784,
+      "grad_norm": 1.0036488771438599,
+      "learning_rate": 2.2705082032813126e-05,
+      "loss": 0.5017,
+      "step": 13655
+    },
+    {
+      "epoch": 17.47968,
+      "grad_norm": 1.0191391706466675,
+      "learning_rate": 2.2703081232492998e-05,
+      "loss": 0.4988,
+      "step": 13656
+    },
+    {
+      "epoch": 17.48096,
+      "grad_norm": 0.9996136426925659,
+      "learning_rate": 2.270108043217287e-05,
+      "loss": 0.5057,
+      "step": 13657
+    },
+    {
+      "epoch": 17.48224,
+      "grad_norm": 1.033124327659607,
+      "learning_rate": 2.2699079631852745e-05,
+      "loss": 0.5378,
+      "step": 13658
+    },
+    {
+      "epoch": 17.48352,
+      "grad_norm": 1.1039026975631714,
+      "learning_rate": 2.2697078831532613e-05,
+      "loss": 0.5528,
+      "step": 13659
+    },
+    {
+      "epoch": 17.4848,
+      "grad_norm": 1.0512938499450684,
+      "learning_rate": 2.2695078031212485e-05,
+      "loss": 0.4989,
+      "step": 13660
+    },
+    {
+      "epoch": 17.48608,
+      "grad_norm": 1.054443120956421,
+      "learning_rate": 2.2693077230892357e-05,
+      "loss": 0.5054,
+      "step": 13661
+    },
+    {
+      "epoch": 17.48736,
+      "grad_norm": 1.0660535097122192,
+      "learning_rate": 2.2691076430572232e-05,
+      "loss": 0.504,
+      "step": 13662
+    },
+    {
+      "epoch": 17.48864,
+      "grad_norm": 0.9800325036048889,
+      "learning_rate": 2.26890756302521e-05,
+      "loss": 0.4377,
+      "step": 13663
+    },
+    {
+      "epoch": 17.48992,
+      "grad_norm": 1.013704538345337,
+      "learning_rate": 2.2687074829931973e-05,
+      "loss": 0.5319,
+      "step": 13664
+    },
+    {
+      "epoch": 17.4912,
+      "grad_norm": 1.0099461078643799,
+      "learning_rate": 2.2685074029611848e-05,
+      "loss": 0.4664,
+      "step": 13665
+    },
+    {
+      "epoch": 17.49248,
+      "grad_norm": 1.0758755207061768,
+      "learning_rate": 2.268307322929172e-05,
+      "loss": 0.5314,
+      "step": 13666
+    },
+    {
+      "epoch": 17.49376,
+      "grad_norm": 1.0263993740081787,
+      "learning_rate": 2.2681072428971588e-05,
+      "loss": 0.4989,
+      "step": 13667
+    },
+    {
+      "epoch": 17.49504,
+      "grad_norm": 1.1413928270339966,
+      "learning_rate": 2.267907162865146e-05,
+      "loss": 0.5251,
+      "step": 13668
+    },
+    {
+      "epoch": 17.49632,
+      "grad_norm": 0.9962292909622192,
+      "learning_rate": 2.2677070828331335e-05,
+      "loss": 0.485,
+      "step": 13669
+    },
+    {
+      "epoch": 17.4976,
+      "grad_norm": 0.9963700771331787,
+      "learning_rate": 2.2675070028011207e-05,
+      "loss": 0.5073,
+      "step": 13670
+    },
+    {
+      "epoch": 17.49888,
+      "grad_norm": 1.0219050645828247,
+      "learning_rate": 2.2673069227691076e-05,
+      "loss": 0.5144,
+      "step": 13671
+    },
+    {
+      "epoch": 17.50016,
+      "grad_norm": 0.9802215695381165,
+      "learning_rate": 2.267106842737095e-05,
+      "loss": 0.5066,
+      "step": 13672
+    },
+    {
+      "epoch": 17.50144,
+      "grad_norm": 0.9986078143119812,
+      "learning_rate": 2.2669067627050823e-05,
+      "loss": 0.4959,
+      "step": 13673
+    },
+    {
+      "epoch": 17.50272,
+      "grad_norm": 1.017099142074585,
+      "learning_rate": 2.2667066826730695e-05,
+      "loss": 0.49,
+      "step": 13674
+    },
+    {
+      "epoch": 17.504,
+      "grad_norm": 1.0355932712554932,
+      "learning_rate": 2.2665066026410563e-05,
+      "loss": 0.5282,
+      "step": 13675
+    },
+    {
+      "epoch": 17.50528,
+      "grad_norm": 0.9949687123298645,
+      "learning_rate": 2.266306522609044e-05,
+      "loss": 0.4931,
+      "step": 13676
+    },
+    {
+      "epoch": 17.50656,
+      "grad_norm": 0.975506603717804,
+      "learning_rate": 2.266106442577031e-05,
+      "loss": 0.525,
+      "step": 13677
+    },
+    {
+      "epoch": 17.50784,
+      "grad_norm": 1.0513207912445068,
+      "learning_rate": 2.2659063625450182e-05,
+      "loss": 0.5377,
+      "step": 13678
+    },
+    {
+      "epoch": 17.50912,
+      "grad_norm": 1.0164129734039307,
+      "learning_rate": 2.2657062825130054e-05,
+      "loss": 0.4787,
+      "step": 13679
+    },
+    {
+      "epoch": 17.5104,
+      "grad_norm": 1.022405982017517,
+      "learning_rate": 2.2655062024809926e-05,
+      "loss": 0.4741,
+      "step": 13680
+    },
+    {
+      "epoch": 17.51168,
+      "grad_norm": 1.0068978071212769,
+      "learning_rate": 2.2653061224489798e-05,
+      "loss": 0.4911,
+      "step": 13681
+    },
+    {
+      "epoch": 17.51296,
+      "grad_norm": 0.9547240138053894,
+      "learning_rate": 2.265106042416967e-05,
+      "loss": 0.4298,
+      "step": 13682
+    },
+    {
+      "epoch": 17.51424,
+      "grad_norm": 0.969224750995636,
+      "learning_rate": 2.264905962384954e-05,
+      "loss": 0.4678,
+      "step": 13683
+    },
+    {
+      "epoch": 17.51552,
+      "grad_norm": 0.9585568904876709,
+      "learning_rate": 2.2647058823529413e-05,
+      "loss": 0.4658,
+      "step": 13684
+    },
+    {
+      "epoch": 17.5168,
+      "grad_norm": 1.0375971794128418,
+      "learning_rate": 2.2645058023209285e-05,
+      "loss": 0.5459,
+      "step": 13685
+    },
+    {
+      "epoch": 17.51808,
+      "grad_norm": 1.1199681758880615,
+      "learning_rate": 2.2643057222889157e-05,
+      "loss": 0.5432,
+      "step": 13686
+    },
+    {
+      "epoch": 17.51936,
+      "grad_norm": 1.0504252910614014,
+      "learning_rate": 2.264105642256903e-05,
+      "loss": 0.531,
+      "step": 13687
+    },
+    {
+      "epoch": 17.52064,
+      "grad_norm": 1.0494788885116577,
+      "learning_rate": 2.26390556222489e-05,
+      "loss": 0.5018,
+      "step": 13688
+    },
+    {
+      "epoch": 17.52192,
+      "grad_norm": 1.0414574146270752,
+      "learning_rate": 2.2637054821928772e-05,
+      "loss": 0.5317,
+      "step": 13689
+    },
+    {
+      "epoch": 17.5232,
+      "grad_norm": 1.058502197265625,
+      "learning_rate": 2.2635054021608644e-05,
+      "loss": 0.5139,
+      "step": 13690
+    },
+    {
+      "epoch": 17.52448,
+      "grad_norm": 1.1048195362091064,
+      "learning_rate": 2.2633053221288516e-05,
+      "loss": 0.5667,
+      "step": 13691
+    },
+    {
+      "epoch": 17.52576,
+      "grad_norm": 1.0545276403427124,
+      "learning_rate": 2.2631052420968388e-05,
+      "loss": 0.4954,
+      "step": 13692
+    },
+    {
+      "epoch": 17.52704,
+      "grad_norm": 1.0869113206863403,
+      "learning_rate": 2.2629051620648263e-05,
+      "loss": 0.564,
+      "step": 13693
+    },
+    {
+      "epoch": 17.52832,
+      "grad_norm": 1.0492844581604004,
+      "learning_rate": 2.2627050820328132e-05,
+      "loss": 0.54,
+      "step": 13694
+    },
+    {
+      "epoch": 17.5296,
+      "grad_norm": 1.0702095031738281,
+      "learning_rate": 2.2625050020008004e-05,
+      "loss": 0.5171,
+      "step": 13695
+    },
+    {
+      "epoch": 17.53088,
+      "grad_norm": 1.0479813814163208,
+      "learning_rate": 2.2623049219687875e-05,
+      "loss": 0.5329,
+      "step": 13696
+    },
+    {
+      "epoch": 17.53216,
+      "grad_norm": 1.0335677862167358,
+      "learning_rate": 2.262104841936775e-05,
+      "loss": 0.5585,
+      "step": 13697
+    },
+    {
+      "epoch": 17.53344,
+      "grad_norm": 1.0604044198989868,
+      "learning_rate": 2.261904761904762e-05,
+      "loss": 0.5967,
+      "step": 13698
+    },
+    {
+      "epoch": 17.53472,
+      "grad_norm": 0.9840775728225708,
+      "learning_rate": 2.261704681872749e-05,
+      "loss": 0.496,
+      "step": 13699
+    },
+    {
+      "epoch": 17.536,
+      "grad_norm": 0.9827370643615723,
+      "learning_rate": 2.2615046018407366e-05,
+      "loss": 0.4696,
+      "step": 13700
+    },
+    {
+      "epoch": 17.53728,
+      "grad_norm": 1.0692596435546875,
+      "learning_rate": 2.2613045218087238e-05,
+      "loss": 0.5455,
+      "step": 13701
+    },
+    {
+      "epoch": 17.53856,
+      "grad_norm": 1.096990704536438,
+      "learning_rate": 2.2611044417767107e-05,
+      "loss": 0.5268,
+      "step": 13702
+    },
+    {
+      "epoch": 17.53984,
+      "grad_norm": 1.0311455726623535,
+      "learning_rate": 2.260904361744698e-05,
+      "loss": 0.5189,
+      "step": 13703
+    },
+    {
+      "epoch": 17.54112,
+      "grad_norm": 0.9897575378417969,
+      "learning_rate": 2.2607042817126854e-05,
+      "loss": 0.4777,
+      "step": 13704
+    },
+    {
+      "epoch": 17.5424,
+      "grad_norm": 1.0446470975875854,
+      "learning_rate": 2.2605042016806726e-05,
+      "loss": 0.5085,
+      "step": 13705
+    },
+    {
+      "epoch": 17.54368,
+      "grad_norm": 1.0381147861480713,
+      "learning_rate": 2.2603041216486594e-05,
+      "loss": 0.5308,
+      "step": 13706
+    },
+    {
+      "epoch": 17.54496,
+      "grad_norm": 1.0593794584274292,
+      "learning_rate": 2.260104041616647e-05,
+      "loss": 0.4676,
+      "step": 13707
+    },
+    {
+      "epoch": 17.54624,
+      "grad_norm": 1.0510727167129517,
+      "learning_rate": 2.259903961584634e-05,
+      "loss": 0.5292,
+      "step": 13708
+    },
+    {
+      "epoch": 17.54752,
+      "grad_norm": 1.047689437866211,
+      "learning_rate": 2.2597038815526213e-05,
+      "loss": 0.5205,
+      "step": 13709
+    },
+    {
+      "epoch": 17.5488,
+      "grad_norm": 1.0579311847686768,
+      "learning_rate": 2.259503801520608e-05,
+      "loss": 0.5061,
+      "step": 13710
+    },
+    {
+      "epoch": 17.55008,
+      "grad_norm": 1.0636060237884521,
+      "learning_rate": 2.2593037214885957e-05,
+      "loss": 0.5463,
+      "step": 13711
+    },
+    {
+      "epoch": 17.55136,
+      "grad_norm": 0.990552544593811,
+      "learning_rate": 2.259103641456583e-05,
+      "loss": 0.4527,
+      "step": 13712
+    },
+    {
+      "epoch": 17.55264,
+      "grad_norm": 1.0721439123153687,
+      "learning_rate": 2.25890356142457e-05,
+      "loss": 0.5183,
+      "step": 13713
+    },
+    {
+      "epoch": 17.55392,
+      "grad_norm": 1.0679312944412231,
+      "learning_rate": 2.258703481392557e-05,
+      "loss": 0.557,
+      "step": 13714
+    },
+    {
+      "epoch": 17.5552,
+      "grad_norm": 0.9733208417892456,
+      "learning_rate": 2.2585034013605444e-05,
+      "loss": 0.507,
+      "step": 13715
+    },
+    {
+      "epoch": 17.55648,
+      "grad_norm": 1.1034499406814575,
+      "learning_rate": 2.2583033213285316e-05,
+      "loss": 0.5612,
+      "step": 13716
+    },
+    {
+      "epoch": 17.557760000000002,
+      "grad_norm": 1.0827926397323608,
+      "learning_rate": 2.2581032412965188e-05,
+      "loss": 0.5139,
+      "step": 13717
+    },
+    {
+      "epoch": 17.55904,
+      "grad_norm": 0.9990882873535156,
+      "learning_rate": 2.257903161264506e-05,
+      "loss": 0.4797,
+      "step": 13718
+    },
+    {
+      "epoch": 17.56032,
+      "grad_norm": 0.9691088199615479,
+      "learning_rate": 2.257703081232493e-05,
+      "loss": 0.4385,
+      "step": 13719
+    },
+    {
+      "epoch": 17.5616,
+      "grad_norm": 1.0070648193359375,
+      "learning_rate": 2.2575030012004803e-05,
+      "loss": 0.4986,
+      "step": 13720
+    },
+    {
+      "epoch": 17.56288,
+      "grad_norm": 1.0735926628112793,
+      "learning_rate": 2.2573029211684675e-05,
+      "loss": 0.5415,
+      "step": 13721
+    },
+    {
+      "epoch": 17.56416,
+      "grad_norm": 1.0106046199798584,
+      "learning_rate": 2.2571028411364547e-05,
+      "loss": 0.4826,
+      "step": 13722
+    },
+    {
+      "epoch": 17.56544,
+      "grad_norm": 1.0311856269836426,
+      "learning_rate": 2.256902761104442e-05,
+      "loss": 0.5534,
+      "step": 13723
+    },
+    {
+      "epoch": 17.56672,
+      "grad_norm": 1.014507532119751,
+      "learning_rate": 2.256702681072429e-05,
+      "loss": 0.4764,
+      "step": 13724
+    },
+    {
+      "epoch": 17.568,
+      "grad_norm": 1.0640530586242676,
+      "learning_rate": 2.2565026010404163e-05,
+      "loss": 0.5272,
+      "step": 13725
+    },
+    {
+      "epoch": 17.56928,
+      "grad_norm": 1.1166555881500244,
+      "learning_rate": 2.2563025210084035e-05,
+      "loss": 0.572,
+      "step": 13726
+    },
+    {
+      "epoch": 17.57056,
+      "grad_norm": 1.061682105064392,
+      "learning_rate": 2.2561024409763906e-05,
+      "loss": 0.5373,
+      "step": 13727
+    },
+    {
+      "epoch": 17.57184,
+      "grad_norm": 1.0302103757858276,
+      "learning_rate": 2.2559023609443778e-05,
+      "loss": 0.4871,
+      "step": 13728
+    },
+    {
+      "epoch": 17.57312,
+      "grad_norm": 1.040071964263916,
+      "learning_rate": 2.255702280912365e-05,
+      "loss": 0.5096,
+      "step": 13729
+    },
+    {
+      "epoch": 17.5744,
+      "grad_norm": 1.0365045070648193,
+      "learning_rate": 2.2555022008803522e-05,
+      "loss": 0.4992,
+      "step": 13730
+    },
+    {
+      "epoch": 17.57568,
+      "grad_norm": 1.0997463464736938,
+      "learning_rate": 2.2553021208483394e-05,
+      "loss": 0.532,
+      "step": 13731
+    },
+    {
+      "epoch": 17.57696,
+      "grad_norm": 1.0069595575332642,
+      "learning_rate": 2.255102040816327e-05,
+      "loss": 0.4886,
+      "step": 13732
+    },
+    {
+      "epoch": 17.57824,
+      "grad_norm": 1.1098484992980957,
+      "learning_rate": 2.2549019607843138e-05,
+      "loss": 0.526,
+      "step": 13733
+    },
+    {
+      "epoch": 17.57952,
+      "grad_norm": 1.034224033355713,
+      "learning_rate": 2.254701880752301e-05,
+      "loss": 0.5107,
+      "step": 13734
+    },
+    {
+      "epoch": 17.5808,
+      "grad_norm": 0.9972882866859436,
+      "learning_rate": 2.254501800720288e-05,
+      "loss": 0.4979,
+      "step": 13735
+    },
+    {
+      "epoch": 17.58208,
+      "grad_norm": 1.033673644065857,
+      "learning_rate": 2.2543017206882756e-05,
+      "loss": 0.491,
+      "step": 13736
+    },
+    {
+      "epoch": 17.58336,
+      "grad_norm": 1.0336534976959229,
+      "learning_rate": 2.2541016406562625e-05,
+      "loss": 0.5155,
+      "step": 13737
+    },
+    {
+      "epoch": 17.58464,
+      "grad_norm": 1.0323609113693237,
+      "learning_rate": 2.2539015606242497e-05,
+      "loss": 0.5079,
+      "step": 13738
+    },
+    {
+      "epoch": 17.58592,
+      "grad_norm": 1.0115076303482056,
+      "learning_rate": 2.2537014805922372e-05,
+      "loss": 0.5356,
+      "step": 13739
+    },
+    {
+      "epoch": 17.5872,
+      "grad_norm": 0.9737477898597717,
+      "learning_rate": 2.2535014005602244e-05,
+      "loss": 0.4799,
+      "step": 13740
+    },
+    {
+      "epoch": 17.58848,
+      "grad_norm": 1.0207149982452393,
+      "learning_rate": 2.2533013205282112e-05,
+      "loss": 0.4796,
+      "step": 13741
+    },
+    {
+      "epoch": 17.58976,
+      "grad_norm": 1.0256177186965942,
+      "learning_rate": 2.2531012404961984e-05,
+      "loss": 0.5228,
+      "step": 13742
+    },
+    {
+      "epoch": 17.59104,
+      "grad_norm": 1.0296961069107056,
+      "learning_rate": 2.252901160464186e-05,
+      "loss": 0.5046,
+      "step": 13743
+    },
+    {
+      "epoch": 17.59232,
+      "grad_norm": 1.0195798873901367,
+      "learning_rate": 2.252701080432173e-05,
+      "loss": 0.4933,
+      "step": 13744
+    },
+    {
+      "epoch": 17.5936,
+      "grad_norm": 1.1169263124465942,
+      "learning_rate": 2.25250100040016e-05,
+      "loss": 0.5553,
+      "step": 13745
+    },
+    {
+      "epoch": 17.59488,
+      "grad_norm": 1.0939453840255737,
+      "learning_rate": 2.2523009203681475e-05,
+      "loss": 0.5122,
+      "step": 13746
+    },
+    {
+      "epoch": 17.59616,
+      "grad_norm": 1.0578433275222778,
+      "learning_rate": 2.2521008403361347e-05,
+      "loss": 0.5133,
+      "step": 13747
+    },
+    {
+      "epoch": 17.59744,
+      "grad_norm": 1.0464441776275635,
+      "learning_rate": 2.251900760304122e-05,
+      "loss": 0.5562,
+      "step": 13748
+    },
+    {
+      "epoch": 17.59872,
+      "grad_norm": 1.0032329559326172,
+      "learning_rate": 2.2517006802721087e-05,
+      "loss": 0.5132,
+      "step": 13749
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 1.0605465173721313,
+      "learning_rate": 2.2515006002400962e-05,
+      "loss": 0.5201,
+      "step": 13750
+    },
+    {
+      "epoch": 17.60128,
+      "grad_norm": 1.0465607643127441,
+      "learning_rate": 2.2513005202080834e-05,
+      "loss": 0.5203,
+      "step": 13751
+    },
+    {
+      "epoch": 17.60256,
+      "grad_norm": 1.0221221446990967,
+      "learning_rate": 2.2511004401760706e-05,
+      "loss": 0.4927,
+      "step": 13752
+    },
+    {
+      "epoch": 17.60384,
+      "grad_norm": 0.970077633857727,
+      "learning_rate": 2.2509003601440578e-05,
+      "loss": 0.4861,
+      "step": 13753
+    },
+    {
+      "epoch": 17.60512,
+      "grad_norm": 1.0048155784606934,
+      "learning_rate": 2.250700280112045e-05,
+      "loss": 0.479,
+      "step": 13754
+    },
+    {
+      "epoch": 17.6064,
+      "grad_norm": 1.0469225645065308,
+      "learning_rate": 2.2505002000800322e-05,
+      "loss": 0.5249,
+      "step": 13755
+    },
+    {
+      "epoch": 17.60768,
+      "grad_norm": 1.0476946830749512,
+      "learning_rate": 2.2503001200480194e-05,
+      "loss": 0.4966,
+      "step": 13756
+    },
+    {
+      "epoch": 17.60896,
+      "grad_norm": 1.0547471046447754,
+      "learning_rate": 2.2501000400160065e-05,
+      "loss": 0.496,
+      "step": 13757
+    },
+    {
+      "epoch": 17.61024,
+      "grad_norm": 1.0567193031311035,
+      "learning_rate": 2.2498999599839937e-05,
+      "loss": 0.5211,
+      "step": 13758
+    },
+    {
+      "epoch": 17.61152,
+      "grad_norm": 1.0392005443572998,
+      "learning_rate": 2.249699879951981e-05,
+      "loss": 0.5055,
+      "step": 13759
+    },
+    {
+      "epoch": 17.6128,
+      "grad_norm": 1.090825080871582,
+      "learning_rate": 2.249499799919968e-05,
+      "loss": 0.5679,
+      "step": 13760
+    },
+    {
+      "epoch": 17.61408,
+      "grad_norm": 1.0225039720535278,
+      "learning_rate": 2.2492997198879553e-05,
+      "loss": 0.4868,
+      "step": 13761
+    },
+    {
+      "epoch": 17.61536,
+      "grad_norm": 1.011189341545105,
+      "learning_rate": 2.2490996398559425e-05,
+      "loss": 0.5217,
+      "step": 13762
+    },
+    {
+      "epoch": 17.61664,
+      "grad_norm": 1.0232630968093872,
+      "learning_rate": 2.2488995598239297e-05,
+      "loss": 0.4865,
+      "step": 13763
+    },
+    {
+      "epoch": 17.61792,
+      "grad_norm": 1.0764962434768677,
+      "learning_rate": 2.248699479791917e-05,
+      "loss": 0.5524,
+      "step": 13764
+    },
+    {
+      "epoch": 17.6192,
+      "grad_norm": 1.0205249786376953,
+      "learning_rate": 2.248499399759904e-05,
+      "loss": 0.4819,
+      "step": 13765
+    },
+    {
+      "epoch": 17.62048,
+      "grad_norm": 1.089855670928955,
+      "learning_rate": 2.2482993197278912e-05,
+      "loss": 0.5628,
+      "step": 13766
+    },
+    {
+      "epoch": 17.62176,
+      "grad_norm": 1.1043068170547485,
+      "learning_rate": 2.2480992396958784e-05,
+      "loss": 0.5314,
+      "step": 13767
+    },
+    {
+      "epoch": 17.62304,
+      "grad_norm": 1.0724689960479736,
+      "learning_rate": 2.2478991596638656e-05,
+      "loss": 0.5377,
+      "step": 13768
+    },
+    {
+      "epoch": 17.62432,
+      "grad_norm": 1.0592753887176514,
+      "learning_rate": 2.2476990796318528e-05,
+      "loss": 0.5109,
+      "step": 13769
+    },
+    {
+      "epoch": 17.6256,
+      "grad_norm": 1.0443251132965088,
+      "learning_rate": 2.24749899959984e-05,
+      "loss": 0.54,
+      "step": 13770
+    },
+    {
+      "epoch": 17.62688,
+      "grad_norm": 1.0627379417419434,
+      "learning_rate": 2.2472989195678275e-05,
+      "loss": 0.5206,
+      "step": 13771
+    },
+    {
+      "epoch": 17.62816,
+      "grad_norm": 1.033038854598999,
+      "learning_rate": 2.2470988395358143e-05,
+      "loss": 0.4665,
+      "step": 13772
+    },
+    {
+      "epoch": 17.62944,
+      "grad_norm": 1.1007275581359863,
+      "learning_rate": 2.2468987595038015e-05,
+      "loss": 0.5408,
+      "step": 13773
+    },
+    {
+      "epoch": 17.63072,
+      "grad_norm": 1.051284670829773,
+      "learning_rate": 2.2466986794717887e-05,
+      "loss": 0.509,
+      "step": 13774
+    },
+    {
+      "epoch": 17.632,
+      "grad_norm": 1.059944748878479,
+      "learning_rate": 2.2464985994397762e-05,
+      "loss": 0.5193,
+      "step": 13775
+    },
+    {
+      "epoch": 17.63328,
+      "grad_norm": 1.0539425611495972,
+      "learning_rate": 2.246298519407763e-05,
+      "loss": 0.5161,
+      "step": 13776
+    },
+    {
+      "epoch": 17.63456,
+      "grad_norm": 0.9573315382003784,
+      "learning_rate": 2.2460984393757503e-05,
+      "loss": 0.4629,
+      "step": 13777
+    },
+    {
+      "epoch": 17.63584,
+      "grad_norm": 1.0628639459609985,
+      "learning_rate": 2.2458983593437378e-05,
+      "loss": 0.5524,
+      "step": 13778
+    },
+    {
+      "epoch": 17.63712,
+      "grad_norm": 0.9728816747665405,
+      "learning_rate": 2.245698279311725e-05,
+      "loss": 0.4881,
+      "step": 13779
+    },
+    {
+      "epoch": 17.6384,
+      "grad_norm": 0.996763288974762,
+      "learning_rate": 2.2454981992797118e-05,
+      "loss": 0.4589,
+      "step": 13780
+    },
+    {
+      "epoch": 17.63968,
+      "grad_norm": 1.0928605794906616,
+      "learning_rate": 2.245298119247699e-05,
+      "loss": 0.5651,
+      "step": 13781
+    },
+    {
+      "epoch": 17.64096,
+      "grad_norm": 1.067138433456421,
+      "learning_rate": 2.2450980392156865e-05,
+      "loss": 0.5257,
+      "step": 13782
+    },
+    {
+      "epoch": 17.64224,
+      "grad_norm": 1.0764461755752563,
+      "learning_rate": 2.2448979591836737e-05,
+      "loss": 0.5539,
+      "step": 13783
+    },
+    {
+      "epoch": 17.64352,
+      "grad_norm": 1.0611051321029663,
+      "learning_rate": 2.2446978791516606e-05,
+      "loss": 0.5756,
+      "step": 13784
+    },
+    {
+      "epoch": 17.6448,
+      "grad_norm": 1.0139261484146118,
+      "learning_rate": 2.244497799119648e-05,
+      "loss": 0.4869,
+      "step": 13785
+    },
+    {
+      "epoch": 17.64608,
+      "grad_norm": 1.0263339281082153,
+      "learning_rate": 2.2442977190876353e-05,
+      "loss": 0.5112,
+      "step": 13786
+    },
+    {
+      "epoch": 17.64736,
+      "grad_norm": 1.041606068611145,
+      "learning_rate": 2.2440976390556225e-05,
+      "loss": 0.5362,
+      "step": 13787
+    },
+    {
+      "epoch": 17.64864,
+      "grad_norm": 1.01603102684021,
+      "learning_rate": 2.2438975590236093e-05,
+      "loss": 0.4863,
+      "step": 13788
+    },
+    {
+      "epoch": 17.64992,
+      "grad_norm": 1.0228776931762695,
+      "learning_rate": 2.2436974789915968e-05,
+      "loss": 0.5174,
+      "step": 13789
+    },
+    {
+      "epoch": 17.6512,
+      "grad_norm": 1.0253472328186035,
+      "learning_rate": 2.243497398959584e-05,
+      "loss": 0.502,
+      "step": 13790
+    },
+    {
+      "epoch": 17.65248,
+      "grad_norm": 1.0483967065811157,
+      "learning_rate": 2.2432973189275712e-05,
+      "loss": 0.5241,
+      "step": 13791
+    },
+    {
+      "epoch": 17.65376,
+      "grad_norm": 0.9999913573265076,
+      "learning_rate": 2.2430972388955584e-05,
+      "loss": 0.5112,
+      "step": 13792
+    },
+    {
+      "epoch": 17.65504,
+      "grad_norm": 0.9996472001075745,
+      "learning_rate": 2.2428971588635456e-05,
+      "loss": 0.4946,
+      "step": 13793
+    },
+    {
+      "epoch": 17.65632,
+      "grad_norm": 1.0371149778366089,
+      "learning_rate": 2.2426970788315328e-05,
+      "loss": 0.5217,
+      "step": 13794
+    },
+    {
+      "epoch": 17.6576,
+      "grad_norm": 1.040626883506775,
+      "learning_rate": 2.24249699879952e-05,
+      "loss": 0.5226,
+      "step": 13795
+    },
+    {
+      "epoch": 17.65888,
+      "grad_norm": 1.0527793169021606,
+      "learning_rate": 2.242296918767507e-05,
+      "loss": 0.5302,
+      "step": 13796
+    },
+    {
+      "epoch": 17.66016,
+      "grad_norm": 1.0551269054412842,
+      "learning_rate": 2.2420968387354943e-05,
+      "loss": 0.524,
+      "step": 13797
+    },
+    {
+      "epoch": 17.66144,
+      "grad_norm": 1.0026994943618774,
+      "learning_rate": 2.2418967587034815e-05,
+      "loss": 0.4582,
+      "step": 13798
+    },
+    {
+      "epoch": 17.66272,
+      "grad_norm": 1.0396666526794434,
+      "learning_rate": 2.2416966786714687e-05,
+      "loss": 0.49,
+      "step": 13799
+    },
+    {
+      "epoch": 17.664,
+      "grad_norm": 1.0443949699401855,
+      "learning_rate": 2.241496598639456e-05,
+      "loss": 0.4863,
+      "step": 13800
+    },
+    {
+      "epoch": 17.66528,
+      "grad_norm": 1.02041757106781,
+      "learning_rate": 2.241296518607443e-05,
+      "loss": 0.5075,
+      "step": 13801
+    },
+    {
+      "epoch": 17.66656,
+      "grad_norm": 1.0015771389007568,
+      "learning_rate": 2.2410964385754302e-05,
+      "loss": 0.4828,
+      "step": 13802
+    },
+    {
+      "epoch": 17.667839999999998,
+      "grad_norm": 1.0166736841201782,
+      "learning_rate": 2.2408963585434174e-05,
+      "loss": 0.4961,
+      "step": 13803
+    },
+    {
+      "epoch": 17.66912,
+      "grad_norm": 1.0742013454437256,
+      "learning_rate": 2.2406962785114046e-05,
+      "loss": 0.5438,
+      "step": 13804
+    },
+    {
+      "epoch": 17.6704,
+      "grad_norm": 1.0339466333389282,
+      "learning_rate": 2.2404961984793918e-05,
+      "loss": 0.5245,
+      "step": 13805
+    },
+    {
+      "epoch": 17.67168,
+      "grad_norm": 1.0903747081756592,
+      "learning_rate": 2.2402961184473793e-05,
+      "loss": 0.5367,
+      "step": 13806
+    },
+    {
+      "epoch": 17.67296,
+      "grad_norm": 1.049980878829956,
+      "learning_rate": 2.2400960384153662e-05,
+      "loss": 0.5121,
+      "step": 13807
+    },
+    {
+      "epoch": 17.67424,
+      "grad_norm": 1.027825117111206,
+      "learning_rate": 2.2398959583833534e-05,
+      "loss": 0.4781,
+      "step": 13808
+    },
+    {
+      "epoch": 17.67552,
+      "grad_norm": 1.0278798341751099,
+      "learning_rate": 2.2396958783513405e-05,
+      "loss": 0.4927,
+      "step": 13809
+    },
+    {
+      "epoch": 17.6768,
+      "grad_norm": 1.1051456928253174,
+      "learning_rate": 2.239495798319328e-05,
+      "loss": 0.5276,
+      "step": 13810
+    },
+    {
+      "epoch": 17.67808,
+      "grad_norm": 1.029605507850647,
+      "learning_rate": 2.239295718287315e-05,
+      "loss": 0.4806,
+      "step": 13811
+    },
+    {
+      "epoch": 17.67936,
+      "grad_norm": 1.0395790338516235,
+      "learning_rate": 2.239095638255302e-05,
+      "loss": 0.4941,
+      "step": 13812
+    },
+    {
+      "epoch": 17.68064,
+      "grad_norm": 1.017499566078186,
+      "learning_rate": 2.2388955582232896e-05,
+      "loss": 0.5287,
+      "step": 13813
+    },
+    {
+      "epoch": 17.68192,
+      "grad_norm": 1.0549243688583374,
+      "learning_rate": 2.2386954781912768e-05,
+      "loss": 0.5473,
+      "step": 13814
+    },
+    {
+      "epoch": 17.6832,
+      "grad_norm": 0.9981441497802734,
+      "learning_rate": 2.2384953981592637e-05,
+      "loss": 0.5042,
+      "step": 13815
+    },
+    {
+      "epoch": 17.68448,
+      "grad_norm": 1.0279474258422852,
+      "learning_rate": 2.238295318127251e-05,
+      "loss": 0.5113,
+      "step": 13816
+    },
+    {
+      "epoch": 17.68576,
+      "grad_norm": 1.023318886756897,
+      "learning_rate": 2.2380952380952384e-05,
+      "loss": 0.5169,
+      "step": 13817
+    },
+    {
+      "epoch": 17.68704,
+      "grad_norm": 1.0410925149917603,
+      "learning_rate": 2.2378951580632256e-05,
+      "loss": 0.5097,
+      "step": 13818
+    },
+    {
+      "epoch": 17.68832,
+      "grad_norm": 1.0423842668533325,
+      "learning_rate": 2.2376950780312124e-05,
+      "loss": 0.5164,
+      "step": 13819
+    },
+    {
+      "epoch": 17.6896,
+      "grad_norm": 1.0338177680969238,
+      "learning_rate": 2.2374949979991996e-05,
+      "loss": 0.5043,
+      "step": 13820
+    },
+    {
+      "epoch": 17.69088,
+      "grad_norm": 1.0916544198989868,
+      "learning_rate": 2.237294917967187e-05,
+      "loss": 0.5311,
+      "step": 13821
+    },
+    {
+      "epoch": 17.69216,
+      "grad_norm": 1.0973963737487793,
+      "learning_rate": 2.2370948379351743e-05,
+      "loss": 0.5184,
+      "step": 13822
+    },
+    {
+      "epoch": 17.69344,
+      "grad_norm": 1.0347501039505005,
+      "learning_rate": 2.236894757903161e-05,
+      "loss": 0.4996,
+      "step": 13823
+    },
+    {
+      "epoch": 17.69472,
+      "grad_norm": 0.9996992349624634,
+      "learning_rate": 2.2366946778711487e-05,
+      "loss": 0.4913,
+      "step": 13824
+    },
+    {
+      "epoch": 17.696,
+      "grad_norm": 1.081697702407837,
+      "learning_rate": 2.236494597839136e-05,
+      "loss": 0.5354,
+      "step": 13825
+    },
+    {
+      "epoch": 17.69728,
+      "grad_norm": 1.0013635158538818,
+      "learning_rate": 2.236294517807123e-05,
+      "loss": 0.5452,
+      "step": 13826
+    },
+    {
+      "epoch": 17.69856,
+      "grad_norm": 1.0059187412261963,
+      "learning_rate": 2.23609443777511e-05,
+      "loss": 0.5179,
+      "step": 13827
+    },
+    {
+      "epoch": 17.699840000000002,
+      "grad_norm": 1.0418435335159302,
+      "learning_rate": 2.2358943577430974e-05,
+      "loss": 0.4985,
+      "step": 13828
+    },
+    {
+      "epoch": 17.70112,
+      "grad_norm": 1.1040401458740234,
+      "learning_rate": 2.2356942777110846e-05,
+      "loss": 0.5672,
+      "step": 13829
+    },
+    {
+      "epoch": 17.7024,
+      "grad_norm": 1.0777002573013306,
+      "learning_rate": 2.2354941976790718e-05,
+      "loss": 0.5346,
+      "step": 13830
+    },
+    {
+      "epoch": 17.70368,
+      "grad_norm": 1.057003378868103,
+      "learning_rate": 2.235294117647059e-05,
+      "loss": 0.529,
+      "step": 13831
+    },
+    {
+      "epoch": 17.70496,
+      "grad_norm": 1.0604835748672485,
+      "learning_rate": 2.235094037615046e-05,
+      "loss": 0.504,
+      "step": 13832
+    },
+    {
+      "epoch": 17.70624,
+      "grad_norm": 1.0369086265563965,
+      "learning_rate": 2.2348939575830333e-05,
+      "loss": 0.512,
+      "step": 13833
+    },
+    {
+      "epoch": 17.70752,
+      "grad_norm": 1.0409032106399536,
+      "learning_rate": 2.2346938775510205e-05,
+      "loss": 0.4881,
+      "step": 13834
+    },
+    {
+      "epoch": 17.7088,
+      "grad_norm": 1.0097156763076782,
+      "learning_rate": 2.2344937975190077e-05,
+      "loss": 0.5367,
+      "step": 13835
+    },
+    {
+      "epoch": 17.71008,
+      "grad_norm": 0.9942651391029358,
+      "learning_rate": 2.234293717486995e-05,
+      "loss": 0.4977,
+      "step": 13836
+    },
+    {
+      "epoch": 17.71136,
+      "grad_norm": 1.0652927160263062,
+      "learning_rate": 2.234093637454982e-05,
+      "loss": 0.5629,
+      "step": 13837
+    },
+    {
+      "epoch": 17.71264,
+      "grad_norm": 1.0204808712005615,
+      "learning_rate": 2.2338935574229693e-05,
+      "loss": 0.4821,
+      "step": 13838
+    },
+    {
+      "epoch": 17.71392,
+      "grad_norm": 1.051592230796814,
+      "learning_rate": 2.2336934773909565e-05,
+      "loss": 0.4978,
+      "step": 13839
+    },
+    {
+      "epoch": 17.7152,
+      "grad_norm": 1.0325385332107544,
+      "learning_rate": 2.2334933973589436e-05,
+      "loss": 0.5587,
+      "step": 13840
+    },
+    {
+      "epoch": 17.71648,
+      "grad_norm": 0.9920241236686707,
+      "learning_rate": 2.2332933173269308e-05,
+      "loss": 0.5126,
+      "step": 13841
+    },
+    {
+      "epoch": 17.71776,
+      "grad_norm": 0.9792672395706177,
+      "learning_rate": 2.233093237294918e-05,
+      "loss": 0.4736,
+      "step": 13842
+    },
+    {
+      "epoch": 17.71904,
+      "grad_norm": 1.0789382457733154,
+      "learning_rate": 2.2328931572629052e-05,
+      "loss": 0.5335,
+      "step": 13843
+    },
+    {
+      "epoch": 17.72032,
+      "grad_norm": 1.0387771129608154,
+      "learning_rate": 2.2326930772308924e-05,
+      "loss": 0.5135,
+      "step": 13844
+    },
+    {
+      "epoch": 17.7216,
+      "grad_norm": 1.0767048597335815,
+      "learning_rate": 2.23249299719888e-05,
+      "loss": 0.5385,
+      "step": 13845
+    },
+    {
+      "epoch": 17.72288,
+      "grad_norm": 1.025465965270996,
+      "learning_rate": 2.2322929171668668e-05,
+      "loss": 0.4867,
+      "step": 13846
+    },
+    {
+      "epoch": 17.72416,
+      "grad_norm": 1.0004117488861084,
+      "learning_rate": 2.232092837134854e-05,
+      "loss": 0.5034,
+      "step": 13847
+    },
+    {
+      "epoch": 17.72544,
+      "grad_norm": 1.0588462352752686,
+      "learning_rate": 2.231892757102841e-05,
+      "loss": 0.5126,
+      "step": 13848
+    },
+    {
+      "epoch": 17.72672,
+      "grad_norm": 1.0342752933502197,
+      "learning_rate": 2.2316926770708286e-05,
+      "loss": 0.5027,
+      "step": 13849
+    },
+    {
+      "epoch": 17.728,
+      "grad_norm": 1.0341858863830566,
+      "learning_rate": 2.2314925970388155e-05,
+      "loss": 0.512,
+      "step": 13850
+    },
+    {
+      "epoch": 17.72928,
+      "grad_norm": 0.9943906664848328,
+      "learning_rate": 2.2312925170068027e-05,
+      "loss": 0.4869,
+      "step": 13851
+    },
+    {
+      "epoch": 17.73056,
+      "grad_norm": 1.0998703241348267,
+      "learning_rate": 2.2310924369747902e-05,
+      "loss": 0.5629,
+      "step": 13852
+    },
+    {
+      "epoch": 17.73184,
+      "grad_norm": 1.0420924425125122,
+      "learning_rate": 2.2308923569427774e-05,
+      "loss": 0.505,
+      "step": 13853
+    },
+    {
+      "epoch": 17.73312,
+      "grad_norm": 1.0193160772323608,
+      "learning_rate": 2.2306922769107642e-05,
+      "loss": 0.5082,
+      "step": 13854
+    },
+    {
+      "epoch": 17.7344,
+      "grad_norm": 1.0729001760482788,
+      "learning_rate": 2.2304921968787514e-05,
+      "loss": 0.4944,
+      "step": 13855
+    },
+    {
+      "epoch": 17.73568,
+      "grad_norm": 0.9842273592948914,
+      "learning_rate": 2.230292116846739e-05,
+      "loss": 0.4526,
+      "step": 13856
+    },
+    {
+      "epoch": 17.73696,
+      "grad_norm": 1.1139209270477295,
+      "learning_rate": 2.230092036814726e-05,
+      "loss": 0.5272,
+      "step": 13857
+    },
+    {
+      "epoch": 17.73824,
+      "grad_norm": 1.0628962516784668,
+      "learning_rate": 2.229891956782713e-05,
+      "loss": 0.5155,
+      "step": 13858
+    },
+    {
+      "epoch": 17.73952,
+      "grad_norm": 1.0056374073028564,
+      "learning_rate": 2.2296918767507005e-05,
+      "loss": 0.4806,
+      "step": 13859
+    },
+    {
+      "epoch": 17.7408,
+      "grad_norm": 1.0062882900238037,
+      "learning_rate": 2.2294917967186877e-05,
+      "loss": 0.5315,
+      "step": 13860
+    },
+    {
+      "epoch": 17.74208,
+      "grad_norm": 1.046491265296936,
+      "learning_rate": 2.229291716686675e-05,
+      "loss": 0.5553,
+      "step": 13861
+    },
+    {
+      "epoch": 17.74336,
+      "grad_norm": 1.0050787925720215,
+      "learning_rate": 2.2290916366546617e-05,
+      "loss": 0.5059,
+      "step": 13862
+    },
+    {
+      "epoch": 17.74464,
+      "grad_norm": 1.0405722856521606,
+      "learning_rate": 2.2288915566226492e-05,
+      "loss": 0.5053,
+      "step": 13863
+    },
+    {
+      "epoch": 17.74592,
+      "grad_norm": 1.0229580402374268,
+      "learning_rate": 2.2286914765906364e-05,
+      "loss": 0.5597,
+      "step": 13864
+    },
+    {
+      "epoch": 17.7472,
+      "grad_norm": 1.025348424911499,
+      "learning_rate": 2.2284913965586236e-05,
+      "loss": 0.5118,
+      "step": 13865
+    },
+    {
+      "epoch": 17.74848,
+      "grad_norm": 1.0150192975997925,
+      "learning_rate": 2.2282913165266108e-05,
+      "loss": 0.4987,
+      "step": 13866
+    },
+    {
+      "epoch": 17.74976,
+      "grad_norm": 1.0555064678192139,
+      "learning_rate": 2.228091236494598e-05,
+      "loss": 0.5402,
+      "step": 13867
+    },
+    {
+      "epoch": 17.75104,
+      "grad_norm": 1.0361968278884888,
+      "learning_rate": 2.2278911564625852e-05,
+      "loss": 0.5342,
+      "step": 13868
+    },
+    {
+      "epoch": 17.75232,
+      "grad_norm": 1.0574556589126587,
+      "learning_rate": 2.2276910764305724e-05,
+      "loss": 0.5591,
+      "step": 13869
+    },
+    {
+      "epoch": 17.7536,
+      "grad_norm": 1.021784782409668,
+      "learning_rate": 2.2274909963985595e-05,
+      "loss": 0.538,
+      "step": 13870
+    },
+    {
+      "epoch": 17.75488,
+      "grad_norm": 0.9862391352653503,
+      "learning_rate": 2.2272909163665467e-05,
+      "loss": 0.4779,
+      "step": 13871
+    },
+    {
+      "epoch": 17.75616,
+      "grad_norm": 1.0507396459579468,
+      "learning_rate": 2.227090836334534e-05,
+      "loss": 0.4927,
+      "step": 13872
+    },
+    {
+      "epoch": 17.75744,
+      "grad_norm": 0.9640161991119385,
+      "learning_rate": 2.226890756302521e-05,
+      "loss": 0.4648,
+      "step": 13873
+    },
+    {
+      "epoch": 17.75872,
+      "grad_norm": 1.0764763355255127,
+      "learning_rate": 2.2266906762705083e-05,
+      "loss": 0.5556,
+      "step": 13874
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 1.0238678455352783,
+      "learning_rate": 2.2264905962384955e-05,
+      "loss": 0.5237,
+      "step": 13875
+    },
+    {
+      "epoch": 17.76128,
+      "grad_norm": 1.0710828304290771,
+      "learning_rate": 2.2262905162064827e-05,
+      "loss": 0.579,
+      "step": 13876
+    },
+    {
+      "epoch": 17.76256,
+      "grad_norm": 1.02402663230896,
+      "learning_rate": 2.22609043617447e-05,
+      "loss": 0.5118,
+      "step": 13877
+    },
+    {
+      "epoch": 17.76384,
+      "grad_norm": 1.037920355796814,
+      "learning_rate": 2.225890356142457e-05,
+      "loss": 0.5287,
+      "step": 13878
+    },
+    {
+      "epoch": 17.76512,
+      "grad_norm": 0.9906246066093445,
+      "learning_rate": 2.2256902761104442e-05,
+      "loss": 0.5074,
+      "step": 13879
+    },
+    {
+      "epoch": 17.7664,
+      "grad_norm": 1.0306636095046997,
+      "learning_rate": 2.2254901960784314e-05,
+      "loss": 0.4908,
+      "step": 13880
+    },
+    {
+      "epoch": 17.76768,
+      "grad_norm": 1.0256168842315674,
+      "learning_rate": 2.2252901160464186e-05,
+      "loss": 0.4749,
+      "step": 13881
+    },
+    {
+      "epoch": 17.76896,
+      "grad_norm": 1.0150201320648193,
+      "learning_rate": 2.2250900360144058e-05,
+      "loss": 0.5489,
+      "step": 13882
+    },
+    {
+      "epoch": 17.77024,
+      "grad_norm": 1.1066970825195312,
+      "learning_rate": 2.224889955982393e-05,
+      "loss": 0.583,
+      "step": 13883
+    },
+    {
+      "epoch": 17.77152,
+      "grad_norm": 0.9935269355773926,
+      "learning_rate": 2.2246898759503805e-05,
+      "loss": 0.471,
+      "step": 13884
+    },
+    {
+      "epoch": 17.7728,
+      "grad_norm": 1.0902072191238403,
+      "learning_rate": 2.2244897959183673e-05,
+      "loss": 0.5516,
+      "step": 13885
+    },
+    {
+      "epoch": 17.77408,
+      "grad_norm": 1.0285429954528809,
+      "learning_rate": 2.2242897158863545e-05,
+      "loss": 0.5101,
+      "step": 13886
+    },
+    {
+      "epoch": 17.77536,
+      "grad_norm": 0.9977599382400513,
+      "learning_rate": 2.2240896358543417e-05,
+      "loss": 0.5045,
+      "step": 13887
+    },
+    {
+      "epoch": 17.77664,
+      "grad_norm": 1.0018473863601685,
+      "learning_rate": 2.2238895558223292e-05,
+      "loss": 0.5293,
+      "step": 13888
+    },
+    {
+      "epoch": 17.77792,
+      "grad_norm": 1.027341604232788,
+      "learning_rate": 2.223689475790316e-05,
+      "loss": 0.5137,
+      "step": 13889
+    },
+    {
+      "epoch": 17.7792,
+      "grad_norm": 1.0242568254470825,
+      "learning_rate": 2.2234893957583033e-05,
+      "loss": 0.5404,
+      "step": 13890
+    },
+    {
+      "epoch": 17.78048,
+      "grad_norm": 0.9905862212181091,
+      "learning_rate": 2.2232893157262908e-05,
+      "loss": 0.5376,
+      "step": 13891
+    },
+    {
+      "epoch": 17.78176,
+      "grad_norm": 0.919065535068512,
+      "learning_rate": 2.223089235694278e-05,
+      "loss": 0.4514,
+      "step": 13892
+    },
+    {
+      "epoch": 17.78304,
+      "grad_norm": 0.9959515333175659,
+      "learning_rate": 2.2228891556622648e-05,
+      "loss": 0.4923,
+      "step": 13893
+    },
+    {
+      "epoch": 17.78432,
+      "grad_norm": 1.0490683317184448,
+      "learning_rate": 2.222689075630252e-05,
+      "loss": 0.4993,
+      "step": 13894
+    },
+    {
+      "epoch": 17.7856,
+      "grad_norm": 1.04204261302948,
+      "learning_rate": 2.2224889955982395e-05,
+      "loss": 0.523,
+      "step": 13895
+    },
+    {
+      "epoch": 17.78688,
+      "grad_norm": 1.001023530960083,
+      "learning_rate": 2.2222889155662267e-05,
+      "loss": 0.524,
+      "step": 13896
+    },
+    {
+      "epoch": 17.78816,
+      "grad_norm": 1.0491348505020142,
+      "learning_rate": 2.2220888355342136e-05,
+      "loss": 0.4963,
+      "step": 13897
+    },
+    {
+      "epoch": 17.78944,
+      "grad_norm": 1.046494960784912,
+      "learning_rate": 2.221888755502201e-05,
+      "loss": 0.5521,
+      "step": 13898
+    },
+    {
+      "epoch": 17.79072,
+      "grad_norm": 1.0331013202667236,
+      "learning_rate": 2.2216886754701883e-05,
+      "loss": 0.5306,
+      "step": 13899
+    },
+    {
+      "epoch": 17.792,
+      "grad_norm": 0.9884205460548401,
+      "learning_rate": 2.2214885954381755e-05,
+      "loss": 0.469,
+      "step": 13900
+    },
+    {
+      "epoch": 17.79328,
+      "grad_norm": 1.1054868698120117,
+      "learning_rate": 2.2212885154061623e-05,
+      "loss": 0.4882,
+      "step": 13901
+    },
+    {
+      "epoch": 17.79456,
+      "grad_norm": 1.0739738941192627,
+      "learning_rate": 2.2210884353741498e-05,
+      "loss": 0.5627,
+      "step": 13902
+    },
+    {
+      "epoch": 17.79584,
+      "grad_norm": 0.9757269024848938,
+      "learning_rate": 2.220888355342137e-05,
+      "loss": 0.474,
+      "step": 13903
+    },
+    {
+      "epoch": 17.79712,
+      "grad_norm": 0.9898340702056885,
+      "learning_rate": 2.2206882753101242e-05,
+      "loss": 0.4801,
+      "step": 13904
+    },
+    {
+      "epoch": 17.7984,
+      "grad_norm": 0.9995481967926025,
+      "learning_rate": 2.2204881952781114e-05,
+      "loss": 0.468,
+      "step": 13905
+    },
+    {
+      "epoch": 17.79968,
+      "grad_norm": 1.0464537143707275,
+      "learning_rate": 2.2202881152460986e-05,
+      "loss": 0.4884,
+      "step": 13906
+    },
+    {
+      "epoch": 17.80096,
+      "grad_norm": 1.0073833465576172,
+      "learning_rate": 2.2200880352140858e-05,
+      "loss": 0.5333,
+      "step": 13907
+    },
+    {
+      "epoch": 17.80224,
+      "grad_norm": 1.025449514389038,
+      "learning_rate": 2.219887955182073e-05,
+      "loss": 0.5314,
+      "step": 13908
+    },
+    {
+      "epoch": 17.80352,
+      "grad_norm": 0.9930814504623413,
+      "learning_rate": 2.21968787515006e-05,
+      "loss": 0.5318,
+      "step": 13909
+    },
+    {
+      "epoch": 17.8048,
+      "grad_norm": 0.9848999977111816,
+      "learning_rate": 2.2194877951180473e-05,
+      "loss": 0.4798,
+      "step": 13910
+    },
+    {
+      "epoch": 17.80608,
+      "grad_norm": 1.0277302265167236,
+      "learning_rate": 2.2192877150860345e-05,
+      "loss": 0.4967,
+      "step": 13911
+    },
+    {
+      "epoch": 17.80736,
+      "grad_norm": 1.0796613693237305,
+      "learning_rate": 2.2190876350540217e-05,
+      "loss": 0.5304,
+      "step": 13912
+    },
+    {
+      "epoch": 17.80864,
+      "grad_norm": 1.055225133895874,
+      "learning_rate": 2.218887555022009e-05,
+      "loss": 0.5225,
+      "step": 13913
+    },
+    {
+      "epoch": 17.809919999999998,
+      "grad_norm": 1.039272665977478,
+      "learning_rate": 2.218687474989996e-05,
+      "loss": 0.4892,
+      "step": 13914
+    },
+    {
+      "epoch": 17.8112,
+      "grad_norm": 1.0163494348526,
+      "learning_rate": 2.2184873949579832e-05,
+      "loss": 0.4807,
+      "step": 13915
+    },
+    {
+      "epoch": 17.81248,
+      "grad_norm": 1.0309205055236816,
+      "learning_rate": 2.2182873149259704e-05,
+      "loss": 0.534,
+      "step": 13916
+    },
+    {
+      "epoch": 17.81376,
+      "grad_norm": 0.9438468813896179,
+      "learning_rate": 2.2180872348939576e-05,
+      "loss": 0.4871,
+      "step": 13917
+    },
+    {
+      "epoch": 17.81504,
+      "grad_norm": 1.0279080867767334,
+      "learning_rate": 2.2178871548619448e-05,
+      "loss": 0.474,
+      "step": 13918
+    },
+    {
+      "epoch": 17.81632,
+      "grad_norm": 1.0761826038360596,
+      "learning_rate": 2.2176870748299323e-05,
+      "loss": 0.53,
+      "step": 13919
+    },
+    {
+      "epoch": 17.8176,
+      "grad_norm": 1.025966763496399,
+      "learning_rate": 2.217486994797919e-05,
+      "loss": 0.505,
+      "step": 13920
+    },
+    {
+      "epoch": 17.81888,
+      "grad_norm": 1.0639162063598633,
+      "learning_rate": 2.2172869147659064e-05,
+      "loss": 0.5237,
+      "step": 13921
+    },
+    {
+      "epoch": 17.82016,
+      "grad_norm": 1.0175074338912964,
+      "learning_rate": 2.2170868347338935e-05,
+      "loss": 0.5159,
+      "step": 13922
+    },
+    {
+      "epoch": 17.82144,
+      "grad_norm": 1.073966145515442,
+      "learning_rate": 2.216886754701881e-05,
+      "loss": 0.5687,
+      "step": 13923
+    },
+    {
+      "epoch": 17.82272,
+      "grad_norm": 1.0590076446533203,
+      "learning_rate": 2.216686674669868e-05,
+      "loss": 0.4935,
+      "step": 13924
+    },
+    {
+      "epoch": 17.824,
+      "grad_norm": 0.9631067514419556,
+      "learning_rate": 2.216486594637855e-05,
+      "loss": 0.4657,
+      "step": 13925
+    },
+    {
+      "epoch": 17.82528,
+      "grad_norm": 1.0229510068893433,
+      "learning_rate": 2.2162865146058426e-05,
+      "loss": 0.4885,
+      "step": 13926
+    },
+    {
+      "epoch": 17.82656,
+      "grad_norm": 1.0246883630752563,
+      "learning_rate": 2.2160864345738298e-05,
+      "loss": 0.5242,
+      "step": 13927
+    },
+    {
+      "epoch": 17.82784,
+      "grad_norm": 1.0278314352035522,
+      "learning_rate": 2.2158863545418167e-05,
+      "loss": 0.5112,
+      "step": 13928
+    },
+    {
+      "epoch": 17.82912,
+      "grad_norm": 1.0408381223678589,
+      "learning_rate": 2.215686274509804e-05,
+      "loss": 0.5122,
+      "step": 13929
+    },
+    {
+      "epoch": 17.8304,
+      "grad_norm": 0.9866704940795898,
+      "learning_rate": 2.2154861944777914e-05,
+      "loss": 0.5143,
+      "step": 13930
+    },
+    {
+      "epoch": 17.83168,
+      "grad_norm": 1.005631446838379,
+      "learning_rate": 2.2152861144457786e-05,
+      "loss": 0.4981,
+      "step": 13931
+    },
+    {
+      "epoch": 17.83296,
+      "grad_norm": 0.9910889267921448,
+      "learning_rate": 2.2150860344137654e-05,
+      "loss": 0.4984,
+      "step": 13932
+    },
+    {
+      "epoch": 17.83424,
+      "grad_norm": 1.0972979068756104,
+      "learning_rate": 2.2148859543817526e-05,
+      "loss": 0.5075,
+      "step": 13933
+    },
+    {
+      "epoch": 17.83552,
+      "grad_norm": 1.0541285276412964,
+      "learning_rate": 2.21468587434974e-05,
+      "loss": 0.484,
+      "step": 13934
+    },
+    {
+      "epoch": 17.8368,
+      "grad_norm": 1.0120394229888916,
+      "learning_rate": 2.2144857943177273e-05,
+      "loss": 0.4929,
+      "step": 13935
+    },
+    {
+      "epoch": 17.83808,
+      "grad_norm": 1.0291119813919067,
+      "learning_rate": 2.214285714285714e-05,
+      "loss": 0.5132,
+      "step": 13936
+    },
+    {
+      "epoch": 17.83936,
+      "grad_norm": 1.034484624862671,
+      "learning_rate": 2.2140856342537017e-05,
+      "loss": 0.4899,
+      "step": 13937
+    },
+    {
+      "epoch": 17.84064,
+      "grad_norm": 1.0646271705627441,
+      "learning_rate": 2.213885554221689e-05,
+      "loss": 0.4747,
+      "step": 13938
+    },
+    {
+      "epoch": 17.841920000000002,
+      "grad_norm": 1.0118857622146606,
+      "learning_rate": 2.213685474189676e-05,
+      "loss": 0.4824,
+      "step": 13939
+    },
+    {
+      "epoch": 17.8432,
+      "grad_norm": 1.0535541772842407,
+      "learning_rate": 2.213485394157663e-05,
+      "loss": 0.4842,
+      "step": 13940
+    },
+    {
+      "epoch": 17.84448,
+      "grad_norm": 1.028926968574524,
+      "learning_rate": 2.2132853141256504e-05,
+      "loss": 0.5243,
+      "step": 13941
+    },
+    {
+      "epoch": 17.84576,
+      "grad_norm": 1.0521934032440186,
+      "learning_rate": 2.2130852340936376e-05,
+      "loss": 0.5343,
+      "step": 13942
+    },
+    {
+      "epoch": 17.84704,
+      "grad_norm": 1.0421346426010132,
+      "learning_rate": 2.2128851540616248e-05,
+      "loss": 0.4903,
+      "step": 13943
+    },
+    {
+      "epoch": 17.84832,
+      "grad_norm": 1.0207431316375732,
+      "learning_rate": 2.212685074029612e-05,
+      "loss": 0.509,
+      "step": 13944
+    },
+    {
+      "epoch": 17.8496,
+      "grad_norm": 1.0445460081100464,
+      "learning_rate": 2.212484993997599e-05,
+      "loss": 0.5096,
+      "step": 13945
+    },
+    {
+      "epoch": 17.85088,
+      "grad_norm": 1.0498018264770508,
+      "learning_rate": 2.2122849139655863e-05,
+      "loss": 0.4767,
+      "step": 13946
+    },
+    {
+      "epoch": 17.85216,
+      "grad_norm": 1.0488752126693726,
+      "learning_rate": 2.2120848339335735e-05,
+      "loss": 0.5048,
+      "step": 13947
+    },
+    {
+      "epoch": 17.85344,
+      "grad_norm": 0.981802225112915,
+      "learning_rate": 2.2118847539015607e-05,
+      "loss": 0.5113,
+      "step": 13948
+    },
+    {
+      "epoch": 17.85472,
+      "grad_norm": 0.988862156867981,
+      "learning_rate": 2.211684673869548e-05,
+      "loss": 0.4887,
+      "step": 13949
+    },
+    {
+      "epoch": 17.856,
+      "grad_norm": 1.0477064847946167,
+      "learning_rate": 2.211484593837535e-05,
+      "loss": 0.5259,
+      "step": 13950
+    },
+    {
+      "epoch": 17.85728,
+      "grad_norm": 1.0353902578353882,
+      "learning_rate": 2.2112845138055223e-05,
+      "loss": 0.5022,
+      "step": 13951
+    },
+    {
+      "epoch": 17.85856,
+      "grad_norm": 1.0380783081054688,
+      "learning_rate": 2.2110844337735094e-05,
+      "loss": 0.5197,
+      "step": 13952
+    },
+    {
+      "epoch": 17.85984,
+      "grad_norm": 0.9977851510047913,
+      "learning_rate": 2.2108843537414966e-05,
+      "loss": 0.4844,
+      "step": 13953
+    },
+    {
+      "epoch": 17.86112,
+      "grad_norm": 1.031292200088501,
+      "learning_rate": 2.2106842737094838e-05,
+      "loss": 0.5341,
+      "step": 13954
+    },
+    {
+      "epoch": 17.8624,
+      "grad_norm": 1.0284069776535034,
+      "learning_rate": 2.210484193677471e-05,
+      "loss": 0.4888,
+      "step": 13955
+    },
+    {
+      "epoch": 17.86368,
+      "grad_norm": 1.0368765592575073,
+      "learning_rate": 2.2102841136454582e-05,
+      "loss": 0.5599,
+      "step": 13956
+    },
+    {
+      "epoch": 17.86496,
+      "grad_norm": 0.9683583378791809,
+      "learning_rate": 2.2100840336134454e-05,
+      "loss": 0.5076,
+      "step": 13957
+    },
+    {
+      "epoch": 17.86624,
+      "grad_norm": 1.0280176401138306,
+      "learning_rate": 2.209883953581433e-05,
+      "loss": 0.5281,
+      "step": 13958
+    },
+    {
+      "epoch": 17.86752,
+      "grad_norm": 1.078447937965393,
+      "learning_rate": 2.2096838735494197e-05,
+      "loss": 0.5285,
+      "step": 13959
+    },
+    {
+      "epoch": 17.8688,
+      "grad_norm": 1.0445505380630493,
+      "learning_rate": 2.209483793517407e-05,
+      "loss": 0.5509,
+      "step": 13960
+    },
+    {
+      "epoch": 17.87008,
+      "grad_norm": 1.0536412000656128,
+      "learning_rate": 2.209283713485394e-05,
+      "loss": 0.5003,
+      "step": 13961
+    },
+    {
+      "epoch": 17.87136,
+      "grad_norm": 0.9787836074829102,
+      "learning_rate": 2.2090836334533816e-05,
+      "loss": 0.4461,
+      "step": 13962
+    },
+    {
+      "epoch": 17.87264,
+      "grad_norm": 1.0095018148422241,
+      "learning_rate": 2.2088835534213685e-05,
+      "loss": 0.4955,
+      "step": 13963
+    },
+    {
+      "epoch": 17.87392,
+      "grad_norm": 1.0202372074127197,
+      "learning_rate": 2.2086834733893557e-05,
+      "loss": 0.5101,
+      "step": 13964
+    },
+    {
+      "epoch": 17.8752,
+      "grad_norm": 1.010644555091858,
+      "learning_rate": 2.2084833933573432e-05,
+      "loss": 0.508,
+      "step": 13965
+    },
+    {
+      "epoch": 17.87648,
+      "grad_norm": 1.0773608684539795,
+      "learning_rate": 2.2082833133253304e-05,
+      "loss": 0.4861,
+      "step": 13966
+    },
+    {
+      "epoch": 17.87776,
+      "grad_norm": 1.0331664085388184,
+      "learning_rate": 2.2080832332933172e-05,
+      "loss": 0.5204,
+      "step": 13967
+    },
+    {
+      "epoch": 17.87904,
+      "grad_norm": 1.081826090812683,
+      "learning_rate": 2.2078831532613044e-05,
+      "loss": 0.5497,
+      "step": 13968
+    },
+    {
+      "epoch": 17.88032,
+      "grad_norm": 1.0755976438522339,
+      "learning_rate": 2.207683073229292e-05,
+      "loss": 0.5408,
+      "step": 13969
+    },
+    {
+      "epoch": 17.8816,
+      "grad_norm": 1.0068060159683228,
+      "learning_rate": 2.207482993197279e-05,
+      "loss": 0.5003,
+      "step": 13970
+    },
+    {
+      "epoch": 17.88288,
+      "grad_norm": 0.9964351058006287,
+      "learning_rate": 2.207282913165266e-05,
+      "loss": 0.4895,
+      "step": 13971
+    },
+    {
+      "epoch": 17.88416,
+      "grad_norm": 1.0196988582611084,
+      "learning_rate": 2.2070828331332535e-05,
+      "loss": 0.5039,
+      "step": 13972
+    },
+    {
+      "epoch": 17.88544,
+      "grad_norm": 1.0305639505386353,
+      "learning_rate": 2.2068827531012407e-05,
+      "loss": 0.5213,
+      "step": 13973
+    },
+    {
+      "epoch": 17.88672,
+      "grad_norm": 1.0133947134017944,
+      "learning_rate": 2.206682673069228e-05,
+      "loss": 0.5076,
+      "step": 13974
+    },
+    {
+      "epoch": 17.888,
+      "grad_norm": 1.0059137344360352,
+      "learning_rate": 2.2064825930372147e-05,
+      "loss": 0.5173,
+      "step": 13975
+    },
+    {
+      "epoch": 17.88928,
+      "grad_norm": 1.0029923915863037,
+      "learning_rate": 2.2062825130052022e-05,
+      "loss": 0.5354,
+      "step": 13976
+    },
+    {
+      "epoch": 17.89056,
+      "grad_norm": 0.9647006392478943,
+      "learning_rate": 2.2060824329731894e-05,
+      "loss": 0.5142,
+      "step": 13977
+    },
+    {
+      "epoch": 17.89184,
+      "grad_norm": 1.0146187543869019,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 0.5561,
+      "step": 13978
+    },
+    {
+      "epoch": 17.89312,
+      "grad_norm": 1.0384979248046875,
+      "learning_rate": 2.2056822729091638e-05,
+      "loss": 0.5139,
+      "step": 13979
+    },
+    {
+      "epoch": 17.8944,
+      "grad_norm": 0.9908726215362549,
+      "learning_rate": 2.205482192877151e-05,
+      "loss": 0.5435,
+      "step": 13980
+    },
+    {
+      "epoch": 17.89568,
+      "grad_norm": 1.0110816955566406,
+      "learning_rate": 2.2052821128451382e-05,
+      "loss": 0.5114,
+      "step": 13981
+    },
+    {
+      "epoch": 17.89696,
+      "grad_norm": 1.0027389526367188,
+      "learning_rate": 2.2050820328131254e-05,
+      "loss": 0.5125,
+      "step": 13982
+    },
+    {
+      "epoch": 17.89824,
+      "grad_norm": 0.9981145262718201,
+      "learning_rate": 2.2048819527811125e-05,
+      "loss": 0.5052,
+      "step": 13983
+    },
+    {
+      "epoch": 17.89952,
+      "grad_norm": 1.0781457424163818,
+      "learning_rate": 2.2046818727490997e-05,
+      "loss": 0.5885,
+      "step": 13984
+    },
+    {
+      "epoch": 17.9008,
+      "grad_norm": 1.0383636951446533,
+      "learning_rate": 2.204481792717087e-05,
+      "loss": 0.5163,
+      "step": 13985
+    },
+    {
+      "epoch": 17.90208,
+      "grad_norm": 1.020270824432373,
+      "learning_rate": 2.2042817126850744e-05,
+      "loss": 0.5207,
+      "step": 13986
+    },
+    {
+      "epoch": 17.90336,
+      "grad_norm": 1.0009729862213135,
+      "learning_rate": 2.2040816326530613e-05,
+      "loss": 0.6061,
+      "step": 13987
+    },
+    {
+      "epoch": 17.90464,
+      "grad_norm": 1.0783268213272095,
+      "learning_rate": 2.2038815526210485e-05,
+      "loss": 0.521,
+      "step": 13988
+    },
+    {
+      "epoch": 17.90592,
+      "grad_norm": 1.0087323188781738,
+      "learning_rate": 2.2036814725890357e-05,
+      "loss": 0.4846,
+      "step": 13989
+    },
+    {
+      "epoch": 17.9072,
+      "grad_norm": 1.005253553390503,
+      "learning_rate": 2.2034813925570232e-05,
+      "loss": 0.5024,
+      "step": 13990
+    },
+    {
+      "epoch": 17.90848,
+      "grad_norm": 1.0598344802856445,
+      "learning_rate": 2.20328131252501e-05,
+      "loss": 0.5176,
+      "step": 13991
+    },
+    {
+      "epoch": 17.90976,
+      "grad_norm": 1.010216474533081,
+      "learning_rate": 2.2030812324929972e-05,
+      "loss": 0.4949,
+      "step": 13992
+    },
+    {
+      "epoch": 17.91104,
+      "grad_norm": 1.0626128911972046,
+      "learning_rate": 2.2028811524609844e-05,
+      "loss": 0.5698,
+      "step": 13993
+    },
+    {
+      "epoch": 17.91232,
+      "grad_norm": 1.0089972019195557,
+      "learning_rate": 2.202681072428972e-05,
+      "loss": 0.4999,
+      "step": 13994
+    },
+    {
+      "epoch": 17.9136,
+      "grad_norm": 1.0127496719360352,
+      "learning_rate": 2.2024809923969588e-05,
+      "loss": 0.5169,
+      "step": 13995
+    },
+    {
+      "epoch": 17.91488,
+      "grad_norm": 1.058738350868225,
+      "learning_rate": 2.202280912364946e-05,
+      "loss": 0.4963,
+      "step": 13996
+    },
+    {
+      "epoch": 17.91616,
+      "grad_norm": 1.0318009853363037,
+      "learning_rate": 2.2020808323329335e-05,
+      "loss": 0.517,
+      "step": 13997
+    },
+    {
+      "epoch": 17.91744,
+      "grad_norm": 1.0412254333496094,
+      "learning_rate": 2.2018807523009207e-05,
+      "loss": 0.5259,
+      "step": 13998
+    },
+    {
+      "epoch": 17.91872,
+      "grad_norm": 1.061491847038269,
+      "learning_rate": 2.2016806722689075e-05,
+      "loss": 0.5451,
+      "step": 13999
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 1.0782477855682373,
+      "learning_rate": 2.2014805922368947e-05,
+      "loss": 0.5462,
+      "step": 14000
+    },
+    {
+      "epoch": 17.92128,
+      "grad_norm": 1.0478885173797607,
+      "learning_rate": 2.2012805122048822e-05,
+      "loss": 0.5489,
+      "step": 14001
+    },
+    {
+      "epoch": 17.92256,
+      "grad_norm": 1.0254915952682495,
+      "learning_rate": 2.2010804321728694e-05,
+      "loss": 0.5149,
+      "step": 14002
+    },
+    {
+      "epoch": 17.92384,
+      "grad_norm": 1.0770113468170166,
+      "learning_rate": 2.2008803521408563e-05,
+      "loss": 0.5494,
+      "step": 14003
+    },
+    {
+      "epoch": 17.92512,
+      "grad_norm": 1.0335642099380493,
+      "learning_rate": 2.2006802721088438e-05,
+      "loss": 0.4951,
+      "step": 14004
+    },
+    {
+      "epoch": 17.9264,
+      "grad_norm": 1.0393280982971191,
+      "learning_rate": 2.200480192076831e-05,
+      "loss": 0.5505,
+      "step": 14005
+    },
+    {
+      "epoch": 17.92768,
+      "grad_norm": 0.9929527044296265,
+      "learning_rate": 2.200280112044818e-05,
+      "loss": 0.4796,
+      "step": 14006
+    },
+    {
+      "epoch": 17.92896,
+      "grad_norm": 0.9717233777046204,
+      "learning_rate": 2.200080032012805e-05,
+      "loss": 0.4864,
+      "step": 14007
+    },
+    {
+      "epoch": 17.93024,
+      "grad_norm": 1.0142488479614258,
+      "learning_rate": 2.1998799519807925e-05,
+      "loss": 0.5637,
+      "step": 14008
+    },
+    {
+      "epoch": 17.93152,
+      "grad_norm": 1.0879509449005127,
+      "learning_rate": 2.1996798719487797e-05,
+      "loss": 0.5425,
+      "step": 14009
+    },
+    {
+      "epoch": 17.9328,
+      "grad_norm": 1.0561232566833496,
+      "learning_rate": 2.199479791916767e-05,
+      "loss": 0.4915,
+      "step": 14010
+    },
+    {
+      "epoch": 17.93408,
+      "grad_norm": 1.0183889865875244,
+      "learning_rate": 2.199279711884754e-05,
+      "loss": 0.4996,
+      "step": 14011
+    },
+    {
+      "epoch": 17.93536,
+      "grad_norm": 0.9836266040802002,
+      "learning_rate": 2.1990796318527413e-05,
+      "loss": 0.4174,
+      "step": 14012
+    },
+    {
+      "epoch": 17.93664,
+      "grad_norm": 1.0205684900283813,
+      "learning_rate": 2.1988795518207285e-05,
+      "loss": 0.5009,
+      "step": 14013
+    },
+    {
+      "epoch": 17.93792,
+      "grad_norm": 1.0839546918869019,
+      "learning_rate": 2.1986794717887156e-05,
+      "loss": 0.5532,
+      "step": 14014
+    },
+    {
+      "epoch": 17.9392,
+      "grad_norm": 0.9612104296684265,
+      "learning_rate": 2.1984793917567028e-05,
+      "loss": 0.4711,
+      "step": 14015
+    },
+    {
+      "epoch": 17.94048,
+      "grad_norm": 1.0516858100891113,
+      "learning_rate": 2.19827931172469e-05,
+      "loss": 0.5215,
+      "step": 14016
+    },
+    {
+      "epoch": 17.94176,
+      "grad_norm": 1.1283824443817139,
+      "learning_rate": 2.1980792316926772e-05,
+      "loss": 0.6009,
+      "step": 14017
+    },
+    {
+      "epoch": 17.94304,
+      "grad_norm": 0.9951545000076294,
+      "learning_rate": 2.1978791516606644e-05,
+      "loss": 0.4589,
+      "step": 14018
+    },
+    {
+      "epoch": 17.94432,
+      "grad_norm": 0.9906835556030273,
+      "learning_rate": 2.1976790716286516e-05,
+      "loss": 0.5127,
+      "step": 14019
+    },
+    {
+      "epoch": 17.9456,
+      "grad_norm": 1.0642714500427246,
+      "learning_rate": 2.1974789915966388e-05,
+      "loss": 0.5101,
+      "step": 14020
+    },
+    {
+      "epoch": 17.94688,
+      "grad_norm": 1.098826289176941,
+      "learning_rate": 2.197278911564626e-05,
+      "loss": 0.5939,
+      "step": 14021
+    },
+    {
+      "epoch": 17.94816,
+      "grad_norm": 1.0480844974517822,
+      "learning_rate": 2.197078831532613e-05,
+      "loss": 0.5587,
+      "step": 14022
+    },
+    {
+      "epoch": 17.94944,
+      "grad_norm": 1.0224511623382568,
+      "learning_rate": 2.1968787515006003e-05,
+      "loss": 0.4883,
+      "step": 14023
+    },
+    {
+      "epoch": 17.95072,
+      "grad_norm": 1.0356502532958984,
+      "learning_rate": 2.1966786714685875e-05,
+      "loss": 0.5223,
+      "step": 14024
+    },
+    {
+      "epoch": 17.951999999999998,
+      "grad_norm": 1.0777286291122437,
+      "learning_rate": 2.196478591436575e-05,
+      "loss": 0.5237,
+      "step": 14025
+    },
+    {
+      "epoch": 17.95328,
+      "grad_norm": 1.1121007204055786,
+      "learning_rate": 2.196278511404562e-05,
+      "loss": 0.5715,
+      "step": 14026
+    },
+    {
+      "epoch": 17.95456,
+      "grad_norm": 1.068606972694397,
+      "learning_rate": 2.196078431372549e-05,
+      "loss": 0.5258,
+      "step": 14027
+    },
+    {
+      "epoch": 17.95584,
+      "grad_norm": 0.943596363067627,
+      "learning_rate": 2.1958783513405362e-05,
+      "loss": 0.4419,
+      "step": 14028
+    },
+    {
+      "epoch": 17.95712,
+      "grad_norm": 0.9647386074066162,
+      "learning_rate": 2.1956782713085238e-05,
+      "loss": 0.5067,
+      "step": 14029
+    },
+    {
+      "epoch": 17.9584,
+      "grad_norm": 0.9991288781166077,
+      "learning_rate": 2.1954781912765106e-05,
+      "loss": 0.4659,
+      "step": 14030
+    },
+    {
+      "epoch": 17.95968,
+      "grad_norm": 1.0615739822387695,
+      "learning_rate": 2.1952781112444978e-05,
+      "loss": 0.5157,
+      "step": 14031
+    },
+    {
+      "epoch": 17.96096,
+      "grad_norm": 1.0784728527069092,
+      "learning_rate": 2.1950780312124853e-05,
+      "loss": 0.5038,
+      "step": 14032
+    },
+    {
+      "epoch": 17.96224,
+      "grad_norm": 1.0348066091537476,
+      "learning_rate": 2.1948779511804725e-05,
+      "loss": 0.5264,
+      "step": 14033
+    },
+    {
+      "epoch": 17.96352,
+      "grad_norm": 0.9429087042808533,
+      "learning_rate": 2.1946778711484594e-05,
+      "loss": 0.4768,
+      "step": 14034
+    },
+    {
+      "epoch": 17.9648,
+      "grad_norm": 1.0070911645889282,
+      "learning_rate": 2.1944777911164465e-05,
+      "loss": 0.491,
+      "step": 14035
+    },
+    {
+      "epoch": 17.96608,
+      "grad_norm": 1.0489453077316284,
+      "learning_rate": 2.194277711084434e-05,
+      "loss": 0.5086,
+      "step": 14036
+    },
+    {
+      "epoch": 17.96736,
+      "grad_norm": 0.9969155788421631,
+      "learning_rate": 2.1940776310524212e-05,
+      "loss": 0.5171,
+      "step": 14037
+    },
+    {
+      "epoch": 17.96864,
+      "grad_norm": 1.0200451612472534,
+      "learning_rate": 2.193877551020408e-05,
+      "loss": 0.4763,
+      "step": 14038
+    },
+    {
+      "epoch": 17.96992,
+      "grad_norm": 1.1074117422103882,
+      "learning_rate": 2.1936774709883956e-05,
+      "loss": 0.5653,
+      "step": 14039
+    },
+    {
+      "epoch": 17.9712,
+      "grad_norm": 0.9458688497543335,
+      "learning_rate": 2.1934773909563828e-05,
+      "loss": 0.4885,
+      "step": 14040
+    },
+    {
+      "epoch": 17.97248,
+      "grad_norm": 0.9838624000549316,
+      "learning_rate": 2.19327731092437e-05,
+      "loss": 0.4665,
+      "step": 14041
+    },
+    {
+      "epoch": 17.97376,
+      "grad_norm": 0.9943659901618958,
+      "learning_rate": 2.193077230892357e-05,
+      "loss": 0.5485,
+      "step": 14042
+    },
+    {
+      "epoch": 17.97504,
+      "grad_norm": 1.0677356719970703,
+      "learning_rate": 2.1928771508603444e-05,
+      "loss": 0.5672,
+      "step": 14043
+    },
+    {
+      "epoch": 17.97632,
+      "grad_norm": 0.9993770122528076,
+      "learning_rate": 2.1926770708283315e-05,
+      "loss": 0.5041,
+      "step": 14044
+    },
+    {
+      "epoch": 17.9776,
+      "grad_norm": 1.0094391107559204,
+      "learning_rate": 2.1924769907963187e-05,
+      "loss": 0.5,
+      "step": 14045
+    },
+    {
+      "epoch": 17.97888,
+      "grad_norm": 0.9990932941436768,
+      "learning_rate": 2.1922769107643056e-05,
+      "loss": 0.5015,
+      "step": 14046
+    },
+    {
+      "epoch": 17.98016,
+      "grad_norm": 0.9945876002311707,
+      "learning_rate": 2.192076830732293e-05,
+      "loss": 0.4928,
+      "step": 14047
+    },
+    {
+      "epoch": 17.98144,
+      "grad_norm": 1.1232936382293701,
+      "learning_rate": 2.1918767507002803e-05,
+      "loss": 0.5627,
+      "step": 14048
+    },
+    {
+      "epoch": 17.98272,
+      "grad_norm": 1.0336296558380127,
+      "learning_rate": 2.1916766706682675e-05,
+      "loss": 0.514,
+      "step": 14049
+    },
+    {
+      "epoch": 17.984,
+      "grad_norm": 1.111663818359375,
+      "learning_rate": 2.1914765906362547e-05,
+      "loss": 0.5868,
+      "step": 14050
+    },
+    {
+      "epoch": 17.98528,
+      "grad_norm": 1.0357799530029297,
+      "learning_rate": 2.191276510604242e-05,
+      "loss": 0.5266,
+      "step": 14051
+    },
+    {
+      "epoch": 17.98656,
+      "grad_norm": 1.0117226839065552,
+      "learning_rate": 2.191076430572229e-05,
+      "loss": 0.5048,
+      "step": 14052
+    },
+    {
+      "epoch": 17.98784,
+      "grad_norm": 0.9796508550643921,
+      "learning_rate": 2.1908763505402162e-05,
+      "loss": 0.5113,
+      "step": 14053
+    },
+    {
+      "epoch": 17.98912,
+      "grad_norm": 1.028061032295227,
+      "learning_rate": 2.1906762705082034e-05,
+      "loss": 0.4864,
+      "step": 14054
+    },
+    {
+      "epoch": 17.9904,
+      "grad_norm": 1.0755497217178345,
+      "learning_rate": 2.1904761904761906e-05,
+      "loss": 0.5686,
+      "step": 14055
+    },
+    {
+      "epoch": 17.99168,
+      "grad_norm": 1.0548936128616333,
+      "learning_rate": 2.1902761104441778e-05,
+      "loss": 0.5476,
+      "step": 14056
+    },
+    {
+      "epoch": 17.99296,
+      "grad_norm": 1.0421359539031982,
+      "learning_rate": 2.190076030412165e-05,
+      "loss": 0.5111,
+      "step": 14057
+    },
+    {
+      "epoch": 17.99424,
+      "grad_norm": 1.0015569925308228,
+      "learning_rate": 2.189875950380152e-05,
+      "loss": 0.5427,
+      "step": 14058
+    },
+    {
+      "epoch": 17.99552,
+      "grad_norm": 1.0161628723144531,
+      "learning_rate": 2.1896758703481393e-05,
+      "loss": 0.5307,
+      "step": 14059
+    },
+    {
+      "epoch": 17.9968,
+      "grad_norm": 1.0471055507659912,
+      "learning_rate": 2.1894757903161265e-05,
+      "loss": 0.5164,
+      "step": 14060
+    },
+    {
+      "epoch": 17.99808,
+      "grad_norm": 1.1203445196151733,
+      "learning_rate": 2.1892757102841137e-05,
+      "loss": 0.557,
+      "step": 14061
+    },
+    {
+      "epoch": 17.99936,
+      "grad_norm": 1.0305187702178955,
+      "learning_rate": 2.189075630252101e-05,
+      "loss": 0.499,
+      "step": 14062
+    },
+    {
+      "epoch": 18.00064,
+      "grad_norm": 2.1932191848754883,
+      "learning_rate": 2.188875550220088e-05,
+      "loss": 0.9041,
+      "step": 14063
+    },
+    {
+      "epoch": 18.00192,
+      "grad_norm": 1.0409256219863892,
+      "learning_rate": 2.1886754701880756e-05,
+      "loss": 0.5222,
+      "step": 14064
+    },
+    {
+      "epoch": 18.0032,
+      "grad_norm": 1.000382900238037,
+      "learning_rate": 2.1884753901560624e-05,
+      "loss": 0.5282,
+      "step": 14065
+    },
+    {
+      "epoch": 18.00448,
+      "grad_norm": 0.9593517780303955,
+      "learning_rate": 2.1882753101240496e-05,
+      "loss": 0.4422,
+      "step": 14066
+    },
+    {
+      "epoch": 18.00576,
+      "grad_norm": 1.0473594665527344,
+      "learning_rate": 2.1880752300920368e-05,
+      "loss": 0.5121,
+      "step": 14067
+    },
+    {
+      "epoch": 18.00704,
+      "grad_norm": 1.0217726230621338,
+      "learning_rate": 2.1878751500600243e-05,
+      "loss": 0.4877,
+      "step": 14068
+    },
+    {
+      "epoch": 18.00832,
+      "grad_norm": 0.9588239789009094,
+      "learning_rate": 2.1876750700280112e-05,
+      "loss": 0.4513,
+      "step": 14069
+    },
+    {
+      "epoch": 18.0096,
+      "grad_norm": 0.9952200651168823,
+      "learning_rate": 2.1874749899959984e-05,
+      "loss": 0.4666,
+      "step": 14070
+    },
+    {
+      "epoch": 18.01088,
+      "grad_norm": 1.0490752458572388,
+      "learning_rate": 2.187274909963986e-05,
+      "loss": 0.5474,
+      "step": 14071
+    },
+    {
+      "epoch": 18.01216,
+      "grad_norm": 1.0104339122772217,
+      "learning_rate": 2.187074829931973e-05,
+      "loss": 0.5175,
+      "step": 14072
+    },
+    {
+      "epoch": 18.01344,
+      "grad_norm": 1.013339877128601,
+      "learning_rate": 2.18687474989996e-05,
+      "loss": 0.5109,
+      "step": 14073
+    },
+    {
+      "epoch": 18.01472,
+      "grad_norm": 1.0241261720657349,
+      "learning_rate": 2.186674669867947e-05,
+      "loss": 0.5022,
+      "step": 14074
+    },
+    {
+      "epoch": 18.016,
+      "grad_norm": 1.0270612239837646,
+      "learning_rate": 2.1864745898359346e-05,
+      "loss": 0.4703,
+      "step": 14075
+    },
+    {
+      "epoch": 18.01728,
+      "grad_norm": 1.0369747877120972,
+      "learning_rate": 2.1862745098039218e-05,
+      "loss": 0.5075,
+      "step": 14076
+    },
+    {
+      "epoch": 18.01856,
+      "grad_norm": 1.053289771080017,
+      "learning_rate": 2.1860744297719087e-05,
+      "loss": 0.5392,
+      "step": 14077
+    },
+    {
+      "epoch": 18.01984,
+      "grad_norm": 1.0217084884643555,
+      "learning_rate": 2.1858743497398962e-05,
+      "loss": 0.4777,
+      "step": 14078
+    },
+    {
+      "epoch": 18.02112,
+      "grad_norm": 0.9881057143211365,
+      "learning_rate": 2.1856742697078834e-05,
+      "loss": 0.4531,
+      "step": 14079
+    },
+    {
+      "epoch": 18.0224,
+      "grad_norm": 1.0281909704208374,
+      "learning_rate": 2.1854741896758706e-05,
+      "loss": 0.528,
+      "step": 14080
+    },
+    {
+      "epoch": 18.02368,
+      "grad_norm": 0.9666361808776855,
+      "learning_rate": 2.1852741096438574e-05,
+      "loss": 0.4326,
+      "step": 14081
+    },
+    {
+      "epoch": 18.02496,
+      "grad_norm": 1.0549180507659912,
+      "learning_rate": 2.185074029611845e-05,
+      "loss": 0.5504,
+      "step": 14082
+    },
+    {
+      "epoch": 18.02624,
+      "grad_norm": 0.9973421692848206,
+      "learning_rate": 2.184873949579832e-05,
+      "loss": 0.4681,
+      "step": 14083
+    },
+    {
+      "epoch": 18.02752,
+      "grad_norm": 1.0544997453689575,
+      "learning_rate": 2.1846738695478193e-05,
+      "loss": 0.4742,
+      "step": 14084
+    },
+    {
+      "epoch": 18.0288,
+      "grad_norm": 1.0692542791366577,
+      "learning_rate": 2.1844737895158065e-05,
+      "loss": 0.5059,
+      "step": 14085
+    },
+    {
+      "epoch": 18.03008,
+      "grad_norm": 1.024929165840149,
+      "learning_rate": 2.1842737094837937e-05,
+      "loss": 0.4643,
+      "step": 14086
+    },
+    {
+      "epoch": 18.03136,
+      "grad_norm": 1.0285801887512207,
+      "learning_rate": 2.184073629451781e-05,
+      "loss": 0.5188,
+      "step": 14087
+    },
+    {
+      "epoch": 18.03264,
+      "grad_norm": 1.0161205530166626,
+      "learning_rate": 2.183873549419768e-05,
+      "loss": 0.485,
+      "step": 14088
+    },
+    {
+      "epoch": 18.03392,
+      "grad_norm": 1.0766185522079468,
+      "learning_rate": 2.1836734693877552e-05,
+      "loss": 0.5295,
+      "step": 14089
+    },
+    {
+      "epoch": 18.0352,
+      "grad_norm": 1.038028359413147,
+      "learning_rate": 2.1834733893557424e-05,
+      "loss": 0.4863,
+      "step": 14090
+    },
+    {
+      "epoch": 18.03648,
+      "grad_norm": 1.0519113540649414,
+      "learning_rate": 2.1832733093237296e-05,
+      "loss": 0.5356,
+      "step": 14091
+    },
+    {
+      "epoch": 18.03776,
+      "grad_norm": 0.9999911189079285,
+      "learning_rate": 2.1830732292917168e-05,
+      "loss": 0.455,
+      "step": 14092
+    },
+    {
+      "epoch": 18.03904,
+      "grad_norm": 1.067815899848938,
+      "learning_rate": 2.182873149259704e-05,
+      "loss": 0.4824,
+      "step": 14093
+    },
+    {
+      "epoch": 18.04032,
+      "grad_norm": 1.0944994688034058,
+      "learning_rate": 2.1826730692276912e-05,
+      "loss": 0.468,
+      "step": 14094
+    },
+    {
+      "epoch": 18.0416,
+      "grad_norm": 1.105892300605774,
+      "learning_rate": 2.1824729891956784e-05,
+      "loss": 0.5392,
+      "step": 14095
+    },
+    {
+      "epoch": 18.04288,
+      "grad_norm": 1.038599967956543,
+      "learning_rate": 2.1822729091636655e-05,
+      "loss": 0.5009,
+      "step": 14096
+    },
+    {
+      "epoch": 18.04416,
+      "grad_norm": 1.0022287368774414,
+      "learning_rate": 2.1820728291316527e-05,
+      "loss": 0.473,
+      "step": 14097
+    },
+    {
+      "epoch": 18.04544,
+      "grad_norm": 1.0601282119750977,
+      "learning_rate": 2.18187274909964e-05,
+      "loss": 0.5105,
+      "step": 14098
+    },
+    {
+      "epoch": 18.04672,
+      "grad_norm": 1.0425713062286377,
+      "learning_rate": 2.181672669067627e-05,
+      "loss": 0.448,
+      "step": 14099
+    },
+    {
+      "epoch": 18.048,
+      "grad_norm": 1.0768316984176636,
+      "learning_rate": 2.1814725890356143e-05,
+      "loss": 0.5035,
+      "step": 14100
+    },
+    {
+      "epoch": 18.04928,
+      "grad_norm": 1.0370937585830688,
+      "learning_rate": 2.1812725090036015e-05,
+      "loss": 0.4883,
+      "step": 14101
+    },
+    {
+      "epoch": 18.05056,
+      "grad_norm": 1.0263327360153198,
+      "learning_rate": 2.1810724289715887e-05,
+      "loss": 0.5022,
+      "step": 14102
+    },
+    {
+      "epoch": 18.05184,
+      "grad_norm": 0.9880037307739258,
+      "learning_rate": 2.1808723489395762e-05,
+      "loss": 0.4505,
+      "step": 14103
+    },
+    {
+      "epoch": 18.05312,
+      "grad_norm": 1.0344130992889404,
+      "learning_rate": 2.180672268907563e-05,
+      "loss": 0.5191,
+      "step": 14104
+    },
+    {
+      "epoch": 18.0544,
+      "grad_norm": 1.0637491941452026,
+      "learning_rate": 2.1804721888755502e-05,
+      "loss": 0.4676,
+      "step": 14105
+    },
+    {
+      "epoch": 18.05568,
+      "grad_norm": 1.0038073062896729,
+      "learning_rate": 2.1802721088435374e-05,
+      "loss": 0.4869,
+      "step": 14106
+    },
+    {
+      "epoch": 18.05696,
+      "grad_norm": 1.0341646671295166,
+      "learning_rate": 2.180072028811525e-05,
+      "loss": 0.4734,
+      "step": 14107
+    },
+    {
+      "epoch": 18.05824,
+      "grad_norm": 1.0689383745193481,
+      "learning_rate": 2.1798719487795118e-05,
+      "loss": 0.4993,
+      "step": 14108
+    },
+    {
+      "epoch": 18.05952,
+      "grad_norm": 1.0641204118728638,
+      "learning_rate": 2.179671868747499e-05,
+      "loss": 0.5123,
+      "step": 14109
+    },
+    {
+      "epoch": 18.0608,
+      "grad_norm": 1.047566533088684,
+      "learning_rate": 2.1794717887154865e-05,
+      "loss": 0.5061,
+      "step": 14110
+    },
+    {
+      "epoch": 18.06208,
+      "grad_norm": 1.0506558418273926,
+      "learning_rate": 2.1792717086834737e-05,
+      "loss": 0.4856,
+      "step": 14111
+    },
+    {
+      "epoch": 18.06336,
+      "grad_norm": 0.985125720500946,
+      "learning_rate": 2.1790716286514605e-05,
+      "loss": 0.4718,
+      "step": 14112
+    },
+    {
+      "epoch": 18.06464,
+      "grad_norm": 1.012773036956787,
+      "learning_rate": 2.1788715486194477e-05,
+      "loss": 0.4976,
+      "step": 14113
+    },
+    {
+      "epoch": 18.06592,
+      "grad_norm": 0.9665741920471191,
+      "learning_rate": 2.1786714685874352e-05,
+      "loss": 0.4588,
+      "step": 14114
+    },
+    {
+      "epoch": 18.0672,
+      "grad_norm": 1.0558841228485107,
+      "learning_rate": 2.1784713885554224e-05,
+      "loss": 0.5345,
+      "step": 14115
+    },
+    {
+      "epoch": 18.06848,
+      "grad_norm": 1.1132677793502808,
+      "learning_rate": 2.1782713085234093e-05,
+      "loss": 0.5223,
+      "step": 14116
+    },
+    {
+      "epoch": 18.06976,
+      "grad_norm": 0.9984009265899658,
+      "learning_rate": 2.1780712284913968e-05,
+      "loss": 0.4445,
+      "step": 14117
+    },
+    {
+      "epoch": 18.07104,
+      "grad_norm": 1.0193313360214233,
+      "learning_rate": 2.177871148459384e-05,
+      "loss": 0.4674,
+      "step": 14118
+    },
+    {
+      "epoch": 18.07232,
+      "grad_norm": 1.009688138961792,
+      "learning_rate": 2.177671068427371e-05,
+      "loss": 0.449,
+      "step": 14119
+    },
+    {
+      "epoch": 18.0736,
+      "grad_norm": 0.9733455777168274,
+      "learning_rate": 2.177470988395358e-05,
+      "loss": 0.4378,
+      "step": 14120
+    },
+    {
+      "epoch": 18.07488,
+      "grad_norm": 1.0891609191894531,
+      "learning_rate": 2.1772709083633455e-05,
+      "loss": 0.4709,
+      "step": 14121
+    },
+    {
+      "epoch": 18.07616,
+      "grad_norm": 1.1113479137420654,
+      "learning_rate": 2.1770708283313327e-05,
+      "loss": 0.5089,
+      "step": 14122
+    },
+    {
+      "epoch": 18.07744,
+      "grad_norm": 1.1308077573776245,
+      "learning_rate": 2.17687074829932e-05,
+      "loss": 0.5187,
+      "step": 14123
+    },
+    {
+      "epoch": 18.07872,
+      "grad_norm": 1.1170744895935059,
+      "learning_rate": 2.176670668267307e-05,
+      "loss": 0.528,
+      "step": 14124
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 1.0902800559997559,
+      "learning_rate": 2.1764705882352943e-05,
+      "loss": 0.5186,
+      "step": 14125
+    },
+    {
+      "epoch": 18.08128,
+      "grad_norm": 1.0982624292373657,
+      "learning_rate": 2.1762705082032815e-05,
+      "loss": 0.5318,
+      "step": 14126
+    },
+    {
+      "epoch": 18.08256,
+      "grad_norm": 1.066246747970581,
+      "learning_rate": 2.1760704281712686e-05,
+      "loss": 0.4535,
+      "step": 14127
+    },
+    {
+      "epoch": 18.08384,
+      "grad_norm": 1.0559955835342407,
+      "learning_rate": 2.1758703481392558e-05,
+      "loss": 0.5188,
+      "step": 14128
+    },
+    {
+      "epoch": 18.08512,
+      "grad_norm": 1.0126625299453735,
+      "learning_rate": 2.175670268107243e-05,
+      "loss": 0.4805,
+      "step": 14129
+    },
+    {
+      "epoch": 18.0864,
+      "grad_norm": 1.0629305839538574,
+      "learning_rate": 2.1754701880752302e-05,
+      "loss": 0.492,
+      "step": 14130
+    },
+    {
+      "epoch": 18.08768,
+      "grad_norm": 1.0466065406799316,
+      "learning_rate": 2.1752701080432174e-05,
+      "loss": 0.5372,
+      "step": 14131
+    },
+    {
+      "epoch": 18.08896,
+      "grad_norm": 1.072338581085205,
+      "learning_rate": 2.1750700280112046e-05,
+      "loss": 0.5614,
+      "step": 14132
+    },
+    {
+      "epoch": 18.09024,
+      "grad_norm": 1.0793755054473877,
+      "learning_rate": 2.1748699479791918e-05,
+      "loss": 0.5204,
+      "step": 14133
+    },
+    {
+      "epoch": 18.09152,
+      "grad_norm": 1.063680648803711,
+      "learning_rate": 2.174669867947179e-05,
+      "loss": 0.4683,
+      "step": 14134
+    },
+    {
+      "epoch": 18.0928,
+      "grad_norm": 1.070969820022583,
+      "learning_rate": 2.174469787915166e-05,
+      "loss": 0.515,
+      "step": 14135
+    },
+    {
+      "epoch": 18.09408,
+      "grad_norm": 1.0573210716247559,
+      "learning_rate": 2.1742697078831533e-05,
+      "loss": 0.5124,
+      "step": 14136
+    },
+    {
+      "epoch": 18.09536,
+      "grad_norm": 0.9679538607597351,
+      "learning_rate": 2.1740696278511405e-05,
+      "loss": 0.4349,
+      "step": 14137
+    },
+    {
+      "epoch": 18.09664,
+      "grad_norm": 1.0116899013519287,
+      "learning_rate": 2.173869547819128e-05,
+      "loss": 0.5204,
+      "step": 14138
+    },
+    {
+      "epoch": 18.09792,
+      "grad_norm": 1.0702815055847168,
+      "learning_rate": 2.173669467787115e-05,
+      "loss": 0.5164,
+      "step": 14139
+    },
+    {
+      "epoch": 18.0992,
+      "grad_norm": 1.0475645065307617,
+      "learning_rate": 2.173469387755102e-05,
+      "loss": 0.5105,
+      "step": 14140
+    },
+    {
+      "epoch": 18.10048,
+      "grad_norm": 1.0068633556365967,
+      "learning_rate": 2.1732693077230892e-05,
+      "loss": 0.4967,
+      "step": 14141
+    },
+    {
+      "epoch": 18.10176,
+      "grad_norm": 1.0416619777679443,
+      "learning_rate": 2.1730692276910768e-05,
+      "loss": 0.4917,
+      "step": 14142
+    },
+    {
+      "epoch": 18.10304,
+      "grad_norm": 0.996741533279419,
+      "learning_rate": 2.1728691476590636e-05,
+      "loss": 0.4628,
+      "step": 14143
+    },
+    {
+      "epoch": 18.10432,
+      "grad_norm": 1.0638675689697266,
+      "learning_rate": 2.1726690676270508e-05,
+      "loss": 0.5081,
+      "step": 14144
+    },
+    {
+      "epoch": 18.1056,
+      "grad_norm": 1.0904065370559692,
+      "learning_rate": 2.1724689875950383e-05,
+      "loss": 0.4946,
+      "step": 14145
+    },
+    {
+      "epoch": 18.10688,
+      "grad_norm": 1.032350778579712,
+      "learning_rate": 2.1722689075630255e-05,
+      "loss": 0.4844,
+      "step": 14146
+    },
+    {
+      "epoch": 18.10816,
+      "grad_norm": 1.0421230792999268,
+      "learning_rate": 2.1720688275310124e-05,
+      "loss": 0.4894,
+      "step": 14147
+    },
+    {
+      "epoch": 18.10944,
+      "grad_norm": 1.130836009979248,
+      "learning_rate": 2.1718687474989995e-05,
+      "loss": 0.5482,
+      "step": 14148
+    },
+    {
+      "epoch": 18.11072,
+      "grad_norm": 1.1370742321014404,
+      "learning_rate": 2.171668667466987e-05,
+      "loss": 0.5494,
+      "step": 14149
+    },
+    {
+      "epoch": 18.112,
+      "grad_norm": 1.0708554983139038,
+      "learning_rate": 2.1714685874349742e-05,
+      "loss": 0.5181,
+      "step": 14150
+    },
+    {
+      "epoch": 18.11328,
+      "grad_norm": 1.0703370571136475,
+      "learning_rate": 2.171268507402961e-05,
+      "loss": 0.5118,
+      "step": 14151
+    },
+    {
+      "epoch": 18.11456,
+      "grad_norm": 1.0429341793060303,
+      "learning_rate": 2.1710684273709486e-05,
+      "loss": 0.4991,
+      "step": 14152
+    },
+    {
+      "epoch": 18.11584,
+      "grad_norm": 1.0283002853393555,
+      "learning_rate": 2.1708683473389358e-05,
+      "loss": 0.481,
+      "step": 14153
+    },
+    {
+      "epoch": 18.11712,
+      "grad_norm": 1.0293543338775635,
+      "learning_rate": 2.170668267306923e-05,
+      "loss": 0.5489,
+      "step": 14154
+    },
+    {
+      "epoch": 18.1184,
+      "grad_norm": 1.0159963369369507,
+      "learning_rate": 2.17046818727491e-05,
+      "loss": 0.4763,
+      "step": 14155
+    },
+    {
+      "epoch": 18.11968,
+      "grad_norm": 1.0897349119186401,
+      "learning_rate": 2.1702681072428974e-05,
+      "loss": 0.5192,
+      "step": 14156
+    },
+    {
+      "epoch": 18.12096,
+      "grad_norm": 1.015059471130371,
+      "learning_rate": 2.1700680272108845e-05,
+      "loss": 0.4625,
+      "step": 14157
+    },
+    {
+      "epoch": 18.12224,
+      "grad_norm": 1.0573114156723022,
+      "learning_rate": 2.1698679471788717e-05,
+      "loss": 0.5278,
+      "step": 14158
+    },
+    {
+      "epoch": 18.12352,
+      "grad_norm": 1.0435305833816528,
+      "learning_rate": 2.1696678671468586e-05,
+      "loss": 0.4909,
+      "step": 14159
+    },
+    {
+      "epoch": 18.1248,
+      "grad_norm": 1.0256880521774292,
+      "learning_rate": 2.169467787114846e-05,
+      "loss": 0.4836,
+      "step": 14160
+    },
+    {
+      "epoch": 18.12608,
+      "grad_norm": 1.0937752723693848,
+      "learning_rate": 2.1692677070828333e-05,
+      "loss": 0.5833,
+      "step": 14161
+    },
+    {
+      "epoch": 18.12736,
+      "grad_norm": 1.005741834640503,
+      "learning_rate": 2.1690676270508205e-05,
+      "loss": 0.4886,
+      "step": 14162
+    },
+    {
+      "epoch": 18.12864,
+      "grad_norm": 1.0671494007110596,
+      "learning_rate": 2.1688675470188077e-05,
+      "loss": 0.4829,
+      "step": 14163
+    },
+    {
+      "epoch": 18.12992,
+      "grad_norm": 1.0546166896820068,
+      "learning_rate": 2.168667466986795e-05,
+      "loss": 0.5001,
+      "step": 14164
+    },
+    {
+      "epoch": 18.1312,
+      "grad_norm": 1.0181262493133545,
+      "learning_rate": 2.168467386954782e-05,
+      "loss": 0.4707,
+      "step": 14165
+    },
+    {
+      "epoch": 18.13248,
+      "grad_norm": 1.008811116218567,
+      "learning_rate": 2.1682673069227692e-05,
+      "loss": 0.4869,
+      "step": 14166
+    },
+    {
+      "epoch": 18.13376,
+      "grad_norm": 1.00252366065979,
+      "learning_rate": 2.1680672268907564e-05,
+      "loss": 0.4692,
+      "step": 14167
+    },
+    {
+      "epoch": 18.13504,
+      "grad_norm": 1.0217723846435547,
+      "learning_rate": 2.1678671468587436e-05,
+      "loss": 0.5281,
+      "step": 14168
+    },
+    {
+      "epoch": 18.13632,
+      "grad_norm": 1.0290299654006958,
+      "learning_rate": 2.1676670668267308e-05,
+      "loss": 0.4796,
+      "step": 14169
+    },
+    {
+      "epoch": 18.1376,
+      "grad_norm": 1.0708261728286743,
+      "learning_rate": 2.167466986794718e-05,
+      "loss": 0.4746,
+      "step": 14170
+    },
+    {
+      "epoch": 18.13888,
+      "grad_norm": 1.0332971811294556,
+      "learning_rate": 2.167266906762705e-05,
+      "loss": 0.485,
+      "step": 14171
+    },
+    {
+      "epoch": 18.14016,
+      "grad_norm": 1.0188629627227783,
+      "learning_rate": 2.1670668267306923e-05,
+      "loss": 0.4689,
+      "step": 14172
+    },
+    {
+      "epoch": 18.14144,
+      "grad_norm": 1.0185388326644897,
+      "learning_rate": 2.1668667466986795e-05,
+      "loss": 0.5077,
+      "step": 14173
+    },
+    {
+      "epoch": 18.14272,
+      "grad_norm": 1.0127745866775513,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 0.4879,
+      "step": 14174
+    },
+    {
+      "epoch": 18.144,
+      "grad_norm": 1.0742979049682617,
+      "learning_rate": 2.166466586634654e-05,
+      "loss": 0.5216,
+      "step": 14175
+    },
+    {
+      "epoch": 18.14528,
+      "grad_norm": 1.0509237051010132,
+      "learning_rate": 2.166266506602641e-05,
+      "loss": 0.4848,
+      "step": 14176
+    },
+    {
+      "epoch": 18.14656,
+      "grad_norm": 1.0471779108047485,
+      "learning_rate": 2.1660664265706286e-05,
+      "loss": 0.5013,
+      "step": 14177
+    },
+    {
+      "epoch": 18.14784,
+      "grad_norm": 1.0413541793823242,
+      "learning_rate": 2.1658663465386154e-05,
+      "loss": 0.4817,
+      "step": 14178
+    },
+    {
+      "epoch": 18.14912,
+      "grad_norm": 1.007503867149353,
+      "learning_rate": 2.1656662665066026e-05,
+      "loss": 0.4622,
+      "step": 14179
+    },
+    {
+      "epoch": 18.1504,
+      "grad_norm": 1.0402361154556274,
+      "learning_rate": 2.1654661864745898e-05,
+      "loss": 0.5028,
+      "step": 14180
+    },
+    {
+      "epoch": 18.15168,
+      "grad_norm": 1.0889134407043457,
+      "learning_rate": 2.1652661064425773e-05,
+      "loss": 0.5201,
+      "step": 14181
+    },
+    {
+      "epoch": 18.15296,
+      "grad_norm": 1.0945775508880615,
+      "learning_rate": 2.1650660264105642e-05,
+      "loss": 0.4916,
+      "step": 14182
+    },
+    {
+      "epoch": 18.15424,
+      "grad_norm": 1.025062918663025,
+      "learning_rate": 2.1648659463785514e-05,
+      "loss": 0.5115,
+      "step": 14183
+    },
+    {
+      "epoch": 18.15552,
+      "grad_norm": 1.0531753301620483,
+      "learning_rate": 2.164665866346539e-05,
+      "loss": 0.4737,
+      "step": 14184
+    },
+    {
+      "epoch": 18.1568,
+      "grad_norm": 1.0863885879516602,
+      "learning_rate": 2.164465786314526e-05,
+      "loss": 0.5113,
+      "step": 14185
+    },
+    {
+      "epoch": 18.158079999999998,
+      "grad_norm": 1.0436323881149292,
+      "learning_rate": 2.164265706282513e-05,
+      "loss": 0.5154,
+      "step": 14186
+    },
+    {
+      "epoch": 18.15936,
+      "grad_norm": 1.0325742959976196,
+      "learning_rate": 2.1640656262505e-05,
+      "loss": 0.4873,
+      "step": 14187
+    },
+    {
+      "epoch": 18.16064,
+      "grad_norm": 1.0843673944473267,
+      "learning_rate": 2.1638655462184876e-05,
+      "loss": 0.4948,
+      "step": 14188
+    },
+    {
+      "epoch": 18.16192,
+      "grad_norm": 1.0292965173721313,
+      "learning_rate": 2.1636654661864748e-05,
+      "loss": 0.5001,
+      "step": 14189
+    },
+    {
+      "epoch": 18.1632,
+      "grad_norm": 1.010935664176941,
+      "learning_rate": 2.1634653861544617e-05,
+      "loss": 0.4492,
+      "step": 14190
+    },
+    {
+      "epoch": 18.16448,
+      "grad_norm": 1.0702733993530273,
+      "learning_rate": 2.1632653061224492e-05,
+      "loss": 0.486,
+      "step": 14191
+    },
+    {
+      "epoch": 18.16576,
+      "grad_norm": 1.0180492401123047,
+      "learning_rate": 2.1630652260904364e-05,
+      "loss": 0.4948,
+      "step": 14192
+    },
+    {
+      "epoch": 18.16704,
+      "grad_norm": 1.0777477025985718,
+      "learning_rate": 2.1628651460584236e-05,
+      "loss": 0.5253,
+      "step": 14193
+    },
+    {
+      "epoch": 18.16832,
+      "grad_norm": 1.03486168384552,
+      "learning_rate": 2.1626650660264104e-05,
+      "loss": 0.4609,
+      "step": 14194
+    },
+    {
+      "epoch": 18.1696,
+      "grad_norm": 1.0693778991699219,
+      "learning_rate": 2.162464985994398e-05,
+      "loss": 0.5002,
+      "step": 14195
+    },
+    {
+      "epoch": 18.17088,
+      "grad_norm": 1.0229696035385132,
+      "learning_rate": 2.162264905962385e-05,
+      "loss": 0.4825,
+      "step": 14196
+    },
+    {
+      "epoch": 18.17216,
+      "grad_norm": 1.0002144575119019,
+      "learning_rate": 2.1620648259303723e-05,
+      "loss": 0.4239,
+      "step": 14197
+    },
+    {
+      "epoch": 18.17344,
+      "grad_norm": 1.0359293222427368,
+      "learning_rate": 2.1618647458983595e-05,
+      "loss": 0.4582,
+      "step": 14198
+    },
+    {
+      "epoch": 18.17472,
+      "grad_norm": 1.0600343942642212,
+      "learning_rate": 2.1616646658663467e-05,
+      "loss": 0.5052,
+      "step": 14199
+    },
+    {
+      "epoch": 18.176,
+      "grad_norm": 1.0161858797073364,
+      "learning_rate": 2.161464585834334e-05,
+      "loss": 0.4748,
+      "step": 14200
+    },
+    {
+      "epoch": 18.17728,
+      "grad_norm": 1.0281199216842651,
+      "learning_rate": 2.161264505802321e-05,
+      "loss": 0.4505,
+      "step": 14201
+    },
+    {
+      "epoch": 18.17856,
+      "grad_norm": 1.0303138494491577,
+      "learning_rate": 2.1610644257703082e-05,
+      "loss": 0.5022,
+      "step": 14202
+    },
+    {
+      "epoch": 18.17984,
+      "grad_norm": 1.0298250913619995,
+      "learning_rate": 2.1608643457382954e-05,
+      "loss": 0.4955,
+      "step": 14203
+    },
+    {
+      "epoch": 18.18112,
+      "grad_norm": 1.0383046865463257,
+      "learning_rate": 2.1606642657062826e-05,
+      "loss": 0.503,
+      "step": 14204
+    },
+    {
+      "epoch": 18.1824,
+      "grad_norm": 1.0609245300292969,
+      "learning_rate": 2.1604641856742698e-05,
+      "loss": 0.4975,
+      "step": 14205
+    },
+    {
+      "epoch": 18.18368,
+      "grad_norm": 1.0384618043899536,
+      "learning_rate": 2.160264105642257e-05,
+      "loss": 0.5371,
+      "step": 14206
+    },
+    {
+      "epoch": 18.18496,
+      "grad_norm": 1.044183373451233,
+      "learning_rate": 2.160064025610244e-05,
+      "loss": 0.452,
+      "step": 14207
+    },
+    {
+      "epoch": 18.18624,
+      "grad_norm": 1.0944687128067017,
+      "learning_rate": 2.1598639455782314e-05,
+      "loss": 0.5241,
+      "step": 14208
+    },
+    {
+      "epoch": 18.18752,
+      "grad_norm": 1.0087506771087646,
+      "learning_rate": 2.1596638655462185e-05,
+      "loss": 0.4625,
+      "step": 14209
+    },
+    {
+      "epoch": 18.1888,
+      "grad_norm": 1.0070358514785767,
+      "learning_rate": 2.1594637855142057e-05,
+      "loss": 0.4719,
+      "step": 14210
+    },
+    {
+      "epoch": 18.19008,
+      "grad_norm": 1.0301635265350342,
+      "learning_rate": 2.159263705482193e-05,
+      "loss": 0.448,
+      "step": 14211
+    },
+    {
+      "epoch": 18.19136,
+      "grad_norm": 1.077388048171997,
+      "learning_rate": 2.15906362545018e-05,
+      "loss": 0.5245,
+      "step": 14212
+    },
+    {
+      "epoch": 18.19264,
+      "grad_norm": 1.0530251264572144,
+      "learning_rate": 2.1588635454181673e-05,
+      "loss": 0.5032,
+      "step": 14213
+    },
+    {
+      "epoch": 18.19392,
+      "grad_norm": 1.058570384979248,
+      "learning_rate": 2.1586634653861545e-05,
+      "loss": 0.5114,
+      "step": 14214
+    },
+    {
+      "epoch": 18.1952,
+      "grad_norm": 1.0303319692611694,
+      "learning_rate": 2.1584633853541417e-05,
+      "loss": 0.4486,
+      "step": 14215
+    },
+    {
+      "epoch": 18.19648,
+      "grad_norm": 1.0652695894241333,
+      "learning_rate": 2.1582633053221292e-05,
+      "loss": 0.5292,
+      "step": 14216
+    },
+    {
+      "epoch": 18.19776,
+      "grad_norm": 1.0437886714935303,
+      "learning_rate": 2.158063225290116e-05,
+      "loss": 0.5014,
+      "step": 14217
+    },
+    {
+      "epoch": 18.19904,
+      "grad_norm": 1.0610305070877075,
+      "learning_rate": 2.1578631452581032e-05,
+      "loss": 0.5458,
+      "step": 14218
+    },
+    {
+      "epoch": 18.20032,
+      "grad_norm": 0.995061457157135,
+      "learning_rate": 2.1576630652260904e-05,
+      "loss": 0.4919,
+      "step": 14219
+    },
+    {
+      "epoch": 18.2016,
+      "grad_norm": 1.0778111219406128,
+      "learning_rate": 2.157462985194078e-05,
+      "loss": 0.5306,
+      "step": 14220
+    },
+    {
+      "epoch": 18.20288,
+      "grad_norm": 1.0350569486618042,
+      "learning_rate": 2.1572629051620648e-05,
+      "loss": 0.5296,
+      "step": 14221
+    },
+    {
+      "epoch": 18.20416,
+      "grad_norm": 1.0401970148086548,
+      "learning_rate": 2.157062825130052e-05,
+      "loss": 0.5086,
+      "step": 14222
+    },
+    {
+      "epoch": 18.20544,
+      "grad_norm": 1.0326118469238281,
+      "learning_rate": 2.1568627450980395e-05,
+      "loss": 0.5022,
+      "step": 14223
+    },
+    {
+      "epoch": 18.20672,
+      "grad_norm": 1.0625585317611694,
+      "learning_rate": 2.1566626650660267e-05,
+      "loss": 0.5218,
+      "step": 14224
+    },
+    {
+      "epoch": 18.208,
+      "grad_norm": 1.0629734992980957,
+      "learning_rate": 2.1564625850340135e-05,
+      "loss": 0.5225,
+      "step": 14225
+    },
+    {
+      "epoch": 18.20928,
+      "grad_norm": 1.0436673164367676,
+      "learning_rate": 2.1562625050020007e-05,
+      "loss": 0.4921,
+      "step": 14226
+    },
+    {
+      "epoch": 18.21056,
+      "grad_norm": 1.0012400150299072,
+      "learning_rate": 2.1560624249699882e-05,
+      "loss": 0.4427,
+      "step": 14227
+    },
+    {
+      "epoch": 18.21184,
+      "grad_norm": 1.1108235120773315,
+      "learning_rate": 2.1558623449379754e-05,
+      "loss": 0.5484,
+      "step": 14228
+    },
+    {
+      "epoch": 18.21312,
+      "grad_norm": 1.080520510673523,
+      "learning_rate": 2.1556622649059623e-05,
+      "loss": 0.5021,
+      "step": 14229
+    },
+    {
+      "epoch": 18.2144,
+      "grad_norm": 1.0559675693511963,
+      "learning_rate": 2.1554621848739498e-05,
+      "loss": 0.4703,
+      "step": 14230
+    },
+    {
+      "epoch": 18.21568,
+      "grad_norm": 1.0659006834030151,
+      "learning_rate": 2.155262104841937e-05,
+      "loss": 0.5139,
+      "step": 14231
+    },
+    {
+      "epoch": 18.21696,
+      "grad_norm": 1.0316319465637207,
+      "learning_rate": 2.155062024809924e-05,
+      "loss": 0.4966,
+      "step": 14232
+    },
+    {
+      "epoch": 18.21824,
+      "grad_norm": 1.0869457721710205,
+      "learning_rate": 2.154861944777911e-05,
+      "loss": 0.5206,
+      "step": 14233
+    },
+    {
+      "epoch": 18.21952,
+      "grad_norm": 0.9864334464073181,
+      "learning_rate": 2.1546618647458985e-05,
+      "loss": 0.4524,
+      "step": 14234
+    },
+    {
+      "epoch": 18.2208,
+      "grad_norm": 1.0641202926635742,
+      "learning_rate": 2.1544617847138857e-05,
+      "loss": 0.5093,
+      "step": 14235
+    },
+    {
+      "epoch": 18.22208,
+      "grad_norm": 0.9992152452468872,
+      "learning_rate": 2.154261704681873e-05,
+      "loss": 0.448,
+      "step": 14236
+    },
+    {
+      "epoch": 18.22336,
+      "grad_norm": 1.1175810098648071,
+      "learning_rate": 2.15406162464986e-05,
+      "loss": 0.4918,
+      "step": 14237
+    },
+    {
+      "epoch": 18.22464,
+      "grad_norm": 1.0763076543807983,
+      "learning_rate": 2.1538615446178473e-05,
+      "loss": 0.4894,
+      "step": 14238
+    },
+    {
+      "epoch": 18.22592,
+      "grad_norm": 1.0875471830368042,
+      "learning_rate": 2.1536614645858344e-05,
+      "loss": 0.5458,
+      "step": 14239
+    },
+    {
+      "epoch": 18.2272,
+      "grad_norm": 1.0565011501312256,
+      "learning_rate": 2.1534613845538216e-05,
+      "loss": 0.488,
+      "step": 14240
+    },
+    {
+      "epoch": 18.22848,
+      "grad_norm": 1.0078827142715454,
+      "learning_rate": 2.1532613045218088e-05,
+      "loss": 0.4727,
+      "step": 14241
+    },
+    {
+      "epoch": 18.22976,
+      "grad_norm": 1.0515302419662476,
+      "learning_rate": 2.153061224489796e-05,
+      "loss": 0.4798,
+      "step": 14242
+    },
+    {
+      "epoch": 18.23104,
+      "grad_norm": 1.0603551864624023,
+      "learning_rate": 2.1528611444577832e-05,
+      "loss": 0.5047,
+      "step": 14243
+    },
+    {
+      "epoch": 18.23232,
+      "grad_norm": 1.032323956489563,
+      "learning_rate": 2.1526610644257704e-05,
+      "loss": 0.4825,
+      "step": 14244
+    },
+    {
+      "epoch": 18.2336,
+      "grad_norm": 1.0973374843597412,
+      "learning_rate": 2.1524609843937576e-05,
+      "loss": 0.4856,
+      "step": 14245
+    },
+    {
+      "epoch": 18.23488,
+      "grad_norm": 1.1034042835235596,
+      "learning_rate": 2.1522609043617447e-05,
+      "loss": 0.5269,
+      "step": 14246
+    },
+    {
+      "epoch": 18.23616,
+      "grad_norm": 1.0152696371078491,
+      "learning_rate": 2.152060824329732e-05,
+      "loss": 0.47,
+      "step": 14247
+    },
+    {
+      "epoch": 18.23744,
+      "grad_norm": 1.0043752193450928,
+      "learning_rate": 2.151860744297719e-05,
+      "loss": 0.5078,
+      "step": 14248
+    },
+    {
+      "epoch": 18.23872,
+      "grad_norm": 1.1099114418029785,
+      "learning_rate": 2.1516606642657063e-05,
+      "loss": 0.528,
+      "step": 14249
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 1.0465238094329834,
+      "learning_rate": 2.1514605842336935e-05,
+      "loss": 0.5168,
+      "step": 14250
+    },
+    {
+      "epoch": 18.24128,
+      "grad_norm": 1.0789815187454224,
+      "learning_rate": 2.151260504201681e-05,
+      "loss": 0.5096,
+      "step": 14251
+    },
+    {
+      "epoch": 18.24256,
+      "grad_norm": 1.1060971021652222,
+      "learning_rate": 2.151060424169668e-05,
+      "loss": 0.5233,
+      "step": 14252
+    },
+    {
+      "epoch": 18.24384,
+      "grad_norm": 1.0800647735595703,
+      "learning_rate": 2.150860344137655e-05,
+      "loss": 0.5079,
+      "step": 14253
+    },
+    {
+      "epoch": 18.24512,
+      "grad_norm": 0.9907492995262146,
+      "learning_rate": 2.1506602641056422e-05,
+      "loss": 0.4908,
+      "step": 14254
+    },
+    {
+      "epoch": 18.2464,
+      "grad_norm": 1.0035676956176758,
+      "learning_rate": 2.1504601840736298e-05,
+      "loss": 0.4876,
+      "step": 14255
+    },
+    {
+      "epoch": 18.24768,
+      "grad_norm": 1.0252283811569214,
+      "learning_rate": 2.1502601040416166e-05,
+      "loss": 0.4663,
+      "step": 14256
+    },
+    {
+      "epoch": 18.24896,
+      "grad_norm": 1.023335337638855,
+      "learning_rate": 2.1500600240096038e-05,
+      "loss": 0.494,
+      "step": 14257
+    },
+    {
+      "epoch": 18.25024,
+      "grad_norm": 1.0962717533111572,
+      "learning_rate": 2.1498599439775913e-05,
+      "loss": 0.5021,
+      "step": 14258
+    },
+    {
+      "epoch": 18.25152,
+      "grad_norm": 1.0230556726455688,
+      "learning_rate": 2.1496598639455785e-05,
+      "loss": 0.5067,
+      "step": 14259
+    },
+    {
+      "epoch": 18.2528,
+      "grad_norm": 0.9827935099601746,
+      "learning_rate": 2.1494597839135653e-05,
+      "loss": 0.4315,
+      "step": 14260
+    },
+    {
+      "epoch": 18.25408,
+      "grad_norm": 1.0110362768173218,
+      "learning_rate": 2.1492597038815525e-05,
+      "loss": 0.4864,
+      "step": 14261
+    },
+    {
+      "epoch": 18.25536,
+      "grad_norm": 1.0713372230529785,
+      "learning_rate": 2.14905962384954e-05,
+      "loss": 0.5581,
+      "step": 14262
+    },
+    {
+      "epoch": 18.25664,
+      "grad_norm": 1.0925161838531494,
+      "learning_rate": 2.1488595438175272e-05,
+      "loss": 0.4843,
+      "step": 14263
+    },
+    {
+      "epoch": 18.25792,
+      "grad_norm": 1.0575942993164062,
+      "learning_rate": 2.148659463785514e-05,
+      "loss": 0.5034,
+      "step": 14264
+    },
+    {
+      "epoch": 18.2592,
+      "grad_norm": 1.080422282218933,
+      "learning_rate": 2.1484593837535016e-05,
+      "loss": 0.5236,
+      "step": 14265
+    },
+    {
+      "epoch": 18.26048,
+      "grad_norm": 1.0370043516159058,
+      "learning_rate": 2.1482593037214888e-05,
+      "loss": 0.4743,
+      "step": 14266
+    },
+    {
+      "epoch": 18.26176,
+      "grad_norm": 1.0181622505187988,
+      "learning_rate": 2.148059223689476e-05,
+      "loss": 0.449,
+      "step": 14267
+    },
+    {
+      "epoch": 18.26304,
+      "grad_norm": 1.0741147994995117,
+      "learning_rate": 2.147859143657463e-05,
+      "loss": 0.544,
+      "step": 14268
+    },
+    {
+      "epoch": 18.26432,
+      "grad_norm": 1.0276820659637451,
+      "learning_rate": 2.1476590636254504e-05,
+      "loss": 0.4783,
+      "step": 14269
+    },
+    {
+      "epoch": 18.2656,
+      "grad_norm": 1.094807744026184,
+      "learning_rate": 2.1474589835934375e-05,
+      "loss": 0.488,
+      "step": 14270
+    },
+    {
+      "epoch": 18.26688,
+      "grad_norm": 1.0752545595169067,
+      "learning_rate": 2.1472589035614247e-05,
+      "loss": 0.4985,
+      "step": 14271
+    },
+    {
+      "epoch": 18.26816,
+      "grad_norm": 1.1067184209823608,
+      "learning_rate": 2.1470588235294116e-05,
+      "loss": 0.4627,
+      "step": 14272
+    },
+    {
+      "epoch": 18.26944,
+      "grad_norm": 1.0492066144943237,
+      "learning_rate": 2.146858743497399e-05,
+      "loss": 0.5197,
+      "step": 14273
+    },
+    {
+      "epoch": 18.27072,
+      "grad_norm": 1.0525888204574585,
+      "learning_rate": 2.1466586634653863e-05,
+      "loss": 0.5238,
+      "step": 14274
+    },
+    {
+      "epoch": 18.272,
+      "grad_norm": 1.0903679132461548,
+      "learning_rate": 2.1464585834333735e-05,
+      "loss": 0.5231,
+      "step": 14275
+    },
+    {
+      "epoch": 18.27328,
+      "grad_norm": 1.1250532865524292,
+      "learning_rate": 2.1462585034013607e-05,
+      "loss": 0.5518,
+      "step": 14276
+    },
+    {
+      "epoch": 18.27456,
+      "grad_norm": 1.0479706525802612,
+      "learning_rate": 2.146058423369348e-05,
+      "loss": 0.4735,
+      "step": 14277
+    },
+    {
+      "epoch": 18.27584,
+      "grad_norm": 1.100510597229004,
+      "learning_rate": 2.145858343337335e-05,
+      "loss": 0.5421,
+      "step": 14278
+    },
+    {
+      "epoch": 18.27712,
+      "grad_norm": 1.0894125699996948,
+      "learning_rate": 2.1456582633053222e-05,
+      "loss": 0.4891,
+      "step": 14279
+    },
+    {
+      "epoch": 18.2784,
+      "grad_norm": 1.0481358766555786,
+      "learning_rate": 2.1454581832733094e-05,
+      "loss": 0.4968,
+      "step": 14280
+    },
+    {
+      "epoch": 18.27968,
+      "grad_norm": 1.0981334447860718,
+      "learning_rate": 2.1452581032412966e-05,
+      "loss": 0.506,
+      "step": 14281
+    },
+    {
+      "epoch": 18.28096,
+      "grad_norm": 1.0324424505233765,
+      "learning_rate": 2.1450580232092838e-05,
+      "loss": 0.5348,
+      "step": 14282
+    },
+    {
+      "epoch": 18.28224,
+      "grad_norm": 0.9970609545707703,
+      "learning_rate": 2.144857943177271e-05,
+      "loss": 0.5035,
+      "step": 14283
+    },
+    {
+      "epoch": 18.28352,
+      "grad_norm": 1.0144636631011963,
+      "learning_rate": 2.144657863145258e-05,
+      "loss": 0.499,
+      "step": 14284
+    },
+    {
+      "epoch": 18.2848,
+      "grad_norm": 1.009353756904602,
+      "learning_rate": 2.1444577831132453e-05,
+      "loss": 0.5351,
+      "step": 14285
+    },
+    {
+      "epoch": 18.28608,
+      "grad_norm": 0.9811288118362427,
+      "learning_rate": 2.1442577030812325e-05,
+      "loss": 0.4582,
+      "step": 14286
+    },
+    {
+      "epoch": 18.28736,
+      "grad_norm": 1.0266668796539307,
+      "learning_rate": 2.1440576230492197e-05,
+      "loss": 0.4908,
+      "step": 14287
+    },
+    {
+      "epoch": 18.28864,
+      "grad_norm": 1.0499911308288574,
+      "learning_rate": 2.143857543017207e-05,
+      "loss": 0.5122,
+      "step": 14288
+    },
+    {
+      "epoch": 18.28992,
+      "grad_norm": 1.0684877634048462,
+      "learning_rate": 2.143657462985194e-05,
+      "loss": 0.5143,
+      "step": 14289
+    },
+    {
+      "epoch": 18.2912,
+      "grad_norm": 1.0018022060394287,
+      "learning_rate": 2.1434573829531816e-05,
+      "loss": 0.4595,
+      "step": 14290
+    },
+    {
+      "epoch": 18.29248,
+      "grad_norm": 1.0234333276748657,
+      "learning_rate": 2.1432573029211684e-05,
+      "loss": 0.4952,
+      "step": 14291
+    },
+    {
+      "epoch": 18.29376,
+      "grad_norm": 1.015306830406189,
+      "learning_rate": 2.1430572228891556e-05,
+      "loss": 0.4971,
+      "step": 14292
+    },
+    {
+      "epoch": 18.29504,
+      "grad_norm": 1.069213628768921,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.5289,
+      "step": 14293
+    },
+    {
+      "epoch": 18.29632,
+      "grad_norm": 1.0142141580581665,
+      "learning_rate": 2.1426570628251303e-05,
+      "loss": 0.502,
+      "step": 14294
+    },
+    {
+      "epoch": 18.2976,
+      "grad_norm": 1.0849502086639404,
+      "learning_rate": 2.1424569827931172e-05,
+      "loss": 0.4753,
+      "step": 14295
+    },
+    {
+      "epoch": 18.29888,
+      "grad_norm": 1.010981798171997,
+      "learning_rate": 2.1422569027611044e-05,
+      "loss": 0.4907,
+      "step": 14296
+    },
+    {
+      "epoch": 18.300159999999998,
+      "grad_norm": 1.0261670351028442,
+      "learning_rate": 2.142056822729092e-05,
+      "loss": 0.4917,
+      "step": 14297
+    },
+    {
+      "epoch": 18.30144,
+      "grad_norm": 1.057773232460022,
+      "learning_rate": 2.141856742697079e-05,
+      "loss": 0.5051,
+      "step": 14298
+    },
+    {
+      "epoch": 18.30272,
+      "grad_norm": 1.0207065343856812,
+      "learning_rate": 2.141656662665066e-05,
+      "loss": 0.4841,
+      "step": 14299
+    },
+    {
+      "epoch": 18.304,
+      "grad_norm": 1.052815318107605,
+      "learning_rate": 2.141456582633053e-05,
+      "loss": 0.4918,
+      "step": 14300
+    },
+    {
+      "epoch": 18.30528,
+      "grad_norm": 1.0421321392059326,
+      "learning_rate": 2.1412565026010406e-05,
+      "loss": 0.5357,
+      "step": 14301
+    },
+    {
+      "epoch": 18.30656,
+      "grad_norm": 1.0197725296020508,
+      "learning_rate": 2.1410564225690278e-05,
+      "loss": 0.4925,
+      "step": 14302
+    },
+    {
+      "epoch": 18.30784,
+      "grad_norm": 1.0184986591339111,
+      "learning_rate": 2.1408563425370147e-05,
+      "loss": 0.4907,
+      "step": 14303
+    },
+    {
+      "epoch": 18.30912,
+      "grad_norm": 1.0371520519256592,
+      "learning_rate": 2.1406562625050022e-05,
+      "loss": 0.469,
+      "step": 14304
+    },
+    {
+      "epoch": 18.3104,
+      "grad_norm": 1.0472283363342285,
+      "learning_rate": 2.1404561824729894e-05,
+      "loss": 0.4778,
+      "step": 14305
+    },
+    {
+      "epoch": 18.31168,
+      "grad_norm": 1.0578492879867554,
+      "learning_rate": 2.1402561024409766e-05,
+      "loss": 0.5058,
+      "step": 14306
+    },
+    {
+      "epoch": 18.31296,
+      "grad_norm": 1.044542670249939,
+      "learning_rate": 2.1400560224089634e-05,
+      "loss": 0.5086,
+      "step": 14307
+    },
+    {
+      "epoch": 18.31424,
+      "grad_norm": 1.0305312871932983,
+      "learning_rate": 2.139855942376951e-05,
+      "loss": 0.4927,
+      "step": 14308
+    },
+    {
+      "epoch": 18.31552,
+      "grad_norm": 1.1355465650558472,
+      "learning_rate": 2.139655862344938e-05,
+      "loss": 0.5151,
+      "step": 14309
+    },
+    {
+      "epoch": 18.3168,
+      "grad_norm": 1.0777387619018555,
+      "learning_rate": 2.1394557823129253e-05,
+      "loss": 0.5034,
+      "step": 14310
+    },
+    {
+      "epoch": 18.31808,
+      "grad_norm": 1.1117095947265625,
+      "learning_rate": 2.1392557022809125e-05,
+      "loss": 0.5293,
+      "step": 14311
+    },
+    {
+      "epoch": 18.31936,
+      "grad_norm": 1.0630356073379517,
+      "learning_rate": 2.1390556222488997e-05,
+      "loss": 0.5152,
+      "step": 14312
+    },
+    {
+      "epoch": 18.32064,
+      "grad_norm": 1.0686274766921997,
+      "learning_rate": 2.138855542216887e-05,
+      "loss": 0.4792,
+      "step": 14313
+    },
+    {
+      "epoch": 18.32192,
+      "grad_norm": 1.0616755485534668,
+      "learning_rate": 2.138655462184874e-05,
+      "loss": 0.4932,
+      "step": 14314
+    },
+    {
+      "epoch": 18.3232,
+      "grad_norm": 1.079272985458374,
+      "learning_rate": 2.1384553821528612e-05,
+      "loss": 0.4785,
+      "step": 14315
+    },
+    {
+      "epoch": 18.32448,
+      "grad_norm": 1.0683273077011108,
+      "learning_rate": 2.1382553021208484e-05,
+      "loss": 0.517,
+      "step": 14316
+    },
+    {
+      "epoch": 18.32576,
+      "grad_norm": 1.0856717824935913,
+      "learning_rate": 2.1380552220888356e-05,
+      "loss": 0.489,
+      "step": 14317
+    },
+    {
+      "epoch": 18.32704,
+      "grad_norm": 1.0788426399230957,
+      "learning_rate": 2.137855142056823e-05,
+      "loss": 0.4987,
+      "step": 14318
+    },
+    {
+      "epoch": 18.32832,
+      "grad_norm": 1.0643631219863892,
+      "learning_rate": 2.13765506202481e-05,
+      "loss": 0.4891,
+      "step": 14319
+    },
+    {
+      "epoch": 18.3296,
+      "grad_norm": 1.1336820125579834,
+      "learning_rate": 2.137454981992797e-05,
+      "loss": 0.5683,
+      "step": 14320
+    },
+    {
+      "epoch": 18.33088,
+      "grad_norm": 1.0837833881378174,
+      "learning_rate": 2.1372549019607844e-05,
+      "loss": 0.4934,
+      "step": 14321
+    },
+    {
+      "epoch": 18.332160000000002,
+      "grad_norm": 1.0664664506912231,
+      "learning_rate": 2.137054821928772e-05,
+      "loss": 0.5126,
+      "step": 14322
+    },
+    {
+      "epoch": 18.33344,
+      "grad_norm": 1.0915263891220093,
+      "learning_rate": 2.1368547418967587e-05,
+      "loss": 0.5302,
+      "step": 14323
+    },
+    {
+      "epoch": 18.33472,
+      "grad_norm": 1.0284439325332642,
+      "learning_rate": 2.136654661864746e-05,
+      "loss": 0.4844,
+      "step": 14324
+    },
+    {
+      "epoch": 18.336,
+      "grad_norm": 1.0181630849838257,
+      "learning_rate": 2.136454581832733e-05,
+      "loss": 0.5208,
+      "step": 14325
+    },
+    {
+      "epoch": 18.33728,
+      "grad_norm": 1.0594981908798218,
+      "learning_rate": 2.1362545018007206e-05,
+      "loss": 0.5249,
+      "step": 14326
+    },
+    {
+      "epoch": 18.33856,
+      "grad_norm": 1.0511082410812378,
+      "learning_rate": 2.1360544217687075e-05,
+      "loss": 0.5147,
+      "step": 14327
+    },
+    {
+      "epoch": 18.33984,
+      "grad_norm": 1.042351484298706,
+      "learning_rate": 2.1358543417366947e-05,
+      "loss": 0.498,
+      "step": 14328
+    },
+    {
+      "epoch": 18.34112,
+      "grad_norm": 1.043094515800476,
+      "learning_rate": 2.1356542617046822e-05,
+      "loss": 0.4971,
+      "step": 14329
+    },
+    {
+      "epoch": 18.3424,
+      "grad_norm": 1.0453187227249146,
+      "learning_rate": 2.1354541816726694e-05,
+      "loss": 0.5092,
+      "step": 14330
+    },
+    {
+      "epoch": 18.34368,
+      "grad_norm": 1.0761491060256958,
+      "learning_rate": 2.1352541016406562e-05,
+      "loss": 0.5334,
+      "step": 14331
+    },
+    {
+      "epoch": 18.34496,
+      "grad_norm": 0.980190098285675,
+      "learning_rate": 2.1350540216086434e-05,
+      "loss": 0.4569,
+      "step": 14332
+    },
+    {
+      "epoch": 18.34624,
+      "grad_norm": 1.0307832956314087,
+      "learning_rate": 2.134853941576631e-05,
+      "loss": 0.4986,
+      "step": 14333
+    },
+    {
+      "epoch": 18.34752,
+      "grad_norm": 1.0275218486785889,
+      "learning_rate": 2.134653861544618e-05,
+      "loss": 0.4659,
+      "step": 14334
+    },
+    {
+      "epoch": 18.3488,
+      "grad_norm": 1.1118059158325195,
+      "learning_rate": 2.134453781512605e-05,
+      "loss": 0.5302,
+      "step": 14335
+    },
+    {
+      "epoch": 18.35008,
+      "grad_norm": 1.1193236112594604,
+      "learning_rate": 2.1342537014805925e-05,
+      "loss": 0.5402,
+      "step": 14336
+    },
+    {
+      "epoch": 18.35136,
+      "grad_norm": 1.0722352266311646,
+      "learning_rate": 2.1340536214485797e-05,
+      "loss": 0.458,
+      "step": 14337
+    },
+    {
+      "epoch": 18.35264,
+      "grad_norm": 1.0813183784484863,
+      "learning_rate": 2.133853541416567e-05,
+      "loss": 0.5093,
+      "step": 14338
+    },
+    {
+      "epoch": 18.35392,
+      "grad_norm": 1.0807271003723145,
+      "learning_rate": 2.1336534613845537e-05,
+      "loss": 0.5005,
+      "step": 14339
+    },
+    {
+      "epoch": 18.3552,
+      "grad_norm": 1.1022156476974487,
+      "learning_rate": 2.1334533813525412e-05,
+      "loss": 0.5259,
+      "step": 14340
+    },
+    {
+      "epoch": 18.35648,
+      "grad_norm": 1.0361143350601196,
+      "learning_rate": 2.1332533013205284e-05,
+      "loss": 0.5219,
+      "step": 14341
+    },
+    {
+      "epoch": 18.35776,
+      "grad_norm": 1.035565972328186,
+      "learning_rate": 2.1330532212885156e-05,
+      "loss": 0.4797,
+      "step": 14342
+    },
+    {
+      "epoch": 18.35904,
+      "grad_norm": 1.088816523551941,
+      "learning_rate": 2.1328531412565028e-05,
+      "loss": 0.4918,
+      "step": 14343
+    },
+    {
+      "epoch": 18.36032,
+      "grad_norm": 1.0977002382278442,
+      "learning_rate": 2.13265306122449e-05,
+      "loss": 0.5464,
+      "step": 14344
+    },
+    {
+      "epoch": 18.3616,
+      "grad_norm": 1.0365880727767944,
+      "learning_rate": 2.132452981192477e-05,
+      "loss": 0.5105,
+      "step": 14345
+    },
+    {
+      "epoch": 18.36288,
+      "grad_norm": 1.0541160106658936,
+      "learning_rate": 2.1322529011604643e-05,
+      "loss": 0.5156,
+      "step": 14346
+    },
+    {
+      "epoch": 18.36416,
+      "grad_norm": 1.0975985527038574,
+      "learning_rate": 2.1320528211284515e-05,
+      "loss": 0.5529,
+      "step": 14347
+    },
+    {
+      "epoch": 18.36544,
+      "grad_norm": 1.104960322380066,
+      "learning_rate": 2.1318527410964387e-05,
+      "loss": 0.5088,
+      "step": 14348
+    },
+    {
+      "epoch": 18.36672,
+      "grad_norm": 1.0740171670913696,
+      "learning_rate": 2.131652661064426e-05,
+      "loss": 0.508,
+      "step": 14349
+    },
+    {
+      "epoch": 18.368,
+      "grad_norm": 1.1651798486709595,
+      "learning_rate": 2.131452581032413e-05,
+      "loss": 0.5813,
+      "step": 14350
+    },
+    {
+      "epoch": 18.36928,
+      "grad_norm": 1.0417823791503906,
+      "learning_rate": 2.1312525010004003e-05,
+      "loss": 0.5102,
+      "step": 14351
+    },
+    {
+      "epoch": 18.37056,
+      "grad_norm": 1.0664454698562622,
+      "learning_rate": 2.1310524209683874e-05,
+      "loss": 0.544,
+      "step": 14352
+    },
+    {
+      "epoch": 18.37184,
+      "grad_norm": 1.0032341480255127,
+      "learning_rate": 2.1308523409363746e-05,
+      "loss": 0.4847,
+      "step": 14353
+    },
+    {
+      "epoch": 18.37312,
+      "grad_norm": 1.0495922565460205,
+      "learning_rate": 2.1306522609043618e-05,
+      "loss": 0.4937,
+      "step": 14354
+    },
+    {
+      "epoch": 18.3744,
+      "grad_norm": 1.0448967218399048,
+      "learning_rate": 2.130452180872349e-05,
+      "loss": 0.5093,
+      "step": 14355
+    },
+    {
+      "epoch": 18.37568,
+      "grad_norm": 1.141594648361206,
+      "learning_rate": 2.1302521008403362e-05,
+      "loss": 0.5038,
+      "step": 14356
+    },
+    {
+      "epoch": 18.37696,
+      "grad_norm": 1.0529148578643799,
+      "learning_rate": 2.1300520208083237e-05,
+      "loss": 0.5381,
+      "step": 14357
+    },
+    {
+      "epoch": 18.37824,
+      "grad_norm": 1.064096450805664,
+      "learning_rate": 2.1298519407763106e-05,
+      "loss": 0.4849,
+      "step": 14358
+    },
+    {
+      "epoch": 18.37952,
+      "grad_norm": 1.0783796310424805,
+      "learning_rate": 2.1296518607442977e-05,
+      "loss": 0.483,
+      "step": 14359
+    },
+    {
+      "epoch": 18.3808,
+      "grad_norm": 1.0386112928390503,
+      "learning_rate": 2.129451780712285e-05,
+      "loss": 0.5142,
+      "step": 14360
+    },
+    {
+      "epoch": 18.38208,
+      "grad_norm": 1.104622721672058,
+      "learning_rate": 2.1292517006802725e-05,
+      "loss": 0.5947,
+      "step": 14361
+    },
+    {
+      "epoch": 18.38336,
+      "grad_norm": 1.0468897819519043,
+      "learning_rate": 2.1290516206482593e-05,
+      "loss": 0.4647,
+      "step": 14362
+    },
+    {
+      "epoch": 18.38464,
+      "grad_norm": 1.0939141511917114,
+      "learning_rate": 2.1288515406162465e-05,
+      "loss": 0.5201,
+      "step": 14363
+    },
+    {
+      "epoch": 18.38592,
+      "grad_norm": 1.0154157876968384,
+      "learning_rate": 2.128651460584234e-05,
+      "loss": 0.4947,
+      "step": 14364
+    },
+    {
+      "epoch": 18.3872,
+      "grad_norm": 1.0433145761489868,
+      "learning_rate": 2.1284513805522212e-05,
+      "loss": 0.4675,
+      "step": 14365
+    },
+    {
+      "epoch": 18.38848,
+      "grad_norm": 1.0545381307601929,
+      "learning_rate": 2.128251300520208e-05,
+      "loss": 0.5372,
+      "step": 14366
+    },
+    {
+      "epoch": 18.38976,
+      "grad_norm": 1.0403685569763184,
+      "learning_rate": 2.1280512204881952e-05,
+      "loss": 0.5044,
+      "step": 14367
+    },
+    {
+      "epoch": 18.39104,
+      "grad_norm": 1.0547093152999878,
+      "learning_rate": 2.1278511404561828e-05,
+      "loss": 0.5217,
+      "step": 14368
+    },
+    {
+      "epoch": 18.39232,
+      "grad_norm": 1.0452656745910645,
+      "learning_rate": 2.12765106042417e-05,
+      "loss": 0.5011,
+      "step": 14369
+    },
+    {
+      "epoch": 18.3936,
+      "grad_norm": 1.124005675315857,
+      "learning_rate": 2.1274509803921568e-05,
+      "loss": 0.5562,
+      "step": 14370
+    },
+    {
+      "epoch": 18.39488,
+      "grad_norm": 1.0460633039474487,
+      "learning_rate": 2.1272509003601443e-05,
+      "loss": 0.4712,
+      "step": 14371
+    },
+    {
+      "epoch": 18.39616,
+      "grad_norm": 1.0718797445297241,
+      "learning_rate": 2.1270508203281315e-05,
+      "loss": 0.5197,
+      "step": 14372
+    },
+    {
+      "epoch": 18.39744,
+      "grad_norm": 1.036122441291809,
+      "learning_rate": 2.1268507402961187e-05,
+      "loss": 0.4993,
+      "step": 14373
+    },
+    {
+      "epoch": 18.39872,
+      "grad_norm": 1.087517499923706,
+      "learning_rate": 2.1266506602641055e-05,
+      "loss": 0.5288,
+      "step": 14374
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 1.0326744318008423,
+      "learning_rate": 2.126450580232093e-05,
+      "loss": 0.4968,
+      "step": 14375
+    },
+    {
+      "epoch": 18.40128,
+      "grad_norm": 1.0811477899551392,
+      "learning_rate": 2.1262505002000802e-05,
+      "loss": 0.5151,
+      "step": 14376
+    },
+    {
+      "epoch": 18.40256,
+      "grad_norm": 1.0682767629623413,
+      "learning_rate": 2.1260504201680674e-05,
+      "loss": 0.5454,
+      "step": 14377
+    },
+    {
+      "epoch": 18.40384,
+      "grad_norm": 1.0771913528442383,
+      "learning_rate": 2.1258503401360543e-05,
+      "loss": 0.5331,
+      "step": 14378
+    },
+    {
+      "epoch": 18.40512,
+      "grad_norm": 1.0536373853683472,
+      "learning_rate": 2.1256502601040418e-05,
+      "loss": 0.5086,
+      "step": 14379
+    },
+    {
+      "epoch": 18.4064,
+      "grad_norm": 1.0189549922943115,
+      "learning_rate": 2.125450180072029e-05,
+      "loss": 0.497,
+      "step": 14380
+    },
+    {
+      "epoch": 18.40768,
+      "grad_norm": 1.0349353551864624,
+      "learning_rate": 2.1252501000400162e-05,
+      "loss": 0.4787,
+      "step": 14381
+    },
+    {
+      "epoch": 18.40896,
+      "grad_norm": 1.0494370460510254,
+      "learning_rate": 2.1250500200080034e-05,
+      "loss": 0.5013,
+      "step": 14382
+    },
+    {
+      "epoch": 18.41024,
+      "grad_norm": 1.0900458097457886,
+      "learning_rate": 2.1248499399759905e-05,
+      "loss": 0.5222,
+      "step": 14383
+    },
+    {
+      "epoch": 18.41152,
+      "grad_norm": 1.0162975788116455,
+      "learning_rate": 2.1246498599439777e-05,
+      "loss": 0.4884,
+      "step": 14384
+    },
+    {
+      "epoch": 18.4128,
+      "grad_norm": 1.079209804534912,
+      "learning_rate": 2.124449779911965e-05,
+      "loss": 0.5416,
+      "step": 14385
+    },
+    {
+      "epoch": 18.41408,
+      "grad_norm": 1.0865408182144165,
+      "learning_rate": 2.124249699879952e-05,
+      "loss": 0.5044,
+      "step": 14386
+    },
+    {
+      "epoch": 18.41536,
+      "grad_norm": 1.0583715438842773,
+      "learning_rate": 2.1240496198479393e-05,
+      "loss": 0.5338,
+      "step": 14387
+    },
+    {
+      "epoch": 18.41664,
+      "grad_norm": 1.0370880365371704,
+      "learning_rate": 2.1238495398159265e-05,
+      "loss": 0.4833,
+      "step": 14388
+    },
+    {
+      "epoch": 18.41792,
+      "grad_norm": 1.0636683702468872,
+      "learning_rate": 2.1236494597839137e-05,
+      "loss": 0.4915,
+      "step": 14389
+    },
+    {
+      "epoch": 18.4192,
+      "grad_norm": 1.079131841659546,
+      "learning_rate": 2.123449379751901e-05,
+      "loss": 0.4931,
+      "step": 14390
+    },
+    {
+      "epoch": 18.42048,
+      "grad_norm": 1.0714484453201294,
+      "learning_rate": 2.123249299719888e-05,
+      "loss": 0.4701,
+      "step": 14391
+    },
+    {
+      "epoch": 18.42176,
+      "grad_norm": 1.1398820877075195,
+      "learning_rate": 2.1230492196878752e-05,
+      "loss": 0.5147,
+      "step": 14392
+    },
+    {
+      "epoch": 18.42304,
+      "grad_norm": 1.0825884342193604,
+      "learning_rate": 2.1228491396558624e-05,
+      "loss": 0.4938,
+      "step": 14393
+    },
+    {
+      "epoch": 18.42432,
+      "grad_norm": 1.0806691646575928,
+      "learning_rate": 2.1226490596238496e-05,
+      "loss": 0.5118,
+      "step": 14394
+    },
+    {
+      "epoch": 18.4256,
+      "grad_norm": 1.0657216310501099,
+      "learning_rate": 2.1224489795918368e-05,
+      "loss": 0.5168,
+      "step": 14395
+    },
+    {
+      "epoch": 18.42688,
+      "grad_norm": 1.0217489004135132,
+      "learning_rate": 2.1222488995598243e-05,
+      "loss": 0.4725,
+      "step": 14396
+    },
+    {
+      "epoch": 18.42816,
+      "grad_norm": 1.0732911825180054,
+      "learning_rate": 2.122048819527811e-05,
+      "loss": 0.5417,
+      "step": 14397
+    },
+    {
+      "epoch": 18.42944,
+      "grad_norm": 1.091342568397522,
+      "learning_rate": 2.1218487394957983e-05,
+      "loss": 0.5097,
+      "step": 14398
+    },
+    {
+      "epoch": 18.43072,
+      "grad_norm": 1.0096945762634277,
+      "learning_rate": 2.1216486594637855e-05,
+      "loss": 0.4584,
+      "step": 14399
+    },
+    {
+      "epoch": 18.432,
+      "grad_norm": 0.9952658414840698,
+      "learning_rate": 2.121448579431773e-05,
+      "loss": 0.4642,
+      "step": 14400
+    },
+    {
+      "epoch": 18.43328,
+      "grad_norm": 1.108870506286621,
+      "learning_rate": 2.12124849939976e-05,
+      "loss": 0.5275,
+      "step": 14401
+    },
+    {
+      "epoch": 18.43456,
+      "grad_norm": 1.1087889671325684,
+      "learning_rate": 2.121048419367747e-05,
+      "loss": 0.5089,
+      "step": 14402
+    },
+    {
+      "epoch": 18.43584,
+      "grad_norm": 1.057420253753662,
+      "learning_rate": 2.1208483393357346e-05,
+      "loss": 0.513,
+      "step": 14403
+    },
+    {
+      "epoch": 18.43712,
+      "grad_norm": 1.1254922151565552,
+      "learning_rate": 2.1206482593037218e-05,
+      "loss": 0.5404,
+      "step": 14404
+    },
+    {
+      "epoch": 18.4384,
+      "grad_norm": 1.0429891347885132,
+      "learning_rate": 2.1204481792717086e-05,
+      "loss": 0.4935,
+      "step": 14405
+    },
+    {
+      "epoch": 18.43968,
+      "grad_norm": 1.089011549949646,
+      "learning_rate": 2.1202480992396958e-05,
+      "loss": 0.5318,
+      "step": 14406
+    },
+    {
+      "epoch": 18.44096,
+      "grad_norm": 1.080314040184021,
+      "learning_rate": 2.1200480192076833e-05,
+      "loss": 0.5239,
+      "step": 14407
+    },
+    {
+      "epoch": 18.44224,
+      "grad_norm": 1.047875165939331,
+      "learning_rate": 2.1198479391756705e-05,
+      "loss": 0.4918,
+      "step": 14408
+    },
+    {
+      "epoch": 18.44352,
+      "grad_norm": 1.0228180885314941,
+      "learning_rate": 2.1196478591436574e-05,
+      "loss": 0.4958,
+      "step": 14409
+    },
+    {
+      "epoch": 18.4448,
+      "grad_norm": 1.0188748836517334,
+      "learning_rate": 2.119447779111645e-05,
+      "loss": 0.4798,
+      "step": 14410
+    },
+    {
+      "epoch": 18.44608,
+      "grad_norm": 1.0608867406845093,
+      "learning_rate": 2.119247699079632e-05,
+      "loss": 0.5117,
+      "step": 14411
+    },
+    {
+      "epoch": 18.44736,
+      "grad_norm": 1.0850874185562134,
+      "learning_rate": 2.1190476190476193e-05,
+      "loss": 0.5156,
+      "step": 14412
+    },
+    {
+      "epoch": 18.44864,
+      "grad_norm": 1.093931794166565,
+      "learning_rate": 2.118847539015606e-05,
+      "loss": 0.5186,
+      "step": 14413
+    },
+    {
+      "epoch": 18.44992,
+      "grad_norm": 1.0857752561569214,
+      "learning_rate": 2.1186474589835936e-05,
+      "loss": 0.5215,
+      "step": 14414
+    },
+    {
+      "epoch": 18.4512,
+      "grad_norm": 1.0228488445281982,
+      "learning_rate": 2.1184473789515808e-05,
+      "loss": 0.4972,
+      "step": 14415
+    },
+    {
+      "epoch": 18.45248,
+      "grad_norm": 0.979611337184906,
+      "learning_rate": 2.118247298919568e-05,
+      "loss": 0.4884,
+      "step": 14416
+    },
+    {
+      "epoch": 18.45376,
+      "grad_norm": 1.0769535303115845,
+      "learning_rate": 2.1180472188875552e-05,
+      "loss": 0.53,
+      "step": 14417
+    },
+    {
+      "epoch": 18.45504,
+      "grad_norm": 1.103134274482727,
+      "learning_rate": 2.1178471388555424e-05,
+      "loss": 0.5321,
+      "step": 14418
+    },
+    {
+      "epoch": 18.45632,
+      "grad_norm": 1.0389997959136963,
+      "learning_rate": 2.1176470588235296e-05,
+      "loss": 0.4775,
+      "step": 14419
+    },
+    {
+      "epoch": 18.4576,
+      "grad_norm": 1.014994740486145,
+      "learning_rate": 2.1174469787915168e-05,
+      "loss": 0.4521,
+      "step": 14420
+    },
+    {
+      "epoch": 18.45888,
+      "grad_norm": 1.0282139778137207,
+      "learning_rate": 2.117246898759504e-05,
+      "loss": 0.5012,
+      "step": 14421
+    },
+    {
+      "epoch": 18.46016,
+      "grad_norm": 1.0462567806243896,
+      "learning_rate": 2.117046818727491e-05,
+      "loss": 0.5489,
+      "step": 14422
+    },
+    {
+      "epoch": 18.46144,
+      "grad_norm": 1.0593222379684448,
+      "learning_rate": 2.1168467386954783e-05,
+      "loss": 0.528,
+      "step": 14423
+    },
+    {
+      "epoch": 18.46272,
+      "grad_norm": 1.0701686143875122,
+      "learning_rate": 2.1166466586634655e-05,
+      "loss": 0.4987,
+      "step": 14424
+    },
+    {
+      "epoch": 18.464,
+      "grad_norm": 1.0602664947509766,
+      "learning_rate": 2.1164465786314527e-05,
+      "loss": 0.503,
+      "step": 14425
+    },
+    {
+      "epoch": 18.46528,
+      "grad_norm": 1.018227458000183,
+      "learning_rate": 2.11624649859944e-05,
+      "loss": 0.4681,
+      "step": 14426
+    },
+    {
+      "epoch": 18.46656,
+      "grad_norm": 1.078520655632019,
+      "learning_rate": 2.116046418567427e-05,
+      "loss": 0.5599,
+      "step": 14427
+    },
+    {
+      "epoch": 18.46784,
+      "grad_norm": 1.0530571937561035,
+      "learning_rate": 2.1158463385354142e-05,
+      "loss": 0.4804,
+      "step": 14428
+    },
+    {
+      "epoch": 18.46912,
+      "grad_norm": 1.029566764831543,
+      "learning_rate": 2.1156462585034014e-05,
+      "loss": 0.4644,
+      "step": 14429
+    },
+    {
+      "epoch": 18.4704,
+      "grad_norm": 1.0319150686264038,
+      "learning_rate": 2.1154461784713886e-05,
+      "loss": 0.4924,
+      "step": 14430
+    },
+    {
+      "epoch": 18.47168,
+      "grad_norm": 1.114824891090393,
+      "learning_rate": 2.115246098439376e-05,
+      "loss": 0.57,
+      "step": 14431
+    },
+    {
+      "epoch": 18.47296,
+      "grad_norm": 1.093182921409607,
+      "learning_rate": 2.115046018407363e-05,
+      "loss": 0.552,
+      "step": 14432
+    },
+    {
+      "epoch": 18.47424,
+      "grad_norm": 1.050522804260254,
+      "learning_rate": 2.11484593837535e-05,
+      "loss": 0.4954,
+      "step": 14433
+    },
+    {
+      "epoch": 18.47552,
+      "grad_norm": 1.0565167665481567,
+      "learning_rate": 2.1146458583433374e-05,
+      "loss": 0.5101,
+      "step": 14434
+    },
+    {
+      "epoch": 18.4768,
+      "grad_norm": 1.065156102180481,
+      "learning_rate": 2.114445778311325e-05,
+      "loss": 0.5184,
+      "step": 14435
+    },
+    {
+      "epoch": 18.47808,
+      "grad_norm": 0.9988115429878235,
+      "learning_rate": 2.1142456982793117e-05,
+      "loss": 0.4639,
+      "step": 14436
+    },
+    {
+      "epoch": 18.47936,
+      "grad_norm": 1.0403028726577759,
+      "learning_rate": 2.114045618247299e-05,
+      "loss": 0.4858,
+      "step": 14437
+    },
+    {
+      "epoch": 18.48064,
+      "grad_norm": 1.0715668201446533,
+      "learning_rate": 2.113845538215286e-05,
+      "loss": 0.5176,
+      "step": 14438
+    },
+    {
+      "epoch": 18.48192,
+      "grad_norm": 0.9925726652145386,
+      "learning_rate": 2.1136454581832736e-05,
+      "loss": 0.4261,
+      "step": 14439
+    },
+    {
+      "epoch": 18.4832,
+      "grad_norm": 1.02934730052948,
+      "learning_rate": 2.1134453781512605e-05,
+      "loss": 0.491,
+      "step": 14440
+    },
+    {
+      "epoch": 18.48448,
+      "grad_norm": 1.0200088024139404,
+      "learning_rate": 2.1132452981192476e-05,
+      "loss": 0.4953,
+      "step": 14441
+    },
+    {
+      "epoch": 18.48576,
+      "grad_norm": 1.080019235610962,
+      "learning_rate": 2.1130452180872352e-05,
+      "loss": 0.5132,
+      "step": 14442
+    },
+    {
+      "epoch": 18.48704,
+      "grad_norm": 1.0620362758636475,
+      "learning_rate": 2.1128451380552224e-05,
+      "loss": 0.5103,
+      "step": 14443
+    },
+    {
+      "epoch": 18.48832,
+      "grad_norm": 1.0634804964065552,
+      "learning_rate": 2.1126450580232092e-05,
+      "loss": 0.5085,
+      "step": 14444
+    },
+    {
+      "epoch": 18.4896,
+      "grad_norm": 1.110755443572998,
+      "learning_rate": 2.1124449779911964e-05,
+      "loss": 0.532,
+      "step": 14445
+    },
+    {
+      "epoch": 18.49088,
+      "grad_norm": 1.0549490451812744,
+      "learning_rate": 2.112244897959184e-05,
+      "loss": 0.4641,
+      "step": 14446
+    },
+    {
+      "epoch": 18.49216,
+      "grad_norm": 1.0575416088104248,
+      "learning_rate": 2.112044817927171e-05,
+      "loss": 0.4892,
+      "step": 14447
+    },
+    {
+      "epoch": 18.49344,
+      "grad_norm": 1.1010053157806396,
+      "learning_rate": 2.111844737895158e-05,
+      "loss": 0.4996,
+      "step": 14448
+    },
+    {
+      "epoch": 18.49472,
+      "grad_norm": 1.0477495193481445,
+      "learning_rate": 2.1116446578631455e-05,
+      "loss": 0.4811,
+      "step": 14449
+    },
+    {
+      "epoch": 18.496,
+      "grad_norm": 1.125105619430542,
+      "learning_rate": 2.1114445778311327e-05,
+      "loss": 0.5519,
+      "step": 14450
+    },
+    {
+      "epoch": 18.49728,
+      "grad_norm": 1.1108721494674683,
+      "learning_rate": 2.11124449779912e-05,
+      "loss": 0.5117,
+      "step": 14451
+    },
+    {
+      "epoch": 18.49856,
+      "grad_norm": 1.0684090852737427,
+      "learning_rate": 2.1110444177671067e-05,
+      "loss": 0.4912,
+      "step": 14452
+    },
+    {
+      "epoch": 18.49984,
+      "grad_norm": 1.1287049055099487,
+      "learning_rate": 2.1108443377350942e-05,
+      "loss": 0.5469,
+      "step": 14453
+    },
+    {
+      "epoch": 18.50112,
+      "grad_norm": 1.048940658569336,
+      "learning_rate": 2.1106442577030814e-05,
+      "loss": 0.4936,
+      "step": 14454
+    },
+    {
+      "epoch": 18.5024,
+      "grad_norm": 1.0817766189575195,
+      "learning_rate": 2.1104441776710686e-05,
+      "loss": 0.5038,
+      "step": 14455
+    },
+    {
+      "epoch": 18.50368,
+      "grad_norm": 1.0768336057662964,
+      "learning_rate": 2.1102440976390558e-05,
+      "loss": 0.5887,
+      "step": 14456
+    },
+    {
+      "epoch": 18.50496,
+      "grad_norm": 1.0072085857391357,
+      "learning_rate": 2.110044017607043e-05,
+      "loss": 0.478,
+      "step": 14457
+    },
+    {
+      "epoch": 18.50624,
+      "grad_norm": 0.9800832867622375,
+      "learning_rate": 2.10984393757503e-05,
+      "loss": 0.4799,
+      "step": 14458
+    },
+    {
+      "epoch": 18.50752,
+      "grad_norm": 1.042318344116211,
+      "learning_rate": 2.1096438575430173e-05,
+      "loss": 0.5065,
+      "step": 14459
+    },
+    {
+      "epoch": 18.5088,
+      "grad_norm": 1.1013107299804688,
+      "learning_rate": 2.1094437775110045e-05,
+      "loss": 0.6008,
+      "step": 14460
+    },
+    {
+      "epoch": 18.51008,
+      "grad_norm": 0.9804616570472717,
+      "learning_rate": 2.1092436974789917e-05,
+      "loss": 0.4585,
+      "step": 14461
+    },
+    {
+      "epoch": 18.51136,
+      "grad_norm": 1.0614451169967651,
+      "learning_rate": 2.109043617446979e-05,
+      "loss": 0.4966,
+      "step": 14462
+    },
+    {
+      "epoch": 18.51264,
+      "grad_norm": 0.9891878962516785,
+      "learning_rate": 2.108843537414966e-05,
+      "loss": 0.516,
+      "step": 14463
+    },
+    {
+      "epoch": 18.51392,
+      "grad_norm": 1.0645172595977783,
+      "learning_rate": 2.1086434573829533e-05,
+      "loss": 0.5074,
+      "step": 14464
+    },
+    {
+      "epoch": 18.5152,
+      "grad_norm": 1.0641040802001953,
+      "learning_rate": 2.1084433773509404e-05,
+      "loss": 0.5201,
+      "step": 14465
+    },
+    {
+      "epoch": 18.51648,
+      "grad_norm": 0.9872971177101135,
+      "learning_rate": 2.1082432973189276e-05,
+      "loss": 0.4698,
+      "step": 14466
+    },
+    {
+      "epoch": 18.51776,
+      "grad_norm": 1.167199730873108,
+      "learning_rate": 2.1080432172869148e-05,
+      "loss": 0.5102,
+      "step": 14467
+    },
+    {
+      "epoch": 18.51904,
+      "grad_norm": 1.1635416746139526,
+      "learning_rate": 2.107843137254902e-05,
+      "loss": 0.5385,
+      "step": 14468
+    },
+    {
+      "epoch": 18.52032,
+      "grad_norm": 1.1156731843948364,
+      "learning_rate": 2.1076430572228892e-05,
+      "loss": 0.5442,
+      "step": 14469
+    },
+    {
+      "epoch": 18.5216,
+      "grad_norm": 1.0268969535827637,
+      "learning_rate": 2.1074429771908767e-05,
+      "loss": 0.4385,
+      "step": 14470
+    },
+    {
+      "epoch": 18.52288,
+      "grad_norm": 1.0554258823394775,
+      "learning_rate": 2.1072428971588636e-05,
+      "loss": 0.5172,
+      "step": 14471
+    },
+    {
+      "epoch": 18.52416,
+      "grad_norm": 1.07488214969635,
+      "learning_rate": 2.1070428171268507e-05,
+      "loss": 0.5373,
+      "step": 14472
+    },
+    {
+      "epoch": 18.52544,
+      "grad_norm": 0.9701617360115051,
+      "learning_rate": 2.106842737094838e-05,
+      "loss": 0.4699,
+      "step": 14473
+    },
+    {
+      "epoch": 18.52672,
+      "grad_norm": 1.0536491870880127,
+      "learning_rate": 2.1066426570628255e-05,
+      "loss": 0.5075,
+      "step": 14474
+    },
+    {
+      "epoch": 18.528,
+      "grad_norm": 1.1077998876571655,
+      "learning_rate": 2.1064425770308123e-05,
+      "loss": 0.559,
+      "step": 14475
+    },
+    {
+      "epoch": 18.52928,
+      "grad_norm": 1.043217658996582,
+      "learning_rate": 2.1062424969987995e-05,
+      "loss": 0.4921,
+      "step": 14476
+    },
+    {
+      "epoch": 18.53056,
+      "grad_norm": 1.0634691715240479,
+      "learning_rate": 2.106042416966787e-05,
+      "loss": 0.4999,
+      "step": 14477
+    },
+    {
+      "epoch": 18.53184,
+      "grad_norm": 1.024617075920105,
+      "learning_rate": 2.1058423369347742e-05,
+      "loss": 0.5021,
+      "step": 14478
+    },
+    {
+      "epoch": 18.53312,
+      "grad_norm": 1.026159405708313,
+      "learning_rate": 2.105642256902761e-05,
+      "loss": 0.5201,
+      "step": 14479
+    },
+    {
+      "epoch": 18.5344,
+      "grad_norm": 1.085721731185913,
+      "learning_rate": 2.1054421768707482e-05,
+      "loss": 0.5542,
+      "step": 14480
+    },
+    {
+      "epoch": 18.53568,
+      "grad_norm": 1.0542292594909668,
+      "learning_rate": 2.1052420968387358e-05,
+      "loss": 0.5239,
+      "step": 14481
+    },
+    {
+      "epoch": 18.53696,
+      "grad_norm": 1.0141193866729736,
+      "learning_rate": 2.105042016806723e-05,
+      "loss": 0.5116,
+      "step": 14482
+    },
+    {
+      "epoch": 18.538240000000002,
+      "grad_norm": 1.106048822402954,
+      "learning_rate": 2.1048419367747098e-05,
+      "loss": 0.5843,
+      "step": 14483
+    },
+    {
+      "epoch": 18.53952,
+      "grad_norm": 1.0620492696762085,
+      "learning_rate": 2.1046418567426973e-05,
+      "loss": 0.5049,
+      "step": 14484
+    },
+    {
+      "epoch": 18.5408,
+      "grad_norm": 1.0204299688339233,
+      "learning_rate": 2.1044417767106845e-05,
+      "loss": 0.507,
+      "step": 14485
+    },
+    {
+      "epoch": 18.54208,
+      "grad_norm": 1.0955986976623535,
+      "learning_rate": 2.1042416966786717e-05,
+      "loss": 0.5323,
+      "step": 14486
+    },
+    {
+      "epoch": 18.54336,
+      "grad_norm": 1.0099011659622192,
+      "learning_rate": 2.1040416166466585e-05,
+      "loss": 0.5007,
+      "step": 14487
+    },
+    {
+      "epoch": 18.54464,
+      "grad_norm": 0.9762060642242432,
+      "learning_rate": 2.103841536614646e-05,
+      "loss": 0.4201,
+      "step": 14488
+    },
+    {
+      "epoch": 18.54592,
+      "grad_norm": 1.060850739479065,
+      "learning_rate": 2.1036414565826332e-05,
+      "loss": 0.4815,
+      "step": 14489
+    },
+    {
+      "epoch": 18.5472,
+      "grad_norm": 1.077972173690796,
+      "learning_rate": 2.1034413765506204e-05,
+      "loss": 0.5148,
+      "step": 14490
+    },
+    {
+      "epoch": 18.54848,
+      "grad_norm": 1.1198941469192505,
+      "learning_rate": 2.1032412965186073e-05,
+      "loss": 0.5361,
+      "step": 14491
+    },
+    {
+      "epoch": 18.54976,
+      "grad_norm": 1.1299846172332764,
+      "learning_rate": 2.1030412164865948e-05,
+      "loss": 0.5398,
+      "step": 14492
+    },
+    {
+      "epoch": 18.55104,
+      "grad_norm": 0.9961510896682739,
+      "learning_rate": 2.102841136454582e-05,
+      "loss": 0.425,
+      "step": 14493
+    },
+    {
+      "epoch": 18.55232,
+      "grad_norm": 1.0336405038833618,
+      "learning_rate": 2.102641056422569e-05,
+      "loss": 0.5029,
+      "step": 14494
+    },
+    {
+      "epoch": 18.5536,
+      "grad_norm": 1.073876142501831,
+      "learning_rate": 2.1024409763905564e-05,
+      "loss": 0.4941,
+      "step": 14495
+    },
+    {
+      "epoch": 18.55488,
+      "grad_norm": 1.0787442922592163,
+      "learning_rate": 2.1022408963585435e-05,
+      "loss": 0.5394,
+      "step": 14496
+    },
+    {
+      "epoch": 18.55616,
+      "grad_norm": 1.0297425985336304,
+      "learning_rate": 2.1020408163265307e-05,
+      "loss": 0.4841,
+      "step": 14497
+    },
+    {
+      "epoch": 18.55744,
+      "grad_norm": 1.0471930503845215,
+      "learning_rate": 2.101840736294518e-05,
+      "loss": 0.4956,
+      "step": 14498
+    },
+    {
+      "epoch": 18.55872,
+      "grad_norm": 1.0681778192520142,
+      "learning_rate": 2.101640656262505e-05,
+      "loss": 0.5614,
+      "step": 14499
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 1.0257924795150757,
+      "learning_rate": 2.1014405762304923e-05,
+      "loss": 0.4542,
+      "step": 14500
+    },
+    {
+      "epoch": 18.56128,
+      "grad_norm": 1.028685450553894,
+      "learning_rate": 2.1012404961984795e-05,
+      "loss": 0.4824,
+      "step": 14501
+    },
+    {
+      "epoch": 18.56256,
+      "grad_norm": 1.0713627338409424,
+      "learning_rate": 2.1010404161664667e-05,
+      "loss": 0.543,
+      "step": 14502
+    },
+    {
+      "epoch": 18.56384,
+      "grad_norm": 1.0989772081375122,
+      "learning_rate": 2.100840336134454e-05,
+      "loss": 0.5345,
+      "step": 14503
+    },
+    {
+      "epoch": 18.56512,
+      "grad_norm": 1.0844435691833496,
+      "learning_rate": 2.100640256102441e-05,
+      "loss": 0.514,
+      "step": 14504
+    },
+    {
+      "epoch": 18.5664,
+      "grad_norm": 1.0661566257476807,
+      "learning_rate": 2.1004401760704282e-05,
+      "loss": 0.5177,
+      "step": 14505
+    },
+    {
+      "epoch": 18.56768,
+      "grad_norm": 1.0676720142364502,
+      "learning_rate": 2.1002400960384154e-05,
+      "loss": 0.5007,
+      "step": 14506
+    },
+    {
+      "epoch": 18.56896,
+      "grad_norm": 1.0638859272003174,
+      "learning_rate": 2.1000400160064026e-05,
+      "loss": 0.5185,
+      "step": 14507
+    },
+    {
+      "epoch": 18.57024,
+      "grad_norm": 1.020068883895874,
+      "learning_rate": 2.0998399359743898e-05,
+      "loss": 0.4734,
+      "step": 14508
+    },
+    {
+      "epoch": 18.57152,
+      "grad_norm": 1.103440523147583,
+      "learning_rate": 2.0996398559423773e-05,
+      "loss": 0.5404,
+      "step": 14509
+    },
+    {
+      "epoch": 18.5728,
+      "grad_norm": 1.064695119857788,
+      "learning_rate": 2.099439775910364e-05,
+      "loss": 0.5363,
+      "step": 14510
+    },
+    {
+      "epoch": 18.57408,
+      "grad_norm": 1.0438659191131592,
+      "learning_rate": 2.0992396958783513e-05,
+      "loss": 0.534,
+      "step": 14511
+    },
+    {
+      "epoch": 18.57536,
+      "grad_norm": 1.0252081155776978,
+      "learning_rate": 2.0990396158463385e-05,
+      "loss": 0.4815,
+      "step": 14512
+    },
+    {
+      "epoch": 18.57664,
+      "grad_norm": 1.0469212532043457,
+      "learning_rate": 2.098839535814326e-05,
+      "loss": 0.5046,
+      "step": 14513
+    },
+    {
+      "epoch": 18.57792,
+      "grad_norm": 1.011010766029358,
+      "learning_rate": 2.098639455782313e-05,
+      "loss": 0.4726,
+      "step": 14514
+    },
+    {
+      "epoch": 18.5792,
+      "grad_norm": 1.0593057870864868,
+      "learning_rate": 2.0984393757503e-05,
+      "loss": 0.5198,
+      "step": 14515
+    },
+    {
+      "epoch": 18.58048,
+      "grad_norm": 1.0215234756469727,
+      "learning_rate": 2.0982392957182876e-05,
+      "loss": 0.5168,
+      "step": 14516
+    },
+    {
+      "epoch": 18.58176,
+      "grad_norm": 1.0414518117904663,
+      "learning_rate": 2.0980392156862748e-05,
+      "loss": 0.5049,
+      "step": 14517
+    },
+    {
+      "epoch": 18.58304,
+      "grad_norm": 1.1356487274169922,
+      "learning_rate": 2.0978391356542616e-05,
+      "loss": 0.5467,
+      "step": 14518
+    },
+    {
+      "epoch": 18.584319999999998,
+      "grad_norm": 1.0748487710952759,
+      "learning_rate": 2.0976390556222488e-05,
+      "loss": 0.5602,
+      "step": 14519
+    },
+    {
+      "epoch": 18.5856,
+      "grad_norm": 1.0394898653030396,
+      "learning_rate": 2.0974389755902363e-05,
+      "loss": 0.5286,
+      "step": 14520
+    },
+    {
+      "epoch": 18.58688,
+      "grad_norm": 1.0443836450576782,
+      "learning_rate": 2.0972388955582235e-05,
+      "loss": 0.5007,
+      "step": 14521
+    },
+    {
+      "epoch": 18.58816,
+      "grad_norm": 1.083796739578247,
+      "learning_rate": 2.0970388155262104e-05,
+      "loss": 0.5159,
+      "step": 14522
+    },
+    {
+      "epoch": 18.58944,
+      "grad_norm": 1.0455292463302612,
+      "learning_rate": 2.096838735494198e-05,
+      "loss": 0.4783,
+      "step": 14523
+    },
+    {
+      "epoch": 18.59072,
+      "grad_norm": 1.0481806993484497,
+      "learning_rate": 2.096638655462185e-05,
+      "loss": 0.5356,
+      "step": 14524
+    },
+    {
+      "epoch": 18.592,
+      "grad_norm": 0.9991112947463989,
+      "learning_rate": 2.0964385754301723e-05,
+      "loss": 0.479,
+      "step": 14525
+    },
+    {
+      "epoch": 18.59328,
+      "grad_norm": 1.079669713973999,
+      "learning_rate": 2.096238495398159e-05,
+      "loss": 0.5138,
+      "step": 14526
+    },
+    {
+      "epoch": 18.59456,
+      "grad_norm": 1.0523369312286377,
+      "learning_rate": 2.0960384153661466e-05,
+      "loss": 0.4889,
+      "step": 14527
+    },
+    {
+      "epoch": 18.59584,
+      "grad_norm": 1.1188772916793823,
+      "learning_rate": 2.0958383353341338e-05,
+      "loss": 0.5201,
+      "step": 14528
+    },
+    {
+      "epoch": 18.59712,
+      "grad_norm": 1.111080527305603,
+      "learning_rate": 2.095638255302121e-05,
+      "loss": 0.4918,
+      "step": 14529
+    },
+    {
+      "epoch": 18.5984,
+      "grad_norm": 1.019318699836731,
+      "learning_rate": 2.0954381752701082e-05,
+      "loss": 0.4547,
+      "step": 14530
+    },
+    {
+      "epoch": 18.59968,
+      "grad_norm": 1.0236701965332031,
+      "learning_rate": 2.0952380952380954e-05,
+      "loss": 0.4956,
+      "step": 14531
+    },
+    {
+      "epoch": 18.60096,
+      "grad_norm": 1.06638503074646,
+      "learning_rate": 2.0950380152060826e-05,
+      "loss": 0.4941,
+      "step": 14532
+    },
+    {
+      "epoch": 18.60224,
+      "grad_norm": 1.0441734790802002,
+      "learning_rate": 2.0948379351740697e-05,
+      "loss": 0.4714,
+      "step": 14533
+    },
+    {
+      "epoch": 18.60352,
+      "grad_norm": 1.0851494073867798,
+      "learning_rate": 2.094637855142057e-05,
+      "loss": 0.5392,
+      "step": 14534
+    },
+    {
+      "epoch": 18.6048,
+      "grad_norm": 1.0086928606033325,
+      "learning_rate": 2.094437775110044e-05,
+      "loss": 0.4795,
+      "step": 14535
+    },
+    {
+      "epoch": 18.60608,
+      "grad_norm": 1.0636036396026611,
+      "learning_rate": 2.0942376950780313e-05,
+      "loss": 0.5342,
+      "step": 14536
+    },
+    {
+      "epoch": 18.60736,
+      "grad_norm": 1.0684483051300049,
+      "learning_rate": 2.0940376150460185e-05,
+      "loss": 0.5094,
+      "step": 14537
+    },
+    {
+      "epoch": 18.60864,
+      "grad_norm": 1.0535985231399536,
+      "learning_rate": 2.0938375350140057e-05,
+      "loss": 0.4892,
+      "step": 14538
+    },
+    {
+      "epoch": 18.60992,
+      "grad_norm": 1.0609737634658813,
+      "learning_rate": 2.093637454981993e-05,
+      "loss": 0.5016,
+      "step": 14539
+    },
+    {
+      "epoch": 18.6112,
+      "grad_norm": 1.0694406032562256,
+      "learning_rate": 2.09343737494998e-05,
+      "loss": 0.5134,
+      "step": 14540
+    },
+    {
+      "epoch": 18.61248,
+      "grad_norm": 1.0116039514541626,
+      "learning_rate": 2.0932372949179672e-05,
+      "loss": 0.4585,
+      "step": 14541
+    },
+    {
+      "epoch": 18.61376,
+      "grad_norm": 1.125255823135376,
+      "learning_rate": 2.0930372148859544e-05,
+      "loss": 0.5355,
+      "step": 14542
+    },
+    {
+      "epoch": 18.61504,
+      "grad_norm": 1.111659288406372,
+      "learning_rate": 2.0928371348539416e-05,
+      "loss": 0.5312,
+      "step": 14543
+    },
+    {
+      "epoch": 18.61632,
+      "grad_norm": 1.1183501482009888,
+      "learning_rate": 2.092637054821929e-05,
+      "loss": 0.5408,
+      "step": 14544
+    },
+    {
+      "epoch": 18.6176,
+      "grad_norm": 1.0673730373382568,
+      "learning_rate": 2.092436974789916e-05,
+      "loss": 0.5031,
+      "step": 14545
+    },
+    {
+      "epoch": 18.61888,
+      "grad_norm": 1.039713978767395,
+      "learning_rate": 2.092236894757903e-05,
+      "loss": 0.5115,
+      "step": 14546
+    },
+    {
+      "epoch": 18.62016,
+      "grad_norm": 1.0131070613861084,
+      "learning_rate": 2.0920368147258903e-05,
+      "loss": 0.4742,
+      "step": 14547
+    },
+    {
+      "epoch": 18.62144,
+      "grad_norm": 1.0833195447921753,
+      "learning_rate": 2.091836734693878e-05,
+      "loss": 0.5393,
+      "step": 14548
+    },
+    {
+      "epoch": 18.62272,
+      "grad_norm": 1.193190097808838,
+      "learning_rate": 2.0916366546618647e-05,
+      "loss": 0.5043,
+      "step": 14549
+    },
+    {
+      "epoch": 18.624,
+      "grad_norm": 1.0585458278656006,
+      "learning_rate": 2.091436574629852e-05,
+      "loss": 0.4841,
+      "step": 14550
+    },
+    {
+      "epoch": 18.62528,
+      "grad_norm": 1.0493146181106567,
+      "learning_rate": 2.091236494597839e-05,
+      "loss": 0.5084,
+      "step": 14551
+    },
+    {
+      "epoch": 18.62656,
+      "grad_norm": 1.0405904054641724,
+      "learning_rate": 2.0910364145658266e-05,
+      "loss": 0.5273,
+      "step": 14552
+    },
+    {
+      "epoch": 18.62784,
+      "grad_norm": 1.0221456289291382,
+      "learning_rate": 2.0908363345338135e-05,
+      "loss": 0.4846,
+      "step": 14553
+    },
+    {
+      "epoch": 18.62912,
+      "grad_norm": 1.0705711841583252,
+      "learning_rate": 2.0906362545018006e-05,
+      "loss": 0.5271,
+      "step": 14554
+    },
+    {
+      "epoch": 18.6304,
+      "grad_norm": 1.0186147689819336,
+      "learning_rate": 2.0904361744697882e-05,
+      "loss": 0.4886,
+      "step": 14555
+    },
+    {
+      "epoch": 18.63168,
+      "grad_norm": 1.0867236852645874,
+      "learning_rate": 2.0902360944377754e-05,
+      "loss": 0.517,
+      "step": 14556
+    },
+    {
+      "epoch": 18.63296,
+      "grad_norm": 1.061066746711731,
+      "learning_rate": 2.0900360144057622e-05,
+      "loss": 0.5189,
+      "step": 14557
+    },
+    {
+      "epoch": 18.63424,
+      "grad_norm": 1.0720415115356445,
+      "learning_rate": 2.0898359343737494e-05,
+      "loss": 0.5382,
+      "step": 14558
+    },
+    {
+      "epoch": 18.63552,
+      "grad_norm": 1.061384916305542,
+      "learning_rate": 2.089635854341737e-05,
+      "loss": 0.5331,
+      "step": 14559
+    },
+    {
+      "epoch": 18.6368,
+      "grad_norm": 1.0454514026641846,
+      "learning_rate": 2.089435774309724e-05,
+      "loss": 0.4793,
+      "step": 14560
+    },
+    {
+      "epoch": 18.63808,
+      "grad_norm": 1.0604616403579712,
+      "learning_rate": 2.089235694277711e-05,
+      "loss": 0.4637,
+      "step": 14561
+    },
+    {
+      "epoch": 18.63936,
+      "grad_norm": 1.0945936441421509,
+      "learning_rate": 2.0890356142456985e-05,
+      "loss": 0.5413,
+      "step": 14562
+    },
+    {
+      "epoch": 18.64064,
+      "grad_norm": 1.0144293308258057,
+      "learning_rate": 2.0888355342136857e-05,
+      "loss": 0.5071,
+      "step": 14563
+    },
+    {
+      "epoch": 18.64192,
+      "grad_norm": 1.0849329233169556,
+      "learning_rate": 2.088635454181673e-05,
+      "loss": 0.5343,
+      "step": 14564
+    },
+    {
+      "epoch": 18.6432,
+      "grad_norm": 0.9900054931640625,
+      "learning_rate": 2.0884353741496597e-05,
+      "loss": 0.4709,
+      "step": 14565
+    },
+    {
+      "epoch": 18.64448,
+      "grad_norm": 1.0568320751190186,
+      "learning_rate": 2.0882352941176472e-05,
+      "loss": 0.4829,
+      "step": 14566
+    },
+    {
+      "epoch": 18.64576,
+      "grad_norm": 1.1182734966278076,
+      "learning_rate": 2.0880352140856344e-05,
+      "loss": 0.5742,
+      "step": 14567
+    },
+    {
+      "epoch": 18.64704,
+      "grad_norm": 1.1429896354675293,
+      "learning_rate": 2.0878351340536216e-05,
+      "loss": 0.5731,
+      "step": 14568
+    },
+    {
+      "epoch": 18.64832,
+      "grad_norm": 1.0689339637756348,
+      "learning_rate": 2.0876350540216088e-05,
+      "loss": 0.496,
+      "step": 14569
+    },
+    {
+      "epoch": 18.6496,
+      "grad_norm": 1.0403820276260376,
+      "learning_rate": 2.087434973989596e-05,
+      "loss": 0.4286,
+      "step": 14570
+    },
+    {
+      "epoch": 18.65088,
+      "grad_norm": 1.0727622509002686,
+      "learning_rate": 2.087234893957583e-05,
+      "loss": 0.5124,
+      "step": 14571
+    },
+    {
+      "epoch": 18.65216,
+      "grad_norm": 1.0290236473083496,
+      "learning_rate": 2.0870348139255703e-05,
+      "loss": 0.4925,
+      "step": 14572
+    },
+    {
+      "epoch": 18.65344,
+      "grad_norm": 1.0462875366210938,
+      "learning_rate": 2.0868347338935575e-05,
+      "loss": 0.4666,
+      "step": 14573
+    },
+    {
+      "epoch": 18.65472,
+      "grad_norm": 1.0664476156234741,
+      "learning_rate": 2.0866346538615447e-05,
+      "loss": 0.5112,
+      "step": 14574
+    },
+    {
+      "epoch": 18.656,
+      "grad_norm": 1.0481314659118652,
+      "learning_rate": 2.086434573829532e-05,
+      "loss": 0.4763,
+      "step": 14575
+    },
+    {
+      "epoch": 18.65728,
+      "grad_norm": 1.0135314464569092,
+      "learning_rate": 2.086234493797519e-05,
+      "loss": 0.4885,
+      "step": 14576
+    },
+    {
+      "epoch": 18.65856,
+      "grad_norm": 0.9977083206176758,
+      "learning_rate": 2.0860344137655063e-05,
+      "loss": 0.4949,
+      "step": 14577
+    },
+    {
+      "epoch": 18.65984,
+      "grad_norm": 1.043670892715454,
+      "learning_rate": 2.0858343337334934e-05,
+      "loss": 0.5031,
+      "step": 14578
+    },
+    {
+      "epoch": 18.66112,
+      "grad_norm": 1.0433337688446045,
+      "learning_rate": 2.0856342537014806e-05,
+      "loss": 0.5061,
+      "step": 14579
+    },
+    {
+      "epoch": 18.6624,
+      "grad_norm": 1.006962537765503,
+      "learning_rate": 2.0854341736694678e-05,
+      "loss": 0.5039,
+      "step": 14580
+    },
+    {
+      "epoch": 18.66368,
+      "grad_norm": 1.046007513999939,
+      "learning_rate": 2.085234093637455e-05,
+      "loss": 0.5434,
+      "step": 14581
+    },
+    {
+      "epoch": 18.66496,
+      "grad_norm": 1.0263768434524536,
+      "learning_rate": 2.0850340136054422e-05,
+      "loss": 0.5117,
+      "step": 14582
+    },
+    {
+      "epoch": 18.66624,
+      "grad_norm": 1.0261712074279785,
+      "learning_rate": 2.0848339335734297e-05,
+      "loss": 0.4705,
+      "step": 14583
+    },
+    {
+      "epoch": 18.66752,
+      "grad_norm": 1.0606975555419922,
+      "learning_rate": 2.0846338535414166e-05,
+      "loss": 0.4931,
+      "step": 14584
+    },
+    {
+      "epoch": 18.6688,
+      "grad_norm": 1.0955727100372314,
+      "learning_rate": 2.0844337735094037e-05,
+      "loss": 0.5149,
+      "step": 14585
+    },
+    {
+      "epoch": 18.67008,
+      "grad_norm": 0.9943707585334778,
+      "learning_rate": 2.084233693477391e-05,
+      "loss": 0.4668,
+      "step": 14586
+    },
+    {
+      "epoch": 18.67136,
+      "grad_norm": 1.082153081893921,
+      "learning_rate": 2.0840336134453785e-05,
+      "loss": 0.5313,
+      "step": 14587
+    },
+    {
+      "epoch": 18.67264,
+      "grad_norm": 1.0203741788864136,
+      "learning_rate": 2.0838335334133653e-05,
+      "loss": 0.4759,
+      "step": 14588
+    },
+    {
+      "epoch": 18.67392,
+      "grad_norm": 1.0411763191223145,
+      "learning_rate": 2.0836334533813525e-05,
+      "loss": 0.5068,
+      "step": 14589
+    },
+    {
+      "epoch": 18.6752,
+      "grad_norm": 1.0571635961532593,
+      "learning_rate": 2.08343337334934e-05,
+      "loss": 0.5386,
+      "step": 14590
+    },
+    {
+      "epoch": 18.67648,
+      "grad_norm": 1.0805469751358032,
+      "learning_rate": 2.0832332933173272e-05,
+      "loss": 0.5289,
+      "step": 14591
+    },
+    {
+      "epoch": 18.67776,
+      "grad_norm": 1.01336669921875,
+      "learning_rate": 2.083033213285314e-05,
+      "loss": 0.4634,
+      "step": 14592
+    },
+    {
+      "epoch": 18.67904,
+      "grad_norm": 1.0346022844314575,
+      "learning_rate": 2.0828331332533012e-05,
+      "loss": 0.504,
+      "step": 14593
+    },
+    {
+      "epoch": 18.680320000000002,
+      "grad_norm": 1.0586795806884766,
+      "learning_rate": 2.0826330532212888e-05,
+      "loss": 0.4977,
+      "step": 14594
+    },
+    {
+      "epoch": 18.6816,
+      "grad_norm": 1.0512717962265015,
+      "learning_rate": 2.082432973189276e-05,
+      "loss": 0.4652,
+      "step": 14595
+    },
+    {
+      "epoch": 18.68288,
+      "grad_norm": 1.0366772413253784,
+      "learning_rate": 2.0822328931572628e-05,
+      "loss": 0.4688,
+      "step": 14596
+    },
+    {
+      "epoch": 18.68416,
+      "grad_norm": 1.096342921257019,
+      "learning_rate": 2.0820328131252503e-05,
+      "loss": 0.5036,
+      "step": 14597
+    },
+    {
+      "epoch": 18.68544,
+      "grad_norm": 1.0564435720443726,
+      "learning_rate": 2.0818327330932375e-05,
+      "loss": 0.4801,
+      "step": 14598
+    },
+    {
+      "epoch": 18.68672,
+      "grad_norm": 1.0218154191970825,
+      "learning_rate": 2.0816326530612247e-05,
+      "loss": 0.5102,
+      "step": 14599
+    },
+    {
+      "epoch": 18.688,
+      "grad_norm": 1.0529659986495972,
+      "learning_rate": 2.0814325730292115e-05,
+      "loss": 0.5007,
+      "step": 14600
+    },
+    {
+      "epoch": 18.68928,
+      "grad_norm": 1.0413590669631958,
+      "learning_rate": 2.081232492997199e-05,
+      "loss": 0.4678,
+      "step": 14601
+    },
+    {
+      "epoch": 18.69056,
+      "grad_norm": 1.0885825157165527,
+      "learning_rate": 2.0810324129651862e-05,
+      "loss": 0.5116,
+      "step": 14602
+    },
+    {
+      "epoch": 18.69184,
+      "grad_norm": 1.0246473550796509,
+      "learning_rate": 2.0808323329331734e-05,
+      "loss": 0.4818,
+      "step": 14603
+    },
+    {
+      "epoch": 18.69312,
+      "grad_norm": 1.1003787517547607,
+      "learning_rate": 2.0806322529011603e-05,
+      "loss": 0.5168,
+      "step": 14604
+    },
+    {
+      "epoch": 18.6944,
+      "grad_norm": 1.0684752464294434,
+      "learning_rate": 2.0804321728691478e-05,
+      "loss": 0.4637,
+      "step": 14605
+    },
+    {
+      "epoch": 18.69568,
+      "grad_norm": 1.1062880754470825,
+      "learning_rate": 2.080232092837135e-05,
+      "loss": 0.5151,
+      "step": 14606
+    },
+    {
+      "epoch": 18.69696,
+      "grad_norm": 1.0668532848358154,
+      "learning_rate": 2.080032012805122e-05,
+      "loss": 0.5465,
+      "step": 14607
+    },
+    {
+      "epoch": 18.69824,
+      "grad_norm": 1.0295997858047485,
+      "learning_rate": 2.0798319327731094e-05,
+      "loss": 0.4645,
+      "step": 14608
+    },
+    {
+      "epoch": 18.69952,
+      "grad_norm": 1.020206093788147,
+      "learning_rate": 2.0796318527410965e-05,
+      "loss": 0.4971,
+      "step": 14609
+    },
+    {
+      "epoch": 18.7008,
+      "grad_norm": 1.0235021114349365,
+      "learning_rate": 2.0794317727090837e-05,
+      "loss": 0.5198,
+      "step": 14610
+    },
+    {
+      "epoch": 18.70208,
+      "grad_norm": 1.0551282167434692,
+      "learning_rate": 2.079231692677071e-05,
+      "loss": 0.5083,
+      "step": 14611
+    },
+    {
+      "epoch": 18.70336,
+      "grad_norm": 1.0582573413848877,
+      "learning_rate": 2.079031612645058e-05,
+      "loss": 0.5274,
+      "step": 14612
+    },
+    {
+      "epoch": 18.70464,
+      "grad_norm": 1.0594959259033203,
+      "learning_rate": 2.0788315326130453e-05,
+      "loss": 0.4954,
+      "step": 14613
+    },
+    {
+      "epoch": 18.70592,
+      "grad_norm": 1.0503864288330078,
+      "learning_rate": 2.0786314525810325e-05,
+      "loss": 0.5453,
+      "step": 14614
+    },
+    {
+      "epoch": 18.7072,
+      "grad_norm": 1.0637823343276978,
+      "learning_rate": 2.0784313725490197e-05,
+      "loss": 0.5736,
+      "step": 14615
+    },
+    {
+      "epoch": 18.70848,
+      "grad_norm": 1.0056936740875244,
+      "learning_rate": 2.078231292517007e-05,
+      "loss": 0.4663,
+      "step": 14616
+    },
+    {
+      "epoch": 18.70976,
+      "grad_norm": 1.081607699394226,
+      "learning_rate": 2.078031212484994e-05,
+      "loss": 0.5333,
+      "step": 14617
+    },
+    {
+      "epoch": 18.71104,
+      "grad_norm": 1.1000036001205444,
+      "learning_rate": 2.0778311324529812e-05,
+      "loss": 0.5505,
+      "step": 14618
+    },
+    {
+      "epoch": 18.71232,
+      "grad_norm": 1.0323201417922974,
+      "learning_rate": 2.0776310524209684e-05,
+      "loss": 0.4896,
+      "step": 14619
+    },
+    {
+      "epoch": 18.7136,
+      "grad_norm": 1.0627686977386475,
+      "learning_rate": 2.0774309723889556e-05,
+      "loss": 0.5104,
+      "step": 14620
+    },
+    {
+      "epoch": 18.71488,
+      "grad_norm": 1.0364760160446167,
+      "learning_rate": 2.0772308923569428e-05,
+      "loss": 0.5054,
+      "step": 14621
+    },
+    {
+      "epoch": 18.71616,
+      "grad_norm": 1.0415117740631104,
+      "learning_rate": 2.0770308123249303e-05,
+      "loss": 0.4796,
+      "step": 14622
+    },
+    {
+      "epoch": 18.71744,
+      "grad_norm": 1.0772327184677124,
+      "learning_rate": 2.076830732292917e-05,
+      "loss": 0.527,
+      "step": 14623
+    },
+    {
+      "epoch": 18.71872,
+      "grad_norm": 1.140191674232483,
+      "learning_rate": 2.0766306522609043e-05,
+      "loss": 0.5161,
+      "step": 14624
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 1.1282105445861816,
+      "learning_rate": 2.0764305722288915e-05,
+      "loss": 0.5677,
+      "step": 14625
+    },
+    {
+      "epoch": 18.72128,
+      "grad_norm": 1.060404658317566,
+      "learning_rate": 2.076230492196879e-05,
+      "loss": 0.5116,
+      "step": 14626
+    },
+    {
+      "epoch": 18.72256,
+      "grad_norm": 1.0670077800750732,
+      "learning_rate": 2.076030412164866e-05,
+      "loss": 0.4953,
+      "step": 14627
+    },
+    {
+      "epoch": 18.72384,
+      "grad_norm": 1.036635398864746,
+      "learning_rate": 2.075830332132853e-05,
+      "loss": 0.4991,
+      "step": 14628
+    },
+    {
+      "epoch": 18.72512,
+      "grad_norm": 1.0683354139328003,
+      "learning_rate": 2.0756302521008406e-05,
+      "loss": 0.5511,
+      "step": 14629
+    },
+    {
+      "epoch": 18.7264,
+      "grad_norm": 1.094277024269104,
+      "learning_rate": 2.0754301720688278e-05,
+      "loss": 0.5233,
+      "step": 14630
+    },
+    {
+      "epoch": 18.72768,
+      "grad_norm": 1.093468189239502,
+      "learning_rate": 2.0752300920368146e-05,
+      "loss": 0.5616,
+      "step": 14631
+    },
+    {
+      "epoch": 18.72896,
+      "grad_norm": 1.0831431150436401,
+      "learning_rate": 2.0750300120048018e-05,
+      "loss": 0.5404,
+      "step": 14632
+    },
+    {
+      "epoch": 18.73024,
+      "grad_norm": 1.1620758771896362,
+      "learning_rate": 2.0748299319727893e-05,
+      "loss": 0.5539,
+      "step": 14633
+    },
+    {
+      "epoch": 18.73152,
+      "grad_norm": 1.077682614326477,
+      "learning_rate": 2.0746298519407765e-05,
+      "loss": 0.4842,
+      "step": 14634
+    },
+    {
+      "epoch": 18.7328,
+      "grad_norm": 1.0813980102539062,
+      "learning_rate": 2.0744297719087634e-05,
+      "loss": 0.4992,
+      "step": 14635
+    },
+    {
+      "epoch": 18.73408,
+      "grad_norm": 1.1097936630249023,
+      "learning_rate": 2.074229691876751e-05,
+      "loss": 0.569,
+      "step": 14636
+    },
+    {
+      "epoch": 18.73536,
+      "grad_norm": 1.0540353059768677,
+      "learning_rate": 2.074029611844738e-05,
+      "loss": 0.4809,
+      "step": 14637
+    },
+    {
+      "epoch": 18.73664,
+      "grad_norm": 1.0522795915603638,
+      "learning_rate": 2.0738295318127253e-05,
+      "loss": 0.5611,
+      "step": 14638
+    },
+    {
+      "epoch": 18.73792,
+      "grad_norm": 1.0796175003051758,
+      "learning_rate": 2.073629451780712e-05,
+      "loss": 0.522,
+      "step": 14639
+    },
+    {
+      "epoch": 18.7392,
+      "grad_norm": 1.0437400341033936,
+      "learning_rate": 2.0734293717486996e-05,
+      "loss": 0.503,
+      "step": 14640
+    },
+    {
+      "epoch": 18.74048,
+      "grad_norm": 1.0430114269256592,
+      "learning_rate": 2.0732292917166868e-05,
+      "loss": 0.4834,
+      "step": 14641
+    },
+    {
+      "epoch": 18.74176,
+      "grad_norm": 1.015177845954895,
+      "learning_rate": 2.073029211684674e-05,
+      "loss": 0.4529,
+      "step": 14642
+    },
+    {
+      "epoch": 18.74304,
+      "grad_norm": 1.0426579713821411,
+      "learning_rate": 2.0728291316526612e-05,
+      "loss": 0.4777,
+      "step": 14643
+    },
+    {
+      "epoch": 18.74432,
+      "grad_norm": 1.106436848640442,
+      "learning_rate": 2.0726290516206484e-05,
+      "loss": 0.5592,
+      "step": 14644
+    },
+    {
+      "epoch": 18.7456,
+      "grad_norm": 1.096053123474121,
+      "learning_rate": 2.0724289715886356e-05,
+      "loss": 0.5066,
+      "step": 14645
+    },
+    {
+      "epoch": 18.74688,
+      "grad_norm": 1.077257752418518,
+      "learning_rate": 2.0722288915566227e-05,
+      "loss": 0.4936,
+      "step": 14646
+    },
+    {
+      "epoch": 18.74816,
+      "grad_norm": 1.0427172183990479,
+      "learning_rate": 2.07202881152461e-05,
+      "loss": 0.5277,
+      "step": 14647
+    },
+    {
+      "epoch": 18.74944,
+      "grad_norm": 1.0899598598480225,
+      "learning_rate": 2.071828731492597e-05,
+      "loss": 0.5181,
+      "step": 14648
+    },
+    {
+      "epoch": 18.75072,
+      "grad_norm": 1.05521559715271,
+      "learning_rate": 2.0716286514605843e-05,
+      "loss": 0.5036,
+      "step": 14649
+    },
+    {
+      "epoch": 18.752,
+      "grad_norm": 1.0131468772888184,
+      "learning_rate": 2.0714285714285718e-05,
+      "loss": 0.4956,
+      "step": 14650
+    },
+    {
+      "epoch": 18.75328,
+      "grad_norm": 1.0845153331756592,
+      "learning_rate": 2.0712284913965587e-05,
+      "loss": 0.5164,
+      "step": 14651
+    },
+    {
+      "epoch": 18.75456,
+      "grad_norm": 1.0508294105529785,
+      "learning_rate": 2.071028411364546e-05,
+      "loss": 0.4958,
+      "step": 14652
+    },
+    {
+      "epoch": 18.75584,
+      "grad_norm": 1.0884485244750977,
+      "learning_rate": 2.070828331332533e-05,
+      "loss": 0.5257,
+      "step": 14653
+    },
+    {
+      "epoch": 18.75712,
+      "grad_norm": 1.0835343599319458,
+      "learning_rate": 2.0706282513005206e-05,
+      "loss": 0.5102,
+      "step": 14654
+    },
+    {
+      "epoch": 18.7584,
+      "grad_norm": 1.0350970029830933,
+      "learning_rate": 2.0704281712685074e-05,
+      "loss": 0.4823,
+      "step": 14655
+    },
+    {
+      "epoch": 18.75968,
+      "grad_norm": 1.0485219955444336,
+      "learning_rate": 2.0702280912364946e-05,
+      "loss": 0.4848,
+      "step": 14656
+    },
+    {
+      "epoch": 18.76096,
+      "grad_norm": 1.053740382194519,
+      "learning_rate": 2.0700280112044818e-05,
+      "loss": 0.5062,
+      "step": 14657
+    },
+    {
+      "epoch": 18.76224,
+      "grad_norm": 1.0165501832962036,
+      "learning_rate": 2.0698279311724693e-05,
+      "loss": 0.4833,
+      "step": 14658
+    },
+    {
+      "epoch": 18.76352,
+      "grad_norm": 1.0042489767074585,
+      "learning_rate": 2.069627851140456e-05,
+      "loss": 0.4763,
+      "step": 14659
+    },
+    {
+      "epoch": 18.7648,
+      "grad_norm": 1.03883695602417,
+      "learning_rate": 2.0694277711084433e-05,
+      "loss": 0.5079,
+      "step": 14660
+    },
+    {
+      "epoch": 18.76608,
+      "grad_norm": 1.0785133838653564,
+      "learning_rate": 2.069227691076431e-05,
+      "loss": 0.5352,
+      "step": 14661
+    },
+    {
+      "epoch": 18.76736,
+      "grad_norm": 1.0216038227081299,
+      "learning_rate": 2.069027611044418e-05,
+      "loss": 0.4591,
+      "step": 14662
+    },
+    {
+      "epoch": 18.76864,
+      "grad_norm": 1.0611780881881714,
+      "learning_rate": 2.068827531012405e-05,
+      "loss": 0.5309,
+      "step": 14663
+    },
+    {
+      "epoch": 18.76992,
+      "grad_norm": 1.055109977722168,
+      "learning_rate": 2.068627450980392e-05,
+      "loss": 0.5455,
+      "step": 14664
+    },
+    {
+      "epoch": 18.7712,
+      "grad_norm": 1.064369559288025,
+      "learning_rate": 2.0684273709483796e-05,
+      "loss": 0.5099,
+      "step": 14665
+    },
+    {
+      "epoch": 18.77248,
+      "grad_norm": 1.0237936973571777,
+      "learning_rate": 2.0682272909163668e-05,
+      "loss": 0.4916,
+      "step": 14666
+    },
+    {
+      "epoch": 18.77376,
+      "grad_norm": 1.1315503120422363,
+      "learning_rate": 2.0680272108843536e-05,
+      "loss": 0.5178,
+      "step": 14667
+    },
+    {
+      "epoch": 18.77504,
+      "grad_norm": 1.0587965250015259,
+      "learning_rate": 2.0678271308523412e-05,
+      "loss": 0.5022,
+      "step": 14668
+    },
+    {
+      "epoch": 18.77632,
+      "grad_norm": 1.0879102945327759,
+      "learning_rate": 2.0676270508203284e-05,
+      "loss": 0.4688,
+      "step": 14669
+    },
+    {
+      "epoch": 18.7776,
+      "grad_norm": 1.0429378747940063,
+      "learning_rate": 2.0674269707883155e-05,
+      "loss": 0.483,
+      "step": 14670
+    },
+    {
+      "epoch": 18.77888,
+      "grad_norm": 1.0255515575408936,
+      "learning_rate": 2.0672268907563024e-05,
+      "loss": 0.49,
+      "step": 14671
+    },
+    {
+      "epoch": 18.78016,
+      "grad_norm": 1.1541240215301514,
+      "learning_rate": 2.06702681072429e-05,
+      "loss": 0.5842,
+      "step": 14672
+    },
+    {
+      "epoch": 18.78144,
+      "grad_norm": 1.0616282224655151,
+      "learning_rate": 2.066826730692277e-05,
+      "loss": 0.4856,
+      "step": 14673
+    },
+    {
+      "epoch": 18.78272,
+      "grad_norm": 1.0617082118988037,
+      "learning_rate": 2.0666266506602643e-05,
+      "loss": 0.4908,
+      "step": 14674
+    },
+    {
+      "epoch": 18.784,
+      "grad_norm": 1.1243386268615723,
+      "learning_rate": 2.0664265706282515e-05,
+      "loss": 0.5216,
+      "step": 14675
+    },
+    {
+      "epoch": 18.78528,
+      "grad_norm": 1.103163480758667,
+      "learning_rate": 2.0662264905962387e-05,
+      "loss": 0.5228,
+      "step": 14676
+    },
+    {
+      "epoch": 18.78656,
+      "grad_norm": 1.0311192274093628,
+      "learning_rate": 2.066026410564226e-05,
+      "loss": 0.5204,
+      "step": 14677
+    },
+    {
+      "epoch": 18.78784,
+      "grad_norm": 1.1256968975067139,
+      "learning_rate": 2.065826330532213e-05,
+      "loss": 0.5192,
+      "step": 14678
+    },
+    {
+      "epoch": 18.78912,
+      "grad_norm": 1.0992389917373657,
+      "learning_rate": 2.0656262505002002e-05,
+      "loss": 0.5015,
+      "step": 14679
+    },
+    {
+      "epoch": 18.790399999999998,
+      "grad_norm": 1.0932484865188599,
+      "learning_rate": 2.0654261704681874e-05,
+      "loss": 0.5712,
+      "step": 14680
+    },
+    {
+      "epoch": 18.79168,
+      "grad_norm": 1.0706785917282104,
+      "learning_rate": 2.0652260904361746e-05,
+      "loss": 0.5222,
+      "step": 14681
+    },
+    {
+      "epoch": 18.79296,
+      "grad_norm": 1.0165241956710815,
+      "learning_rate": 2.0650260104041618e-05,
+      "loss": 0.4922,
+      "step": 14682
+    },
+    {
+      "epoch": 18.79424,
+      "grad_norm": 1.0056289434432983,
+      "learning_rate": 2.064825930372149e-05,
+      "loss": 0.4751,
+      "step": 14683
+    },
+    {
+      "epoch": 18.79552,
+      "grad_norm": 1.0656893253326416,
+      "learning_rate": 2.064625850340136e-05,
+      "loss": 0.4966,
+      "step": 14684
+    },
+    {
+      "epoch": 18.7968,
+      "grad_norm": 1.0573127269744873,
+      "learning_rate": 2.0644257703081233e-05,
+      "loss": 0.4845,
+      "step": 14685
+    },
+    {
+      "epoch": 18.79808,
+      "grad_norm": 1.0371816158294678,
+      "learning_rate": 2.0642256902761105e-05,
+      "loss": 0.4863,
+      "step": 14686
+    },
+    {
+      "epoch": 18.79936,
+      "grad_norm": 1.0891573429107666,
+      "learning_rate": 2.0640256102440977e-05,
+      "loss": 0.5613,
+      "step": 14687
+    },
+    {
+      "epoch": 18.80064,
+      "grad_norm": 1.053895115852356,
+      "learning_rate": 2.063825530212085e-05,
+      "loss": 0.5053,
+      "step": 14688
+    },
+    {
+      "epoch": 18.80192,
+      "grad_norm": 1.0640217065811157,
+      "learning_rate": 2.0636254501800724e-05,
+      "loss": 0.5254,
+      "step": 14689
+    },
+    {
+      "epoch": 18.8032,
+      "grad_norm": 1.1202378273010254,
+      "learning_rate": 2.0634253701480593e-05,
+      "loss": 0.5121,
+      "step": 14690
+    },
+    {
+      "epoch": 18.80448,
+      "grad_norm": 1.0097864866256714,
+      "learning_rate": 2.0632252901160464e-05,
+      "loss": 0.4796,
+      "step": 14691
+    },
+    {
+      "epoch": 18.80576,
+      "grad_norm": 1.1119987964630127,
+      "learning_rate": 2.0630252100840336e-05,
+      "loss": 0.5685,
+      "step": 14692
+    },
+    {
+      "epoch": 18.80704,
+      "grad_norm": 1.0607649087905884,
+      "learning_rate": 2.062825130052021e-05,
+      "loss": 0.519,
+      "step": 14693
+    },
+    {
+      "epoch": 18.80832,
+      "grad_norm": 1.0458481311798096,
+      "learning_rate": 2.062625050020008e-05,
+      "loss": 0.4971,
+      "step": 14694
+    },
+    {
+      "epoch": 18.8096,
+      "grad_norm": 1.0886461734771729,
+      "learning_rate": 2.0624249699879952e-05,
+      "loss": 0.5249,
+      "step": 14695
+    },
+    {
+      "epoch": 18.81088,
+      "grad_norm": 1.0727638006210327,
+      "learning_rate": 2.0622248899559827e-05,
+      "loss": 0.5262,
+      "step": 14696
+    },
+    {
+      "epoch": 18.81216,
+      "grad_norm": 1.00741446018219,
+      "learning_rate": 2.06202480992397e-05,
+      "loss": 0.5311,
+      "step": 14697
+    },
+    {
+      "epoch": 18.81344,
+      "grad_norm": 1.0804330110549927,
+      "learning_rate": 2.0618247298919567e-05,
+      "loss": 0.547,
+      "step": 14698
+    },
+    {
+      "epoch": 18.81472,
+      "grad_norm": 1.0355829000473022,
+      "learning_rate": 2.061624649859944e-05,
+      "loss": 0.5129,
+      "step": 14699
+    },
+    {
+      "epoch": 18.816,
+      "grad_norm": 1.0258690118789673,
+      "learning_rate": 2.0614245698279315e-05,
+      "loss": 0.4925,
+      "step": 14700
+    },
+    {
+      "epoch": 18.81728,
+      "grad_norm": 1.0901426076889038,
+      "learning_rate": 2.0612244897959186e-05,
+      "loss": 0.4997,
+      "step": 14701
+    },
+    {
+      "epoch": 18.81856,
+      "grad_norm": 1.072962760925293,
+      "learning_rate": 2.0610244097639055e-05,
+      "loss": 0.5183,
+      "step": 14702
+    },
+    {
+      "epoch": 18.81984,
+      "grad_norm": 1.004193902015686,
+      "learning_rate": 2.060824329731893e-05,
+      "loss": 0.5215,
+      "step": 14703
+    },
+    {
+      "epoch": 18.82112,
+      "grad_norm": 1.0526185035705566,
+      "learning_rate": 2.0606242496998802e-05,
+      "loss": 0.5038,
+      "step": 14704
+    },
+    {
+      "epoch": 18.822400000000002,
+      "grad_norm": 1.0493122339248657,
+      "learning_rate": 2.0604241696678674e-05,
+      "loss": 0.5235,
+      "step": 14705
+    },
+    {
+      "epoch": 18.82368,
+      "grad_norm": 0.9919229745864868,
+      "learning_rate": 2.0602240896358542e-05,
+      "loss": 0.496,
+      "step": 14706
+    },
+    {
+      "epoch": 18.82496,
+      "grad_norm": 1.027870535850525,
+      "learning_rate": 2.0600240096038418e-05,
+      "loss": 0.4837,
+      "step": 14707
+    },
+    {
+      "epoch": 18.82624,
+      "grad_norm": 1.0324792861938477,
+      "learning_rate": 2.059823929571829e-05,
+      "loss": 0.5041,
+      "step": 14708
+    },
+    {
+      "epoch": 18.82752,
+      "grad_norm": 1.0613089799880981,
+      "learning_rate": 2.059623849539816e-05,
+      "loss": 0.5293,
+      "step": 14709
+    },
+    {
+      "epoch": 18.8288,
+      "grad_norm": 1.0319715738296509,
+      "learning_rate": 2.0594237695078033e-05,
+      "loss": 0.51,
+      "step": 14710
+    },
+    {
+      "epoch": 18.83008,
+      "grad_norm": 1.0587095022201538,
+      "learning_rate": 2.0592236894757905e-05,
+      "loss": 0.5019,
+      "step": 14711
+    },
+    {
+      "epoch": 18.83136,
+      "grad_norm": 1.0132025480270386,
+      "learning_rate": 2.0590236094437777e-05,
+      "loss": 0.4797,
+      "step": 14712
+    },
+    {
+      "epoch": 18.83264,
+      "grad_norm": 0.9891220331192017,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 0.4877,
+      "step": 14713
+    },
+    {
+      "epoch": 18.83392,
+      "grad_norm": 1.0485026836395264,
+      "learning_rate": 2.058623449379752e-05,
+      "loss": 0.5021,
+      "step": 14714
+    },
+    {
+      "epoch": 18.8352,
+      "grad_norm": 1.0996410846710205,
+      "learning_rate": 2.0584233693477392e-05,
+      "loss": 0.5152,
+      "step": 14715
+    },
+    {
+      "epoch": 18.83648,
+      "grad_norm": 1.065560221672058,
+      "learning_rate": 2.0582232893157264e-05,
+      "loss": 0.5149,
+      "step": 14716
+    },
+    {
+      "epoch": 18.83776,
+      "grad_norm": 1.12687349319458,
+      "learning_rate": 2.0580232092837136e-05,
+      "loss": 0.5378,
+      "step": 14717
+    },
+    {
+      "epoch": 18.83904,
+      "grad_norm": 1.1263388395309448,
+      "learning_rate": 2.0578231292517008e-05,
+      "loss": 0.5573,
+      "step": 14718
+    },
+    {
+      "epoch": 18.84032,
+      "grad_norm": 1.0918395519256592,
+      "learning_rate": 2.057623049219688e-05,
+      "loss": 0.5072,
+      "step": 14719
+    },
+    {
+      "epoch": 18.8416,
+      "grad_norm": 1.0265675783157349,
+      "learning_rate": 2.057422969187675e-05,
+      "loss": 0.506,
+      "step": 14720
+    },
+    {
+      "epoch": 18.84288,
+      "grad_norm": 0.9929207563400269,
+      "learning_rate": 2.0572228891556623e-05,
+      "loss": 0.4728,
+      "step": 14721
+    },
+    {
+      "epoch": 18.84416,
+      "grad_norm": 1.0715802907943726,
+      "learning_rate": 2.0570228091236495e-05,
+      "loss": 0.4897,
+      "step": 14722
+    },
+    {
+      "epoch": 18.84544,
+      "grad_norm": 0.9983776211738586,
+      "learning_rate": 2.0568227290916367e-05,
+      "loss": 0.481,
+      "step": 14723
+    },
+    {
+      "epoch": 18.84672,
+      "grad_norm": 1.0175827741622925,
+      "learning_rate": 2.056622649059624e-05,
+      "loss": 0.5271,
+      "step": 14724
+    },
+    {
+      "epoch": 18.848,
+      "grad_norm": 1.0712765455245972,
+      "learning_rate": 2.056422569027611e-05,
+      "loss": 0.5686,
+      "step": 14725
+    },
+    {
+      "epoch": 18.84928,
+      "grad_norm": 1.0037692785263062,
+      "learning_rate": 2.0562224889955983e-05,
+      "loss": 0.4798,
+      "step": 14726
+    },
+    {
+      "epoch": 18.85056,
+      "grad_norm": 0.9978422522544861,
+      "learning_rate": 2.0560224089635855e-05,
+      "loss": 0.44,
+      "step": 14727
+    },
+    {
+      "epoch": 18.85184,
+      "grad_norm": 1.021209955215454,
+      "learning_rate": 2.055822328931573e-05,
+      "loss": 0.4934,
+      "step": 14728
+    },
+    {
+      "epoch": 18.85312,
+      "grad_norm": 1.0505516529083252,
+      "learning_rate": 2.05562224889956e-05,
+      "loss": 0.5043,
+      "step": 14729
+    },
+    {
+      "epoch": 18.8544,
+      "grad_norm": 1.0562058687210083,
+      "learning_rate": 2.055422168867547e-05,
+      "loss": 0.5147,
+      "step": 14730
+    },
+    {
+      "epoch": 18.85568,
+      "grad_norm": 1.0676435232162476,
+      "learning_rate": 2.0552220888355342e-05,
+      "loss": 0.5092,
+      "step": 14731
+    },
+    {
+      "epoch": 18.85696,
+      "grad_norm": 1.0419723987579346,
+      "learning_rate": 2.0550220088035217e-05,
+      "loss": 0.529,
+      "step": 14732
+    },
+    {
+      "epoch": 18.85824,
+      "grad_norm": 1.0583957433700562,
+      "learning_rate": 2.0548219287715086e-05,
+      "loss": 0.4971,
+      "step": 14733
+    },
+    {
+      "epoch": 18.85952,
+      "grad_norm": 1.0169494152069092,
+      "learning_rate": 2.0546218487394958e-05,
+      "loss": 0.508,
+      "step": 14734
+    },
+    {
+      "epoch": 18.8608,
+      "grad_norm": 1.019565224647522,
+      "learning_rate": 2.0544217687074833e-05,
+      "loss": 0.4782,
+      "step": 14735
+    },
+    {
+      "epoch": 18.86208,
+      "grad_norm": 1.087472915649414,
+      "learning_rate": 2.0542216886754705e-05,
+      "loss": 0.5106,
+      "step": 14736
+    },
+    {
+      "epoch": 18.86336,
+      "grad_norm": 1.0848655700683594,
+      "learning_rate": 2.0540216086434573e-05,
+      "loss": 0.4834,
+      "step": 14737
+    },
+    {
+      "epoch": 18.86464,
+      "grad_norm": 1.049041509628296,
+      "learning_rate": 2.0538215286114445e-05,
+      "loss": 0.5162,
+      "step": 14738
+    },
+    {
+      "epoch": 18.86592,
+      "grad_norm": 1.0465503931045532,
+      "learning_rate": 2.053621448579432e-05,
+      "loss": 0.5309,
+      "step": 14739
+    },
+    {
+      "epoch": 18.8672,
+      "grad_norm": 1.0960649251937866,
+      "learning_rate": 2.0534213685474192e-05,
+      "loss": 0.4926,
+      "step": 14740
+    },
+    {
+      "epoch": 18.86848,
+      "grad_norm": 1.0588141679763794,
+      "learning_rate": 2.053221288515406e-05,
+      "loss": 0.5142,
+      "step": 14741
+    },
+    {
+      "epoch": 18.86976,
+      "grad_norm": 1.0418102741241455,
+      "learning_rate": 2.0530212084833936e-05,
+      "loss": 0.4991,
+      "step": 14742
+    },
+    {
+      "epoch": 18.87104,
+      "grad_norm": 1.0588616132736206,
+      "learning_rate": 2.0528211284513808e-05,
+      "loss": 0.5113,
+      "step": 14743
+    },
+    {
+      "epoch": 18.87232,
+      "grad_norm": 1.0481520891189575,
+      "learning_rate": 2.052621048419368e-05,
+      "loss": 0.474,
+      "step": 14744
+    },
+    {
+      "epoch": 18.8736,
+      "grad_norm": 1.0571566820144653,
+      "learning_rate": 2.0524209683873548e-05,
+      "loss": 0.5072,
+      "step": 14745
+    },
+    {
+      "epoch": 18.87488,
+      "grad_norm": 1.0196195840835571,
+      "learning_rate": 2.0522208883553423e-05,
+      "loss": 0.535,
+      "step": 14746
+    },
+    {
+      "epoch": 18.87616,
+      "grad_norm": 1.1058493852615356,
+      "learning_rate": 2.0520208083233295e-05,
+      "loss": 0.5374,
+      "step": 14747
+    },
+    {
+      "epoch": 18.87744,
+      "grad_norm": 1.0693318843841553,
+      "learning_rate": 2.0518207282913167e-05,
+      "loss": 0.4748,
+      "step": 14748
+    },
+    {
+      "epoch": 18.87872,
+      "grad_norm": 1.0735576152801514,
+      "learning_rate": 2.051620648259304e-05,
+      "loss": 0.5005,
+      "step": 14749
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 1.0688385963439941,
+      "learning_rate": 2.051420568227291e-05,
+      "loss": 0.5158,
+      "step": 14750
+    },
+    {
+      "epoch": 18.88128,
+      "grad_norm": 1.0036412477493286,
+      "learning_rate": 2.0512204881952783e-05,
+      "loss": 0.5041,
+      "step": 14751
+    },
+    {
+      "epoch": 18.88256,
+      "grad_norm": 1.089450478553772,
+      "learning_rate": 2.0510204081632654e-05,
+      "loss": 0.5485,
+      "step": 14752
+    },
+    {
+      "epoch": 18.88384,
+      "grad_norm": 1.050765872001648,
+      "learning_rate": 2.0508203281312526e-05,
+      "loss": 0.496,
+      "step": 14753
+    },
+    {
+      "epoch": 18.88512,
+      "grad_norm": 1.0228477716445923,
+      "learning_rate": 2.0506202480992398e-05,
+      "loss": 0.5171,
+      "step": 14754
+    },
+    {
+      "epoch": 18.8864,
+      "grad_norm": 1.0970298051834106,
+      "learning_rate": 2.050420168067227e-05,
+      "loss": 0.5334,
+      "step": 14755
+    },
+    {
+      "epoch": 18.88768,
+      "grad_norm": 1.0475279092788696,
+      "learning_rate": 2.0502200880352142e-05,
+      "loss": 0.4951,
+      "step": 14756
+    },
+    {
+      "epoch": 18.88896,
+      "grad_norm": 1.0471172332763672,
+      "learning_rate": 2.0500200080032014e-05,
+      "loss": 0.5046,
+      "step": 14757
+    },
+    {
+      "epoch": 18.89024,
+      "grad_norm": 1.0250929594039917,
+      "learning_rate": 2.0498199279711886e-05,
+      "loss": 0.4712,
+      "step": 14758
+    },
+    {
+      "epoch": 18.89152,
+      "grad_norm": 1.0119768381118774,
+      "learning_rate": 2.0496198479391757e-05,
+      "loss": 0.5076,
+      "step": 14759
+    },
+    {
+      "epoch": 18.8928,
+      "grad_norm": 1.1097235679626465,
+      "learning_rate": 2.049419767907163e-05,
+      "loss": 0.5243,
+      "step": 14760
+    },
+    {
+      "epoch": 18.89408,
+      "grad_norm": 1.0308051109313965,
+      "learning_rate": 2.04921968787515e-05,
+      "loss": 0.5459,
+      "step": 14761
+    },
+    {
+      "epoch": 18.89536,
+      "grad_norm": 1.027444839477539,
+      "learning_rate": 2.0490196078431373e-05,
+      "loss": 0.532,
+      "step": 14762
+    },
+    {
+      "epoch": 18.89664,
+      "grad_norm": 1.0802733898162842,
+      "learning_rate": 2.0488195278111248e-05,
+      "loss": 0.5034,
+      "step": 14763
+    },
+    {
+      "epoch": 18.89792,
+      "grad_norm": 1.0943602323532104,
+      "learning_rate": 2.0486194477791117e-05,
+      "loss": 0.5108,
+      "step": 14764
+    },
+    {
+      "epoch": 18.8992,
+      "grad_norm": 1.0663763284683228,
+      "learning_rate": 2.048419367747099e-05,
+      "loss": 0.4789,
+      "step": 14765
+    },
+    {
+      "epoch": 18.90048,
+      "grad_norm": 1.0287508964538574,
+      "learning_rate": 2.048219287715086e-05,
+      "loss": 0.4964,
+      "step": 14766
+    },
+    {
+      "epoch": 18.90176,
+      "grad_norm": 1.0889275074005127,
+      "learning_rate": 2.0480192076830736e-05,
+      "loss": 0.513,
+      "step": 14767
+    },
+    {
+      "epoch": 18.90304,
+      "grad_norm": 1.0352449417114258,
+      "learning_rate": 2.0478191276510604e-05,
+      "loss": 0.4871,
+      "step": 14768
+    },
+    {
+      "epoch": 18.90432,
+      "grad_norm": 1.0806697607040405,
+      "learning_rate": 2.0476190476190476e-05,
+      "loss": 0.5069,
+      "step": 14769
+    },
+    {
+      "epoch": 18.9056,
+      "grad_norm": 1.0483494997024536,
+      "learning_rate": 2.0474189675870348e-05,
+      "loss": 0.493,
+      "step": 14770
+    },
+    {
+      "epoch": 18.90688,
+      "grad_norm": 0.9899161458015442,
+      "learning_rate": 2.0472188875550223e-05,
+      "loss": 0.4689,
+      "step": 14771
+    },
+    {
+      "epoch": 18.90816,
+      "grad_norm": 0.9991376399993896,
+      "learning_rate": 2.047018807523009e-05,
+      "loss": 0.441,
+      "step": 14772
+    },
+    {
+      "epoch": 18.90944,
+      "grad_norm": 1.0828230381011963,
+      "learning_rate": 2.0468187274909963e-05,
+      "loss": 0.4842,
+      "step": 14773
+    },
+    {
+      "epoch": 18.91072,
+      "grad_norm": 1.0319201946258545,
+      "learning_rate": 2.046618647458984e-05,
+      "loss": 0.4966,
+      "step": 14774
+    },
+    {
+      "epoch": 18.912,
+      "grad_norm": 1.1114946603775024,
+      "learning_rate": 2.046418567426971e-05,
+      "loss": 0.5483,
+      "step": 14775
+    },
+    {
+      "epoch": 18.91328,
+      "grad_norm": 1.1752065420150757,
+      "learning_rate": 2.046218487394958e-05,
+      "loss": 0.5939,
+      "step": 14776
+    },
+    {
+      "epoch": 18.91456,
+      "grad_norm": 1.0463483333587646,
+      "learning_rate": 2.046018407362945e-05,
+      "loss": 0.509,
+      "step": 14777
+    },
+    {
+      "epoch": 18.91584,
+      "grad_norm": 1.0437871217727661,
+      "learning_rate": 2.0458183273309326e-05,
+      "loss": 0.4781,
+      "step": 14778
+    },
+    {
+      "epoch": 18.91712,
+      "grad_norm": 1.054076910018921,
+      "learning_rate": 2.0456182472989198e-05,
+      "loss": 0.5121,
+      "step": 14779
+    },
+    {
+      "epoch": 18.9184,
+      "grad_norm": 1.0515813827514648,
+      "learning_rate": 2.0454181672669066e-05,
+      "loss": 0.5137,
+      "step": 14780
+    },
+    {
+      "epoch": 18.91968,
+      "grad_norm": 1.0603755712509155,
+      "learning_rate": 2.045218087234894e-05,
+      "loss": 0.4898,
+      "step": 14781
+    },
+    {
+      "epoch": 18.92096,
+      "grad_norm": 1.0575968027114868,
+      "learning_rate": 2.0450180072028814e-05,
+      "loss": 0.497,
+      "step": 14782
+    },
+    {
+      "epoch": 18.92224,
+      "grad_norm": 1.0913869142532349,
+      "learning_rate": 2.0448179271708685e-05,
+      "loss": 0.5125,
+      "step": 14783
+    },
+    {
+      "epoch": 18.92352,
+      "grad_norm": 1.018833041191101,
+      "learning_rate": 2.0446178471388554e-05,
+      "loss": 0.4733,
+      "step": 14784
+    },
+    {
+      "epoch": 18.9248,
+      "grad_norm": 1.0295926332473755,
+      "learning_rate": 2.044417767106843e-05,
+      "loss": 0.4817,
+      "step": 14785
+    },
+    {
+      "epoch": 18.92608,
+      "grad_norm": 1.0770633220672607,
+      "learning_rate": 2.04421768707483e-05,
+      "loss": 0.4856,
+      "step": 14786
+    },
+    {
+      "epoch": 18.92736,
+      "grad_norm": 1.0353119373321533,
+      "learning_rate": 2.0440176070428173e-05,
+      "loss": 0.5007,
+      "step": 14787
+    },
+    {
+      "epoch": 18.92864,
+      "grad_norm": 1.082660436630249,
+      "learning_rate": 2.0438175270108045e-05,
+      "loss": 0.5117,
+      "step": 14788
+    },
+    {
+      "epoch": 18.92992,
+      "grad_norm": 1.0934934616088867,
+      "learning_rate": 2.0436174469787917e-05,
+      "loss": 0.5001,
+      "step": 14789
+    },
+    {
+      "epoch": 18.9312,
+      "grad_norm": 1.0217633247375488,
+      "learning_rate": 2.043417366946779e-05,
+      "loss": 0.5105,
+      "step": 14790
+    },
+    {
+      "epoch": 18.932479999999998,
+      "grad_norm": 0.9955496788024902,
+      "learning_rate": 2.043217286914766e-05,
+      "loss": 0.4753,
+      "step": 14791
+    },
+    {
+      "epoch": 18.93376,
+      "grad_norm": 1.0870311260223389,
+      "learning_rate": 2.0430172068827532e-05,
+      "loss": 0.5214,
+      "step": 14792
+    },
+    {
+      "epoch": 18.93504,
+      "grad_norm": 1.0062708854675293,
+      "learning_rate": 2.0428171268507404e-05,
+      "loss": 0.471,
+      "step": 14793
+    },
+    {
+      "epoch": 18.93632,
+      "grad_norm": 1.0651724338531494,
+      "learning_rate": 2.0426170468187276e-05,
+      "loss": 0.5747,
+      "step": 14794
+    },
+    {
+      "epoch": 18.9376,
+      "grad_norm": 0.9868699908256531,
+      "learning_rate": 2.0424169667867148e-05,
+      "loss": 0.4606,
+      "step": 14795
+    },
+    {
+      "epoch": 18.93888,
+      "grad_norm": 1.0331077575683594,
+      "learning_rate": 2.042216886754702e-05,
+      "loss": 0.5139,
+      "step": 14796
+    },
+    {
+      "epoch": 18.94016,
+      "grad_norm": 1.0682381391525269,
+      "learning_rate": 2.042016806722689e-05,
+      "loss": 0.4899,
+      "step": 14797
+    },
+    {
+      "epoch": 18.94144,
+      "grad_norm": 1.1549947261810303,
+      "learning_rate": 2.0418167266906763e-05,
+      "loss": 0.5366,
+      "step": 14798
+    },
+    {
+      "epoch": 18.94272,
+      "grad_norm": 1.0490201711654663,
+      "learning_rate": 2.0416166466586635e-05,
+      "loss": 0.5038,
+      "step": 14799
+    },
+    {
+      "epoch": 18.944,
+      "grad_norm": 1.097759485244751,
+      "learning_rate": 2.0414165666266507e-05,
+      "loss": 0.5136,
+      "step": 14800
+    },
+    {
+      "epoch": 18.94528,
+      "grad_norm": 1.0571385622024536,
+      "learning_rate": 2.041216486594638e-05,
+      "loss": 0.4652,
+      "step": 14801
+    },
+    {
+      "epoch": 18.94656,
+      "grad_norm": 1.071684718132019,
+      "learning_rate": 2.0410164065626254e-05,
+      "loss": 0.5625,
+      "step": 14802
+    },
+    {
+      "epoch": 18.94784,
+      "grad_norm": 1.0939278602600098,
+      "learning_rate": 2.0408163265306123e-05,
+      "loss": 0.5401,
+      "step": 14803
+    },
+    {
+      "epoch": 18.94912,
+      "grad_norm": 1.045653223991394,
+      "learning_rate": 2.0406162464985994e-05,
+      "loss": 0.5289,
+      "step": 14804
+    },
+    {
+      "epoch": 18.9504,
+      "grad_norm": 1.070016622543335,
+      "learning_rate": 2.0404161664665866e-05,
+      "loss": 0.542,
+      "step": 14805
+    },
+    {
+      "epoch": 18.95168,
+      "grad_norm": 0.992372989654541,
+      "learning_rate": 2.040216086434574e-05,
+      "loss": 0.4731,
+      "step": 14806
+    },
+    {
+      "epoch": 18.95296,
+      "grad_norm": 1.062925100326538,
+      "learning_rate": 2.040016006402561e-05,
+      "loss": 0.5159,
+      "step": 14807
+    },
+    {
+      "epoch": 18.95424,
+      "grad_norm": 1.0661933422088623,
+      "learning_rate": 2.0398159263705482e-05,
+      "loss": 0.5155,
+      "step": 14808
+    },
+    {
+      "epoch": 18.95552,
+      "grad_norm": 1.0385395288467407,
+      "learning_rate": 2.0396158463385357e-05,
+      "loss": 0.4745,
+      "step": 14809
+    },
+    {
+      "epoch": 18.9568,
+      "grad_norm": 1.0102157592773438,
+      "learning_rate": 2.039415766306523e-05,
+      "loss": 0.4902,
+      "step": 14810
+    },
+    {
+      "epoch": 18.95808,
+      "grad_norm": 1.0179786682128906,
+      "learning_rate": 2.0392156862745097e-05,
+      "loss": 0.4641,
+      "step": 14811
+    },
+    {
+      "epoch": 18.95936,
+      "grad_norm": 1.062633991241455,
+      "learning_rate": 2.039015606242497e-05,
+      "loss": 0.5406,
+      "step": 14812
+    },
+    {
+      "epoch": 18.96064,
+      "grad_norm": 1.0476435422897339,
+      "learning_rate": 2.0388155262104844e-05,
+      "loss": 0.5176,
+      "step": 14813
+    },
+    {
+      "epoch": 18.96192,
+      "grad_norm": 1.0489333868026733,
+      "learning_rate": 2.0386154461784716e-05,
+      "loss": 0.5355,
+      "step": 14814
+    },
+    {
+      "epoch": 18.9632,
+      "grad_norm": 1.0147716999053955,
+      "learning_rate": 2.0384153661464585e-05,
+      "loss": 0.5249,
+      "step": 14815
+    },
+    {
+      "epoch": 18.964480000000002,
+      "grad_norm": 1.0759072303771973,
+      "learning_rate": 2.038215286114446e-05,
+      "loss": 0.4902,
+      "step": 14816
+    },
+    {
+      "epoch": 18.96576,
+      "grad_norm": 1.095579743385315,
+      "learning_rate": 2.0380152060824332e-05,
+      "loss": 0.5045,
+      "step": 14817
+    },
+    {
+      "epoch": 18.96704,
+      "grad_norm": 1.0580558776855469,
+      "learning_rate": 2.0378151260504204e-05,
+      "loss": 0.5302,
+      "step": 14818
+    },
+    {
+      "epoch": 18.96832,
+      "grad_norm": 1.0755257606506348,
+      "learning_rate": 2.0376150460184072e-05,
+      "loss": 0.5294,
+      "step": 14819
+    },
+    {
+      "epoch": 18.9696,
+      "grad_norm": 1.0983492136001587,
+      "learning_rate": 2.0374149659863947e-05,
+      "loss": 0.5071,
+      "step": 14820
+    },
+    {
+      "epoch": 18.97088,
+      "grad_norm": 1.050110936164856,
+      "learning_rate": 2.037214885954382e-05,
+      "loss": 0.4988,
+      "step": 14821
+    },
+    {
+      "epoch": 18.97216,
+      "grad_norm": 1.052294135093689,
+      "learning_rate": 2.037014805922369e-05,
+      "loss": 0.5414,
+      "step": 14822
+    },
+    {
+      "epoch": 18.97344,
+      "grad_norm": 1.1228139400482178,
+      "learning_rate": 2.036814725890356e-05,
+      "loss": 0.535,
+      "step": 14823
+    },
+    {
+      "epoch": 18.97472,
+      "grad_norm": 1.0272384881973267,
+      "learning_rate": 2.0366146458583435e-05,
+      "loss": 0.4597,
+      "step": 14824
+    },
+    {
+      "epoch": 18.976,
+      "grad_norm": 1.0405932664871216,
+      "learning_rate": 2.0364145658263307e-05,
+      "loss": 0.4597,
+      "step": 14825
+    },
+    {
+      "epoch": 18.97728,
+      "grad_norm": 1.1259437799453735,
+      "learning_rate": 2.036214485794318e-05,
+      "loss": 0.5336,
+      "step": 14826
+    },
+    {
+      "epoch": 18.97856,
+      "grad_norm": 1.0937503576278687,
+      "learning_rate": 2.036014405762305e-05,
+      "loss": 0.4824,
+      "step": 14827
+    },
+    {
+      "epoch": 18.97984,
+      "grad_norm": 1.0712480545043945,
+      "learning_rate": 2.0358143257302922e-05,
+      "loss": 0.5119,
+      "step": 14828
+    },
+    {
+      "epoch": 18.98112,
+      "grad_norm": 1.1139438152313232,
+      "learning_rate": 2.0356142456982794e-05,
+      "loss": 0.5282,
+      "step": 14829
+    },
+    {
+      "epoch": 18.9824,
+      "grad_norm": 1.103958010673523,
+      "learning_rate": 2.0354141656662666e-05,
+      "loss": 0.5595,
+      "step": 14830
+    },
+    {
+      "epoch": 18.98368,
+      "grad_norm": 1.0063518285751343,
+      "learning_rate": 2.0352140856342538e-05,
+      "loss": 0.4573,
+      "step": 14831
+    },
+    {
+      "epoch": 18.98496,
+      "grad_norm": 1.033522605895996,
+      "learning_rate": 2.035014005602241e-05,
+      "loss": 0.4847,
+      "step": 14832
+    },
+    {
+      "epoch": 18.98624,
+      "grad_norm": 1.0984948873519897,
+      "learning_rate": 2.034813925570228e-05,
+      "loss": 0.5332,
+      "step": 14833
+    },
+    {
+      "epoch": 18.98752,
+      "grad_norm": 1.1030117273330688,
+      "learning_rate": 2.0346138455382153e-05,
+      "loss": 0.538,
+      "step": 14834
+    },
+    {
+      "epoch": 18.9888,
+      "grad_norm": 1.097759485244751,
+      "learning_rate": 2.0344137655062025e-05,
+      "loss": 0.5191,
+      "step": 14835
+    },
+    {
+      "epoch": 18.99008,
+      "grad_norm": 1.111032485961914,
+      "learning_rate": 2.0342136854741897e-05,
+      "loss": 0.4796,
+      "step": 14836
+    },
+    {
+      "epoch": 18.99136,
+      "grad_norm": 1.1036945581436157,
+      "learning_rate": 2.034013605442177e-05,
+      "loss": 0.5219,
+      "step": 14837
+    },
+    {
+      "epoch": 18.99264,
+      "grad_norm": 1.1179360151290894,
+      "learning_rate": 2.033813525410164e-05,
+      "loss": 0.5537,
+      "step": 14838
+    },
+    {
+      "epoch": 18.99392,
+      "grad_norm": 1.048097848892212,
+      "learning_rate": 2.0336134453781513e-05,
+      "loss": 0.5053,
+      "step": 14839
+    },
+    {
+      "epoch": 18.9952,
+      "grad_norm": 1.0695916414260864,
+      "learning_rate": 2.0334133653461385e-05,
+      "loss": 0.5526,
+      "step": 14840
+    },
+    {
+      "epoch": 18.99648,
+      "grad_norm": 1.0661946535110474,
+      "learning_rate": 2.033213285314126e-05,
+      "loss": 0.5055,
+      "step": 14841
+    },
+    {
+      "epoch": 18.99776,
+      "grad_norm": 1.0300498008728027,
+      "learning_rate": 2.033013205282113e-05,
+      "loss": 0.4859,
+      "step": 14842
+    },
+    {
+      "epoch": 18.99904,
+      "grad_norm": 1.1209156513214111,
+      "learning_rate": 2.0328131252501e-05,
+      "loss": 0.5599,
+      "step": 14843
+    },
+    {
+      "epoch": 19.00032,
+      "grad_norm": Infinity,
+      "learning_rate": 2.0328131252501e-05,
+      "loss": 0.9375,
+      "step": 14844
+    },
+    {
+      "epoch": 19.0016,
+      "grad_norm": 0.9681119918823242,
+      "learning_rate": 2.0326130452180872e-05,
+      "loss": 0.501,
+      "step": 14845
+    },
+    {
+      "epoch": 19.00288,
+      "grad_norm": 1.0563879013061523,
+      "learning_rate": 2.0324129651860747e-05,
+      "loss": 0.4601,
+      "step": 14846
+    },
+    {
+      "epoch": 19.00416,
+      "grad_norm": 1.0467544794082642,
+      "learning_rate": 2.0322128851540616e-05,
+      "loss": 0.5335,
+      "step": 14847
+    },
+    {
+      "epoch": 19.00544,
+      "grad_norm": 1.031529188156128,
+      "learning_rate": 2.0320128051220488e-05,
+      "loss": 0.463,
+      "step": 14848
+    },
+    {
+      "epoch": 19.00672,
+      "grad_norm": 1.036137580871582,
+      "learning_rate": 2.0318127250900363e-05,
+      "loss": 0.4777,
+      "step": 14849
+    },
+    {
+      "epoch": 19.008,
+      "grad_norm": 1.0127391815185547,
+      "learning_rate": 2.0316126450580235e-05,
+      "loss": 0.4845,
+      "step": 14850
+    },
+    {
+      "epoch": 19.00928,
+      "grad_norm": 1.0757827758789062,
+      "learning_rate": 2.0314125650260103e-05,
+      "loss": 0.5228,
+      "step": 14851
+    },
+    {
+      "epoch": 19.01056,
+      "grad_norm": 1.035569190979004,
+      "learning_rate": 2.0312124849939975e-05,
+      "loss": 0.4674,
+      "step": 14852
+    },
+    {
+      "epoch": 19.01184,
+      "grad_norm": 1.0188312530517578,
+      "learning_rate": 2.031012404961985e-05,
+      "loss": 0.486,
+      "step": 14853
+    },
+    {
+      "epoch": 19.01312,
+      "grad_norm": 1.087680697441101,
+      "learning_rate": 2.0308123249299722e-05,
+      "loss": 0.4803,
+      "step": 14854
+    },
+    {
+      "epoch": 19.0144,
+      "grad_norm": 1.090846061706543,
+      "learning_rate": 2.030612244897959e-05,
+      "loss": 0.473,
+      "step": 14855
+    },
+    {
+      "epoch": 19.01568,
+      "grad_norm": 1.1319047212600708,
+      "learning_rate": 2.0304121648659466e-05,
+      "loss": 0.5024,
+      "step": 14856
+    },
+    {
+      "epoch": 19.01696,
+      "grad_norm": 1.1232948303222656,
+      "learning_rate": 2.0302120848339338e-05,
+      "loss": 0.5176,
+      "step": 14857
+    },
+    {
+      "epoch": 19.01824,
+      "grad_norm": 1.08295476436615,
+      "learning_rate": 2.030012004801921e-05,
+      "loss": 0.4821,
+      "step": 14858
+    },
+    {
+      "epoch": 19.01952,
+      "grad_norm": 1.0505188703536987,
+      "learning_rate": 2.0298119247699078e-05,
+      "loss": 0.4906,
+      "step": 14859
+    },
+    {
+      "epoch": 19.0208,
+      "grad_norm": 1.0271227359771729,
+      "learning_rate": 2.0296118447378953e-05,
+      "loss": 0.4796,
+      "step": 14860
+    },
+    {
+      "epoch": 19.02208,
+      "grad_norm": 1.0881240367889404,
+      "learning_rate": 2.0294117647058825e-05,
+      "loss": 0.4729,
+      "step": 14861
+    },
+    {
+      "epoch": 19.02336,
+      "grad_norm": 1.1086496114730835,
+      "learning_rate": 2.0292116846738697e-05,
+      "loss": 0.4996,
+      "step": 14862
+    },
+    {
+      "epoch": 19.02464,
+      "grad_norm": 1.087936520576477,
+      "learning_rate": 2.029011604641857e-05,
+      "loss": 0.4926,
+      "step": 14863
+    },
+    {
+      "epoch": 19.02592,
+      "grad_norm": 1.062474250793457,
+      "learning_rate": 2.028811524609844e-05,
+      "loss": 0.495,
+      "step": 14864
+    },
+    {
+      "epoch": 19.0272,
+      "grad_norm": 1.049436092376709,
+      "learning_rate": 2.0286114445778313e-05,
+      "loss": 0.4552,
+      "step": 14865
+    },
+    {
+      "epoch": 19.02848,
+      "grad_norm": 1.0556187629699707,
+      "learning_rate": 2.0284113645458184e-05,
+      "loss": 0.4789,
+      "step": 14866
+    },
+    {
+      "epoch": 19.02976,
+      "grad_norm": 0.9977188110351562,
+      "learning_rate": 2.0282112845138056e-05,
+      "loss": 0.4788,
+      "step": 14867
+    },
+    {
+      "epoch": 19.03104,
+      "grad_norm": 1.0484732389450073,
+      "learning_rate": 2.0280112044817928e-05,
+      "loss": 0.4596,
+      "step": 14868
+    },
+    {
+      "epoch": 19.03232,
+      "grad_norm": 1.135587453842163,
+      "learning_rate": 2.02781112444978e-05,
+      "loss": 0.4965,
+      "step": 14869
+    },
+    {
+      "epoch": 19.0336,
+      "grad_norm": 1.0908023118972778,
+      "learning_rate": 2.0276110444177672e-05,
+      "loss": 0.4776,
+      "step": 14870
+    },
+    {
+      "epoch": 19.03488,
+      "grad_norm": 1.1048251390457153,
+      "learning_rate": 2.0274109643857544e-05,
+      "loss": 0.5342,
+      "step": 14871
+    },
+    {
+      "epoch": 19.03616,
+      "grad_norm": 1.1822048425674438,
+      "learning_rate": 2.0272108843537416e-05,
+      "loss": 0.5458,
+      "step": 14872
+    },
+    {
+      "epoch": 19.03744,
+      "grad_norm": 1.0560060739517212,
+      "learning_rate": 2.0270108043217287e-05,
+      "loss": 0.477,
+      "step": 14873
+    },
+    {
+      "epoch": 19.03872,
+      "grad_norm": 1.0473270416259766,
+      "learning_rate": 2.026810724289716e-05,
+      "loss": 0.519,
+      "step": 14874
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 1.0593985319137573,
+      "learning_rate": 2.026610644257703e-05,
+      "loss": 0.4724,
+      "step": 14875
+    },
+    {
+      "epoch": 19.04128,
+      "grad_norm": 1.1364285945892334,
+      "learning_rate": 2.0264105642256903e-05,
+      "loss": 0.4902,
+      "step": 14876
+    },
+    {
+      "epoch": 19.04256,
+      "grad_norm": 1.1284868717193604,
+      "learning_rate": 2.0262104841936778e-05,
+      "loss": 0.5167,
+      "step": 14877
+    },
+    {
+      "epoch": 19.04384,
+      "grad_norm": 1.0828670263290405,
+      "learning_rate": 2.0260104041616647e-05,
+      "loss": 0.4799,
+      "step": 14878
+    },
+    {
+      "epoch": 19.04512,
+      "grad_norm": 1.0774503946304321,
+      "learning_rate": 2.025810324129652e-05,
+      "loss": 0.516,
+      "step": 14879
+    },
+    {
+      "epoch": 19.0464,
+      "grad_norm": 1.074471354484558,
+      "learning_rate": 2.025610244097639e-05,
+      "loss": 0.5001,
+      "step": 14880
+    },
+    {
+      "epoch": 19.04768,
+      "grad_norm": 1.0600991249084473,
+      "learning_rate": 2.0254101640656266e-05,
+      "loss": 0.475,
+      "step": 14881
+    },
+    {
+      "epoch": 19.04896,
+      "grad_norm": 1.0403833389282227,
+      "learning_rate": 2.0252100840336134e-05,
+      "loss": 0.4729,
+      "step": 14882
+    },
+    {
+      "epoch": 19.05024,
+      "grad_norm": 1.0715597867965698,
+      "learning_rate": 2.0250100040016006e-05,
+      "loss": 0.5526,
+      "step": 14883
+    },
+    {
+      "epoch": 19.05152,
+      "grad_norm": 1.0039360523223877,
+      "learning_rate": 2.0248099239695878e-05,
+      "loss": 0.4799,
+      "step": 14884
+    },
+    {
+      "epoch": 19.0528,
+      "grad_norm": 0.990351676940918,
+      "learning_rate": 2.0246098439375753e-05,
+      "loss": 0.4442,
+      "step": 14885
+    },
+    {
+      "epoch": 19.05408,
+      "grad_norm": 1.0488803386688232,
+      "learning_rate": 2.024409763905562e-05,
+      "loss": 0.4884,
+      "step": 14886
+    },
+    {
+      "epoch": 19.05536,
+      "grad_norm": 1.0708200931549072,
+      "learning_rate": 2.0242096838735493e-05,
+      "loss": 0.5011,
+      "step": 14887
+    },
+    {
+      "epoch": 19.05664,
+      "grad_norm": 1.0574580430984497,
+      "learning_rate": 2.024009603841537e-05,
+      "loss": 0.4919,
+      "step": 14888
+    },
+    {
+      "epoch": 19.05792,
+      "grad_norm": 1.0483922958374023,
+      "learning_rate": 2.023809523809524e-05,
+      "loss": 0.5121,
+      "step": 14889
+    },
+    {
+      "epoch": 19.0592,
+      "grad_norm": 1.039752721786499,
+      "learning_rate": 2.023609443777511e-05,
+      "loss": 0.4642,
+      "step": 14890
+    },
+    {
+      "epoch": 19.06048,
+      "grad_norm": 1.0242259502410889,
+      "learning_rate": 2.023409363745498e-05,
+      "loss": 0.4452,
+      "step": 14891
+    },
+    {
+      "epoch": 19.06176,
+      "grad_norm": 1.0347176790237427,
+      "learning_rate": 2.0232092837134856e-05,
+      "loss": 0.5101,
+      "step": 14892
+    },
+    {
+      "epoch": 19.06304,
+      "grad_norm": 1.0924898386001587,
+      "learning_rate": 2.0230092036814728e-05,
+      "loss": 0.4904,
+      "step": 14893
+    },
+    {
+      "epoch": 19.06432,
+      "grad_norm": 1.0558232069015503,
+      "learning_rate": 2.0228091236494596e-05,
+      "loss": 0.4689,
+      "step": 14894
+    },
+    {
+      "epoch": 19.0656,
+      "grad_norm": 1.082149863243103,
+      "learning_rate": 2.022609043617447e-05,
+      "loss": 0.5111,
+      "step": 14895
+    },
+    {
+      "epoch": 19.06688,
+      "grad_norm": 1.0433650016784668,
+      "learning_rate": 2.0224089635854344e-05,
+      "loss": 0.4458,
+      "step": 14896
+    },
+    {
+      "epoch": 19.06816,
+      "grad_norm": 1.053171157836914,
+      "learning_rate": 2.0222088835534215e-05,
+      "loss": 0.4724,
+      "step": 14897
+    },
+    {
+      "epoch": 19.06944,
+      "grad_norm": 1.0355948209762573,
+      "learning_rate": 2.0220088035214084e-05,
+      "loss": 0.5265,
+      "step": 14898
+    },
+    {
+      "epoch": 19.07072,
+      "grad_norm": 1.0293118953704834,
+      "learning_rate": 2.021808723489396e-05,
+      "loss": 0.4963,
+      "step": 14899
+    },
+    {
+      "epoch": 19.072,
+      "grad_norm": 1.065530776977539,
+      "learning_rate": 2.021608643457383e-05,
+      "loss": 0.4682,
+      "step": 14900
+    },
+    {
+      "epoch": 19.07328,
+      "grad_norm": 1.0934525728225708,
+      "learning_rate": 2.0214085634253703e-05,
+      "loss": 0.4689,
+      "step": 14901
+    },
+    {
+      "epoch": 19.07456,
+      "grad_norm": 0.997524082660675,
+      "learning_rate": 2.0212084833933575e-05,
+      "loss": 0.4693,
+      "step": 14902
+    },
+    {
+      "epoch": 19.07584,
+      "grad_norm": 1.1037408113479614,
+      "learning_rate": 2.0210084033613447e-05,
+      "loss": 0.5157,
+      "step": 14903
+    },
+    {
+      "epoch": 19.07712,
+      "grad_norm": 1.0279779434204102,
+      "learning_rate": 2.020808323329332e-05,
+      "loss": 0.4996,
+      "step": 14904
+    },
+    {
+      "epoch": 19.0784,
+      "grad_norm": 1.0512694120407104,
+      "learning_rate": 2.020608243297319e-05,
+      "loss": 0.4589,
+      "step": 14905
+    },
+    {
+      "epoch": 19.07968,
+      "grad_norm": 1.048998475074768,
+      "learning_rate": 2.0204081632653062e-05,
+      "loss": 0.4715,
+      "step": 14906
+    },
+    {
+      "epoch": 19.08096,
+      "grad_norm": 1.1140145063400269,
+      "learning_rate": 2.0202080832332934e-05,
+      "loss": 0.5188,
+      "step": 14907
+    },
+    {
+      "epoch": 19.08224,
+      "grad_norm": 1.1645013093948364,
+      "learning_rate": 2.0200080032012806e-05,
+      "loss": 0.5306,
+      "step": 14908
+    },
+    {
+      "epoch": 19.08352,
+      "grad_norm": 1.1035782098770142,
+      "learning_rate": 2.0198079231692678e-05,
+      "loss": 0.5419,
+      "step": 14909
+    },
+    {
+      "epoch": 19.0848,
+      "grad_norm": 1.1250030994415283,
+      "learning_rate": 2.019607843137255e-05,
+      "loss": 0.4853,
+      "step": 14910
+    },
+    {
+      "epoch": 19.08608,
+      "grad_norm": 1.019386887550354,
+      "learning_rate": 2.019407763105242e-05,
+      "loss": 0.4822,
+      "step": 14911
+    },
+    {
+      "epoch": 19.08736,
+      "grad_norm": 1.0403432846069336,
+      "learning_rate": 2.0192076830732293e-05,
+      "loss": 0.4773,
+      "step": 14912
+    },
+    {
+      "epoch": 19.08864,
+      "grad_norm": 1.073562741279602,
+      "learning_rate": 2.0190076030412165e-05,
+      "loss": 0.5333,
+      "step": 14913
+    },
+    {
+      "epoch": 19.08992,
+      "grad_norm": 1.0300809144973755,
+      "learning_rate": 2.0188075230092037e-05,
+      "loss": 0.4463,
+      "step": 14914
+    },
+    {
+      "epoch": 19.0912,
+      "grad_norm": 1.0957305431365967,
+      "learning_rate": 2.018607442977191e-05,
+      "loss": 0.5052,
+      "step": 14915
+    },
+    {
+      "epoch": 19.09248,
+      "grad_norm": 1.155129313468933,
+      "learning_rate": 2.0184073629451784e-05,
+      "loss": 0.5286,
+      "step": 14916
+    },
+    {
+      "epoch": 19.09376,
+      "grad_norm": 1.0918629169464111,
+      "learning_rate": 2.0182072829131653e-05,
+      "loss": 0.5577,
+      "step": 14917
+    },
+    {
+      "epoch": 19.09504,
+      "grad_norm": 1.0622458457946777,
+      "learning_rate": 2.0180072028811524e-05,
+      "loss": 0.4989,
+      "step": 14918
+    },
+    {
+      "epoch": 19.09632,
+      "grad_norm": 1.0885812044143677,
+      "learning_rate": 2.0178071228491396e-05,
+      "loss": 0.522,
+      "step": 14919
+    },
+    {
+      "epoch": 19.0976,
+      "grad_norm": 1.0788886547088623,
+      "learning_rate": 2.017607042817127e-05,
+      "loss": 0.5262,
+      "step": 14920
+    },
+    {
+      "epoch": 19.09888,
+      "grad_norm": 1.0732629299163818,
+      "learning_rate": 2.017406962785114e-05,
+      "loss": 0.4768,
+      "step": 14921
+    },
+    {
+      "epoch": 19.10016,
+      "grad_norm": 1.0360231399536133,
+      "learning_rate": 2.0172068827531012e-05,
+      "loss": 0.4685,
+      "step": 14922
+    },
+    {
+      "epoch": 19.10144,
+      "grad_norm": 1.0716302394866943,
+      "learning_rate": 2.0170068027210887e-05,
+      "loss": 0.497,
+      "step": 14923
+    },
+    {
+      "epoch": 19.10272,
+      "grad_norm": 1.0575581789016724,
+      "learning_rate": 2.016806722689076e-05,
+      "loss": 0.4984,
+      "step": 14924
+    },
+    {
+      "epoch": 19.104,
+      "grad_norm": 1.0178310871124268,
+      "learning_rate": 2.0166066426570627e-05,
+      "loss": 0.4571,
+      "step": 14925
+    },
+    {
+      "epoch": 19.10528,
+      "grad_norm": 1.0144866704940796,
+      "learning_rate": 2.01640656262505e-05,
+      "loss": 0.4566,
+      "step": 14926
+    },
+    {
+      "epoch": 19.10656,
+      "grad_norm": 1.0763506889343262,
+      "learning_rate": 2.0162064825930374e-05,
+      "loss": 0.4842,
+      "step": 14927
+    },
+    {
+      "epoch": 19.10784,
+      "grad_norm": 1.0165890455245972,
+      "learning_rate": 2.0160064025610246e-05,
+      "loss": 0.4678,
+      "step": 14928
+    },
+    {
+      "epoch": 19.10912,
+      "grad_norm": 1.0530028343200684,
+      "learning_rate": 2.0158063225290115e-05,
+      "loss": 0.4977,
+      "step": 14929
+    },
+    {
+      "epoch": 19.1104,
+      "grad_norm": 1.109286904335022,
+      "learning_rate": 2.015606242496999e-05,
+      "loss": 0.5079,
+      "step": 14930
+    },
+    {
+      "epoch": 19.11168,
+      "grad_norm": 1.081906795501709,
+      "learning_rate": 2.0154061624649862e-05,
+      "loss": 0.4783,
+      "step": 14931
+    },
+    {
+      "epoch": 19.11296,
+      "grad_norm": 1.0961225032806396,
+      "learning_rate": 2.0152060824329734e-05,
+      "loss": 0.5479,
+      "step": 14932
+    },
+    {
+      "epoch": 19.11424,
+      "grad_norm": 1.105312705039978,
+      "learning_rate": 2.0150060024009602e-05,
+      "loss": 0.5046,
+      "step": 14933
+    },
+    {
+      "epoch": 19.11552,
+      "grad_norm": 1.146582841873169,
+      "learning_rate": 2.0148059223689477e-05,
+      "loss": 0.5447,
+      "step": 14934
+    },
+    {
+      "epoch": 19.1168,
+      "grad_norm": 1.0598522424697876,
+      "learning_rate": 2.014605842336935e-05,
+      "loss": 0.4873,
+      "step": 14935
+    },
+    {
+      "epoch": 19.11808,
+      "grad_norm": 1.0884562730789185,
+      "learning_rate": 2.014405762304922e-05,
+      "loss": 0.49,
+      "step": 14936
+    },
+    {
+      "epoch": 19.11936,
+      "grad_norm": 1.1092275381088257,
+      "learning_rate": 2.014205682272909e-05,
+      "loss": 0.5157,
+      "step": 14937
+    },
+    {
+      "epoch": 19.12064,
+      "grad_norm": 1.0648193359375,
+      "learning_rate": 2.0140056022408965e-05,
+      "loss": 0.4773,
+      "step": 14938
+    },
+    {
+      "epoch": 19.12192,
+      "grad_norm": 1.0710915327072144,
+      "learning_rate": 2.0138055222088837e-05,
+      "loss": 0.5003,
+      "step": 14939
+    },
+    {
+      "epoch": 19.1232,
+      "grad_norm": 1.0918967723846436,
+      "learning_rate": 2.013605442176871e-05,
+      "loss": 0.4891,
+      "step": 14940
+    },
+    {
+      "epoch": 19.12448,
+      "grad_norm": 1.0287532806396484,
+      "learning_rate": 2.013405362144858e-05,
+      "loss": 0.4392,
+      "step": 14941
+    },
+    {
+      "epoch": 19.12576,
+      "grad_norm": 1.050429105758667,
+      "learning_rate": 2.0132052821128452e-05,
+      "loss": 0.4599,
+      "step": 14942
+    },
+    {
+      "epoch": 19.12704,
+      "grad_norm": 1.1075915098190308,
+      "learning_rate": 2.0130052020808324e-05,
+      "loss": 0.5304,
+      "step": 14943
+    },
+    {
+      "epoch": 19.12832,
+      "grad_norm": 1.040389060974121,
+      "learning_rate": 2.0128051220488196e-05,
+      "loss": 0.5047,
+      "step": 14944
+    },
+    {
+      "epoch": 19.1296,
+      "grad_norm": 1.0519245862960815,
+      "learning_rate": 2.0126050420168068e-05,
+      "loss": 0.4767,
+      "step": 14945
+    },
+    {
+      "epoch": 19.13088,
+      "grad_norm": 1.1048765182495117,
+      "learning_rate": 2.012404961984794e-05,
+      "loss": 0.5269,
+      "step": 14946
+    },
+    {
+      "epoch": 19.13216,
+      "grad_norm": 1.0859326124191284,
+      "learning_rate": 2.012204881952781e-05,
+      "loss": 0.5006,
+      "step": 14947
+    },
+    {
+      "epoch": 19.13344,
+      "grad_norm": 1.0915594100952148,
+      "learning_rate": 2.0120048019207683e-05,
+      "loss": 0.4768,
+      "step": 14948
+    },
+    {
+      "epoch": 19.13472,
+      "grad_norm": 1.1056619882583618,
+      "learning_rate": 2.0118047218887555e-05,
+      "loss": 0.5421,
+      "step": 14949
+    },
+    {
+      "epoch": 19.136,
+      "grad_norm": 1.124221920967102,
+      "learning_rate": 2.0116046418567427e-05,
+      "loss": 0.4826,
+      "step": 14950
+    },
+    {
+      "epoch": 19.13728,
+      "grad_norm": 1.0759928226470947,
+      "learning_rate": 2.01140456182473e-05,
+      "loss": 0.5218,
+      "step": 14951
+    },
+    {
+      "epoch": 19.13856,
+      "grad_norm": 1.0738402605056763,
+      "learning_rate": 2.011204481792717e-05,
+      "loss": 0.4804,
+      "step": 14952
+    },
+    {
+      "epoch": 19.13984,
+      "grad_norm": 1.070008397102356,
+      "learning_rate": 2.0110044017607043e-05,
+      "loss": 0.5434,
+      "step": 14953
+    },
+    {
+      "epoch": 19.14112,
+      "grad_norm": 1.1064351797103882,
+      "learning_rate": 2.0108043217286915e-05,
+      "loss": 0.5187,
+      "step": 14954
+    },
+    {
+      "epoch": 19.1424,
+      "grad_norm": 1.0281625986099243,
+      "learning_rate": 2.010604241696679e-05,
+      "loss": 0.4896,
+      "step": 14955
+    },
+    {
+      "epoch": 19.14368,
+      "grad_norm": 1.088550329208374,
+      "learning_rate": 2.010404161664666e-05,
+      "loss": 0.5228,
+      "step": 14956
+    },
+    {
+      "epoch": 19.14496,
+      "grad_norm": 1.1894739866256714,
+      "learning_rate": 2.010204081632653e-05,
+      "loss": 0.5813,
+      "step": 14957
+    },
+    {
+      "epoch": 19.14624,
+      "grad_norm": 1.1209919452667236,
+      "learning_rate": 2.0100040016006402e-05,
+      "loss": 0.483,
+      "step": 14958
+    },
+    {
+      "epoch": 19.14752,
+      "grad_norm": 1.055732011795044,
+      "learning_rate": 2.0098039215686277e-05,
+      "loss": 0.5229,
+      "step": 14959
+    },
+    {
+      "epoch": 19.1488,
+      "grad_norm": 1.0904511213302612,
+      "learning_rate": 2.0096038415366146e-05,
+      "loss": 0.521,
+      "step": 14960
+    },
+    {
+      "epoch": 19.15008,
+      "grad_norm": 1.0930554866790771,
+      "learning_rate": 2.0094037615046018e-05,
+      "loss": 0.4945,
+      "step": 14961
+    },
+    {
+      "epoch": 19.15136,
+      "grad_norm": 1.0355908870697021,
+      "learning_rate": 2.0092036814725893e-05,
+      "loss": 0.5065,
+      "step": 14962
+    },
+    {
+      "epoch": 19.15264,
+      "grad_norm": 1.0164462327957153,
+      "learning_rate": 2.0090036014405765e-05,
+      "loss": 0.4642,
+      "step": 14963
+    },
+    {
+      "epoch": 19.15392,
+      "grad_norm": 1.0890769958496094,
+      "learning_rate": 2.0088035214085633e-05,
+      "loss": 0.4984,
+      "step": 14964
+    },
+    {
+      "epoch": 19.1552,
+      "grad_norm": 1.0881518125534058,
+      "learning_rate": 2.0086034413765505e-05,
+      "loss": 0.5328,
+      "step": 14965
+    },
+    {
+      "epoch": 19.15648,
+      "grad_norm": 1.0909157991409302,
+      "learning_rate": 2.008403361344538e-05,
+      "loss": 0.4777,
+      "step": 14966
+    },
+    {
+      "epoch": 19.15776,
+      "grad_norm": 1.0730060338974,
+      "learning_rate": 2.0082032813125252e-05,
+      "loss": 0.4747,
+      "step": 14967
+    },
+    {
+      "epoch": 19.15904,
+      "grad_norm": 1.0788432359695435,
+      "learning_rate": 2.008003201280512e-05,
+      "loss": 0.4639,
+      "step": 14968
+    },
+    {
+      "epoch": 19.16032,
+      "grad_norm": 1.0355757474899292,
+      "learning_rate": 2.0078031212484996e-05,
+      "loss": 0.4812,
+      "step": 14969
+    },
+    {
+      "epoch": 19.1616,
+      "grad_norm": 1.1290099620819092,
+      "learning_rate": 2.0076030412164868e-05,
+      "loss": 0.5178,
+      "step": 14970
+    },
+    {
+      "epoch": 19.16288,
+      "grad_norm": 1.0541107654571533,
+      "learning_rate": 2.007402961184474e-05,
+      "loss": 0.4936,
+      "step": 14971
+    },
+    {
+      "epoch": 19.16416,
+      "grad_norm": 1.0474649667739868,
+      "learning_rate": 2.0072028811524608e-05,
+      "loss": 0.5262,
+      "step": 14972
+    },
+    {
+      "epoch": 19.16544,
+      "grad_norm": 1.0278681516647339,
+      "learning_rate": 2.0070028011204483e-05,
+      "loss": 0.4955,
+      "step": 14973
+    },
+    {
+      "epoch": 19.16672,
+      "grad_norm": 1.0625478029251099,
+      "learning_rate": 2.0068027210884355e-05,
+      "loss": 0.4842,
+      "step": 14974
+    },
+    {
+      "epoch": 19.168,
+      "grad_norm": 1.0970510244369507,
+      "learning_rate": 2.0066026410564227e-05,
+      "loss": 0.5195,
+      "step": 14975
+    },
+    {
+      "epoch": 19.16928,
+      "grad_norm": 1.0882428884506226,
+      "learning_rate": 2.00640256102441e-05,
+      "loss": 0.5226,
+      "step": 14976
+    },
+    {
+      "epoch": 19.17056,
+      "grad_norm": 1.0383604764938354,
+      "learning_rate": 2.006202480992397e-05,
+      "loss": 0.4514,
+      "step": 14977
+    },
+    {
+      "epoch": 19.17184,
+      "grad_norm": 1.1663484573364258,
+      "learning_rate": 2.0060024009603843e-05,
+      "loss": 0.5614,
+      "step": 14978
+    },
+    {
+      "epoch": 19.17312,
+      "grad_norm": 1.0208680629730225,
+      "learning_rate": 2.0058023209283714e-05,
+      "loss": 0.4491,
+      "step": 14979
+    },
+    {
+      "epoch": 19.1744,
+      "grad_norm": 1.0704641342163086,
+      "learning_rate": 2.0056022408963586e-05,
+      "loss": 0.5215,
+      "step": 14980
+    },
+    {
+      "epoch": 19.17568,
+      "grad_norm": 1.1021260023117065,
+      "learning_rate": 2.0054021608643458e-05,
+      "loss": 0.4645,
+      "step": 14981
+    },
+    {
+      "epoch": 19.17696,
+      "grad_norm": 1.0811761617660522,
+      "learning_rate": 2.005202080832333e-05,
+      "loss": 0.5263,
+      "step": 14982
+    },
+    {
+      "epoch": 19.17824,
+      "grad_norm": 1.1112900972366333,
+      "learning_rate": 2.0050020008003205e-05,
+      "loss": 0.4977,
+      "step": 14983
+    },
+    {
+      "epoch": 19.17952,
+      "grad_norm": 1.1169459819793701,
+      "learning_rate": 2.0048019207683074e-05,
+      "loss": 0.4697,
+      "step": 14984
+    },
+    {
+      "epoch": 19.1808,
+      "grad_norm": 1.056740164756775,
+      "learning_rate": 2.0046018407362946e-05,
+      "loss": 0.4776,
+      "step": 14985
+    },
+    {
+      "epoch": 19.18208,
+      "grad_norm": 1.0898573398590088,
+      "learning_rate": 2.0044017607042817e-05,
+      "loss": 0.5211,
+      "step": 14986
+    },
+    {
+      "epoch": 19.18336,
+      "grad_norm": 1.0396350622177124,
+      "learning_rate": 2.0042016806722693e-05,
+      "loss": 0.4809,
+      "step": 14987
+    },
+    {
+      "epoch": 19.18464,
+      "grad_norm": 1.1024669408798218,
+      "learning_rate": 2.004001600640256e-05,
+      "loss": 0.4961,
+      "step": 14988
+    },
+    {
+      "epoch": 19.18592,
+      "grad_norm": 1.1737849712371826,
+      "learning_rate": 2.0038015206082433e-05,
+      "loss": 0.5079,
+      "step": 14989
+    },
+    {
+      "epoch": 19.1872,
+      "grad_norm": 1.1294634342193604,
+      "learning_rate": 2.0036014405762308e-05,
+      "loss": 0.5342,
+      "step": 14990
+    },
+    {
+      "epoch": 19.18848,
+      "grad_norm": 1.021791696548462,
+      "learning_rate": 2.003401360544218e-05,
+      "loss": 0.497,
+      "step": 14991
+    },
+    {
+      "epoch": 19.18976,
+      "grad_norm": 1.0725295543670654,
+      "learning_rate": 2.003201280512205e-05,
+      "loss": 0.4671,
+      "step": 14992
+    },
+    {
+      "epoch": 19.19104,
+      "grad_norm": 1.098502278327942,
+      "learning_rate": 2.003001200480192e-05,
+      "loss": 0.5108,
+      "step": 14993
+    },
+    {
+      "epoch": 19.19232,
+      "grad_norm": 1.060045599937439,
+      "learning_rate": 2.0028011204481796e-05,
+      "loss": 0.4673,
+      "step": 14994
+    },
+    {
+      "epoch": 19.1936,
+      "grad_norm": 1.056288242340088,
+      "learning_rate": 2.0026010404161667e-05,
+      "loss": 0.4897,
+      "step": 14995
+    },
+    {
+      "epoch": 19.19488,
+      "grad_norm": 1.0819458961486816,
+      "learning_rate": 2.0024009603841536e-05,
+      "loss": 0.5011,
+      "step": 14996
+    },
+    {
+      "epoch": 19.19616,
+      "grad_norm": 1.0354195833206177,
+      "learning_rate": 2.0022008803521408e-05,
+      "loss": 0.4803,
+      "step": 14997
+    },
+    {
+      "epoch": 19.19744,
+      "grad_norm": 1.0913057327270508,
+      "learning_rate": 2.0020008003201283e-05,
+      "loss": 0.4986,
+      "step": 14998
+    },
+    {
+      "epoch": 19.19872,
+      "grad_norm": 1.101427674293518,
+      "learning_rate": 2.0018007202881155e-05,
+      "loss": 0.4715,
+      "step": 14999
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 0.9858012199401855,
+      "learning_rate": 2.0016006402561023e-05,
+      "loss": 0.4406,
+      "step": 15000
+    },
+    {
+      "epoch": 19.20128,
+      "grad_norm": 1.187526822090149,
+      "learning_rate": 2.00140056022409e-05,
+      "loss": 0.5555,
+      "step": 15001
+    },
+    {
+      "epoch": 19.20256,
+      "grad_norm": 1.087991714477539,
+      "learning_rate": 2.001200480192077e-05,
+      "loss": 0.5228,
+      "step": 15002
+    },
+    {
+      "epoch": 19.20384,
+      "grad_norm": 1.0227059125900269,
+      "learning_rate": 2.0010004001600642e-05,
+      "loss": 0.4722,
+      "step": 15003
+    },
+    {
+      "epoch": 19.20512,
+      "grad_norm": 1.0761233568191528,
+      "learning_rate": 2.000800320128051e-05,
+      "loss": 0.5027,
+      "step": 15004
+    },
+    {
+      "epoch": 19.2064,
+      "grad_norm": 1.0522879362106323,
+      "learning_rate": 2.0006002400960386e-05,
+      "loss": 0.5058,
+      "step": 15005
+    },
+    {
+      "epoch": 19.20768,
+      "grad_norm": 1.104921579360962,
+      "learning_rate": 2.0004001600640258e-05,
+      "loss": 0.4954,
+      "step": 15006
+    },
+    {
+      "epoch": 19.20896,
+      "grad_norm": 1.0269452333450317,
+      "learning_rate": 2.000200080032013e-05,
+      "loss": 0.4919,
+      "step": 15007
+    },
+    {
+      "epoch": 19.21024,
+      "grad_norm": 1.1063481569290161,
+      "learning_rate": 2e-05,
+      "loss": 0.5197,
+      "step": 15008
+    },
+    {
+      "epoch": 19.21152,
+      "grad_norm": 1.0551246404647827,
+      "learning_rate": 1.9997999199679873e-05,
+      "loss": 0.5092,
+      "step": 15009
+    },
+    {
+      "epoch": 19.2128,
+      "grad_norm": 1.0208076238632202,
+      "learning_rate": 1.9995998399359745e-05,
+      "loss": 0.4658,
+      "step": 15010
+    },
+    {
+      "epoch": 19.21408,
+      "grad_norm": 1.0514510869979858,
+      "learning_rate": 1.9993997599039617e-05,
+      "loss": 0.4883,
+      "step": 15011
+    },
+    {
+      "epoch": 19.21536,
+      "grad_norm": 1.0635521411895752,
+      "learning_rate": 1.999199679871949e-05,
+      "loss": 0.4738,
+      "step": 15012
+    },
+    {
+      "epoch": 19.21664,
+      "grad_norm": 1.0084950923919678,
+      "learning_rate": 1.998999599839936e-05,
+      "loss": 0.4461,
+      "step": 15013
+    },
+    {
+      "epoch": 19.21792,
+      "grad_norm": 1.0877736806869507,
+      "learning_rate": 1.9987995198079233e-05,
+      "loss": 0.5056,
+      "step": 15014
+    },
+    {
+      "epoch": 19.2192,
+      "grad_norm": 1.0358392000198364,
+      "learning_rate": 1.9985994397759105e-05,
+      "loss": 0.4948,
+      "step": 15015
+    },
+    {
+      "epoch": 19.22048,
+      "grad_norm": 1.0706851482391357,
+      "learning_rate": 1.9983993597438976e-05,
+      "loss": 0.5167,
+      "step": 15016
+    },
+    {
+      "epoch": 19.22176,
+      "grad_norm": 1.0379141569137573,
+      "learning_rate": 1.998199279711885e-05,
+      "loss": 0.5038,
+      "step": 15017
+    },
+    {
+      "epoch": 19.22304,
+      "grad_norm": 1.045020580291748,
+      "learning_rate": 1.997999199679872e-05,
+      "loss": 0.4624,
+      "step": 15018
+    },
+    {
+      "epoch": 19.22432,
+      "grad_norm": 1.0447577238082886,
+      "learning_rate": 1.9977991196478592e-05,
+      "loss": 0.494,
+      "step": 15019
+    },
+    {
+      "epoch": 19.2256,
+      "grad_norm": 1.057941198348999,
+      "learning_rate": 1.9975990396158464e-05,
+      "loss": 0.4793,
+      "step": 15020
+    },
+    {
+      "epoch": 19.22688,
+      "grad_norm": 1.0662115812301636,
+      "learning_rate": 1.9973989595838336e-05,
+      "loss": 0.4976,
+      "step": 15021
+    },
+    {
+      "epoch": 19.22816,
+      "grad_norm": 1.040855050086975,
+      "learning_rate": 1.997198879551821e-05,
+      "loss": 0.4681,
+      "step": 15022
+    },
+    {
+      "epoch": 19.22944,
+      "grad_norm": 1.0539730787277222,
+      "learning_rate": 1.996998799519808e-05,
+      "loss": 0.4807,
+      "step": 15023
+    },
+    {
+      "epoch": 19.23072,
+      "grad_norm": 1.0343507528305054,
+      "learning_rate": 1.996798719487795e-05,
+      "loss": 0.4931,
+      "step": 15024
+    },
+    {
+      "epoch": 19.232,
+      "grad_norm": 1.0538980960845947,
+      "learning_rate": 1.9965986394557823e-05,
+      "loss": 0.4921,
+      "step": 15025
+    },
+    {
+      "epoch": 19.23328,
+      "grad_norm": 1.0909112691879272,
+      "learning_rate": 1.99639855942377e-05,
+      "loss": 0.4804,
+      "step": 15026
+    },
+    {
+      "epoch": 19.23456,
+      "grad_norm": 0.9821348786354065,
+      "learning_rate": 1.9961984793917567e-05,
+      "loss": 0.445,
+      "step": 15027
+    },
+    {
+      "epoch": 19.23584,
+      "grad_norm": 0.9998717904090881,
+      "learning_rate": 1.995998399359744e-05,
+      "loss": 0.4519,
+      "step": 15028
+    },
+    {
+      "epoch": 19.23712,
+      "grad_norm": 1.1057769060134888,
+      "learning_rate": 1.9957983193277314e-05,
+      "loss": 0.5291,
+      "step": 15029
+    },
+    {
+      "epoch": 19.2384,
+      "grad_norm": 1.1260607242584229,
+      "learning_rate": 1.9955982392957186e-05,
+      "loss": 0.4936,
+      "step": 15030
+    },
+    {
+      "epoch": 19.23968,
+      "grad_norm": 1.0743900537490845,
+      "learning_rate": 1.9953981592637054e-05,
+      "loss": 0.5174,
+      "step": 15031
+    },
+    {
+      "epoch": 19.24096,
+      "grad_norm": 1.0485973358154297,
+      "learning_rate": 1.9951980792316926e-05,
+      "loss": 0.4953,
+      "step": 15032
+    },
+    {
+      "epoch": 19.24224,
+      "grad_norm": 1.104485273361206,
+      "learning_rate": 1.99499799919968e-05,
+      "loss": 0.5035,
+      "step": 15033
+    },
+    {
+      "epoch": 19.24352,
+      "grad_norm": 1.1098350286483765,
+      "learning_rate": 1.9947979191676673e-05,
+      "loss": 0.5428,
+      "step": 15034
+    },
+    {
+      "epoch": 19.2448,
+      "grad_norm": 1.1056627035140991,
+      "learning_rate": 1.9945978391356542e-05,
+      "loss": 0.5034,
+      "step": 15035
+    },
+    {
+      "epoch": 19.24608,
+      "grad_norm": 1.1753102540969849,
+      "learning_rate": 1.9943977591036417e-05,
+      "loss": 0.559,
+      "step": 15036
+    },
+    {
+      "epoch": 19.24736,
+      "grad_norm": 1.1483614444732666,
+      "learning_rate": 1.994197679071629e-05,
+      "loss": 0.5179,
+      "step": 15037
+    },
+    {
+      "epoch": 19.24864,
+      "grad_norm": 1.0651777982711792,
+      "learning_rate": 1.993997599039616e-05,
+      "loss": 0.4794,
+      "step": 15038
+    },
+    {
+      "epoch": 19.24992,
+      "grad_norm": 1.1044447422027588,
+      "learning_rate": 1.993797519007603e-05,
+      "loss": 0.4939,
+      "step": 15039
+    },
+    {
+      "epoch": 19.2512,
+      "grad_norm": 1.1543649435043335,
+      "learning_rate": 1.9935974389755904e-05,
+      "loss": 0.5337,
+      "step": 15040
+    },
+    {
+      "epoch": 19.25248,
+      "grad_norm": 1.0800881385803223,
+      "learning_rate": 1.9933973589435776e-05,
+      "loss": 0.4684,
+      "step": 15041
+    },
+    {
+      "epoch": 19.25376,
+      "grad_norm": 1.0753213167190552,
+      "learning_rate": 1.9931972789115648e-05,
+      "loss": 0.5192,
+      "step": 15042
+    },
+    {
+      "epoch": 19.25504,
+      "grad_norm": 1.0312221050262451,
+      "learning_rate": 1.992997198879552e-05,
+      "loss": 0.484,
+      "step": 15043
+    },
+    {
+      "epoch": 19.25632,
+      "grad_norm": 1.0647677183151245,
+      "learning_rate": 1.9927971188475392e-05,
+      "loss": 0.4847,
+      "step": 15044
+    },
+    {
+      "epoch": 19.2576,
+      "grad_norm": 1.090179443359375,
+      "learning_rate": 1.9925970388155264e-05,
+      "loss": 0.4858,
+      "step": 15045
+    },
+    {
+      "epoch": 19.25888,
+      "grad_norm": 1.055359125137329,
+      "learning_rate": 1.9923969587835136e-05,
+      "loss": 0.46,
+      "step": 15046
+    },
+    {
+      "epoch": 19.26016,
+      "grad_norm": 1.0518981218338013,
+      "learning_rate": 1.9921968787515007e-05,
+      "loss": 0.4574,
+      "step": 15047
+    },
+    {
+      "epoch": 19.26144,
+      "grad_norm": 1.1169776916503906,
+      "learning_rate": 1.991996798719488e-05,
+      "loss": 0.4914,
+      "step": 15048
+    },
+    {
+      "epoch": 19.26272,
+      "grad_norm": 1.026633620262146,
+      "learning_rate": 1.991796718687475e-05,
+      "loss": 0.4853,
+      "step": 15049
+    },
+    {
+      "epoch": 19.264,
+      "grad_norm": 1.0704096555709839,
+      "learning_rate": 1.9915966386554623e-05,
+      "loss": 0.5028,
+      "step": 15050
+    },
+    {
+      "epoch": 19.26528,
+      "grad_norm": 1.075821042060852,
+      "learning_rate": 1.9913965586234495e-05,
+      "loss": 0.4707,
+      "step": 15051
+    },
+    {
+      "epoch": 19.26656,
+      "grad_norm": 1.025315523147583,
+      "learning_rate": 1.9911964785914367e-05,
+      "loss": 0.4675,
+      "step": 15052
+    },
+    {
+      "epoch": 19.26784,
+      "grad_norm": 1.0183051824569702,
+      "learning_rate": 1.990996398559424e-05,
+      "loss": 0.4757,
+      "step": 15053
+    },
+    {
+      "epoch": 19.26912,
+      "grad_norm": 1.0708404779434204,
+      "learning_rate": 1.990796318527411e-05,
+      "loss": 0.5087,
+      "step": 15054
+    },
+    {
+      "epoch": 19.2704,
+      "grad_norm": 1.080012559890747,
+      "learning_rate": 1.9905962384953982e-05,
+      "loss": 0.5115,
+      "step": 15055
+    },
+    {
+      "epoch": 19.27168,
+      "grad_norm": 1.0551676750183105,
+      "learning_rate": 1.9903961584633854e-05,
+      "loss": 0.4924,
+      "step": 15056
+    },
+    {
+      "epoch": 19.27296,
+      "grad_norm": 1.0453039407730103,
+      "learning_rate": 1.9901960784313726e-05,
+      "loss": 0.501,
+      "step": 15057
+    },
+    {
+      "epoch": 19.27424,
+      "grad_norm": 1.0461045503616333,
+      "learning_rate": 1.9899959983993598e-05,
+      "loss": 0.4755,
+      "step": 15058
+    },
+    {
+      "epoch": 19.27552,
+      "grad_norm": 1.0412358045578003,
+      "learning_rate": 1.989795918367347e-05,
+      "loss": 0.4581,
+      "step": 15059
+    },
+    {
+      "epoch": 19.2768,
+      "grad_norm": 1.027969479560852,
+      "learning_rate": 1.989595838335334e-05,
+      "loss": 0.4646,
+      "step": 15060
+    },
+    {
+      "epoch": 19.27808,
+      "grad_norm": 1.1006345748901367,
+      "learning_rate": 1.9893957583033217e-05,
+      "loss": 0.4882,
+      "step": 15061
+    },
+    {
+      "epoch": 19.27936,
+      "grad_norm": 1.0730010271072388,
+      "learning_rate": 1.9891956782713085e-05,
+      "loss": 0.4754,
+      "step": 15062
+    },
+    {
+      "epoch": 19.28064,
+      "grad_norm": 0.998073399066925,
+      "learning_rate": 1.9889955982392957e-05,
+      "loss": 0.4791,
+      "step": 15063
+    },
+    {
+      "epoch": 19.28192,
+      "grad_norm": 1.0365452766418457,
+      "learning_rate": 1.988795518207283e-05,
+      "loss": 0.4787,
+      "step": 15064
+    },
+    {
+      "epoch": 19.2832,
+      "grad_norm": 1.0752921104431152,
+      "learning_rate": 1.9885954381752704e-05,
+      "loss": 0.5146,
+      "step": 15065
+    },
+    {
+      "epoch": 19.28448,
+      "grad_norm": 0.9972426295280457,
+      "learning_rate": 1.9883953581432573e-05,
+      "loss": 0.4596,
+      "step": 15066
+    },
+    {
+      "epoch": 19.28576,
+      "grad_norm": 1.1111432313919067,
+      "learning_rate": 1.9881952781112445e-05,
+      "loss": 0.4724,
+      "step": 15067
+    },
+    {
+      "epoch": 19.28704,
+      "grad_norm": 1.0094146728515625,
+      "learning_rate": 1.987995198079232e-05,
+      "loss": 0.4806,
+      "step": 15068
+    },
+    {
+      "epoch": 19.28832,
+      "grad_norm": 1.0732576847076416,
+      "learning_rate": 1.987795118047219e-05,
+      "loss": 0.4957,
+      "step": 15069
+    },
+    {
+      "epoch": 19.2896,
+      "grad_norm": 1.131595253944397,
+      "learning_rate": 1.987595038015206e-05,
+      "loss": 0.518,
+      "step": 15070
+    },
+    {
+      "epoch": 19.29088,
+      "grad_norm": 1.0386974811553955,
+      "learning_rate": 1.9873949579831932e-05,
+      "loss": 0.4767,
+      "step": 15071
+    },
+    {
+      "epoch": 19.29216,
+      "grad_norm": 1.0035613775253296,
+      "learning_rate": 1.9871948779511807e-05,
+      "loss": 0.4594,
+      "step": 15072
+    },
+    {
+      "epoch": 19.29344,
+      "grad_norm": 1.0777531862258911,
+      "learning_rate": 1.986994797919168e-05,
+      "loss": 0.4983,
+      "step": 15073
+    },
+    {
+      "epoch": 19.29472,
+      "grad_norm": 1.0419715642929077,
+      "learning_rate": 1.9867947178871548e-05,
+      "loss": 0.4667,
+      "step": 15074
+    },
+    {
+      "epoch": 19.296,
+      "grad_norm": 1.090698003768921,
+      "learning_rate": 1.9865946378551423e-05,
+      "loss": 0.5553,
+      "step": 15075
+    },
+    {
+      "epoch": 19.29728,
+      "grad_norm": 1.0526976585388184,
+      "learning_rate": 1.9863945578231295e-05,
+      "loss": 0.5112,
+      "step": 15076
+    },
+    {
+      "epoch": 19.29856,
+      "grad_norm": 1.0763030052185059,
+      "learning_rate": 1.9861944777911167e-05,
+      "loss": 0.4721,
+      "step": 15077
+    },
+    {
+      "epoch": 19.29984,
+      "grad_norm": 1.032845377922058,
+      "learning_rate": 1.9859943977591035e-05,
+      "loss": 0.5084,
+      "step": 15078
+    },
+    {
+      "epoch": 19.30112,
+      "grad_norm": 1.0398248434066772,
+      "learning_rate": 1.985794317727091e-05,
+      "loss": 0.4519,
+      "step": 15079
+    },
+    {
+      "epoch": 19.3024,
+      "grad_norm": 1.0678154230117798,
+      "learning_rate": 1.9855942376950782e-05,
+      "loss": 0.5367,
+      "step": 15080
+    },
+    {
+      "epoch": 19.30368,
+      "grad_norm": 1.0743778944015503,
+      "learning_rate": 1.9853941576630654e-05,
+      "loss": 0.4997,
+      "step": 15081
+    },
+    {
+      "epoch": 19.30496,
+      "grad_norm": 1.0013165473937988,
+      "learning_rate": 1.9851940776310526e-05,
+      "loss": 0.4498,
+      "step": 15082
+    },
+    {
+      "epoch": 19.30624,
+      "grad_norm": 1.052975058555603,
+      "learning_rate": 1.9849939975990398e-05,
+      "loss": 0.4922,
+      "step": 15083
+    },
+    {
+      "epoch": 19.30752,
+      "grad_norm": 1.0732128620147705,
+      "learning_rate": 1.984793917567027e-05,
+      "loss": 0.4821,
+      "step": 15084
+    },
+    {
+      "epoch": 19.3088,
+      "grad_norm": 1.0122027397155762,
+      "learning_rate": 1.984593837535014e-05,
+      "loss": 0.4702,
+      "step": 15085
+    },
+    {
+      "epoch": 19.31008,
+      "grad_norm": 1.1038622856140137,
+      "learning_rate": 1.9843937575030013e-05,
+      "loss": 0.5435,
+      "step": 15086
+    },
+    {
+      "epoch": 19.31136,
+      "grad_norm": 1.1267859935760498,
+      "learning_rate": 1.9841936774709885e-05,
+      "loss": 0.5358,
+      "step": 15087
+    },
+    {
+      "epoch": 19.31264,
+      "grad_norm": 1.109407663345337,
+      "learning_rate": 1.9839935974389757e-05,
+      "loss": 0.4993,
+      "step": 15088
+    },
+    {
+      "epoch": 19.31392,
+      "grad_norm": 1.0586917400360107,
+      "learning_rate": 1.983793517406963e-05,
+      "loss": 0.541,
+      "step": 15089
+    },
+    {
+      "epoch": 19.3152,
+      "grad_norm": 1.0636937618255615,
+      "learning_rate": 1.98359343737495e-05,
+      "loss": 0.5048,
+      "step": 15090
+    },
+    {
+      "epoch": 19.31648,
+      "grad_norm": 1.0607376098632812,
+      "learning_rate": 1.9833933573429373e-05,
+      "loss": 0.485,
+      "step": 15091
+    },
+    {
+      "epoch": 19.31776,
+      "grad_norm": 1.100722074508667,
+      "learning_rate": 1.9831932773109244e-05,
+      "loss": 0.5225,
+      "step": 15092
+    },
+    {
+      "epoch": 19.31904,
+      "grad_norm": 1.06232750415802,
+      "learning_rate": 1.9829931972789116e-05,
+      "loss": 0.5353,
+      "step": 15093
+    },
+    {
+      "epoch": 19.32032,
+      "grad_norm": 1.0231244564056396,
+      "learning_rate": 1.9827931172468988e-05,
+      "loss": 0.5138,
+      "step": 15094
+    },
+    {
+      "epoch": 19.3216,
+      "grad_norm": 1.1289072036743164,
+      "learning_rate": 1.982593037214886e-05,
+      "loss": 0.5095,
+      "step": 15095
+    },
+    {
+      "epoch": 19.32288,
+      "grad_norm": 1.0838286876678467,
+      "learning_rate": 1.9823929571828735e-05,
+      "loss": 0.5244,
+      "step": 15096
+    },
+    {
+      "epoch": 19.32416,
+      "grad_norm": 0.9947683811187744,
+      "learning_rate": 1.9821928771508604e-05,
+      "loss": 0.4371,
+      "step": 15097
+    },
+    {
+      "epoch": 19.32544,
+      "grad_norm": 0.9954429864883423,
+      "learning_rate": 1.9819927971188476e-05,
+      "loss": 0.4654,
+      "step": 15098
+    },
+    {
+      "epoch": 19.32672,
+      "grad_norm": 0.9967473149299622,
+      "learning_rate": 1.9817927170868347e-05,
+      "loss": 0.4414,
+      "step": 15099
+    },
+    {
+      "epoch": 19.328,
+      "grad_norm": 1.0796631574630737,
+      "learning_rate": 1.9815926370548223e-05,
+      "loss": 0.4671,
+      "step": 15100
+    },
+    {
+      "epoch": 19.32928,
+      "grad_norm": 1.1318116188049316,
+      "learning_rate": 1.981392557022809e-05,
+      "loss": 0.487,
+      "step": 15101
+    },
+    {
+      "epoch": 19.33056,
+      "grad_norm": 1.1136324405670166,
+      "learning_rate": 1.9811924769907963e-05,
+      "loss": 0.502,
+      "step": 15102
+    },
+    {
+      "epoch": 19.33184,
+      "grad_norm": 1.093737244606018,
+      "learning_rate": 1.9809923969587835e-05,
+      "loss": 0.5176,
+      "step": 15103
+    },
+    {
+      "epoch": 19.33312,
+      "grad_norm": 1.1789205074310303,
+      "learning_rate": 1.980792316926771e-05,
+      "loss": 0.494,
+      "step": 15104
+    },
+    {
+      "epoch": 19.3344,
+      "grad_norm": 1.1485955715179443,
+      "learning_rate": 1.980592236894758e-05,
+      "loss": 0.5493,
+      "step": 15105
+    },
+    {
+      "epoch": 19.33568,
+      "grad_norm": 1.1409004926681519,
+      "learning_rate": 1.980392156862745e-05,
+      "loss": 0.5134,
+      "step": 15106
+    },
+    {
+      "epoch": 19.33696,
+      "grad_norm": 1.1257143020629883,
+      "learning_rate": 1.9801920768307326e-05,
+      "loss": 0.5045,
+      "step": 15107
+    },
+    {
+      "epoch": 19.33824,
+      "grad_norm": 1.071676254272461,
+      "learning_rate": 1.9799919967987197e-05,
+      "loss": 0.4601,
+      "step": 15108
+    },
+    {
+      "epoch": 19.33952,
+      "grad_norm": 1.0796996355056763,
+      "learning_rate": 1.9797919167667066e-05,
+      "loss": 0.5009,
+      "step": 15109
+    },
+    {
+      "epoch": 19.3408,
+      "grad_norm": 1.1070489883422852,
+      "learning_rate": 1.9795918367346938e-05,
+      "loss": 0.5274,
+      "step": 15110
+    },
+    {
+      "epoch": 19.34208,
+      "grad_norm": 1.0755043029785156,
+      "learning_rate": 1.9793917567026813e-05,
+      "loss": 0.5438,
+      "step": 15111
+    },
+    {
+      "epoch": 19.34336,
+      "grad_norm": 1.077373743057251,
+      "learning_rate": 1.9791916766706685e-05,
+      "loss": 0.4636,
+      "step": 15112
+    },
+    {
+      "epoch": 19.34464,
+      "grad_norm": 1.0781618356704712,
+      "learning_rate": 1.9789915966386553e-05,
+      "loss": 0.5117,
+      "step": 15113
+    },
+    {
+      "epoch": 19.34592,
+      "grad_norm": 1.0397669076919556,
+      "learning_rate": 1.978791516606643e-05,
+      "loss": 0.4753,
+      "step": 15114
+    },
+    {
+      "epoch": 19.3472,
+      "grad_norm": 1.0098005533218384,
+      "learning_rate": 1.97859143657463e-05,
+      "loss": 0.4595,
+      "step": 15115
+    },
+    {
+      "epoch": 19.34848,
+      "grad_norm": 1.054352879524231,
+      "learning_rate": 1.9783913565426172e-05,
+      "loss": 0.5363,
+      "step": 15116
+    },
+    {
+      "epoch": 19.34976,
+      "grad_norm": 1.1141222715377808,
+      "learning_rate": 1.978191276510604e-05,
+      "loss": 0.5044,
+      "step": 15117
+    },
+    {
+      "epoch": 19.35104,
+      "grad_norm": 1.093037486076355,
+      "learning_rate": 1.9779911964785916e-05,
+      "loss": 0.5358,
+      "step": 15118
+    },
+    {
+      "epoch": 19.35232,
+      "grad_norm": 1.0386536121368408,
+      "learning_rate": 1.9777911164465788e-05,
+      "loss": 0.4794,
+      "step": 15119
+    },
+    {
+      "epoch": 19.3536,
+      "grad_norm": 1.0772889852523804,
+      "learning_rate": 1.977591036414566e-05,
+      "loss": 0.4993,
+      "step": 15120
+    },
+    {
+      "epoch": 19.35488,
+      "grad_norm": 1.0881242752075195,
+      "learning_rate": 1.977390956382553e-05,
+      "loss": 0.5044,
+      "step": 15121
+    },
+    {
+      "epoch": 19.35616,
+      "grad_norm": 1.0567618608474731,
+      "learning_rate": 1.9771908763505403e-05,
+      "loss": 0.497,
+      "step": 15122
+    },
+    {
+      "epoch": 19.35744,
+      "grad_norm": 1.0052014589309692,
+      "learning_rate": 1.9769907963185275e-05,
+      "loss": 0.4805,
+      "step": 15123
+    },
+    {
+      "epoch": 19.35872,
+      "grad_norm": 1.0756340026855469,
+      "learning_rate": 1.9767907162865147e-05,
+      "loss": 0.4962,
+      "step": 15124
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 1.0965667963027954,
+      "learning_rate": 1.976590636254502e-05,
+      "loss": 0.4453,
+      "step": 15125
+    },
+    {
+      "epoch": 19.36128,
+      "grad_norm": 1.1271300315856934,
+      "learning_rate": 1.976390556222489e-05,
+      "loss": 0.4642,
+      "step": 15126
+    },
+    {
+      "epoch": 19.36256,
+      "grad_norm": 1.1453865766525269,
+      "learning_rate": 1.9761904761904763e-05,
+      "loss": 0.5407,
+      "step": 15127
+    },
+    {
+      "epoch": 19.36384,
+      "grad_norm": 1.1263774633407593,
+      "learning_rate": 1.9759903961584635e-05,
+      "loss": 0.5207,
+      "step": 15128
+    },
+    {
+      "epoch": 19.36512,
+      "grad_norm": 1.1071971654891968,
+      "learning_rate": 1.9757903161264506e-05,
+      "loss": 0.5006,
+      "step": 15129
+    },
+    {
+      "epoch": 19.3664,
+      "grad_norm": 1.0227842330932617,
+      "learning_rate": 1.975590236094438e-05,
+      "loss": 0.4618,
+      "step": 15130
+    },
+    {
+      "epoch": 19.36768,
+      "grad_norm": 1.094109058380127,
+      "learning_rate": 1.975390156062425e-05,
+      "loss": 0.4808,
+      "step": 15131
+    },
+    {
+      "epoch": 19.36896,
+      "grad_norm": 1.1060868501663208,
+      "learning_rate": 1.9751900760304122e-05,
+      "loss": 0.4996,
+      "step": 15132
+    },
+    {
+      "epoch": 19.37024,
+      "grad_norm": 1.075696587562561,
+      "learning_rate": 1.9749899959983994e-05,
+      "loss": 0.5082,
+      "step": 15133
+    },
+    {
+      "epoch": 19.37152,
+      "grad_norm": 1.0798813104629517,
+      "learning_rate": 1.9747899159663866e-05,
+      "loss": 0.4954,
+      "step": 15134
+    },
+    {
+      "epoch": 19.3728,
+      "grad_norm": 1.1250864267349243,
+      "learning_rate": 1.974589835934374e-05,
+      "loss": 0.5539,
+      "step": 15135
+    },
+    {
+      "epoch": 19.37408,
+      "grad_norm": 1.0421397686004639,
+      "learning_rate": 1.974389755902361e-05,
+      "loss": 0.4999,
+      "step": 15136
+    },
+    {
+      "epoch": 19.37536,
+      "grad_norm": 1.0826746225357056,
+      "learning_rate": 1.974189675870348e-05,
+      "loss": 0.5566,
+      "step": 15137
+    },
+    {
+      "epoch": 19.37664,
+      "grad_norm": 1.078742504119873,
+      "learning_rate": 1.9739895958383353e-05,
+      "loss": 0.5022,
+      "step": 15138
+    },
+    {
+      "epoch": 19.37792,
+      "grad_norm": 1.0308053493499756,
+      "learning_rate": 1.973789515806323e-05,
+      "loss": 0.4586,
+      "step": 15139
+    },
+    {
+      "epoch": 19.3792,
+      "grad_norm": 1.0350420475006104,
+      "learning_rate": 1.9735894357743097e-05,
+      "loss": 0.5138,
+      "step": 15140
+    },
+    {
+      "epoch": 19.38048,
+      "grad_norm": 1.1027657985687256,
+      "learning_rate": 1.973389355742297e-05,
+      "loss": 0.538,
+      "step": 15141
+    },
+    {
+      "epoch": 19.38176,
+      "grad_norm": 1.0577800273895264,
+      "learning_rate": 1.9731892757102844e-05,
+      "loss": 0.522,
+      "step": 15142
+    },
+    {
+      "epoch": 19.38304,
+      "grad_norm": 1.043058156967163,
+      "learning_rate": 1.9729891956782716e-05,
+      "loss": 0.467,
+      "step": 15143
+    },
+    {
+      "epoch": 19.38432,
+      "grad_norm": 1.0484575033187866,
+      "learning_rate": 1.9727891156462584e-05,
+      "loss": 0.524,
+      "step": 15144
+    },
+    {
+      "epoch": 19.3856,
+      "grad_norm": 1.0222641229629517,
+      "learning_rate": 1.9725890356142456e-05,
+      "loss": 0.5035,
+      "step": 15145
+    },
+    {
+      "epoch": 19.38688,
+      "grad_norm": 0.9348340630531311,
+      "learning_rate": 1.972388955582233e-05,
+      "loss": 0.4127,
+      "step": 15146
+    },
+    {
+      "epoch": 19.38816,
+      "grad_norm": 1.0297220945358276,
+      "learning_rate": 1.9721888755502203e-05,
+      "loss": 0.4698,
+      "step": 15147
+    },
+    {
+      "epoch": 19.38944,
+      "grad_norm": 1.1119678020477295,
+      "learning_rate": 1.9719887955182072e-05,
+      "loss": 0.5864,
+      "step": 15148
+    },
+    {
+      "epoch": 19.39072,
+      "grad_norm": 1.0301072597503662,
+      "learning_rate": 1.9717887154861947e-05,
+      "loss": 0.4696,
+      "step": 15149
+    },
+    {
+      "epoch": 19.392,
+      "grad_norm": 1.0419156551361084,
+      "learning_rate": 1.971588635454182e-05,
+      "loss": 0.4809,
+      "step": 15150
+    },
+    {
+      "epoch": 19.39328,
+      "grad_norm": 1.1309643983840942,
+      "learning_rate": 1.971388555422169e-05,
+      "loss": 0.5239,
+      "step": 15151
+    },
+    {
+      "epoch": 19.39456,
+      "grad_norm": 1.0294731855392456,
+      "learning_rate": 1.971188475390156e-05,
+      "loss": 0.4653,
+      "step": 15152
+    },
+    {
+      "epoch": 19.39584,
+      "grad_norm": 1.0680971145629883,
+      "learning_rate": 1.9709883953581434e-05,
+      "loss": 0.5049,
+      "step": 15153
+    },
+    {
+      "epoch": 19.39712,
+      "grad_norm": 1.0883816480636597,
+      "learning_rate": 1.9707883153261306e-05,
+      "loss": 0.4632,
+      "step": 15154
+    },
+    {
+      "epoch": 19.3984,
+      "grad_norm": 1.0957615375518799,
+      "learning_rate": 1.9705882352941178e-05,
+      "loss": 0.4984,
+      "step": 15155
+    },
+    {
+      "epoch": 19.39968,
+      "grad_norm": 1.0659648180007935,
+      "learning_rate": 1.970388155262105e-05,
+      "loss": 0.5045,
+      "step": 15156
+    },
+    {
+      "epoch": 19.40096,
+      "grad_norm": 1.0844171047210693,
+      "learning_rate": 1.9701880752300922e-05,
+      "loss": 0.4732,
+      "step": 15157
+    },
+    {
+      "epoch": 19.40224,
+      "grad_norm": 1.1332968473434448,
+      "learning_rate": 1.9699879951980794e-05,
+      "loss": 0.5328,
+      "step": 15158
+    },
+    {
+      "epoch": 19.40352,
+      "grad_norm": 1.1378060579299927,
+      "learning_rate": 1.9697879151660666e-05,
+      "loss": 0.5104,
+      "step": 15159
+    },
+    {
+      "epoch": 19.4048,
+      "grad_norm": 1.137235164642334,
+      "learning_rate": 1.9695878351340537e-05,
+      "loss": 0.5345,
+      "step": 15160
+    },
+    {
+      "epoch": 19.40608,
+      "grad_norm": 1.0622361898422241,
+      "learning_rate": 1.969387755102041e-05,
+      "loss": 0.4613,
+      "step": 15161
+    },
+    {
+      "epoch": 19.40736,
+      "grad_norm": 1.0199161767959595,
+      "learning_rate": 1.969187675070028e-05,
+      "loss": 0.4496,
+      "step": 15162
+    },
+    {
+      "epoch": 19.40864,
+      "grad_norm": 1.075202226638794,
+      "learning_rate": 1.9689875950380153e-05,
+      "loss": 0.521,
+      "step": 15163
+    },
+    {
+      "epoch": 19.40992,
+      "grad_norm": 1.0620388984680176,
+      "learning_rate": 1.9687875150060025e-05,
+      "loss": 0.4758,
+      "step": 15164
+    },
+    {
+      "epoch": 19.4112,
+      "grad_norm": 1.073757529258728,
+      "learning_rate": 1.9685874349739897e-05,
+      "loss": 0.4867,
+      "step": 15165
+    },
+    {
+      "epoch": 19.41248,
+      "grad_norm": 1.063657522201538,
+      "learning_rate": 1.968387354941977e-05,
+      "loss": 0.5093,
+      "step": 15166
+    },
+    {
+      "epoch": 19.41376,
+      "grad_norm": 1.1278645992279053,
+      "learning_rate": 1.968187274909964e-05,
+      "loss": 0.4889,
+      "step": 15167
+    },
+    {
+      "epoch": 19.41504,
+      "grad_norm": 1.0807961225509644,
+      "learning_rate": 1.9679871948779512e-05,
+      "loss": 0.5224,
+      "step": 15168
+    },
+    {
+      "epoch": 19.41632,
+      "grad_norm": 1.0827388763427734,
+      "learning_rate": 1.9677871148459384e-05,
+      "loss": 0.5286,
+      "step": 15169
+    },
+    {
+      "epoch": 19.4176,
+      "grad_norm": 1.0594857931137085,
+      "learning_rate": 1.9675870348139256e-05,
+      "loss": 0.4851,
+      "step": 15170
+    },
+    {
+      "epoch": 19.41888,
+      "grad_norm": 1.0610501766204834,
+      "learning_rate": 1.9673869547819128e-05,
+      "loss": 0.4941,
+      "step": 15171
+    },
+    {
+      "epoch": 19.42016,
+      "grad_norm": 1.06877863407135,
+      "learning_rate": 1.9671868747499e-05,
+      "loss": 0.4593,
+      "step": 15172
+    },
+    {
+      "epoch": 19.42144,
+      "grad_norm": 1.0728821754455566,
+      "learning_rate": 1.966986794717887e-05,
+      "loss": 0.5086,
+      "step": 15173
+    },
+    {
+      "epoch": 19.422719999999998,
+      "grad_norm": 1.062900185585022,
+      "learning_rate": 1.9667867146858747e-05,
+      "loss": 0.4906,
+      "step": 15174
+    },
+    {
+      "epoch": 19.424,
+      "grad_norm": 1.0392909049987793,
+      "learning_rate": 1.9665866346538615e-05,
+      "loss": 0.4578,
+      "step": 15175
+    },
+    {
+      "epoch": 19.42528,
+      "grad_norm": 1.0775206089019775,
+      "learning_rate": 1.9663865546218487e-05,
+      "loss": 0.4938,
+      "step": 15176
+    },
+    {
+      "epoch": 19.42656,
+      "grad_norm": 1.0761048793792725,
+      "learning_rate": 1.966186474589836e-05,
+      "loss": 0.5025,
+      "step": 15177
+    },
+    {
+      "epoch": 19.42784,
+      "grad_norm": 1.1409354209899902,
+      "learning_rate": 1.9659863945578234e-05,
+      "loss": 0.545,
+      "step": 15178
+    },
+    {
+      "epoch": 19.42912,
+      "grad_norm": 1.0487301349639893,
+      "learning_rate": 1.9657863145258103e-05,
+      "loss": 0.4613,
+      "step": 15179
+    },
+    {
+      "epoch": 19.4304,
+      "grad_norm": 1.12277090549469,
+      "learning_rate": 1.9655862344937975e-05,
+      "loss": 0.5116,
+      "step": 15180
+    },
+    {
+      "epoch": 19.43168,
+      "grad_norm": 1.0349456071853638,
+      "learning_rate": 1.965386154461785e-05,
+      "loss": 0.4875,
+      "step": 15181
+    },
+    {
+      "epoch": 19.43296,
+      "grad_norm": 1.0730128288269043,
+      "learning_rate": 1.965186074429772e-05,
+      "loss": 0.4988,
+      "step": 15182
+    },
+    {
+      "epoch": 19.43424,
+      "grad_norm": 1.111451268196106,
+      "learning_rate": 1.964985994397759e-05,
+      "loss": 0.4995,
+      "step": 15183
+    },
+    {
+      "epoch": 19.43552,
+      "grad_norm": 1.0789837837219238,
+      "learning_rate": 1.9647859143657462e-05,
+      "loss": 0.5073,
+      "step": 15184
+    },
+    {
+      "epoch": 19.4368,
+      "grad_norm": 1.0748100280761719,
+      "learning_rate": 1.9645858343337337e-05,
+      "loss": 0.4898,
+      "step": 15185
+    },
+    {
+      "epoch": 19.43808,
+      "grad_norm": 1.031887173652649,
+      "learning_rate": 1.964385754301721e-05,
+      "loss": 0.4586,
+      "step": 15186
+    },
+    {
+      "epoch": 19.43936,
+      "grad_norm": 1.1339378356933594,
+      "learning_rate": 1.9641856742697078e-05,
+      "loss": 0.5464,
+      "step": 15187
+    },
+    {
+      "epoch": 19.44064,
+      "grad_norm": 1.0379207134246826,
+      "learning_rate": 1.9639855942376953e-05,
+      "loss": 0.4321,
+      "step": 15188
+    },
+    {
+      "epoch": 19.44192,
+      "grad_norm": 1.0902132987976074,
+      "learning_rate": 1.9637855142056825e-05,
+      "loss": 0.5081,
+      "step": 15189
+    },
+    {
+      "epoch": 19.4432,
+      "grad_norm": 1.0853605270385742,
+      "learning_rate": 1.9635854341736697e-05,
+      "loss": 0.4636,
+      "step": 15190
+    },
+    {
+      "epoch": 19.44448,
+      "grad_norm": 1.0282881259918213,
+      "learning_rate": 1.9633853541416565e-05,
+      "loss": 0.5072,
+      "step": 15191
+    },
+    {
+      "epoch": 19.44576,
+      "grad_norm": 1.1072465181350708,
+      "learning_rate": 1.963185274109644e-05,
+      "loss": 0.5032,
+      "step": 15192
+    },
+    {
+      "epoch": 19.44704,
+      "grad_norm": 1.0915943384170532,
+      "learning_rate": 1.9629851940776312e-05,
+      "loss": 0.5344,
+      "step": 15193
+    },
+    {
+      "epoch": 19.44832,
+      "grad_norm": 1.1000624895095825,
+      "learning_rate": 1.9627851140456184e-05,
+      "loss": 0.449,
+      "step": 15194
+    },
+    {
+      "epoch": 19.4496,
+      "grad_norm": 1.1032449007034302,
+      "learning_rate": 1.9625850340136056e-05,
+      "loss": 0.5057,
+      "step": 15195
+    },
+    {
+      "epoch": 19.45088,
+      "grad_norm": 1.0936930179595947,
+      "learning_rate": 1.9623849539815928e-05,
+      "loss": 0.5342,
+      "step": 15196
+    },
+    {
+      "epoch": 19.45216,
+      "grad_norm": 1.1406618356704712,
+      "learning_rate": 1.96218487394958e-05,
+      "loss": 0.514,
+      "step": 15197
+    },
+    {
+      "epoch": 19.45344,
+      "grad_norm": 1.1138516664505005,
+      "learning_rate": 1.961984793917567e-05,
+      "loss": 0.4917,
+      "step": 15198
+    },
+    {
+      "epoch": 19.454720000000002,
+      "grad_norm": 1.051370620727539,
+      "learning_rate": 1.9617847138855543e-05,
+      "loss": 0.497,
+      "step": 15199
+    },
+    {
+      "epoch": 19.456,
+      "grad_norm": 1.126387357711792,
+      "learning_rate": 1.9615846338535415e-05,
+      "loss": 0.5121,
+      "step": 15200
+    },
+    {
+      "epoch": 19.45728,
+      "grad_norm": 1.0682495832443237,
+      "learning_rate": 1.9613845538215287e-05,
+      "loss": 0.5013,
+      "step": 15201
+    },
+    {
+      "epoch": 19.45856,
+      "grad_norm": 1.1218750476837158,
+      "learning_rate": 1.961184473789516e-05,
+      "loss": 0.5307,
+      "step": 15202
+    },
+    {
+      "epoch": 19.45984,
+      "grad_norm": 1.0630134344100952,
+      "learning_rate": 1.960984393757503e-05,
+      "loss": 0.4964,
+      "step": 15203
+    },
+    {
+      "epoch": 19.46112,
+      "grad_norm": 1.0802773237228394,
+      "learning_rate": 1.9607843137254903e-05,
+      "loss": 0.493,
+      "step": 15204
+    },
+    {
+      "epoch": 19.4624,
+      "grad_norm": 1.061515212059021,
+      "learning_rate": 1.9605842336934774e-05,
+      "loss": 0.5217,
+      "step": 15205
+    },
+    {
+      "epoch": 19.46368,
+      "grad_norm": 1.0091540813446045,
+      "learning_rate": 1.9603841536614646e-05,
+      "loss": 0.4742,
+      "step": 15206
+    },
+    {
+      "epoch": 19.46496,
+      "grad_norm": 1.0679512023925781,
+      "learning_rate": 1.9601840736294518e-05,
+      "loss": 0.5139,
+      "step": 15207
+    },
+    {
+      "epoch": 19.46624,
+      "grad_norm": 1.019675850868225,
+      "learning_rate": 1.959983993597439e-05,
+      "loss": 0.4558,
+      "step": 15208
+    },
+    {
+      "epoch": 19.46752,
+      "grad_norm": 1.1162679195404053,
+      "learning_rate": 1.9597839135654265e-05,
+      "loss": 0.4691,
+      "step": 15209
+    },
+    {
+      "epoch": 19.4688,
+      "grad_norm": 1.0296701192855835,
+      "learning_rate": 1.9595838335334134e-05,
+      "loss": 0.4811,
+      "step": 15210
+    },
+    {
+      "epoch": 19.47008,
+      "grad_norm": 1.0854653120040894,
+      "learning_rate": 1.9593837535014005e-05,
+      "loss": 0.5192,
+      "step": 15211
+    },
+    {
+      "epoch": 19.47136,
+      "grad_norm": 1.088734745979309,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 0.507,
+      "step": 15212
+    },
+    {
+      "epoch": 19.47264,
+      "grad_norm": 1.077596664428711,
+      "learning_rate": 1.9589835934373753e-05,
+      "loss": 0.4915,
+      "step": 15213
+    },
+    {
+      "epoch": 19.47392,
+      "grad_norm": 1.1279915571212769,
+      "learning_rate": 1.958783513405362e-05,
+      "loss": 0.5085,
+      "step": 15214
+    },
+    {
+      "epoch": 19.4752,
+      "grad_norm": 1.0682718753814697,
+      "learning_rate": 1.9585834333733493e-05,
+      "loss": 0.4905,
+      "step": 15215
+    },
+    {
+      "epoch": 19.47648,
+      "grad_norm": 1.0742157697677612,
+      "learning_rate": 1.9583833533413365e-05,
+      "loss": 0.4806,
+      "step": 15216
+    },
+    {
+      "epoch": 19.47776,
+      "grad_norm": 1.042440414428711,
+      "learning_rate": 1.958183273309324e-05,
+      "loss": 0.5013,
+      "step": 15217
+    },
+    {
+      "epoch": 19.47904,
+      "grad_norm": 1.071058750152588,
+      "learning_rate": 1.957983193277311e-05,
+      "loss": 0.5188,
+      "step": 15218
+    },
+    {
+      "epoch": 19.48032,
+      "grad_norm": 1.0596333742141724,
+      "learning_rate": 1.957783113245298e-05,
+      "loss": 0.5147,
+      "step": 15219
+    },
+    {
+      "epoch": 19.4816,
+      "grad_norm": 1.0863871574401855,
+      "learning_rate": 1.9575830332132856e-05,
+      "loss": 0.5211,
+      "step": 15220
+    },
+    {
+      "epoch": 19.48288,
+      "grad_norm": 1.0600311756134033,
+      "learning_rate": 1.9573829531812727e-05,
+      "loss": 0.4686,
+      "step": 15221
+    },
+    {
+      "epoch": 19.48416,
+      "grad_norm": 1.0868990421295166,
+      "learning_rate": 1.9571828731492596e-05,
+      "loss": 0.5152,
+      "step": 15222
+    },
+    {
+      "epoch": 19.48544,
+      "grad_norm": 1.065700888633728,
+      "learning_rate": 1.9569827931172468e-05,
+      "loss": 0.4934,
+      "step": 15223
+    },
+    {
+      "epoch": 19.48672,
+      "grad_norm": 1.0733686685562134,
+      "learning_rate": 1.9567827130852343e-05,
+      "loss": 0.4662,
+      "step": 15224
+    },
+    {
+      "epoch": 19.488,
+      "grad_norm": 1.0808594226837158,
+      "learning_rate": 1.9565826330532215e-05,
+      "loss": 0.497,
+      "step": 15225
+    },
+    {
+      "epoch": 19.48928,
+      "grad_norm": 1.066211223602295,
+      "learning_rate": 1.9563825530212083e-05,
+      "loss": 0.549,
+      "step": 15226
+    },
+    {
+      "epoch": 19.49056,
+      "grad_norm": 1.0072721242904663,
+      "learning_rate": 1.956182472989196e-05,
+      "loss": 0.4312,
+      "step": 15227
+    },
+    {
+      "epoch": 19.49184,
+      "grad_norm": 1.107842206954956,
+      "learning_rate": 1.955982392957183e-05,
+      "loss": 0.4793,
+      "step": 15228
+    },
+    {
+      "epoch": 19.49312,
+      "grad_norm": 1.1411617994308472,
+      "learning_rate": 1.9557823129251702e-05,
+      "loss": 0.534,
+      "step": 15229
+    },
+    {
+      "epoch": 19.4944,
+      "grad_norm": 1.0612369775772095,
+      "learning_rate": 1.955582232893157e-05,
+      "loss": 0.5087,
+      "step": 15230
+    },
+    {
+      "epoch": 19.49568,
+      "grad_norm": 1.0148245096206665,
+      "learning_rate": 1.9553821528611446e-05,
+      "loss": 0.4518,
+      "step": 15231
+    },
+    {
+      "epoch": 19.49696,
+      "grad_norm": 1.133119821548462,
+      "learning_rate": 1.9551820728291318e-05,
+      "loss": 0.5301,
+      "step": 15232
+    },
+    {
+      "epoch": 19.49824,
+      "grad_norm": 1.0392405986785889,
+      "learning_rate": 1.954981992797119e-05,
+      "loss": 0.4912,
+      "step": 15233
+    },
+    {
+      "epoch": 19.49952,
+      "grad_norm": 1.0500421524047852,
+      "learning_rate": 1.954781912765106e-05,
+      "loss": 0.4866,
+      "step": 15234
+    },
+    {
+      "epoch": 19.5008,
+      "grad_norm": 1.0911407470703125,
+      "learning_rate": 1.9545818327330933e-05,
+      "loss": 0.5114,
+      "step": 15235
+    },
+    {
+      "epoch": 19.50208,
+      "grad_norm": 1.0419782400131226,
+      "learning_rate": 1.9543817527010805e-05,
+      "loss": 0.4925,
+      "step": 15236
+    },
+    {
+      "epoch": 19.50336,
+      "grad_norm": 1.1287118196487427,
+      "learning_rate": 1.9541816726690677e-05,
+      "loss": 0.5297,
+      "step": 15237
+    },
+    {
+      "epoch": 19.50464,
+      "grad_norm": 1.0417295694351196,
+      "learning_rate": 1.953981592637055e-05,
+      "loss": 0.4834,
+      "step": 15238
+    },
+    {
+      "epoch": 19.50592,
+      "grad_norm": 1.098235011100769,
+      "learning_rate": 1.953781512605042e-05,
+      "loss": 0.5075,
+      "step": 15239
+    },
+    {
+      "epoch": 19.5072,
+      "grad_norm": 1.0871201753616333,
+      "learning_rate": 1.9535814325730293e-05,
+      "loss": 0.4985,
+      "step": 15240
+    },
+    {
+      "epoch": 19.50848,
+      "grad_norm": 1.1194583177566528,
+      "learning_rate": 1.9533813525410165e-05,
+      "loss": 0.5293,
+      "step": 15241
+    },
+    {
+      "epoch": 19.50976,
+      "grad_norm": 1.0736286640167236,
+      "learning_rate": 1.9531812725090036e-05,
+      "loss": 0.488,
+      "step": 15242
+    },
+    {
+      "epoch": 19.51104,
+      "grad_norm": 1.0650372505187988,
+      "learning_rate": 1.9529811924769908e-05,
+      "loss": 0.4828,
+      "step": 15243
+    },
+    {
+      "epoch": 19.51232,
+      "grad_norm": 1.0762523412704468,
+      "learning_rate": 1.952781112444978e-05,
+      "loss": 0.5154,
+      "step": 15244
+    },
+    {
+      "epoch": 19.5136,
+      "grad_norm": 1.1131715774536133,
+      "learning_rate": 1.9525810324129652e-05,
+      "loss": 0.4885,
+      "step": 15245
+    },
+    {
+      "epoch": 19.51488,
+      "grad_norm": 1.0484470129013062,
+      "learning_rate": 1.9523809523809524e-05,
+      "loss": 0.4878,
+      "step": 15246
+    },
+    {
+      "epoch": 19.51616,
+      "grad_norm": 1.0681511163711548,
+      "learning_rate": 1.9521808723489396e-05,
+      "loss": 0.5208,
+      "step": 15247
+    },
+    {
+      "epoch": 19.51744,
+      "grad_norm": 1.0821027755737305,
+      "learning_rate": 1.951980792316927e-05,
+      "loss": 0.53,
+      "step": 15248
+    },
+    {
+      "epoch": 19.51872,
+      "grad_norm": 1.14222252368927,
+      "learning_rate": 1.951780712284914e-05,
+      "loss": 0.5255,
+      "step": 15249
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 1.0315016508102417,
+      "learning_rate": 1.951580632252901e-05,
+      "loss": 0.4661,
+      "step": 15250
+    },
+    {
+      "epoch": 19.52128,
+      "grad_norm": 1.0619421005249023,
+      "learning_rate": 1.9513805522208883e-05,
+      "loss": 0.5088,
+      "step": 15251
+    },
+    {
+      "epoch": 19.52256,
+      "grad_norm": 1.1743899583816528,
+      "learning_rate": 1.951180472188876e-05,
+      "loss": 0.5224,
+      "step": 15252
+    },
+    {
+      "epoch": 19.52384,
+      "grad_norm": 1.0858947038650513,
+      "learning_rate": 1.9509803921568627e-05,
+      "loss": 0.4706,
+      "step": 15253
+    },
+    {
+      "epoch": 19.52512,
+      "grad_norm": 1.1015316247940063,
+      "learning_rate": 1.95078031212485e-05,
+      "loss": 0.5343,
+      "step": 15254
+    },
+    {
+      "epoch": 19.5264,
+      "grad_norm": 1.0959560871124268,
+      "learning_rate": 1.9505802320928374e-05,
+      "loss": 0.4969,
+      "step": 15255
+    },
+    {
+      "epoch": 19.52768,
+      "grad_norm": 1.0881457328796387,
+      "learning_rate": 1.9503801520608246e-05,
+      "loss": 0.5351,
+      "step": 15256
+    },
+    {
+      "epoch": 19.52896,
+      "grad_norm": 1.0661768913269043,
+      "learning_rate": 1.9501800720288114e-05,
+      "loss": 0.4861,
+      "step": 15257
+    },
+    {
+      "epoch": 19.53024,
+      "grad_norm": 1.1012307405471802,
+      "learning_rate": 1.9499799919967986e-05,
+      "loss": 0.5191,
+      "step": 15258
+    },
+    {
+      "epoch": 19.53152,
+      "grad_norm": 1.0949671268463135,
+      "learning_rate": 1.949779911964786e-05,
+      "loss": 0.5407,
+      "step": 15259
+    },
+    {
+      "epoch": 19.5328,
+      "grad_norm": 1.0634304285049438,
+      "learning_rate": 1.9495798319327733e-05,
+      "loss": 0.4695,
+      "step": 15260
+    },
+    {
+      "epoch": 19.53408,
+      "grad_norm": 1.0616800785064697,
+      "learning_rate": 1.9493797519007602e-05,
+      "loss": 0.4984,
+      "step": 15261
+    },
+    {
+      "epoch": 19.53536,
+      "grad_norm": 1.1313798427581787,
+      "learning_rate": 1.9491796718687477e-05,
+      "loss": 0.5477,
+      "step": 15262
+    },
+    {
+      "epoch": 19.53664,
+      "grad_norm": 1.0650761127471924,
+      "learning_rate": 1.948979591836735e-05,
+      "loss": 0.5243,
+      "step": 15263
+    },
+    {
+      "epoch": 19.53792,
+      "grad_norm": 1.0772321224212646,
+      "learning_rate": 1.948779511804722e-05,
+      "loss": 0.4764,
+      "step": 15264
+    },
+    {
+      "epoch": 19.5392,
+      "grad_norm": 1.0623836517333984,
+      "learning_rate": 1.948579431772709e-05,
+      "loss": 0.5006,
+      "step": 15265
+    },
+    {
+      "epoch": 19.54048,
+      "grad_norm": 0.9955456852912903,
+      "learning_rate": 1.9483793517406964e-05,
+      "loss": 0.4675,
+      "step": 15266
+    },
+    {
+      "epoch": 19.54176,
+      "grad_norm": 1.0686817169189453,
+      "learning_rate": 1.9481792717086836e-05,
+      "loss": 0.5182,
+      "step": 15267
+    },
+    {
+      "epoch": 19.54304,
+      "grad_norm": 1.1312791109085083,
+      "learning_rate": 1.9479791916766708e-05,
+      "loss": 0.5157,
+      "step": 15268
+    },
+    {
+      "epoch": 19.54432,
+      "grad_norm": 1.1050148010253906,
+      "learning_rate": 1.947779111644658e-05,
+      "loss": 0.5053,
+      "step": 15269
+    },
+    {
+      "epoch": 19.5456,
+      "grad_norm": 1.0860576629638672,
+      "learning_rate": 1.9475790316126452e-05,
+      "loss": 0.4846,
+      "step": 15270
+    },
+    {
+      "epoch": 19.54688,
+      "grad_norm": 1.0629215240478516,
+      "learning_rate": 1.9473789515806324e-05,
+      "loss": 0.5131,
+      "step": 15271
+    },
+    {
+      "epoch": 19.54816,
+      "grad_norm": 1.05284583568573,
+      "learning_rate": 1.9471788715486196e-05,
+      "loss": 0.4852,
+      "step": 15272
+    },
+    {
+      "epoch": 19.54944,
+      "grad_norm": 1.0399121046066284,
+      "learning_rate": 1.9469787915166067e-05,
+      "loss": 0.4853,
+      "step": 15273
+    },
+    {
+      "epoch": 19.55072,
+      "grad_norm": 1.0178717374801636,
+      "learning_rate": 1.946778711484594e-05,
+      "loss": 0.4905,
+      "step": 15274
+    },
+    {
+      "epoch": 19.552,
+      "grad_norm": 1.0732840299606323,
+      "learning_rate": 1.946578631452581e-05,
+      "loss": 0.5017,
+      "step": 15275
+    },
+    {
+      "epoch": 19.55328,
+      "grad_norm": 1.1350263357162476,
+      "learning_rate": 1.9463785514205683e-05,
+      "loss": 0.5214,
+      "step": 15276
+    },
+    {
+      "epoch": 19.55456,
+      "grad_norm": 1.050378441810608,
+      "learning_rate": 1.9461784713885555e-05,
+      "loss": 0.4801,
+      "step": 15277
+    },
+    {
+      "epoch": 19.55584,
+      "grad_norm": 1.0464566946029663,
+      "learning_rate": 1.9459783913565427e-05,
+      "loss": 0.4978,
+      "step": 15278
+    },
+    {
+      "epoch": 19.55712,
+      "grad_norm": 1.0169779062271118,
+      "learning_rate": 1.94577831132453e-05,
+      "loss": 0.4731,
+      "step": 15279
+    },
+    {
+      "epoch": 19.5584,
+      "grad_norm": 1.0549458265304565,
+      "learning_rate": 1.945578231292517e-05,
+      "loss": 0.498,
+      "step": 15280
+    },
+    {
+      "epoch": 19.55968,
+      "grad_norm": 1.0301837921142578,
+      "learning_rate": 1.9453781512605042e-05,
+      "loss": 0.5407,
+      "step": 15281
+    },
+    {
+      "epoch": 19.56096,
+      "grad_norm": 1.0875555276870728,
+      "learning_rate": 1.9451780712284914e-05,
+      "loss": 0.509,
+      "step": 15282
+    },
+    {
+      "epoch": 19.56224,
+      "grad_norm": 1.072562336921692,
+      "learning_rate": 1.9449779911964786e-05,
+      "loss": 0.4804,
+      "step": 15283
+    },
+    {
+      "epoch": 19.56352,
+      "grad_norm": 1.0816986560821533,
+      "learning_rate": 1.9447779111644658e-05,
+      "loss": 0.5124,
+      "step": 15284
+    },
+    {
+      "epoch": 19.564799999999998,
+      "grad_norm": 1.0577350854873657,
+      "learning_rate": 1.944577831132453e-05,
+      "loss": 0.5007,
+      "step": 15285
+    },
+    {
+      "epoch": 19.56608,
+      "grad_norm": 1.0548467636108398,
+      "learning_rate": 1.94437775110044e-05,
+      "loss": 0.4651,
+      "step": 15286
+    },
+    {
+      "epoch": 19.56736,
+      "grad_norm": 1.0653223991394043,
+      "learning_rate": 1.9441776710684277e-05,
+      "loss": 0.4804,
+      "step": 15287
+    },
+    {
+      "epoch": 19.56864,
+      "grad_norm": 1.1180814504623413,
+      "learning_rate": 1.9439775910364145e-05,
+      "loss": 0.4905,
+      "step": 15288
+    },
+    {
+      "epoch": 19.56992,
+      "grad_norm": 1.0680838823318481,
+      "learning_rate": 1.9437775110044017e-05,
+      "loss": 0.4631,
+      "step": 15289
+    },
+    {
+      "epoch": 19.5712,
+      "grad_norm": 1.106482982635498,
+      "learning_rate": 1.943577430972389e-05,
+      "loss": 0.4835,
+      "step": 15290
+    },
+    {
+      "epoch": 19.57248,
+      "grad_norm": 1.105340838432312,
+      "learning_rate": 1.9433773509403764e-05,
+      "loss": 0.5132,
+      "step": 15291
+    },
+    {
+      "epoch": 19.57376,
+      "grad_norm": 1.0390416383743286,
+      "learning_rate": 1.9431772709083633e-05,
+      "loss": 0.4297,
+      "step": 15292
+    },
+    {
+      "epoch": 19.57504,
+      "grad_norm": 1.0809773206710815,
+      "learning_rate": 1.9429771908763505e-05,
+      "loss": 0.5219,
+      "step": 15293
+    },
+    {
+      "epoch": 19.57632,
+      "grad_norm": 0.9835126996040344,
+      "learning_rate": 1.942777110844338e-05,
+      "loss": 0.4615,
+      "step": 15294
+    },
+    {
+      "epoch": 19.5776,
+      "grad_norm": 1.0638185739517212,
+      "learning_rate": 1.942577030812325e-05,
+      "loss": 0.511,
+      "step": 15295
+    },
+    {
+      "epoch": 19.57888,
+      "grad_norm": 1.027719259262085,
+      "learning_rate": 1.942376950780312e-05,
+      "loss": 0.4703,
+      "step": 15296
+    },
+    {
+      "epoch": 19.58016,
+      "grad_norm": 1.035800814628601,
+      "learning_rate": 1.9421768707482992e-05,
+      "loss": 0.4904,
+      "step": 15297
+    },
+    {
+      "epoch": 19.58144,
+      "grad_norm": 1.0650843381881714,
+      "learning_rate": 1.9419767907162867e-05,
+      "loss": 0.512,
+      "step": 15298
+    },
+    {
+      "epoch": 19.58272,
+      "grad_norm": 1.0700833797454834,
+      "learning_rate": 1.941776710684274e-05,
+      "loss": 0.494,
+      "step": 15299
+    },
+    {
+      "epoch": 19.584,
+      "grad_norm": 1.0548943281173706,
+      "learning_rate": 1.9415766306522608e-05,
+      "loss": 0.4706,
+      "step": 15300
+    },
+    {
+      "epoch": 19.58528,
+      "grad_norm": 1.083795189857483,
+      "learning_rate": 1.9413765506202483e-05,
+      "loss": 0.4939,
+      "step": 15301
+    },
+    {
+      "epoch": 19.58656,
+      "grad_norm": 1.0601279735565186,
+      "learning_rate": 1.9411764705882355e-05,
+      "loss": 0.4932,
+      "step": 15302
+    },
+    {
+      "epoch": 19.58784,
+      "grad_norm": 1.109287977218628,
+      "learning_rate": 1.9409763905562226e-05,
+      "loss": 0.5211,
+      "step": 15303
+    },
+    {
+      "epoch": 19.58912,
+      "grad_norm": 1.0838240385055542,
+      "learning_rate": 1.9407763105242095e-05,
+      "loss": 0.4992,
+      "step": 15304
+    },
+    {
+      "epoch": 19.5904,
+      "grad_norm": 1.0001522302627563,
+      "learning_rate": 1.940576230492197e-05,
+      "loss": 0.4945,
+      "step": 15305
+    },
+    {
+      "epoch": 19.59168,
+      "grad_norm": 1.0223277807235718,
+      "learning_rate": 1.9403761504601842e-05,
+      "loss": 0.4509,
+      "step": 15306
+    },
+    {
+      "epoch": 19.59296,
+      "grad_norm": 1.0435092449188232,
+      "learning_rate": 1.9401760704281714e-05,
+      "loss": 0.4684,
+      "step": 15307
+    },
+    {
+      "epoch": 19.59424,
+      "grad_norm": 1.0610374212265015,
+      "learning_rate": 1.9399759903961586e-05,
+      "loss": 0.5055,
+      "step": 15308
+    },
+    {
+      "epoch": 19.59552,
+      "grad_norm": 1.1668132543563843,
+      "learning_rate": 1.9397759103641458e-05,
+      "loss": 0.5376,
+      "step": 15309
+    },
+    {
+      "epoch": 19.5968,
+      "grad_norm": 1.0593454837799072,
+      "learning_rate": 1.939575830332133e-05,
+      "loss": 0.4956,
+      "step": 15310
+    },
+    {
+      "epoch": 19.59808,
+      "grad_norm": 1.0181503295898438,
+      "learning_rate": 1.93937575030012e-05,
+      "loss": 0.4587,
+      "step": 15311
+    },
+    {
+      "epoch": 19.59936,
+      "grad_norm": 1.0744041204452515,
+      "learning_rate": 1.9391756702681073e-05,
+      "loss": 0.4853,
+      "step": 15312
+    },
+    {
+      "epoch": 19.60064,
+      "grad_norm": 1.075843334197998,
+      "learning_rate": 1.9389755902360945e-05,
+      "loss": 0.5086,
+      "step": 15313
+    },
+    {
+      "epoch": 19.60192,
+      "grad_norm": 1.0893293619155884,
+      "learning_rate": 1.9387755102040817e-05,
+      "loss": 0.5133,
+      "step": 15314
+    },
+    {
+      "epoch": 19.6032,
+      "grad_norm": 1.1128917932510376,
+      "learning_rate": 1.9385754301720692e-05,
+      "loss": 0.524,
+      "step": 15315
+    },
+    {
+      "epoch": 19.60448,
+      "grad_norm": 1.0819975137710571,
+      "learning_rate": 1.938375350140056e-05,
+      "loss": 0.4789,
+      "step": 15316
+    },
+    {
+      "epoch": 19.60576,
+      "grad_norm": 1.0744059085845947,
+      "learning_rate": 1.9381752701080432e-05,
+      "loss": 0.4927,
+      "step": 15317
+    },
+    {
+      "epoch": 19.60704,
+      "grad_norm": 1.0915199518203735,
+      "learning_rate": 1.9379751900760304e-05,
+      "loss": 0.5235,
+      "step": 15318
+    },
+    {
+      "epoch": 19.60832,
+      "grad_norm": 1.0718075037002563,
+      "learning_rate": 1.937775110044018e-05,
+      "loss": 0.4807,
+      "step": 15319
+    },
+    {
+      "epoch": 19.6096,
+      "grad_norm": 1.1020848751068115,
+      "learning_rate": 1.9375750300120048e-05,
+      "loss": 0.5598,
+      "step": 15320
+    },
+    {
+      "epoch": 19.61088,
+      "grad_norm": 1.1813325881958008,
+      "learning_rate": 1.937374949979992e-05,
+      "loss": 0.5096,
+      "step": 15321
+    },
+    {
+      "epoch": 19.61216,
+      "grad_norm": 1.097054362297058,
+      "learning_rate": 1.9371748699479795e-05,
+      "loss": 0.4994,
+      "step": 15322
+    },
+    {
+      "epoch": 19.61344,
+      "grad_norm": 1.1227426528930664,
+      "learning_rate": 1.9369747899159667e-05,
+      "loss": 0.517,
+      "step": 15323
+    },
+    {
+      "epoch": 19.61472,
+      "grad_norm": 1.0609750747680664,
+      "learning_rate": 1.9367747098839535e-05,
+      "loss": 0.4834,
+      "step": 15324
+    },
+    {
+      "epoch": 19.616,
+      "grad_norm": 1.1133859157562256,
+      "learning_rate": 1.9365746298519407e-05,
+      "loss": 0.5274,
+      "step": 15325
+    },
+    {
+      "epoch": 19.61728,
+      "grad_norm": 1.035995602607727,
+      "learning_rate": 1.9363745498199283e-05,
+      "loss": 0.4571,
+      "step": 15326
+    },
+    {
+      "epoch": 19.61856,
+      "grad_norm": 1.0632821321487427,
+      "learning_rate": 1.9361744697879154e-05,
+      "loss": 0.4931,
+      "step": 15327
+    },
+    {
+      "epoch": 19.61984,
+      "grad_norm": 1.1032476425170898,
+      "learning_rate": 1.9359743897559023e-05,
+      "loss": 0.4829,
+      "step": 15328
+    },
+    {
+      "epoch": 19.62112,
+      "grad_norm": 1.1157118082046509,
+      "learning_rate": 1.9357743097238895e-05,
+      "loss": 0.5142,
+      "step": 15329
+    },
+    {
+      "epoch": 19.6224,
+      "grad_norm": 1.0863420963287354,
+      "learning_rate": 1.935574229691877e-05,
+      "loss": 0.5191,
+      "step": 15330
+    },
+    {
+      "epoch": 19.62368,
+      "grad_norm": 1.044219970703125,
+      "learning_rate": 1.9353741496598642e-05,
+      "loss": 0.4757,
+      "step": 15331
+    },
+    {
+      "epoch": 19.62496,
+      "grad_norm": 1.0467369556427002,
+      "learning_rate": 1.935174069627851e-05,
+      "loss": 0.5285,
+      "step": 15332
+    },
+    {
+      "epoch": 19.62624,
+      "grad_norm": 1.1456965208053589,
+      "learning_rate": 1.9349739895958386e-05,
+      "loss": 0.5265,
+      "step": 15333
+    },
+    {
+      "epoch": 19.62752,
+      "grad_norm": 1.1731479167938232,
+      "learning_rate": 1.9347739095638257e-05,
+      "loss": 0.5279,
+      "step": 15334
+    },
+    {
+      "epoch": 19.6288,
+      "grad_norm": 1.0172317028045654,
+      "learning_rate": 1.934573829531813e-05,
+      "loss": 0.4484,
+      "step": 15335
+    },
+    {
+      "epoch": 19.63008,
+      "grad_norm": 1.009181022644043,
+      "learning_rate": 1.9343737494997998e-05,
+      "loss": 0.4591,
+      "step": 15336
+    },
+    {
+      "epoch": 19.63136,
+      "grad_norm": 1.1361030340194702,
+      "learning_rate": 1.9341736694677873e-05,
+      "loss": 0.5232,
+      "step": 15337
+    },
+    {
+      "epoch": 19.63264,
+      "grad_norm": 1.0451850891113281,
+      "learning_rate": 1.9339735894357745e-05,
+      "loss": 0.4996,
+      "step": 15338
+    },
+    {
+      "epoch": 19.63392,
+      "grad_norm": 1.0648033618927002,
+      "learning_rate": 1.9337735094037617e-05,
+      "loss": 0.5315,
+      "step": 15339
+    },
+    {
+      "epoch": 19.6352,
+      "grad_norm": 1.0276292562484741,
+      "learning_rate": 1.933573429371749e-05,
+      "loss": 0.4351,
+      "step": 15340
+    },
+    {
+      "epoch": 19.63648,
+      "grad_norm": 1.0635597705841064,
+      "learning_rate": 1.933373349339736e-05,
+      "loss": 0.5164,
+      "step": 15341
+    },
+    {
+      "epoch": 19.63776,
+      "grad_norm": 1.0696072578430176,
+      "learning_rate": 1.9331732693077232e-05,
+      "loss": 0.4814,
+      "step": 15342
+    },
+    {
+      "epoch": 19.63904,
+      "grad_norm": 1.0718597173690796,
+      "learning_rate": 1.9329731892757104e-05,
+      "loss": 0.5205,
+      "step": 15343
+    },
+    {
+      "epoch": 19.64032,
+      "grad_norm": 1.0730284452438354,
+      "learning_rate": 1.9327731092436976e-05,
+      "loss": 0.5375,
+      "step": 15344
+    },
+    {
+      "epoch": 19.6416,
+      "grad_norm": 1.0293726921081543,
+      "learning_rate": 1.9325730292116848e-05,
+      "loss": 0.4832,
+      "step": 15345
+    },
+    {
+      "epoch": 19.64288,
+      "grad_norm": 1.0186909437179565,
+      "learning_rate": 1.932372949179672e-05,
+      "loss": 0.4701,
+      "step": 15346
+    },
+    {
+      "epoch": 19.64416,
+      "grad_norm": 1.0497504472732544,
+      "learning_rate": 1.932172869147659e-05,
+      "loss": 0.4562,
+      "step": 15347
+    },
+    {
+      "epoch": 19.64544,
+      "grad_norm": 1.1007732152938843,
+      "learning_rate": 1.9319727891156463e-05,
+      "loss": 0.5467,
+      "step": 15348
+    },
+    {
+      "epoch": 19.64672,
+      "grad_norm": 1.1249552965164185,
+      "learning_rate": 1.9317727090836335e-05,
+      "loss": 0.4875,
+      "step": 15349
+    },
+    {
+      "epoch": 19.648,
+      "grad_norm": 1.0523337125778198,
+      "learning_rate": 1.9315726290516207e-05,
+      "loss": 0.4888,
+      "step": 15350
+    },
+    {
+      "epoch": 19.64928,
+      "grad_norm": 1.131332516670227,
+      "learning_rate": 1.931372549019608e-05,
+      "loss": 0.5072,
+      "step": 15351
+    },
+    {
+      "epoch": 19.65056,
+      "grad_norm": 1.137013554573059,
+      "learning_rate": 1.931172468987595e-05,
+      "loss": 0.4821,
+      "step": 15352
+    },
+    {
+      "epoch": 19.65184,
+      "grad_norm": 1.0466101169586182,
+      "learning_rate": 1.9309723889555823e-05,
+      "loss": 0.4602,
+      "step": 15353
+    },
+    {
+      "epoch": 19.65312,
+      "grad_norm": 0.9879499673843384,
+      "learning_rate": 1.9307723089235698e-05,
+      "loss": 0.4225,
+      "step": 15354
+    },
+    {
+      "epoch": 19.6544,
+      "grad_norm": 1.0248045921325684,
+      "learning_rate": 1.9305722288915566e-05,
+      "loss": 0.4888,
+      "step": 15355
+    },
+    {
+      "epoch": 19.65568,
+      "grad_norm": 1.0874507427215576,
+      "learning_rate": 1.9303721488595438e-05,
+      "loss": 0.5084,
+      "step": 15356
+    },
+    {
+      "epoch": 19.65696,
+      "grad_norm": 1.0725950002670288,
+      "learning_rate": 1.930172068827531e-05,
+      "loss": 0.5525,
+      "step": 15357
+    },
+    {
+      "epoch": 19.65824,
+      "grad_norm": 1.0607742071151733,
+      "learning_rate": 1.9299719887955185e-05,
+      "loss": 0.5004,
+      "step": 15358
+    },
+    {
+      "epoch": 19.65952,
+      "grad_norm": 1.1186076402664185,
+      "learning_rate": 1.9297719087635054e-05,
+      "loss": 0.5398,
+      "step": 15359
+    },
+    {
+      "epoch": 19.660800000000002,
+      "grad_norm": 1.0648303031921387,
+      "learning_rate": 1.9295718287314926e-05,
+      "loss": 0.4627,
+      "step": 15360
+    },
+    {
+      "epoch": 19.66208,
+      "grad_norm": 1.0890082120895386,
+      "learning_rate": 1.92937174869948e-05,
+      "loss": 0.5041,
+      "step": 15361
+    },
+    {
+      "epoch": 19.66336,
+      "grad_norm": 1.079013705253601,
+      "learning_rate": 1.9291716686674673e-05,
+      "loss": 0.5148,
+      "step": 15362
+    },
+    {
+      "epoch": 19.66464,
+      "grad_norm": 1.1085528135299683,
+      "learning_rate": 1.928971588635454e-05,
+      "loss": 0.534,
+      "step": 15363
+    },
+    {
+      "epoch": 19.66592,
+      "grad_norm": 1.06136953830719,
+      "learning_rate": 1.9287715086034413e-05,
+      "loss": 0.4668,
+      "step": 15364
+    },
+    {
+      "epoch": 19.6672,
+      "grad_norm": 1.0138074159622192,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 0.4922,
+      "step": 15365
+    },
+    {
+      "epoch": 19.66848,
+      "grad_norm": 1.0837033987045288,
+      "learning_rate": 1.928371348539416e-05,
+      "loss": 0.5442,
+      "step": 15366
+    },
+    {
+      "epoch": 19.66976,
+      "grad_norm": 1.1243685483932495,
+      "learning_rate": 1.928171268507403e-05,
+      "loss": 0.4993,
+      "step": 15367
+    },
+    {
+      "epoch": 19.67104,
+      "grad_norm": 1.0624139308929443,
+      "learning_rate": 1.9279711884753904e-05,
+      "loss": 0.4684,
+      "step": 15368
+    },
+    {
+      "epoch": 19.67232,
+      "grad_norm": 1.0839632749557495,
+      "learning_rate": 1.9277711084433776e-05,
+      "loss": 0.4872,
+      "step": 15369
+    },
+    {
+      "epoch": 19.6736,
+      "grad_norm": 1.078340768814087,
+      "learning_rate": 1.9275710284113648e-05,
+      "loss": 0.4678,
+      "step": 15370
+    },
+    {
+      "epoch": 19.67488,
+      "grad_norm": 1.0516289472579956,
+      "learning_rate": 1.9273709483793516e-05,
+      "loss": 0.4567,
+      "step": 15371
+    },
+    {
+      "epoch": 19.67616,
+      "grad_norm": 1.0610841512680054,
+      "learning_rate": 1.927170868347339e-05,
+      "loss": 0.4731,
+      "step": 15372
+    },
+    {
+      "epoch": 19.67744,
+      "grad_norm": 1.0811412334442139,
+      "learning_rate": 1.9269707883153263e-05,
+      "loss": 0.4978,
+      "step": 15373
+    },
+    {
+      "epoch": 19.67872,
+      "grad_norm": 1.0839446783065796,
+      "learning_rate": 1.9267707082833135e-05,
+      "loss": 0.4874,
+      "step": 15374
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 1.06401526927948,
+      "learning_rate": 1.9265706282513007e-05,
+      "loss": 0.508,
+      "step": 15375
+    },
+    {
+      "epoch": 19.68128,
+      "grad_norm": 1.0685803890228271,
+      "learning_rate": 1.926370548219288e-05,
+      "loss": 0.507,
+      "step": 15376
+    },
+    {
+      "epoch": 19.68256,
+      "grad_norm": 1.075534462928772,
+      "learning_rate": 1.926170468187275e-05,
+      "loss": 0.5059,
+      "step": 15377
+    },
+    {
+      "epoch": 19.68384,
+      "grad_norm": 1.0783944129943848,
+      "learning_rate": 1.9259703881552623e-05,
+      "loss": 0.4845,
+      "step": 15378
+    },
+    {
+      "epoch": 19.68512,
+      "grad_norm": 1.0071983337402344,
+      "learning_rate": 1.9257703081232494e-05,
+      "loss": 0.4961,
+      "step": 15379
+    },
+    {
+      "epoch": 19.6864,
+      "grad_norm": 1.0298351049423218,
+      "learning_rate": 1.9255702280912366e-05,
+      "loss": 0.512,
+      "step": 15380
+    },
+    {
+      "epoch": 19.68768,
+      "grad_norm": 1.0420536994934082,
+      "learning_rate": 1.9253701480592238e-05,
+      "loss": 0.5143,
+      "step": 15381
+    },
+    {
+      "epoch": 19.68896,
+      "grad_norm": 1.153039813041687,
+      "learning_rate": 1.925170068027211e-05,
+      "loss": 0.5164,
+      "step": 15382
+    },
+    {
+      "epoch": 19.69024,
+      "grad_norm": 1.0602798461914062,
+      "learning_rate": 1.9249699879951982e-05,
+      "loss": 0.5358,
+      "step": 15383
+    },
+    {
+      "epoch": 19.69152,
+      "grad_norm": 1.0772628784179688,
+      "learning_rate": 1.9247699079631854e-05,
+      "loss": 0.4898,
+      "step": 15384
+    },
+    {
+      "epoch": 19.6928,
+      "grad_norm": 1.1267281770706177,
+      "learning_rate": 1.9245698279311726e-05,
+      "loss": 0.5324,
+      "step": 15385
+    },
+    {
+      "epoch": 19.69408,
+      "grad_norm": 1.059523344039917,
+      "learning_rate": 1.9243697478991597e-05,
+      "loss": 0.4948,
+      "step": 15386
+    },
+    {
+      "epoch": 19.69536,
+      "grad_norm": 1.0987871885299683,
+      "learning_rate": 1.924169667867147e-05,
+      "loss": 0.4947,
+      "step": 15387
+    },
+    {
+      "epoch": 19.69664,
+      "grad_norm": 1.0897448062896729,
+      "learning_rate": 1.923969587835134e-05,
+      "loss": 0.5271,
+      "step": 15388
+    },
+    {
+      "epoch": 19.69792,
+      "grad_norm": 1.0368708372116089,
+      "learning_rate": 1.9237695078031213e-05,
+      "loss": 0.4986,
+      "step": 15389
+    },
+    {
+      "epoch": 19.6992,
+      "grad_norm": 1.063676357269287,
+      "learning_rate": 1.9235694277711085e-05,
+      "loss": 0.4886,
+      "step": 15390
+    },
+    {
+      "epoch": 19.70048,
+      "grad_norm": 1.1547781229019165,
+      "learning_rate": 1.9233693477390957e-05,
+      "loss": 0.5081,
+      "step": 15391
+    },
+    {
+      "epoch": 19.70176,
+      "grad_norm": 1.0699111223220825,
+      "learning_rate": 1.923169267707083e-05,
+      "loss": 0.4777,
+      "step": 15392
+    },
+    {
+      "epoch": 19.70304,
+      "grad_norm": 1.0858879089355469,
+      "learning_rate": 1.9229691876750704e-05,
+      "loss": 0.5082,
+      "step": 15393
+    },
+    {
+      "epoch": 19.70432,
+      "grad_norm": 1.145923376083374,
+      "learning_rate": 1.9227691076430572e-05,
+      "loss": 0.5587,
+      "step": 15394
+    },
+    {
+      "epoch": 19.7056,
+      "grad_norm": 1.0954890251159668,
+      "learning_rate": 1.9225690276110444e-05,
+      "loss": 0.5212,
+      "step": 15395
+    },
+    {
+      "epoch": 19.706879999999998,
+      "grad_norm": 1.0601500272750854,
+      "learning_rate": 1.9223689475790316e-05,
+      "loss": 0.4863,
+      "step": 15396
+    },
+    {
+      "epoch": 19.70816,
+      "grad_norm": 1.0982835292816162,
+      "learning_rate": 1.922168867547019e-05,
+      "loss": 0.4968,
+      "step": 15397
+    },
+    {
+      "epoch": 19.70944,
+      "grad_norm": 1.0785760879516602,
+      "learning_rate": 1.921968787515006e-05,
+      "loss": 0.5218,
+      "step": 15398
+    },
+    {
+      "epoch": 19.71072,
+      "grad_norm": 1.069139003753662,
+      "learning_rate": 1.921768707482993e-05,
+      "loss": 0.4671,
+      "step": 15399
+    },
+    {
+      "epoch": 19.712,
+      "grad_norm": 1.085194706916809,
+      "learning_rate": 1.9215686274509807e-05,
+      "loss": 0.5103,
+      "step": 15400
+    },
+    {
+      "epoch": 19.71328,
+      "grad_norm": 1.1250481605529785,
+      "learning_rate": 1.921368547418968e-05,
+      "loss": 0.5061,
+      "step": 15401
+    },
+    {
+      "epoch": 19.71456,
+      "grad_norm": 1.0836914777755737,
+      "learning_rate": 1.9211684673869547e-05,
+      "loss": 0.4801,
+      "step": 15402
+    },
+    {
+      "epoch": 19.71584,
+      "grad_norm": 1.1382417678833008,
+      "learning_rate": 1.920968387354942e-05,
+      "loss": 0.5333,
+      "step": 15403
+    },
+    {
+      "epoch": 19.71712,
+      "grad_norm": 1.0880017280578613,
+      "learning_rate": 1.9207683073229294e-05,
+      "loss": 0.5191,
+      "step": 15404
+    },
+    {
+      "epoch": 19.7184,
+      "grad_norm": 1.089227557182312,
+      "learning_rate": 1.9205682272909166e-05,
+      "loss": 0.4923,
+      "step": 15405
+    },
+    {
+      "epoch": 19.71968,
+      "grad_norm": 1.0711054801940918,
+      "learning_rate": 1.9203681472589035e-05,
+      "loss": 0.5067,
+      "step": 15406
+    },
+    {
+      "epoch": 19.72096,
+      "grad_norm": 1.0419178009033203,
+      "learning_rate": 1.920168067226891e-05,
+      "loss": 0.4903,
+      "step": 15407
+    },
+    {
+      "epoch": 19.72224,
+      "grad_norm": 1.0852587223052979,
+      "learning_rate": 1.919967987194878e-05,
+      "loss": 0.5343,
+      "step": 15408
+    },
+    {
+      "epoch": 19.72352,
+      "grad_norm": 1.0830180644989014,
+      "learning_rate": 1.9197679071628653e-05,
+      "loss": 0.5289,
+      "step": 15409
+    },
+    {
+      "epoch": 19.7248,
+      "grad_norm": 1.050003170967102,
+      "learning_rate": 1.9195678271308522e-05,
+      "loss": 0.4923,
+      "step": 15410
+    },
+    {
+      "epoch": 19.72608,
+      "grad_norm": 1.1027849912643433,
+      "learning_rate": 1.9193677470988397e-05,
+      "loss": 0.5719,
+      "step": 15411
+    },
+    {
+      "epoch": 19.72736,
+      "grad_norm": 1.0838650465011597,
+      "learning_rate": 1.919167667066827e-05,
+      "loss": 0.5167,
+      "step": 15412
+    },
+    {
+      "epoch": 19.72864,
+      "grad_norm": 1.0345534086227417,
+      "learning_rate": 1.918967587034814e-05,
+      "loss": 0.4831,
+      "step": 15413
+    },
+    {
+      "epoch": 19.72992,
+      "grad_norm": 1.0454968214035034,
+      "learning_rate": 1.9187675070028013e-05,
+      "loss": 0.5205,
+      "step": 15414
+    },
+    {
+      "epoch": 19.7312,
+      "grad_norm": 1.0153090953826904,
+      "learning_rate": 1.9185674269707885e-05,
+      "loss": 0.5044,
+      "step": 15415
+    },
+    {
+      "epoch": 19.73248,
+      "grad_norm": 1.0405809879302979,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 0.4734,
+      "step": 15416
+    },
+    {
+      "epoch": 19.73376,
+      "grad_norm": 1.0736857652664185,
+      "learning_rate": 1.918167266906763e-05,
+      "loss": 0.5145,
+      "step": 15417
+    },
+    {
+      "epoch": 19.73504,
+      "grad_norm": 1.0757614374160767,
+      "learning_rate": 1.91796718687475e-05,
+      "loss": 0.5051,
+      "step": 15418
+    },
+    {
+      "epoch": 19.73632,
+      "grad_norm": 1.074829339981079,
+      "learning_rate": 1.9177671068427372e-05,
+      "loss": 0.5238,
+      "step": 15419
+    },
+    {
+      "epoch": 19.7376,
+      "grad_norm": 1.0481175184249878,
+      "learning_rate": 1.9175670268107244e-05,
+      "loss": 0.501,
+      "step": 15420
+    },
+    {
+      "epoch": 19.73888,
+      "grad_norm": 1.1353001594543457,
+      "learning_rate": 1.9173669467787116e-05,
+      "loss": 0.4727,
+      "step": 15421
+    },
+    {
+      "epoch": 19.74016,
+      "grad_norm": 1.078800082206726,
+      "learning_rate": 1.9171668667466988e-05,
+      "loss": 0.4876,
+      "step": 15422
+    },
+    {
+      "epoch": 19.74144,
+      "grad_norm": 1.0725644826889038,
+      "learning_rate": 1.916966786714686e-05,
+      "loss": 0.489,
+      "step": 15423
+    },
+    {
+      "epoch": 19.74272,
+      "grad_norm": 1.0209654569625854,
+      "learning_rate": 1.916766706682673e-05,
+      "loss": 0.4757,
+      "step": 15424
+    },
+    {
+      "epoch": 19.744,
+      "grad_norm": 1.0331294536590576,
+      "learning_rate": 1.9165666266506603e-05,
+      "loss": 0.5106,
+      "step": 15425
+    },
+    {
+      "epoch": 19.74528,
+      "grad_norm": 1.0943084955215454,
+      "learning_rate": 1.9163665466186475e-05,
+      "loss": 0.4626,
+      "step": 15426
+    },
+    {
+      "epoch": 19.74656,
+      "grad_norm": 1.099189281463623,
+      "learning_rate": 1.9161664665866347e-05,
+      "loss": 0.4718,
+      "step": 15427
+    },
+    {
+      "epoch": 19.74784,
+      "grad_norm": 1.0979981422424316,
+      "learning_rate": 1.9159663865546222e-05,
+      "loss": 0.4987,
+      "step": 15428
+    },
+    {
+      "epoch": 19.74912,
+      "grad_norm": 1.0483365058898926,
+      "learning_rate": 1.915766306522609e-05,
+      "loss": 0.4775,
+      "step": 15429
+    },
+    {
+      "epoch": 19.7504,
+      "grad_norm": 1.0271607637405396,
+      "learning_rate": 1.9155662264905962e-05,
+      "loss": 0.4741,
+      "step": 15430
+    },
+    {
+      "epoch": 19.75168,
+      "grad_norm": 1.0611716508865356,
+      "learning_rate": 1.9153661464585834e-05,
+      "loss": 0.5134,
+      "step": 15431
+    },
+    {
+      "epoch": 19.75296,
+      "grad_norm": 1.0353494882583618,
+      "learning_rate": 1.915166066426571e-05,
+      "loss": 0.4661,
+      "step": 15432
+    },
+    {
+      "epoch": 19.75424,
+      "grad_norm": 1.0064977407455444,
+      "learning_rate": 1.9149659863945578e-05,
+      "loss": 0.4732,
+      "step": 15433
+    },
+    {
+      "epoch": 19.75552,
+      "grad_norm": 1.0152562856674194,
+      "learning_rate": 1.914765906362545e-05,
+      "loss": 0.4847,
+      "step": 15434
+    },
+    {
+      "epoch": 19.7568,
+      "grad_norm": 1.1019326448440552,
+      "learning_rate": 1.9145658263305325e-05,
+      "loss": 0.5283,
+      "step": 15435
+    },
+    {
+      "epoch": 19.75808,
+      "grad_norm": 1.0624425411224365,
+      "learning_rate": 1.9143657462985197e-05,
+      "loss": 0.4991,
+      "step": 15436
+    },
+    {
+      "epoch": 19.75936,
+      "grad_norm": 1.075663685798645,
+      "learning_rate": 1.9141656662665065e-05,
+      "loss": 0.5066,
+      "step": 15437
+    },
+    {
+      "epoch": 19.76064,
+      "grad_norm": 1.1295933723449707,
+      "learning_rate": 1.9139655862344937e-05,
+      "loss": 0.575,
+      "step": 15438
+    },
+    {
+      "epoch": 19.76192,
+      "grad_norm": 1.036911129951477,
+      "learning_rate": 1.9137655062024813e-05,
+      "loss": 0.4768,
+      "step": 15439
+    },
+    {
+      "epoch": 19.7632,
+      "grad_norm": 1.107503056526184,
+      "learning_rate": 1.9135654261704684e-05,
+      "loss": 0.5262,
+      "step": 15440
+    },
+    {
+      "epoch": 19.76448,
+      "grad_norm": 1.0643945932388306,
+      "learning_rate": 1.9133653461384553e-05,
+      "loss": 0.4138,
+      "step": 15441
+    },
+    {
+      "epoch": 19.76576,
+      "grad_norm": 1.0519750118255615,
+      "learning_rate": 1.9131652661064425e-05,
+      "loss": 0.4887,
+      "step": 15442
+    },
+    {
+      "epoch": 19.76704,
+      "grad_norm": 1.0428272485733032,
+      "learning_rate": 1.91296518607443e-05,
+      "loss": 0.4986,
+      "step": 15443
+    },
+    {
+      "epoch": 19.76832,
+      "grad_norm": 1.0842911005020142,
+      "learning_rate": 1.9127651060424172e-05,
+      "loss": 0.5026,
+      "step": 15444
+    },
+    {
+      "epoch": 19.7696,
+      "grad_norm": 1.1030406951904297,
+      "learning_rate": 1.912565026010404e-05,
+      "loss": 0.4867,
+      "step": 15445
+    },
+    {
+      "epoch": 19.77088,
+      "grad_norm": 1.0993808507919312,
+      "learning_rate": 1.9123649459783916e-05,
+      "loss": 0.4984,
+      "step": 15446
+    },
+    {
+      "epoch": 19.77216,
+      "grad_norm": 1.1211193799972534,
+      "learning_rate": 1.9121648659463787e-05,
+      "loss": 0.503,
+      "step": 15447
+    },
+    {
+      "epoch": 19.77344,
+      "grad_norm": 1.1349544525146484,
+      "learning_rate": 1.911964785914366e-05,
+      "loss": 0.5267,
+      "step": 15448
+    },
+    {
+      "epoch": 19.77472,
+      "grad_norm": 1.1778333187103271,
+      "learning_rate": 1.9117647058823528e-05,
+      "loss": 0.5236,
+      "step": 15449
+    },
+    {
+      "epoch": 19.776,
+      "grad_norm": 1.1410446166992188,
+      "learning_rate": 1.9115646258503403e-05,
+      "loss": 0.4975,
+      "step": 15450
+    },
+    {
+      "epoch": 19.77728,
+      "grad_norm": 1.0365501642227173,
+      "learning_rate": 1.9113645458183275e-05,
+      "loss": 0.4599,
+      "step": 15451
+    },
+    {
+      "epoch": 19.77856,
+      "grad_norm": 1.1023868322372437,
+      "learning_rate": 1.9111644657863147e-05,
+      "loss": 0.5273,
+      "step": 15452
+    },
+    {
+      "epoch": 19.77984,
+      "grad_norm": 1.103507399559021,
+      "learning_rate": 1.910964385754302e-05,
+      "loss": 0.4749,
+      "step": 15453
+    },
+    {
+      "epoch": 19.78112,
+      "grad_norm": 1.0735512971878052,
+      "learning_rate": 1.910764305722289e-05,
+      "loss": 0.4875,
+      "step": 15454
+    },
+    {
+      "epoch": 19.7824,
+      "grad_norm": 1.0677052736282349,
+      "learning_rate": 1.9105642256902762e-05,
+      "loss": 0.5173,
+      "step": 15455
+    },
+    {
+      "epoch": 19.78368,
+      "grad_norm": 1.0987626314163208,
+      "learning_rate": 1.9103641456582634e-05,
+      "loss": 0.5271,
+      "step": 15456
+    },
+    {
+      "epoch": 19.78496,
+      "grad_norm": 1.0896143913269043,
+      "learning_rate": 1.9101640656262506e-05,
+      "loss": 0.533,
+      "step": 15457
+    },
+    {
+      "epoch": 19.78624,
+      "grad_norm": 1.1035916805267334,
+      "learning_rate": 1.9099639855942378e-05,
+      "loss": 0.5359,
+      "step": 15458
+    },
+    {
+      "epoch": 19.78752,
+      "grad_norm": 1.1171396970748901,
+      "learning_rate": 1.909763905562225e-05,
+      "loss": 0.4939,
+      "step": 15459
+    },
+    {
+      "epoch": 19.7888,
+      "grad_norm": 1.0793712139129639,
+      "learning_rate": 1.909563825530212e-05,
+      "loss": 0.5262,
+      "step": 15460
+    },
+    {
+      "epoch": 19.79008,
+      "grad_norm": 1.0609002113342285,
+      "learning_rate": 1.9093637454981993e-05,
+      "loss": 0.4997,
+      "step": 15461
+    },
+    {
+      "epoch": 19.79136,
+      "grad_norm": 1.010659098625183,
+      "learning_rate": 1.9091636654661865e-05,
+      "loss": 0.4642,
+      "step": 15462
+    },
+    {
+      "epoch": 19.79264,
+      "grad_norm": 1.043384075164795,
+      "learning_rate": 1.9089635854341737e-05,
+      "loss": 0.5313,
+      "step": 15463
+    },
+    {
+      "epoch": 19.79392,
+      "grad_norm": 1.0543464422225952,
+      "learning_rate": 1.908763505402161e-05,
+      "loss": 0.4636,
+      "step": 15464
+    },
+    {
+      "epoch": 19.7952,
+      "grad_norm": 1.094262957572937,
+      "learning_rate": 1.908563425370148e-05,
+      "loss": 0.5249,
+      "step": 15465
+    },
+    {
+      "epoch": 19.79648,
+      "grad_norm": 1.1226550340652466,
+      "learning_rate": 1.9083633453381353e-05,
+      "loss": 0.55,
+      "step": 15466
+    },
+    {
+      "epoch": 19.79776,
+      "grad_norm": 1.1305582523345947,
+      "learning_rate": 1.9081632653061228e-05,
+      "loss": 0.5418,
+      "step": 15467
+    },
+    {
+      "epoch": 19.79904,
+      "grad_norm": 1.1155486106872559,
+      "learning_rate": 1.9079631852741096e-05,
+      "loss": 0.5226,
+      "step": 15468
+    },
+    {
+      "epoch": 19.80032,
+      "grad_norm": 1.1939311027526855,
+      "learning_rate": 1.9077631052420968e-05,
+      "loss": 0.5122,
+      "step": 15469
+    },
+    {
+      "epoch": 19.8016,
+      "grad_norm": 1.0629066228866577,
+      "learning_rate": 1.907563025210084e-05,
+      "loss": 0.4641,
+      "step": 15470
+    },
+    {
+      "epoch": 19.802880000000002,
+      "grad_norm": 1.05988609790802,
+      "learning_rate": 1.9073629451780715e-05,
+      "loss": 0.5368,
+      "step": 15471
+    },
+    {
+      "epoch": 19.80416,
+      "grad_norm": 1.1533082723617554,
+      "learning_rate": 1.9071628651460584e-05,
+      "loss": 0.578,
+      "step": 15472
+    },
+    {
+      "epoch": 19.80544,
+      "grad_norm": 1.1095813512802124,
+      "learning_rate": 1.9069627851140456e-05,
+      "loss": 0.4966,
+      "step": 15473
+    },
+    {
+      "epoch": 19.80672,
+      "grad_norm": 1.067834734916687,
+      "learning_rate": 1.906762705082033e-05,
+      "loss": 0.5486,
+      "step": 15474
+    },
+    {
+      "epoch": 19.808,
+      "grad_norm": 1.0650924444198608,
+      "learning_rate": 1.9065626250500203e-05,
+      "loss": 0.5034,
+      "step": 15475
+    },
+    {
+      "epoch": 19.80928,
+      "grad_norm": 1.0607784986495972,
+      "learning_rate": 1.906362545018007e-05,
+      "loss": 0.5092,
+      "step": 15476
+    },
+    {
+      "epoch": 19.81056,
+      "grad_norm": 1.0503934621810913,
+      "learning_rate": 1.9061624649859943e-05,
+      "loss": 0.4871,
+      "step": 15477
+    },
+    {
+      "epoch": 19.81184,
+      "grad_norm": 1.1574764251708984,
+      "learning_rate": 1.905962384953982e-05,
+      "loss": 0.5786,
+      "step": 15478
+    },
+    {
+      "epoch": 19.81312,
+      "grad_norm": 1.142246127128601,
+      "learning_rate": 1.905762304921969e-05,
+      "loss": 0.4866,
+      "step": 15479
+    },
+    {
+      "epoch": 19.8144,
+      "grad_norm": 1.0628846883773804,
+      "learning_rate": 1.905562224889956e-05,
+      "loss": 0.4879,
+      "step": 15480
+    },
+    {
+      "epoch": 19.81568,
+      "grad_norm": 1.1084834337234497,
+      "learning_rate": 1.9053621448579434e-05,
+      "loss": 0.5393,
+      "step": 15481
+    },
+    {
+      "epoch": 19.81696,
+      "grad_norm": 1.2051427364349365,
+      "learning_rate": 1.9051620648259306e-05,
+      "loss": 0.5179,
+      "step": 15482
+    },
+    {
+      "epoch": 19.81824,
+      "grad_norm": 1.0791866779327393,
+      "learning_rate": 1.9049619847939178e-05,
+      "loss": 0.463,
+      "step": 15483
+    },
+    {
+      "epoch": 19.81952,
+      "grad_norm": 1.0924525260925293,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.4755,
+      "step": 15484
+    },
+    {
+      "epoch": 19.8208,
+      "grad_norm": 1.0845119953155518,
+      "learning_rate": 1.904561824729892e-05,
+      "loss": 0.5183,
+      "step": 15485
+    },
+    {
+      "epoch": 19.82208,
+      "grad_norm": 1.0582520961761475,
+      "learning_rate": 1.9043617446978793e-05,
+      "loss": 0.4608,
+      "step": 15486
+    },
+    {
+      "epoch": 19.82336,
+      "grad_norm": 1.0958303213119507,
+      "learning_rate": 1.9041616646658665e-05,
+      "loss": 0.4605,
+      "step": 15487
+    },
+    {
+      "epoch": 19.82464,
+      "grad_norm": 1.0015331506729126,
+      "learning_rate": 1.9039615846338537e-05,
+      "loss": 0.4579,
+      "step": 15488
+    },
+    {
+      "epoch": 19.82592,
+      "grad_norm": 1.1324008703231812,
+      "learning_rate": 1.903761504601841e-05,
+      "loss": 0.5423,
+      "step": 15489
+    },
+    {
+      "epoch": 19.8272,
+      "grad_norm": 1.0755281448364258,
+      "learning_rate": 1.903561424569828e-05,
+      "loss": 0.4549,
+      "step": 15490
+    },
+    {
+      "epoch": 19.82848,
+      "grad_norm": 1.1091095209121704,
+      "learning_rate": 1.9033613445378152e-05,
+      "loss": 0.5028,
+      "step": 15491
+    },
+    {
+      "epoch": 19.82976,
+      "grad_norm": 1.0952885150909424,
+      "learning_rate": 1.9031612645058024e-05,
+      "loss": 0.479,
+      "step": 15492
+    },
+    {
+      "epoch": 19.83104,
+      "grad_norm": 1.0651510953903198,
+      "learning_rate": 1.9029611844737896e-05,
+      "loss": 0.5052,
+      "step": 15493
+    },
+    {
+      "epoch": 19.83232,
+      "grad_norm": 1.1032085418701172,
+      "learning_rate": 1.9027611044417768e-05,
+      "loss": 0.5447,
+      "step": 15494
+    },
+    {
+      "epoch": 19.8336,
+      "grad_norm": 1.0811024904251099,
+      "learning_rate": 1.902561024409764e-05,
+      "loss": 0.4788,
+      "step": 15495
+    },
+    {
+      "epoch": 19.83488,
+      "grad_norm": 1.0974313020706177,
+      "learning_rate": 1.9023609443777512e-05,
+      "loss": 0.5162,
+      "step": 15496
+    },
+    {
+      "epoch": 19.83616,
+      "grad_norm": 1.0592910051345825,
+      "learning_rate": 1.9021608643457384e-05,
+      "loss": 0.4786,
+      "step": 15497
+    },
+    {
+      "epoch": 19.83744,
+      "grad_norm": 1.1197059154510498,
+      "learning_rate": 1.9019607843137255e-05,
+      "loss": 0.5087,
+      "step": 15498
+    },
+    {
+      "epoch": 19.83872,
+      "grad_norm": 1.0962718725204468,
+      "learning_rate": 1.9017607042817127e-05,
+      "loss": 0.5393,
+      "step": 15499
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 1.1129200458526611,
+      "learning_rate": 1.9015606242497e-05,
+      "loss": 0.5287,
+      "step": 15500
+    },
+    {
+      "epoch": 19.84128,
+      "grad_norm": 1.049560785293579,
+      "learning_rate": 1.901360544217687e-05,
+      "loss": 0.4888,
+      "step": 15501
+    },
+    {
+      "epoch": 19.84256,
+      "grad_norm": 1.0872515439987183,
+      "learning_rate": 1.9011604641856743e-05,
+      "loss": 0.4652,
+      "step": 15502
+    },
+    {
+      "epoch": 19.84384,
+      "grad_norm": 1.112954020500183,
+      "learning_rate": 1.9009603841536615e-05,
+      "loss": 0.5514,
+      "step": 15503
+    },
+    {
+      "epoch": 19.84512,
+      "grad_norm": 1.0758328437805176,
+      "learning_rate": 1.9007603041216487e-05,
+      "loss": 0.4836,
+      "step": 15504
+    },
+    {
+      "epoch": 19.8464,
+      "grad_norm": 1.0734128952026367,
+      "learning_rate": 1.900560224089636e-05,
+      "loss": 0.5046,
+      "step": 15505
+    },
+    {
+      "epoch": 19.84768,
+      "grad_norm": 1.1305335760116577,
+      "learning_rate": 1.9003601440576234e-05,
+      "loss": 0.5106,
+      "step": 15506
+    },
+    {
+      "epoch": 19.84896,
+      "grad_norm": 1.1036813259124756,
+      "learning_rate": 1.9001600640256102e-05,
+      "loss": 0.5227,
+      "step": 15507
+    },
+    {
+      "epoch": 19.85024,
+      "grad_norm": 1.1305910348892212,
+      "learning_rate": 1.8999599839935974e-05,
+      "loss": 0.5204,
+      "step": 15508
+    },
+    {
+      "epoch": 19.85152,
+      "grad_norm": 1.0744602680206299,
+      "learning_rate": 1.8997599039615846e-05,
+      "loss": 0.4987,
+      "step": 15509
+    },
+    {
+      "epoch": 19.8528,
+      "grad_norm": 1.0705610513687134,
+      "learning_rate": 1.899559823929572e-05,
+      "loss": 0.5388,
+      "step": 15510
+    },
+    {
+      "epoch": 19.85408,
+      "grad_norm": 1.0107085704803467,
+      "learning_rate": 1.899359743897559e-05,
+      "loss": 0.4689,
+      "step": 15511
+    },
+    {
+      "epoch": 19.85536,
+      "grad_norm": 1.0287785530090332,
+      "learning_rate": 1.899159663865546e-05,
+      "loss": 0.4932,
+      "step": 15512
+    },
+    {
+      "epoch": 19.85664,
+      "grad_norm": 1.0263909101486206,
+      "learning_rate": 1.8989595838335337e-05,
+      "loss": 0.4614,
+      "step": 15513
+    },
+    {
+      "epoch": 19.85792,
+      "grad_norm": 1.0720847845077515,
+      "learning_rate": 1.898759503801521e-05,
+      "loss": 0.5024,
+      "step": 15514
+    },
+    {
+      "epoch": 19.8592,
+      "grad_norm": 1.1379327774047852,
+      "learning_rate": 1.8985594237695077e-05,
+      "loss": 0.4958,
+      "step": 15515
+    },
+    {
+      "epoch": 19.86048,
+      "grad_norm": 1.0998201370239258,
+      "learning_rate": 1.898359343737495e-05,
+      "loss": 0.5257,
+      "step": 15516
+    },
+    {
+      "epoch": 19.86176,
+      "grad_norm": 1.068842887878418,
+      "learning_rate": 1.8981592637054824e-05,
+      "loss": 0.4743,
+      "step": 15517
+    },
+    {
+      "epoch": 19.86304,
+      "grad_norm": 1.0820146799087524,
+      "learning_rate": 1.8979591836734696e-05,
+      "loss": 0.5448,
+      "step": 15518
+    },
+    {
+      "epoch": 19.86432,
+      "grad_norm": 1.0579636096954346,
+      "learning_rate": 1.8977591036414564e-05,
+      "loss": 0.5176,
+      "step": 15519
+    },
+    {
+      "epoch": 19.8656,
+      "grad_norm": 1.0649138689041138,
+      "learning_rate": 1.897559023609444e-05,
+      "loss": 0.4429,
+      "step": 15520
+    },
+    {
+      "epoch": 19.86688,
+      "grad_norm": 1.0768613815307617,
+      "learning_rate": 1.897358943577431e-05,
+      "loss": 0.4847,
+      "step": 15521
+    },
+    {
+      "epoch": 19.86816,
+      "grad_norm": 1.0212260484695435,
+      "learning_rate": 1.8971588635454183e-05,
+      "loss": 0.4501,
+      "step": 15522
+    },
+    {
+      "epoch": 19.86944,
+      "grad_norm": 1.0247191190719604,
+      "learning_rate": 1.8969587835134052e-05,
+      "loss": 0.4715,
+      "step": 15523
+    },
+    {
+      "epoch": 19.87072,
+      "grad_norm": 1.0731992721557617,
+      "learning_rate": 1.8967587034813927e-05,
+      "loss": 0.4961,
+      "step": 15524
+    },
+    {
+      "epoch": 19.872,
+      "grad_norm": 1.0773158073425293,
+      "learning_rate": 1.89655862344938e-05,
+      "loss": 0.4927,
+      "step": 15525
+    },
+    {
+      "epoch": 19.87328,
+      "grad_norm": 1.0747638940811157,
+      "learning_rate": 1.896358543417367e-05,
+      "loss": 0.5224,
+      "step": 15526
+    },
+    {
+      "epoch": 19.87456,
+      "grad_norm": 1.0612146854400635,
+      "learning_rate": 1.8961584633853543e-05,
+      "loss": 0.5373,
+      "step": 15527
+    },
+    {
+      "epoch": 19.87584,
+      "grad_norm": 1.1125435829162598,
+      "learning_rate": 1.8959583833533415e-05,
+      "loss": 0.5768,
+      "step": 15528
+    },
+    {
+      "epoch": 19.87712,
+      "grad_norm": 1.0910027027130127,
+      "learning_rate": 1.8957583033213286e-05,
+      "loss": 0.5322,
+      "step": 15529
+    },
+    {
+      "epoch": 19.8784,
+      "grad_norm": 1.1110457181930542,
+      "learning_rate": 1.8955582232893158e-05,
+      "loss": 0.5189,
+      "step": 15530
+    },
+    {
+      "epoch": 19.87968,
+      "grad_norm": 1.0447136163711548,
+      "learning_rate": 1.895358143257303e-05,
+      "loss": 0.4795,
+      "step": 15531
+    },
+    {
+      "epoch": 19.88096,
+      "grad_norm": 1.0677382946014404,
+      "learning_rate": 1.8951580632252902e-05,
+      "loss": 0.4929,
+      "step": 15532
+    },
+    {
+      "epoch": 19.88224,
+      "grad_norm": 1.038959264755249,
+      "learning_rate": 1.8949579831932774e-05,
+      "loss": 0.4779,
+      "step": 15533
+    },
+    {
+      "epoch": 19.88352,
+      "grad_norm": 1.0871193408966064,
+      "learning_rate": 1.8947579031612646e-05,
+      "loss": 0.48,
+      "step": 15534
+    },
+    {
+      "epoch": 19.8848,
+      "grad_norm": 1.1005915403366089,
+      "learning_rate": 1.8945578231292518e-05,
+      "loss": 0.5181,
+      "step": 15535
+    },
+    {
+      "epoch": 19.88608,
+      "grad_norm": 1.0428812503814697,
+      "learning_rate": 1.894357743097239e-05,
+      "loss": 0.4534,
+      "step": 15536
+    },
+    {
+      "epoch": 19.88736,
+      "grad_norm": 1.0860538482666016,
+      "learning_rate": 1.894157663065226e-05,
+      "loss": 0.5012,
+      "step": 15537
+    },
+    {
+      "epoch": 19.88864,
+      "grad_norm": 1.0798299312591553,
+      "learning_rate": 1.8939575830332133e-05,
+      "loss": 0.5091,
+      "step": 15538
+    },
+    {
+      "epoch": 19.88992,
+      "grad_norm": 1.0804452896118164,
+      "learning_rate": 1.8937575030012005e-05,
+      "loss": 0.4978,
+      "step": 15539
+    },
+    {
+      "epoch": 19.8912,
+      "grad_norm": 1.069659948348999,
+      "learning_rate": 1.8935574229691877e-05,
+      "loss": 0.4902,
+      "step": 15540
+    },
+    {
+      "epoch": 19.89248,
+      "grad_norm": 1.1372871398925781,
+      "learning_rate": 1.8933573429371752e-05,
+      "loss": 0.5438,
+      "step": 15541
+    },
+    {
+      "epoch": 19.89376,
+      "grad_norm": 1.079403281211853,
+      "learning_rate": 1.893157262905162e-05,
+      "loss": 0.4882,
+      "step": 15542
+    },
+    {
+      "epoch": 19.89504,
+      "grad_norm": 1.0226234197616577,
+      "learning_rate": 1.8929571828731492e-05,
+      "loss": 0.4323,
+      "step": 15543
+    },
+    {
+      "epoch": 19.89632,
+      "grad_norm": 1.0797722339630127,
+      "learning_rate": 1.8927571028411364e-05,
+      "loss": 0.507,
+      "step": 15544
+    },
+    {
+      "epoch": 19.8976,
+      "grad_norm": 1.0766708850860596,
+      "learning_rate": 1.892557022809124e-05,
+      "loss": 0.5058,
+      "step": 15545
+    },
+    {
+      "epoch": 19.89888,
+      "grad_norm": 1.0884517431259155,
+      "learning_rate": 1.8923569427771108e-05,
+      "loss": 0.5136,
+      "step": 15546
+    },
+    {
+      "epoch": 19.90016,
+      "grad_norm": 1.1171706914901733,
+      "learning_rate": 1.892156862745098e-05,
+      "loss": 0.5082,
+      "step": 15547
+    },
+    {
+      "epoch": 19.90144,
+      "grad_norm": 1.065116286277771,
+      "learning_rate": 1.8919567827130855e-05,
+      "loss": 0.5113,
+      "step": 15548
+    },
+    {
+      "epoch": 19.90272,
+      "grad_norm": 1.0674129724502563,
+      "learning_rate": 1.8917567026810727e-05,
+      "loss": 0.4844,
+      "step": 15549
+    },
+    {
+      "epoch": 19.904,
+      "grad_norm": 1.123174786567688,
+      "learning_rate": 1.8915566226490595e-05,
+      "loss": 0.527,
+      "step": 15550
+    },
+    {
+      "epoch": 19.90528,
+      "grad_norm": 1.113181710243225,
+      "learning_rate": 1.8913565426170467e-05,
+      "loss": 0.5275,
+      "step": 15551
+    },
+    {
+      "epoch": 19.90656,
+      "grad_norm": 1.080833077430725,
+      "learning_rate": 1.8911564625850343e-05,
+      "loss": 0.5218,
+      "step": 15552
+    },
+    {
+      "epoch": 19.90784,
+      "grad_norm": 1.1273305416107178,
+      "learning_rate": 1.8909563825530214e-05,
+      "loss": 0.5157,
+      "step": 15553
+    },
+    {
+      "epoch": 19.90912,
+      "grad_norm": 1.1139534711837769,
+      "learning_rate": 1.8907563025210083e-05,
+      "loss": 0.5408,
+      "step": 15554
+    },
+    {
+      "epoch": 19.9104,
+      "grad_norm": 1.0617810487747192,
+      "learning_rate": 1.8905562224889955e-05,
+      "loss": 0.4972,
+      "step": 15555
+    },
+    {
+      "epoch": 19.91168,
+      "grad_norm": 1.0944743156433105,
+      "learning_rate": 1.890356142456983e-05,
+      "loss": 0.5016,
+      "step": 15556
+    },
+    {
+      "epoch": 19.912959999999998,
+      "grad_norm": 1.0922056436538696,
+      "learning_rate": 1.8901560624249702e-05,
+      "loss": 0.5228,
+      "step": 15557
+    },
+    {
+      "epoch": 19.91424,
+      "grad_norm": 1.060390830039978,
+      "learning_rate": 1.889955982392957e-05,
+      "loss": 0.5006,
+      "step": 15558
+    },
+    {
+      "epoch": 19.91552,
+      "grad_norm": 1.0017435550689697,
+      "learning_rate": 1.8897559023609446e-05,
+      "loss": 0.4723,
+      "step": 15559
+    },
+    {
+      "epoch": 19.9168,
+      "grad_norm": 1.0966525077819824,
+      "learning_rate": 1.8895558223289317e-05,
+      "loss": 0.5312,
+      "step": 15560
+    },
+    {
+      "epoch": 19.91808,
+      "grad_norm": 1.037953495979309,
+      "learning_rate": 1.889355742296919e-05,
+      "loss": 0.5364,
+      "step": 15561
+    },
+    {
+      "epoch": 19.91936,
+      "grad_norm": 1.0990053415298462,
+      "learning_rate": 1.8891556622649058e-05,
+      "loss": 0.515,
+      "step": 15562
+    },
+    {
+      "epoch": 19.92064,
+      "grad_norm": 1.1318585872650146,
+      "learning_rate": 1.8889555822328933e-05,
+      "loss": 0.5202,
+      "step": 15563
+    },
+    {
+      "epoch": 19.92192,
+      "grad_norm": 1.0594831705093384,
+      "learning_rate": 1.8887555022008805e-05,
+      "loss": 0.5013,
+      "step": 15564
+    },
+    {
+      "epoch": 19.9232,
+      "grad_norm": 1.0701175928115845,
+      "learning_rate": 1.8885554221688677e-05,
+      "loss": 0.4641,
+      "step": 15565
+    },
+    {
+      "epoch": 19.92448,
+      "grad_norm": 1.095017433166504,
+      "learning_rate": 1.888355342136855e-05,
+      "loss": 0.5022,
+      "step": 15566
+    },
+    {
+      "epoch": 19.92576,
+      "grad_norm": 1.1008120775222778,
+      "learning_rate": 1.888155262104842e-05,
+      "loss": 0.5144,
+      "step": 15567
+    },
+    {
+      "epoch": 19.92704,
+      "grad_norm": 1.0026847124099731,
+      "learning_rate": 1.8879551820728292e-05,
+      "loss": 0.4596,
+      "step": 15568
+    },
+    {
+      "epoch": 19.92832,
+      "grad_norm": 1.0756937265396118,
+      "learning_rate": 1.8877551020408164e-05,
+      "loss": 0.4879,
+      "step": 15569
+    },
+    {
+      "epoch": 19.9296,
+      "grad_norm": 1.0815765857696533,
+      "learning_rate": 1.8875550220088036e-05,
+      "loss": 0.533,
+      "step": 15570
+    },
+    {
+      "epoch": 19.93088,
+      "grad_norm": 1.0457710027694702,
+      "learning_rate": 1.8873549419767908e-05,
+      "loss": 0.5204,
+      "step": 15571
+    },
+    {
+      "epoch": 19.93216,
+      "grad_norm": 1.008851408958435,
+      "learning_rate": 1.887154861944778e-05,
+      "loss": 0.5,
+      "step": 15572
+    },
+    {
+      "epoch": 19.93344,
+      "grad_norm": 1.0376039743423462,
+      "learning_rate": 1.886954781912765e-05,
+      "loss": 0.5194,
+      "step": 15573
+    },
+    {
+      "epoch": 19.93472,
+      "grad_norm": 1.0573559999465942,
+      "learning_rate": 1.8867547018807523e-05,
+      "loss": 0.4944,
+      "step": 15574
+    },
+    {
+      "epoch": 19.936,
+      "grad_norm": 1.035343885421753,
+      "learning_rate": 1.8865546218487395e-05,
+      "loss": 0.4731,
+      "step": 15575
+    },
+    {
+      "epoch": 19.93728,
+      "grad_norm": 1.033947229385376,
+      "learning_rate": 1.8863545418167267e-05,
+      "loss": 0.4792,
+      "step": 15576
+    },
+    {
+      "epoch": 19.93856,
+      "grad_norm": 1.0770082473754883,
+      "learning_rate": 1.886154461784714e-05,
+      "loss": 0.5264,
+      "step": 15577
+    },
+    {
+      "epoch": 19.93984,
+      "grad_norm": 1.0262564420700073,
+      "learning_rate": 1.885954381752701e-05,
+      "loss": 0.5304,
+      "step": 15578
+    },
+    {
+      "epoch": 19.94112,
+      "grad_norm": 1.0958056449890137,
+      "learning_rate": 1.8857543017206883e-05,
+      "loss": 0.5294,
+      "step": 15579
+    },
+    {
+      "epoch": 19.9424,
+      "grad_norm": 1.1022257804870605,
+      "learning_rate": 1.8855542216886758e-05,
+      "loss": 0.4546,
+      "step": 15580
+    },
+    {
+      "epoch": 19.94368,
+      "grad_norm": 1.0951906442642212,
+      "learning_rate": 1.8853541416566626e-05,
+      "loss": 0.5054,
+      "step": 15581
+    },
+    {
+      "epoch": 19.944960000000002,
+      "grad_norm": 1.0561892986297607,
+      "learning_rate": 1.8851540616246498e-05,
+      "loss": 0.4944,
+      "step": 15582
+    },
+    {
+      "epoch": 19.94624,
+      "grad_norm": 1.084155559539795,
+      "learning_rate": 1.884953981592637e-05,
+      "loss": 0.5098,
+      "step": 15583
+    },
+    {
+      "epoch": 19.94752,
+      "grad_norm": 1.0809601545333862,
+      "learning_rate": 1.8847539015606245e-05,
+      "loss": 0.5657,
+      "step": 15584
+    },
+    {
+      "epoch": 19.9488,
+      "grad_norm": 1.118355393409729,
+      "learning_rate": 1.8845538215286114e-05,
+      "loss": 0.5551,
+      "step": 15585
+    },
+    {
+      "epoch": 19.95008,
+      "grad_norm": 1.0390595197677612,
+      "learning_rate": 1.8843537414965986e-05,
+      "loss": 0.476,
+      "step": 15586
+    },
+    {
+      "epoch": 19.95136,
+      "grad_norm": 1.0548921823501587,
+      "learning_rate": 1.884153661464586e-05,
+      "loss": 0.5211,
+      "step": 15587
+    },
+    {
+      "epoch": 19.95264,
+      "grad_norm": 1.0728892087936401,
+      "learning_rate": 1.8839535814325733e-05,
+      "loss": 0.4795,
+      "step": 15588
+    },
+    {
+      "epoch": 19.95392,
+      "grad_norm": 1.0571420192718506,
+      "learning_rate": 1.88375350140056e-05,
+      "loss": 0.5225,
+      "step": 15589
+    },
+    {
+      "epoch": 19.9552,
+      "grad_norm": 1.0640194416046143,
+      "learning_rate": 1.8835534213685473e-05,
+      "loss": 0.4759,
+      "step": 15590
+    },
+    {
+      "epoch": 19.95648,
+      "grad_norm": 1.0471768379211426,
+      "learning_rate": 1.883353341336535e-05,
+      "loss": 0.4839,
+      "step": 15591
+    },
+    {
+      "epoch": 19.95776,
+      "grad_norm": 1.0839452743530273,
+      "learning_rate": 1.883153261304522e-05,
+      "loss": 0.522,
+      "step": 15592
+    },
+    {
+      "epoch": 19.95904,
+      "grad_norm": 1.0934643745422363,
+      "learning_rate": 1.882953181272509e-05,
+      "loss": 0.5176,
+      "step": 15593
+    },
+    {
+      "epoch": 19.96032,
+      "grad_norm": 1.0967458486557007,
+      "learning_rate": 1.8827531012404964e-05,
+      "loss": 0.5198,
+      "step": 15594
+    },
+    {
+      "epoch": 19.9616,
+      "grad_norm": 1.0761650800704956,
+      "learning_rate": 1.8825530212084836e-05,
+      "loss": 0.5091,
+      "step": 15595
+    },
+    {
+      "epoch": 19.96288,
+      "grad_norm": 1.0557310581207275,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.496,
+      "step": 15596
+    },
+    {
+      "epoch": 19.96416,
+      "grad_norm": 1.0711382627487183,
+      "learning_rate": 1.8821528611444576e-05,
+      "loss": 0.4703,
+      "step": 15597
+    },
+    {
+      "epoch": 19.96544,
+      "grad_norm": 1.1365162134170532,
+      "learning_rate": 1.881952781112445e-05,
+      "loss": 0.4936,
+      "step": 15598
+    },
+    {
+      "epoch": 19.96672,
+      "grad_norm": 1.141715168952942,
+      "learning_rate": 1.8817527010804323e-05,
+      "loss": 0.4871,
+      "step": 15599
+    },
+    {
+      "epoch": 19.968,
+      "grad_norm": 1.146036148071289,
+      "learning_rate": 1.8815526210484195e-05,
+      "loss": 0.5403,
+      "step": 15600
+    },
+    {
+      "epoch": 19.96928,
+      "grad_norm": 1.123439073562622,
+      "learning_rate": 1.8813525410164067e-05,
+      "loss": 0.5526,
+      "step": 15601
+    },
+    {
+      "epoch": 19.97056,
+      "grad_norm": 1.0493305921554565,
+      "learning_rate": 1.881152460984394e-05,
+      "loss": 0.4881,
+      "step": 15602
+    },
+    {
+      "epoch": 19.97184,
+      "grad_norm": 1.0788578987121582,
+      "learning_rate": 1.880952380952381e-05,
+      "loss": 0.481,
+      "step": 15603
+    },
+    {
+      "epoch": 19.97312,
+      "grad_norm": 1.089673399925232,
+      "learning_rate": 1.8807523009203682e-05,
+      "loss": 0.5252,
+      "step": 15604
+    },
+    {
+      "epoch": 19.9744,
+      "grad_norm": 1.0892081260681152,
+      "learning_rate": 1.8805522208883554e-05,
+      "loss": 0.4939,
+      "step": 15605
+    },
+    {
+      "epoch": 19.97568,
+      "grad_norm": 1.0614848136901855,
+      "learning_rate": 1.8803521408563426e-05,
+      "loss": 0.5036,
+      "step": 15606
+    },
+    {
+      "epoch": 19.97696,
+      "grad_norm": 1.028151512145996,
+      "learning_rate": 1.8801520608243298e-05,
+      "loss": 0.4862,
+      "step": 15607
+    },
+    {
+      "epoch": 19.97824,
+      "grad_norm": 1.0651766061782837,
+      "learning_rate": 1.879951980792317e-05,
+      "loss": 0.4842,
+      "step": 15608
+    },
+    {
+      "epoch": 19.97952,
+      "grad_norm": 1.0863500833511353,
+      "learning_rate": 1.8797519007603042e-05,
+      "loss": 0.4905,
+      "step": 15609
+    },
+    {
+      "epoch": 19.9808,
+      "grad_norm": 1.0720291137695312,
+      "learning_rate": 1.8795518207282914e-05,
+      "loss": 0.4991,
+      "step": 15610
+    },
+    {
+      "epoch": 19.98208,
+      "grad_norm": 1.0834654569625854,
+      "learning_rate": 1.8793517406962785e-05,
+      "loss": 0.5061,
+      "step": 15611
+    },
+    {
+      "epoch": 19.98336,
+      "grad_norm": 1.0607835054397583,
+      "learning_rate": 1.8791516606642657e-05,
+      "loss": 0.4861,
+      "step": 15612
+    },
+    {
+      "epoch": 19.98464,
+      "grad_norm": 1.1219713687896729,
+      "learning_rate": 1.878951580632253e-05,
+      "loss": 0.5357,
+      "step": 15613
+    },
+    {
+      "epoch": 19.98592,
+      "grad_norm": 1.078185796737671,
+      "learning_rate": 1.87875150060024e-05,
+      "loss": 0.5251,
+      "step": 15614
+    },
+    {
+      "epoch": 19.9872,
+      "grad_norm": 1.1310948133468628,
+      "learning_rate": 1.8785514205682273e-05,
+      "loss": 0.5558,
+      "step": 15615
+    },
+    {
+      "epoch": 19.98848,
+      "grad_norm": 1.0436378717422485,
+      "learning_rate": 1.8783513405362145e-05,
+      "loss": 0.4855,
+      "step": 15616
+    },
+    {
+      "epoch": 19.98976,
+      "grad_norm": 1.074711799621582,
+      "learning_rate": 1.8781512605042017e-05,
+      "loss": 0.4971,
+      "step": 15617
+    },
+    {
+      "epoch": 19.99104,
+      "grad_norm": 1.123464584350586,
+      "learning_rate": 1.877951180472189e-05,
+      "loss": 0.5288,
+      "step": 15618
+    },
+    {
+      "epoch": 19.99232,
+      "grad_norm": 1.015176773071289,
+      "learning_rate": 1.8777511004401764e-05,
+      "loss": 0.5137,
+      "step": 15619
+    },
+    {
+      "epoch": 19.9936,
+      "grad_norm": 1.0888524055480957,
+      "learning_rate": 1.8775510204081632e-05,
+      "loss": 0.4858,
+      "step": 15620
+    },
+    {
+      "epoch": 19.99488,
+      "grad_norm": 1.0319979190826416,
+      "learning_rate": 1.8773509403761504e-05,
+      "loss": 0.4687,
+      "step": 15621
+    },
+    {
+      "epoch": 19.99616,
+      "grad_norm": 1.116672396659851,
+      "learning_rate": 1.8771508603441376e-05,
+      "loss": 0.5169,
+      "step": 15622
+    },
+    {
+      "epoch": 19.99744,
+      "grad_norm": 1.0477895736694336,
+      "learning_rate": 1.876950780312125e-05,
+      "loss": 0.5392,
+      "step": 15623
+    },
+    {
+      "epoch": 19.99872,
+      "grad_norm": 1.069499135017395,
+      "learning_rate": 1.876750700280112e-05,
+      "loss": 0.4781,
+      "step": 15624
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 2.609760046005249,
+      "learning_rate": 1.876550620248099e-05,
+      "loss": 1.014,
+      "step": 15625
+    },
+    {
+      "epoch": 20.00128,
+      "grad_norm": 1.0402753353118896,
+      "learning_rate": 1.8763505402160867e-05,
+      "loss": 0.4348,
+      "step": 15626
+    },
+    {
+      "epoch": 20.00256,
+      "grad_norm": 1.0104725360870361,
+      "learning_rate": 1.876150460184074e-05,
+      "loss": 0.4641,
+      "step": 15627
+    },
+    {
+      "epoch": 20.00384,
+      "grad_norm": 1.1191972494125366,
+      "learning_rate": 1.8759503801520607e-05,
+      "loss": 0.5,
+      "step": 15628
+    },
+    {
+      "epoch": 20.00512,
+      "grad_norm": 1.0969229936599731,
+      "learning_rate": 1.875750300120048e-05,
+      "loss": 0.5265,
+      "step": 15629
+    },
+    {
+      "epoch": 20.0064,
+      "grad_norm": 1.0212663412094116,
+      "learning_rate": 1.8755502200880354e-05,
+      "loss": 0.4381,
+      "step": 15630
+    },
+    {
+      "epoch": 20.00768,
+      "grad_norm": 1.1064248085021973,
+      "learning_rate": 1.8753501400560226e-05,
+      "loss": 0.5045,
+      "step": 15631
+    },
+    {
+      "epoch": 20.00896,
+      "grad_norm": 1.134947419166565,
+      "learning_rate": 1.8751500600240094e-05,
+      "loss": 0.5008,
+      "step": 15632
+    },
+    {
+      "epoch": 20.01024,
+      "grad_norm": 1.0822664499282837,
+      "learning_rate": 1.874949979991997e-05,
+      "loss": 0.4941,
+      "step": 15633
+    },
+    {
+      "epoch": 20.01152,
+      "grad_norm": 1.0499417781829834,
+      "learning_rate": 1.874749899959984e-05,
+      "loss": 0.4508,
+      "step": 15634
+    },
+    {
+      "epoch": 20.0128,
+      "grad_norm": 1.050943374633789,
+      "learning_rate": 1.8745498199279713e-05,
+      "loss": 0.5099,
+      "step": 15635
+    },
+    {
+      "epoch": 20.01408,
+      "grad_norm": 1.016408085823059,
+      "learning_rate": 1.8743497398959582e-05,
+      "loss": 0.4515,
+      "step": 15636
+    },
+    {
+      "epoch": 20.01536,
+      "grad_norm": 1.0423884391784668,
+      "learning_rate": 1.8741496598639457e-05,
+      "loss": 0.5248,
+      "step": 15637
+    },
+    {
+      "epoch": 20.01664,
+      "grad_norm": 0.9838659763336182,
+      "learning_rate": 1.873949579831933e-05,
+      "loss": 0.4533,
+      "step": 15638
+    },
+    {
+      "epoch": 20.01792,
+      "grad_norm": 1.0517929792404175,
+      "learning_rate": 1.87374949979992e-05,
+      "loss": 0.4678,
+      "step": 15639
+    },
+    {
+      "epoch": 20.0192,
+      "grad_norm": 1.0576074123382568,
+      "learning_rate": 1.8735494197679073e-05,
+      "loss": 0.471,
+      "step": 15640
+    },
+    {
+      "epoch": 20.02048,
+      "grad_norm": 1.119922161102295,
+      "learning_rate": 1.8733493397358945e-05,
+      "loss": 0.5088,
+      "step": 15641
+    },
+    {
+      "epoch": 20.02176,
+      "grad_norm": 1.0552582740783691,
+      "learning_rate": 1.8731492597038816e-05,
+      "loss": 0.477,
+      "step": 15642
+    },
+    {
+      "epoch": 20.02304,
+      "grad_norm": 1.0699156522750854,
+      "learning_rate": 1.8729491796718688e-05,
+      "loss": 0.5483,
+      "step": 15643
+    },
+    {
+      "epoch": 20.02432,
+      "grad_norm": 1.020215630531311,
+      "learning_rate": 1.872749099639856e-05,
+      "loss": 0.4916,
+      "step": 15644
+    },
+    {
+      "epoch": 20.0256,
+      "grad_norm": 1.036571741104126,
+      "learning_rate": 1.8725490196078432e-05,
+      "loss": 0.4523,
+      "step": 15645
+    },
+    {
+      "epoch": 20.02688,
+      "grad_norm": 1.095461130142212,
+      "learning_rate": 1.8723489395758304e-05,
+      "loss": 0.4832,
+      "step": 15646
+    },
+    {
+      "epoch": 20.02816,
+      "grad_norm": 1.053385853767395,
+      "learning_rate": 1.872148859543818e-05,
+      "loss": 0.4627,
+      "step": 15647
+    },
+    {
+      "epoch": 20.02944,
+      "grad_norm": 1.0644296407699585,
+      "learning_rate": 1.8719487795118048e-05,
+      "loss": 0.4749,
+      "step": 15648
+    },
+    {
+      "epoch": 20.03072,
+      "grad_norm": 1.083106517791748,
+      "learning_rate": 1.871748699479792e-05,
+      "loss": 0.4926,
+      "step": 15649
+    },
+    {
+      "epoch": 20.032,
+      "grad_norm": 1.06119704246521,
+      "learning_rate": 1.871548619447779e-05,
+      "loss": 0.4924,
+      "step": 15650
+    },
+    {
+      "epoch": 20.03328,
+      "grad_norm": 1.0661925077438354,
+      "learning_rate": 1.8713485394157667e-05,
+      "loss": 0.4581,
+      "step": 15651
+    },
+    {
+      "epoch": 20.03456,
+      "grad_norm": 1.1228007078170776,
+      "learning_rate": 1.8711484593837535e-05,
+      "loss": 0.5334,
+      "step": 15652
+    },
+    {
+      "epoch": 20.03584,
+      "grad_norm": 1.0884519815444946,
+      "learning_rate": 1.8709483793517407e-05,
+      "loss": 0.4621,
+      "step": 15653
+    },
+    {
+      "epoch": 20.03712,
+      "grad_norm": 1.0557575225830078,
+      "learning_rate": 1.8707482993197282e-05,
+      "loss": 0.4675,
+      "step": 15654
+    },
+    {
+      "epoch": 20.0384,
+      "grad_norm": 1.0507951974868774,
+      "learning_rate": 1.8705482192877154e-05,
+      "loss": 0.4968,
+      "step": 15655
+    },
+    {
+      "epoch": 20.03968,
+      "grad_norm": 1.0624488592147827,
+      "learning_rate": 1.8703481392557022e-05,
+      "loss": 0.4994,
+      "step": 15656
+    },
+    {
+      "epoch": 20.04096,
+      "grad_norm": 1.0612828731536865,
+      "learning_rate": 1.8701480592236894e-05,
+      "loss": 0.4935,
+      "step": 15657
+    },
+    {
+      "epoch": 20.04224,
+      "grad_norm": 1.0293620824813843,
+      "learning_rate": 1.869947979191677e-05,
+      "loss": 0.4321,
+      "step": 15658
+    },
+    {
+      "epoch": 20.04352,
+      "grad_norm": 1.074039101600647,
+      "learning_rate": 1.869747899159664e-05,
+      "loss": 0.474,
+      "step": 15659
+    },
+    {
+      "epoch": 20.0448,
+      "grad_norm": 1.0148733854293823,
+      "learning_rate": 1.869547819127651e-05,
+      "loss": 0.478,
+      "step": 15660
+    },
+    {
+      "epoch": 20.04608,
+      "grad_norm": 1.1181195974349976,
+      "learning_rate": 1.8693477390956382e-05,
+      "loss": 0.5195,
+      "step": 15661
+    },
+    {
+      "epoch": 20.04736,
+      "grad_norm": 1.0959631204605103,
+      "learning_rate": 1.8691476590636257e-05,
+      "loss": 0.5102,
+      "step": 15662
+    },
+    {
+      "epoch": 20.04864,
+      "grad_norm": 1.070541262626648,
+      "learning_rate": 1.868947579031613e-05,
+      "loss": 0.4831,
+      "step": 15663
+    },
+    {
+      "epoch": 20.04992,
+      "grad_norm": 1.032403826713562,
+      "learning_rate": 1.8687474989995997e-05,
+      "loss": 0.4576,
+      "step": 15664
+    },
+    {
+      "epoch": 20.0512,
+      "grad_norm": 0.9710387587547302,
+      "learning_rate": 1.8685474189675873e-05,
+      "loss": 0.464,
+      "step": 15665
+    },
+    {
+      "epoch": 20.05248,
+      "grad_norm": 1.1018273830413818,
+      "learning_rate": 1.8683473389355744e-05,
+      "loss": 0.4972,
+      "step": 15666
+    },
+    {
+      "epoch": 20.05376,
+      "grad_norm": 0.9961934089660645,
+      "learning_rate": 1.8681472589035616e-05,
+      "loss": 0.4921,
+      "step": 15667
+    },
+    {
+      "epoch": 20.05504,
+      "grad_norm": 1.1243575811386108,
+      "learning_rate": 1.8679471788715485e-05,
+      "loss": 0.4818,
+      "step": 15668
+    },
+    {
+      "epoch": 20.05632,
+      "grad_norm": 1.05106520652771,
+      "learning_rate": 1.867747098839536e-05,
+      "loss": 0.4697,
+      "step": 15669
+    },
+    {
+      "epoch": 20.0576,
+      "grad_norm": 1.0926669836044312,
+      "learning_rate": 1.8675470188075232e-05,
+      "loss": 0.5248,
+      "step": 15670
+    },
+    {
+      "epoch": 20.05888,
+      "grad_norm": 1.1060155630111694,
+      "learning_rate": 1.8673469387755104e-05,
+      "loss": 0.4986,
+      "step": 15671
+    },
+    {
+      "epoch": 20.06016,
+      "grad_norm": 1.0984371900558472,
+      "learning_rate": 1.8671468587434976e-05,
+      "loss": 0.5017,
+      "step": 15672
+    },
+    {
+      "epoch": 20.06144,
+      "grad_norm": 1.0190678834915161,
+      "learning_rate": 1.8669467787114847e-05,
+      "loss": 0.4497,
+      "step": 15673
+    },
+    {
+      "epoch": 20.06272,
+      "grad_norm": 1.1226532459259033,
+      "learning_rate": 1.866746698679472e-05,
+      "loss": 0.5061,
+      "step": 15674
+    },
+    {
+      "epoch": 20.064,
+      "grad_norm": 1.016611099243164,
+      "learning_rate": 1.866546618647459e-05,
+      "loss": 0.4425,
+      "step": 15675
+    },
+    {
+      "epoch": 20.06528,
+      "grad_norm": 1.1587402820587158,
+      "learning_rate": 1.8663465386154463e-05,
+      "loss": 0.4797,
+      "step": 15676
+    },
+    {
+      "epoch": 20.06656,
+      "grad_norm": 1.0697815418243408,
+      "learning_rate": 1.8661464585834335e-05,
+      "loss": 0.4623,
+      "step": 15677
+    },
+    {
+      "epoch": 20.06784,
+      "grad_norm": 1.0413057804107666,
+      "learning_rate": 1.8659463785514207e-05,
+      "loss": 0.4997,
+      "step": 15678
+    },
+    {
+      "epoch": 20.06912,
+      "grad_norm": 1.0416499376296997,
+      "learning_rate": 1.865746298519408e-05,
+      "loss": 0.4707,
+      "step": 15679
+    },
+    {
+      "epoch": 20.0704,
+      "grad_norm": 1.034729242324829,
+      "learning_rate": 1.865546218487395e-05,
+      "loss": 0.4702,
+      "step": 15680
+    },
+    {
+      "epoch": 20.07168,
+      "grad_norm": 1.1289206743240356,
+      "learning_rate": 1.8653461384553822e-05,
+      "loss": 0.4843,
+      "step": 15681
+    },
+    {
+      "epoch": 20.07296,
+      "grad_norm": 1.0934054851531982,
+      "learning_rate": 1.8651460584233694e-05,
+      "loss": 0.4927,
+      "step": 15682
+    },
+    {
+      "epoch": 20.07424,
+      "grad_norm": 1.0436177253723145,
+      "learning_rate": 1.8649459783913566e-05,
+      "loss": 0.49,
+      "step": 15683
+    },
+    {
+      "epoch": 20.07552,
+      "grad_norm": 1.0504745244979858,
+      "learning_rate": 1.8647458983593438e-05,
+      "loss": 0.4885,
+      "step": 15684
+    },
+    {
+      "epoch": 20.0768,
+      "grad_norm": 1.0551018714904785,
+      "learning_rate": 1.864545818327331e-05,
+      "loss": 0.4789,
+      "step": 15685
+    },
+    {
+      "epoch": 20.07808,
+      "grad_norm": 1.0579137802124023,
+      "learning_rate": 1.8643457382953185e-05,
+      "loss": 0.4515,
+      "step": 15686
+    },
+    {
+      "epoch": 20.07936,
+      "grad_norm": 1.1196202039718628,
+      "learning_rate": 1.8641456582633053e-05,
+      "loss": 0.5234,
+      "step": 15687
+    },
+    {
+      "epoch": 20.08064,
+      "grad_norm": 1.0107567310333252,
+      "learning_rate": 1.8639455782312925e-05,
+      "loss": 0.4665,
+      "step": 15688
+    },
+    {
+      "epoch": 20.08192,
+      "grad_norm": 1.0458314418792725,
+      "learning_rate": 1.8637454981992797e-05,
+      "loss": 0.4644,
+      "step": 15689
+    },
+    {
+      "epoch": 20.0832,
+      "grad_norm": 1.0873018503189087,
+      "learning_rate": 1.8635454181672672e-05,
+      "loss": 0.4421,
+      "step": 15690
+    },
+    {
+      "epoch": 20.08448,
+      "grad_norm": 1.0975297689437866,
+      "learning_rate": 1.863345338135254e-05,
+      "loss": 0.5099,
+      "step": 15691
+    },
+    {
+      "epoch": 20.08576,
+      "grad_norm": 1.0815826654434204,
+      "learning_rate": 1.8631452581032413e-05,
+      "loss": 0.4522,
+      "step": 15692
+    },
+    {
+      "epoch": 20.087040000000002,
+      "grad_norm": 1.0389448404312134,
+      "learning_rate": 1.8629451780712288e-05,
+      "loss": 0.4587,
+      "step": 15693
+    },
+    {
+      "epoch": 20.08832,
+      "grad_norm": 1.0936323404312134,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.5339,
+      "step": 15694
+    },
+    {
+      "epoch": 20.0896,
+      "grad_norm": 1.065900444984436,
+      "learning_rate": 1.8625450180072028e-05,
+      "loss": 0.4701,
+      "step": 15695
+    },
+    {
+      "epoch": 20.09088,
+      "grad_norm": 1.0863001346588135,
+      "learning_rate": 1.86234493797519e-05,
+      "loss": 0.5019,
+      "step": 15696
+    },
+    {
+      "epoch": 20.09216,
+      "grad_norm": 1.1467665433883667,
+      "learning_rate": 1.8621448579431775e-05,
+      "loss": 0.5244,
+      "step": 15697
+    },
+    {
+      "epoch": 20.09344,
+      "grad_norm": 1.0635558366775513,
+      "learning_rate": 1.8619447779111647e-05,
+      "loss": 0.4646,
+      "step": 15698
+    },
+    {
+      "epoch": 20.09472,
+      "grad_norm": 1.0072330236434937,
+      "learning_rate": 1.8617446978791516e-05,
+      "loss": 0.4558,
+      "step": 15699
+    },
+    {
+      "epoch": 20.096,
+      "grad_norm": 1.0939109325408936,
+      "learning_rate": 1.861544617847139e-05,
+      "loss": 0.5301,
+      "step": 15700
+    },
+    {
+      "epoch": 20.09728,
+      "grad_norm": 1.1008801460266113,
+      "learning_rate": 1.8613445378151263e-05,
+      "loss": 0.4859,
+      "step": 15701
+    },
+    {
+      "epoch": 20.09856,
+      "grad_norm": 1.1168580055236816,
+      "learning_rate": 1.8611444577831135e-05,
+      "loss": 0.4658,
+      "step": 15702
+    },
+    {
+      "epoch": 20.09984,
+      "grad_norm": 1.0580940246582031,
+      "learning_rate": 1.8609443777511003e-05,
+      "loss": 0.4434,
+      "step": 15703
+    },
+    {
+      "epoch": 20.10112,
+      "grad_norm": 1.0258312225341797,
+      "learning_rate": 1.860744297719088e-05,
+      "loss": 0.4644,
+      "step": 15704
+    },
+    {
+      "epoch": 20.1024,
+      "grad_norm": 1.0801990032196045,
+      "learning_rate": 1.860544217687075e-05,
+      "loss": 0.4959,
+      "step": 15705
+    },
+    {
+      "epoch": 20.10368,
+      "grad_norm": 1.0814316272735596,
+      "learning_rate": 1.8603441376550622e-05,
+      "loss": 0.4898,
+      "step": 15706
+    },
+    {
+      "epoch": 20.10496,
+      "grad_norm": 1.116225004196167,
+      "learning_rate": 1.8601440576230494e-05,
+      "loss": 0.4865,
+      "step": 15707
+    },
+    {
+      "epoch": 20.10624,
+      "grad_norm": 1.147642970085144,
+      "learning_rate": 1.8599439775910366e-05,
+      "loss": 0.5007,
+      "step": 15708
+    },
+    {
+      "epoch": 20.10752,
+      "grad_norm": 1.1778041124343872,
+      "learning_rate": 1.8597438975590238e-05,
+      "loss": 0.5123,
+      "step": 15709
+    },
+    {
+      "epoch": 20.1088,
+      "grad_norm": 1.0506484508514404,
+      "learning_rate": 1.859543817527011e-05,
+      "loss": 0.4695,
+      "step": 15710
+    },
+    {
+      "epoch": 20.11008,
+      "grad_norm": 1.0645596981048584,
+      "learning_rate": 1.859343737494998e-05,
+      "loss": 0.4784,
+      "step": 15711
+    },
+    {
+      "epoch": 20.11136,
+      "grad_norm": 1.086012840270996,
+      "learning_rate": 1.8591436574629853e-05,
+      "loss": 0.4975,
+      "step": 15712
+    },
+    {
+      "epoch": 20.11264,
+      "grad_norm": 1.0438339710235596,
+      "learning_rate": 1.8589435774309725e-05,
+      "loss": 0.4973,
+      "step": 15713
+    },
+    {
+      "epoch": 20.11392,
+      "grad_norm": 1.1068918704986572,
+      "learning_rate": 1.8587434973989597e-05,
+      "loss": 0.4846,
+      "step": 15714
+    },
+    {
+      "epoch": 20.1152,
+      "grad_norm": 1.1290284395217896,
+      "learning_rate": 1.858543417366947e-05,
+      "loss": 0.5336,
+      "step": 15715
+    },
+    {
+      "epoch": 20.11648,
+      "grad_norm": 1.1011111736297607,
+      "learning_rate": 1.858343337334934e-05,
+      "loss": 0.4703,
+      "step": 15716
+    },
+    {
+      "epoch": 20.11776,
+      "grad_norm": 1.1040534973144531,
+      "learning_rate": 1.8581432573029212e-05,
+      "loss": 0.4649,
+      "step": 15717
+    },
+    {
+      "epoch": 20.11904,
+      "grad_norm": 1.0694252252578735,
+      "learning_rate": 1.8579431772709084e-05,
+      "loss": 0.4411,
+      "step": 15718
+    },
+    {
+      "epoch": 20.12032,
+      "grad_norm": 1.0621991157531738,
+      "learning_rate": 1.8577430972388956e-05,
+      "loss": 0.4636,
+      "step": 15719
+    },
+    {
+      "epoch": 20.1216,
+      "grad_norm": 1.0658694505691528,
+      "learning_rate": 1.8575430172068828e-05,
+      "loss": 0.4973,
+      "step": 15720
+    },
+    {
+      "epoch": 20.12288,
+      "grad_norm": 1.0541012287139893,
+      "learning_rate": 1.85734293717487e-05,
+      "loss": 0.492,
+      "step": 15721
+    },
+    {
+      "epoch": 20.12416,
+      "grad_norm": 1.139356255531311,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 0.4854,
+      "step": 15722
+    },
+    {
+      "epoch": 20.12544,
+      "grad_norm": 1.0738463401794434,
+      "learning_rate": 1.8569427771108444e-05,
+      "loss": 0.5184,
+      "step": 15723
+    },
+    {
+      "epoch": 20.12672,
+      "grad_norm": 1.042602777481079,
+      "learning_rate": 1.8567426970788315e-05,
+      "loss": 0.4965,
+      "step": 15724
+    },
+    {
+      "epoch": 20.128,
+      "grad_norm": 1.100381851196289,
+      "learning_rate": 1.856542617046819e-05,
+      "loss": 0.52,
+      "step": 15725
+    },
+    {
+      "epoch": 20.12928,
+      "grad_norm": 1.0966511964797974,
+      "learning_rate": 1.856342537014806e-05,
+      "loss": 0.5214,
+      "step": 15726
+    },
+    {
+      "epoch": 20.13056,
+      "grad_norm": 1.1034797430038452,
+      "learning_rate": 1.856142456982793e-05,
+      "loss": 0.4959,
+      "step": 15727
+    },
+    {
+      "epoch": 20.13184,
+      "grad_norm": 1.1695610284805298,
+      "learning_rate": 1.8559423769507803e-05,
+      "loss": 0.5517,
+      "step": 15728
+    },
+    {
+      "epoch": 20.13312,
+      "grad_norm": 1.1129649877548218,
+      "learning_rate": 1.8557422969187678e-05,
+      "loss": 0.4974,
+      "step": 15729
+    },
+    {
+      "epoch": 20.1344,
+      "grad_norm": 1.130495309829712,
+      "learning_rate": 1.8555422168867547e-05,
+      "loss": 0.4873,
+      "step": 15730
+    },
+    {
+      "epoch": 20.13568,
+      "grad_norm": 1.099576473236084,
+      "learning_rate": 1.855342136854742e-05,
+      "loss": 0.4751,
+      "step": 15731
+    },
+    {
+      "epoch": 20.13696,
+      "grad_norm": 1.0953304767608643,
+      "learning_rate": 1.8551420568227294e-05,
+      "loss": 0.5135,
+      "step": 15732
+    },
+    {
+      "epoch": 20.13824,
+      "grad_norm": 1.0430271625518799,
+      "learning_rate": 1.8549419767907166e-05,
+      "loss": 0.4849,
+      "step": 15733
+    },
+    {
+      "epoch": 20.13952,
+      "grad_norm": 1.1032804250717163,
+      "learning_rate": 1.8547418967587034e-05,
+      "loss": 0.5093,
+      "step": 15734
+    },
+    {
+      "epoch": 20.1408,
+      "grad_norm": 1.1103456020355225,
+      "learning_rate": 1.8545418167266906e-05,
+      "loss": 0.5148,
+      "step": 15735
+    },
+    {
+      "epoch": 20.14208,
+      "grad_norm": 1.1591553688049316,
+      "learning_rate": 1.854341736694678e-05,
+      "loss": 0.5335,
+      "step": 15736
+    },
+    {
+      "epoch": 20.14336,
+      "grad_norm": 1.0609407424926758,
+      "learning_rate": 1.8541416566626653e-05,
+      "loss": 0.469,
+      "step": 15737
+    },
+    {
+      "epoch": 20.14464,
+      "grad_norm": 1.0772405862808228,
+      "learning_rate": 1.853941576630652e-05,
+      "loss": 0.522,
+      "step": 15738
+    },
+    {
+      "epoch": 20.14592,
+      "grad_norm": 1.2019360065460205,
+      "learning_rate": 1.8537414965986397e-05,
+      "loss": 0.5215,
+      "step": 15739
+    },
+    {
+      "epoch": 20.1472,
+      "grad_norm": 1.0603306293487549,
+      "learning_rate": 1.853541416566627e-05,
+      "loss": 0.4832,
+      "step": 15740
+    },
+    {
+      "epoch": 20.14848,
+      "grad_norm": 1.1025315523147583,
+      "learning_rate": 1.853341336534614e-05,
+      "loss": 0.5234,
+      "step": 15741
+    },
+    {
+      "epoch": 20.14976,
+      "grad_norm": 1.0431936979293823,
+      "learning_rate": 1.853141256502601e-05,
+      "loss": 0.464,
+      "step": 15742
+    },
+    {
+      "epoch": 20.15104,
+      "grad_norm": 1.0916297435760498,
+      "learning_rate": 1.8529411764705884e-05,
+      "loss": 0.4768,
+      "step": 15743
+    },
+    {
+      "epoch": 20.15232,
+      "grad_norm": 1.0731911659240723,
+      "learning_rate": 1.8527410964385756e-05,
+      "loss": 0.4871,
+      "step": 15744
+    },
+    {
+      "epoch": 20.1536,
+      "grad_norm": 1.060876488685608,
+      "learning_rate": 1.8525410164065628e-05,
+      "loss": 0.4926,
+      "step": 15745
+    },
+    {
+      "epoch": 20.15488,
+      "grad_norm": 1.0614646673202515,
+      "learning_rate": 1.85234093637455e-05,
+      "loss": 0.4415,
+      "step": 15746
+    },
+    {
+      "epoch": 20.15616,
+      "grad_norm": 1.1642632484436035,
+      "learning_rate": 1.852140856342537e-05,
+      "loss": 0.5138,
+      "step": 15747
+    },
+    {
+      "epoch": 20.15744,
+      "grad_norm": 1.0110687017440796,
+      "learning_rate": 1.8519407763105243e-05,
+      "loss": 0.4631,
+      "step": 15748
+    },
+    {
+      "epoch": 20.15872,
+      "grad_norm": 1.130932331085205,
+      "learning_rate": 1.8517406962785115e-05,
+      "loss": 0.4553,
+      "step": 15749
+    },
+    {
+      "epoch": 20.16,
+      "grad_norm": 0.9995177388191223,
+      "learning_rate": 1.8515406162464987e-05,
+      "loss": 0.4915,
+      "step": 15750
+    },
+    {
+      "epoch": 20.16128,
+      "grad_norm": 1.0752531290054321,
+      "learning_rate": 1.851340536214486e-05,
+      "loss": 0.5099,
+      "step": 15751
+    },
+    {
+      "epoch": 20.16256,
+      "grad_norm": 1.0507272481918335,
+      "learning_rate": 1.851140456182473e-05,
+      "loss": 0.4755,
+      "step": 15752
+    },
+    {
+      "epoch": 20.16384,
+      "grad_norm": 1.1398614645004272,
+      "learning_rate": 1.8509403761504603e-05,
+      "loss": 0.5324,
+      "step": 15753
+    },
+    {
+      "epoch": 20.16512,
+      "grad_norm": 0.984160840511322,
+      "learning_rate": 1.8507402961184475e-05,
+      "loss": 0.4188,
+      "step": 15754
+    },
+    {
+      "epoch": 20.1664,
+      "grad_norm": 1.1006124019622803,
+      "learning_rate": 1.8505402160864346e-05,
+      "loss": 0.5033,
+      "step": 15755
+    },
+    {
+      "epoch": 20.16768,
+      "grad_norm": 1.0734970569610596,
+      "learning_rate": 1.8503401360544218e-05,
+      "loss": 0.4785,
+      "step": 15756
+    },
+    {
+      "epoch": 20.16896,
+      "grad_norm": 1.1505695581436157,
+      "learning_rate": 1.850140056022409e-05,
+      "loss": 0.5547,
+      "step": 15757
+    },
+    {
+      "epoch": 20.17024,
+      "grad_norm": 1.147254228591919,
+      "learning_rate": 1.8499399759903962e-05,
+      "loss": 0.5077,
+      "step": 15758
+    },
+    {
+      "epoch": 20.17152,
+      "grad_norm": 1.0394471883773804,
+      "learning_rate": 1.8497398959583834e-05,
+      "loss": 0.465,
+      "step": 15759
+    },
+    {
+      "epoch": 20.1728,
+      "grad_norm": 1.0904779434204102,
+      "learning_rate": 1.849539815926371e-05,
+      "loss": 0.4909,
+      "step": 15760
+    },
+    {
+      "epoch": 20.17408,
+      "grad_norm": 1.2574462890625,
+      "learning_rate": 1.8493397358943578e-05,
+      "loss": 0.4928,
+      "step": 15761
+    },
+    {
+      "epoch": 20.17536,
+      "grad_norm": 1.0924097299575806,
+      "learning_rate": 1.849139655862345e-05,
+      "loss": 0.4794,
+      "step": 15762
+    },
+    {
+      "epoch": 20.17664,
+      "grad_norm": 1.0930646657943726,
+      "learning_rate": 1.848939575830332e-05,
+      "loss": 0.4895,
+      "step": 15763
+    },
+    {
+      "epoch": 20.17792,
+      "grad_norm": 1.0378696918487549,
+      "learning_rate": 1.8487394957983196e-05,
+      "loss": 0.5068,
+      "step": 15764
+    },
+    {
+      "epoch": 20.1792,
+      "grad_norm": 1.121130347251892,
+      "learning_rate": 1.8485394157663065e-05,
+      "loss": 0.49,
+      "step": 15765
+    },
+    {
+      "epoch": 20.18048,
+      "grad_norm": 1.1667362451553345,
+      "learning_rate": 1.8483393357342937e-05,
+      "loss": 0.4976,
+      "step": 15766
+    },
+    {
+      "epoch": 20.18176,
+      "grad_norm": 1.0337846279144287,
+      "learning_rate": 1.8481392557022812e-05,
+      "loss": 0.476,
+      "step": 15767
+    },
+    {
+      "epoch": 20.18304,
+      "grad_norm": 1.0466781854629517,
+      "learning_rate": 1.8479391756702684e-05,
+      "loss": 0.4418,
+      "step": 15768
+    },
+    {
+      "epoch": 20.18432,
+      "grad_norm": 1.1099622249603271,
+      "learning_rate": 1.8477390956382552e-05,
+      "loss": 0.5132,
+      "step": 15769
+    },
+    {
+      "epoch": 20.1856,
+      "grad_norm": 1.1434911489486694,
+      "learning_rate": 1.8475390156062424e-05,
+      "loss": 0.5056,
+      "step": 15770
+    },
+    {
+      "epoch": 20.18688,
+      "grad_norm": 1.0886666774749756,
+      "learning_rate": 1.84733893557423e-05,
+      "loss": 0.447,
+      "step": 15771
+    },
+    {
+      "epoch": 20.18816,
+      "grad_norm": 1.051747441291809,
+      "learning_rate": 1.847138855542217e-05,
+      "loss": 0.4715,
+      "step": 15772
+    },
+    {
+      "epoch": 20.18944,
+      "grad_norm": 1.1362502574920654,
+      "learning_rate": 1.846938775510204e-05,
+      "loss": 0.4843,
+      "step": 15773
+    },
+    {
+      "epoch": 20.19072,
+      "grad_norm": 1.0374624729156494,
+      "learning_rate": 1.846738695478191e-05,
+      "loss": 0.4653,
+      "step": 15774
+    },
+    {
+      "epoch": 20.192,
+      "grad_norm": 1.105280876159668,
+      "learning_rate": 1.8465386154461787e-05,
+      "loss": 0.4934,
+      "step": 15775
+    },
+    {
+      "epoch": 20.19328,
+      "grad_norm": 1.1130280494689941,
+      "learning_rate": 1.846338535414166e-05,
+      "loss": 0.5158,
+      "step": 15776
+    },
+    {
+      "epoch": 20.19456,
+      "grad_norm": 1.0840026140213013,
+      "learning_rate": 1.8461384553821527e-05,
+      "loss": 0.4837,
+      "step": 15777
+    },
+    {
+      "epoch": 20.19584,
+      "grad_norm": 1.0840171575546265,
+      "learning_rate": 1.8459383753501402e-05,
+      "loss": 0.4841,
+      "step": 15778
+    },
+    {
+      "epoch": 20.19712,
+      "grad_norm": 1.044407844543457,
+      "learning_rate": 1.8457382953181274e-05,
+      "loss": 0.4895,
+      "step": 15779
+    },
+    {
+      "epoch": 20.1984,
+      "grad_norm": 1.110978603363037,
+      "learning_rate": 1.8455382152861146e-05,
+      "loss": 0.4946,
+      "step": 15780
+    },
+    {
+      "epoch": 20.19968,
+      "grad_norm": 1.154561161994934,
+      "learning_rate": 1.8453381352541015e-05,
+      "loss": 0.5201,
+      "step": 15781
+    },
+    {
+      "epoch": 20.20096,
+      "grad_norm": 1.1208866834640503,
+      "learning_rate": 1.845138055222089e-05,
+      "loss": 0.4845,
+      "step": 15782
+    },
+    {
+      "epoch": 20.20224,
+      "grad_norm": 1.111349105834961,
+      "learning_rate": 1.8449379751900762e-05,
+      "loss": 0.4799,
+      "step": 15783
+    },
+    {
+      "epoch": 20.20352,
+      "grad_norm": 1.1092686653137207,
+      "learning_rate": 1.8447378951580634e-05,
+      "loss": 0.4641,
+      "step": 15784
+    },
+    {
+      "epoch": 20.2048,
+      "grad_norm": 1.198625922203064,
+      "learning_rate": 1.8445378151260505e-05,
+      "loss": 0.498,
+      "step": 15785
+    },
+    {
+      "epoch": 20.20608,
+      "grad_norm": 1.128433108329773,
+      "learning_rate": 1.8443377350940377e-05,
+      "loss": 0.4709,
+      "step": 15786
+    },
+    {
+      "epoch": 20.20736,
+      "grad_norm": 1.02711820602417,
+      "learning_rate": 1.844137655062025e-05,
+      "loss": 0.4623,
+      "step": 15787
+    },
+    {
+      "epoch": 20.20864,
+      "grad_norm": 1.1294431686401367,
+      "learning_rate": 1.843937575030012e-05,
+      "loss": 0.4966,
+      "step": 15788
+    },
+    {
+      "epoch": 20.20992,
+      "grad_norm": 1.0591932535171509,
+      "learning_rate": 1.8437374949979993e-05,
+      "loss": 0.4802,
+      "step": 15789
+    },
+    {
+      "epoch": 20.2112,
+      "grad_norm": 1.1523479223251343,
+      "learning_rate": 1.8435374149659865e-05,
+      "loss": 0.5155,
+      "step": 15790
+    },
+    {
+      "epoch": 20.21248,
+      "grad_norm": 1.1084343194961548,
+      "learning_rate": 1.8433373349339737e-05,
+      "loss": 0.5098,
+      "step": 15791
+    },
+    {
+      "epoch": 20.21376,
+      "grad_norm": 1.1509208679199219,
+      "learning_rate": 1.843137254901961e-05,
+      "loss": 0.5099,
+      "step": 15792
+    },
+    {
+      "epoch": 20.21504,
+      "grad_norm": 1.1288859844207764,
+      "learning_rate": 1.842937174869948e-05,
+      "loss": 0.482,
+      "step": 15793
+    },
+    {
+      "epoch": 20.21632,
+      "grad_norm": 1.039110779762268,
+      "learning_rate": 1.8427370948379352e-05,
+      "loss": 0.4622,
+      "step": 15794
+    },
+    {
+      "epoch": 20.2176,
+      "grad_norm": 1.1868407726287842,
+      "learning_rate": 1.8425370148059224e-05,
+      "loss": 0.5312,
+      "step": 15795
+    },
+    {
+      "epoch": 20.21888,
+      "grad_norm": 1.0792521238327026,
+      "learning_rate": 1.8423369347739096e-05,
+      "loss": 0.4415,
+      "step": 15796
+    },
+    {
+      "epoch": 20.22016,
+      "grad_norm": 1.0646893978118896,
+      "learning_rate": 1.8421368547418968e-05,
+      "loss": 0.4616,
+      "step": 15797
+    },
+    {
+      "epoch": 20.22144,
+      "grad_norm": 1.0488739013671875,
+      "learning_rate": 1.841936774709884e-05,
+      "loss": 0.4913,
+      "step": 15798
+    },
+    {
+      "epoch": 20.22272,
+      "grad_norm": 1.152377963066101,
+      "learning_rate": 1.8417366946778715e-05,
+      "loss": 0.5146,
+      "step": 15799
+    },
+    {
+      "epoch": 20.224,
+      "grad_norm": 1.0589985847473145,
+      "learning_rate": 1.8415366146458583e-05,
+      "loss": 0.4561,
+      "step": 15800
+    },
+    {
+      "epoch": 20.22528,
+      "grad_norm": 1.1091581583023071,
+      "learning_rate": 1.8413365346138455e-05,
+      "loss": 0.5,
+      "step": 15801
+    },
+    {
+      "epoch": 20.22656,
+      "grad_norm": 1.1588554382324219,
+      "learning_rate": 1.8411364545818327e-05,
+      "loss": 0.4701,
+      "step": 15802
+    },
+    {
+      "epoch": 20.22784,
+      "grad_norm": 1.0531977415084839,
+      "learning_rate": 1.8409363745498202e-05,
+      "loss": 0.4541,
+      "step": 15803
+    },
+    {
+      "epoch": 20.22912,
+      "grad_norm": 1.1038832664489746,
+      "learning_rate": 1.840736294517807e-05,
+      "loss": 0.4773,
+      "step": 15804
+    },
+    {
+      "epoch": 20.2304,
+      "grad_norm": 1.0699522495269775,
+      "learning_rate": 1.8405362144857943e-05,
+      "loss": 0.4831,
+      "step": 15805
+    },
+    {
+      "epoch": 20.23168,
+      "grad_norm": 1.085437297821045,
+      "learning_rate": 1.8403361344537818e-05,
+      "loss": 0.4914,
+      "step": 15806
+    },
+    {
+      "epoch": 20.23296,
+      "grad_norm": 1.104048728942871,
+      "learning_rate": 1.840136054421769e-05,
+      "loss": 0.4827,
+      "step": 15807
+    },
+    {
+      "epoch": 20.23424,
+      "grad_norm": 1.0480152368545532,
+      "learning_rate": 1.8399359743897558e-05,
+      "loss": 0.4724,
+      "step": 15808
+    },
+    {
+      "epoch": 20.23552,
+      "grad_norm": 1.1070879697799683,
+      "learning_rate": 1.839735894357743e-05,
+      "loss": 0.485,
+      "step": 15809
+    },
+    {
+      "epoch": 20.2368,
+      "grad_norm": 1.0816839933395386,
+      "learning_rate": 1.8395358143257305e-05,
+      "loss": 0.4814,
+      "step": 15810
+    },
+    {
+      "epoch": 20.23808,
+      "grad_norm": 1.0380091667175293,
+      "learning_rate": 1.8393357342937177e-05,
+      "loss": 0.483,
+      "step": 15811
+    },
+    {
+      "epoch": 20.23936,
+      "grad_norm": 1.0764389038085938,
+      "learning_rate": 1.8391356542617046e-05,
+      "loss": 0.4797,
+      "step": 15812
+    },
+    {
+      "epoch": 20.24064,
+      "grad_norm": 1.1139464378356934,
+      "learning_rate": 1.838935574229692e-05,
+      "loss": 0.5232,
+      "step": 15813
+    },
+    {
+      "epoch": 20.24192,
+      "grad_norm": 1.0700820684432983,
+      "learning_rate": 1.8387354941976793e-05,
+      "loss": 0.434,
+      "step": 15814
+    },
+    {
+      "epoch": 20.2432,
+      "grad_norm": 1.0398091077804565,
+      "learning_rate": 1.8385354141656665e-05,
+      "loss": 0.4793,
+      "step": 15815
+    },
+    {
+      "epoch": 20.24448,
+      "grad_norm": 1.1150001287460327,
+      "learning_rate": 1.8383353341336533e-05,
+      "loss": 0.4747,
+      "step": 15816
+    },
+    {
+      "epoch": 20.24576,
+      "grad_norm": 1.1604424715042114,
+      "learning_rate": 1.8381352541016408e-05,
+      "loss": 0.5042,
+      "step": 15817
+    },
+    {
+      "epoch": 20.24704,
+      "grad_norm": 1.0625085830688477,
+      "learning_rate": 1.837935174069628e-05,
+      "loss": 0.4773,
+      "step": 15818
+    },
+    {
+      "epoch": 20.24832,
+      "grad_norm": 1.1010764837265015,
+      "learning_rate": 1.8377350940376152e-05,
+      "loss": 0.458,
+      "step": 15819
+    },
+    {
+      "epoch": 20.2496,
+      "grad_norm": 1.1224896907806396,
+      "learning_rate": 1.8375350140056024e-05,
+      "loss": 0.4832,
+      "step": 15820
+    },
+    {
+      "epoch": 20.25088,
+      "grad_norm": 1.1224029064178467,
+      "learning_rate": 1.8373349339735896e-05,
+      "loss": 0.4807,
+      "step": 15821
+    },
+    {
+      "epoch": 20.25216,
+      "grad_norm": 1.0098991394042969,
+      "learning_rate": 1.8371348539415768e-05,
+      "loss": 0.4576,
+      "step": 15822
+    },
+    {
+      "epoch": 20.25344,
+      "grad_norm": 1.065501093864441,
+      "learning_rate": 1.836934773909564e-05,
+      "loss": 0.4252,
+      "step": 15823
+    },
+    {
+      "epoch": 20.25472,
+      "grad_norm": 1.1191290616989136,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 0.4997,
+      "step": 15824
+    },
+    {
+      "epoch": 20.256,
+      "grad_norm": 1.107366919517517,
+      "learning_rate": 1.8365346138455383e-05,
+      "loss": 0.4822,
+      "step": 15825
+    },
+    {
+      "epoch": 20.25728,
+      "grad_norm": 1.0112433433532715,
+      "learning_rate": 1.8363345338135255e-05,
+      "loss": 0.4796,
+      "step": 15826
+    },
+    {
+      "epoch": 20.25856,
+      "grad_norm": 1.03781259059906,
+      "learning_rate": 1.8361344537815127e-05,
+      "loss": 0.4841,
+      "step": 15827
+    },
+    {
+      "epoch": 20.25984,
+      "grad_norm": 1.0869375467300415,
+      "learning_rate": 1.8359343737495e-05,
+      "loss": 0.5392,
+      "step": 15828
+    },
+    {
+      "epoch": 20.26112,
+      "grad_norm": 1.1285794973373413,
+      "learning_rate": 1.835734293717487e-05,
+      "loss": 0.4978,
+      "step": 15829
+    },
+    {
+      "epoch": 20.2624,
+      "grad_norm": 1.1054590940475464,
+      "learning_rate": 1.8355342136854742e-05,
+      "loss": 0.4876,
+      "step": 15830
+    },
+    {
+      "epoch": 20.26368,
+      "grad_norm": 1.0893003940582275,
+      "learning_rate": 1.8353341336534614e-05,
+      "loss": 0.4894,
+      "step": 15831
+    },
+    {
+      "epoch": 20.26496,
+      "grad_norm": 1.1795381307601929,
+      "learning_rate": 1.8351340536214486e-05,
+      "loss": 0.4894,
+      "step": 15832
+    },
+    {
+      "epoch": 20.26624,
+      "grad_norm": 1.0786329507827759,
+      "learning_rate": 1.8349339735894358e-05,
+      "loss": 0.4816,
+      "step": 15833
+    },
+    {
+      "epoch": 20.26752,
+      "grad_norm": 1.0577181577682495,
+      "learning_rate": 1.834733893557423e-05,
+      "loss": 0.4729,
+      "step": 15834
+    },
+    {
+      "epoch": 20.2688,
+      "grad_norm": 1.0976567268371582,
+      "learning_rate": 1.8345338135254102e-05,
+      "loss": 0.5178,
+      "step": 15835
+    },
+    {
+      "epoch": 20.27008,
+      "grad_norm": 1.0734076499938965,
+      "learning_rate": 1.8343337334933974e-05,
+      "loss": 0.4922,
+      "step": 15836
+    },
+    {
+      "epoch": 20.27136,
+      "grad_norm": 1.062116265296936,
+      "learning_rate": 1.8341336534613845e-05,
+      "loss": 0.4958,
+      "step": 15837
+    },
+    {
+      "epoch": 20.27264,
+      "grad_norm": 1.0870201587677002,
+      "learning_rate": 1.833933573429372e-05,
+      "loss": 0.4803,
+      "step": 15838
+    },
+    {
+      "epoch": 20.27392,
+      "grad_norm": 1.0811505317687988,
+      "learning_rate": 1.833733493397359e-05,
+      "loss": 0.5093,
+      "step": 15839
+    },
+    {
+      "epoch": 20.2752,
+      "grad_norm": 1.1151394844055176,
+      "learning_rate": 1.833533413365346e-05,
+      "loss": 0.5054,
+      "step": 15840
+    },
+    {
+      "epoch": 20.27648,
+      "grad_norm": 1.1511766910552979,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.4998,
+      "step": 15841
+    },
+    {
+      "epoch": 20.27776,
+      "grad_norm": 1.0190060138702393,
+      "learning_rate": 1.8331332533013208e-05,
+      "loss": 0.4492,
+      "step": 15842
+    },
+    {
+      "epoch": 20.27904,
+      "grad_norm": 1.0209211111068726,
+      "learning_rate": 1.8329331732693077e-05,
+      "loss": 0.4544,
+      "step": 15843
+    },
+    {
+      "epoch": 20.28032,
+      "grad_norm": 1.0691982507705688,
+      "learning_rate": 1.832733093237295e-05,
+      "loss": 0.4621,
+      "step": 15844
+    },
+    {
+      "epoch": 20.2816,
+      "grad_norm": 1.1807547807693481,
+      "learning_rate": 1.8325330132052824e-05,
+      "loss": 0.5496,
+      "step": 15845
+    },
+    {
+      "epoch": 20.28288,
+      "grad_norm": 1.0593502521514893,
+      "learning_rate": 1.8323329331732696e-05,
+      "loss": 0.439,
+      "step": 15846
+    },
+    {
+      "epoch": 20.28416,
+      "grad_norm": 1.0794156789779663,
+      "learning_rate": 1.8321328531412564e-05,
+      "loss": 0.5034,
+      "step": 15847
+    },
+    {
+      "epoch": 20.28544,
+      "grad_norm": 1.1507887840270996,
+      "learning_rate": 1.8319327731092436e-05,
+      "loss": 0.5477,
+      "step": 15848
+    },
+    {
+      "epoch": 20.28672,
+      "grad_norm": 0.9900954365730286,
+      "learning_rate": 1.831732693077231e-05,
+      "loss": 0.4778,
+      "step": 15849
+    },
+    {
+      "epoch": 20.288,
+      "grad_norm": 1.0622221231460571,
+      "learning_rate": 1.8315326130452183e-05,
+      "loss": 0.4582,
+      "step": 15850
+    },
+    {
+      "epoch": 20.28928,
+      "grad_norm": 1.0999888181686401,
+      "learning_rate": 1.831332533013205e-05,
+      "loss": 0.4872,
+      "step": 15851
+    },
+    {
+      "epoch": 20.29056,
+      "grad_norm": 1.0076533555984497,
+      "learning_rate": 1.8311324529811927e-05,
+      "loss": 0.4649,
+      "step": 15852
+    },
+    {
+      "epoch": 20.29184,
+      "grad_norm": 1.1547635793685913,
+      "learning_rate": 1.83093237294918e-05,
+      "loss": 0.4873,
+      "step": 15853
+    },
+    {
+      "epoch": 20.29312,
+      "grad_norm": 1.1193164587020874,
+      "learning_rate": 1.830732292917167e-05,
+      "loss": 0.4748,
+      "step": 15854
+    },
+    {
+      "epoch": 20.2944,
+      "grad_norm": 1.1040410995483398,
+      "learning_rate": 1.830532212885154e-05,
+      "loss": 0.4852,
+      "step": 15855
+    },
+    {
+      "epoch": 20.29568,
+      "grad_norm": 1.0710041522979736,
+      "learning_rate": 1.8303321328531414e-05,
+      "loss": 0.4598,
+      "step": 15856
+    },
+    {
+      "epoch": 20.29696,
+      "grad_norm": 1.1397336721420288,
+      "learning_rate": 1.8301320528211286e-05,
+      "loss": 0.5641,
+      "step": 15857
+    },
+    {
+      "epoch": 20.29824,
+      "grad_norm": 1.1837016344070435,
+      "learning_rate": 1.8299319727891158e-05,
+      "loss": 0.4928,
+      "step": 15858
+    },
+    {
+      "epoch": 20.29952,
+      "grad_norm": 1.112987756729126,
+      "learning_rate": 1.829731892757103e-05,
+      "loss": 0.4575,
+      "step": 15859
+    },
+    {
+      "epoch": 20.3008,
+      "grad_norm": 1.0400470495224,
+      "learning_rate": 1.82953181272509e-05,
+      "loss": 0.4527,
+      "step": 15860
+    },
+    {
+      "epoch": 20.30208,
+      "grad_norm": 1.0332235097885132,
+      "learning_rate": 1.8293317326930773e-05,
+      "loss": 0.4948,
+      "step": 15861
+    },
+    {
+      "epoch": 20.30336,
+      "grad_norm": 1.1508287191390991,
+      "learning_rate": 1.8291316526610645e-05,
+      "loss": 0.519,
+      "step": 15862
+    },
+    {
+      "epoch": 20.30464,
+      "grad_norm": 1.0921905040740967,
+      "learning_rate": 1.8289315726290517e-05,
+      "loss": 0.465,
+      "step": 15863
+    },
+    {
+      "epoch": 20.30592,
+      "grad_norm": 1.1255881786346436,
+      "learning_rate": 1.828731492597039e-05,
+      "loss": 0.5087,
+      "step": 15864
+    },
+    {
+      "epoch": 20.3072,
+      "grad_norm": 1.13408362865448,
+      "learning_rate": 1.828531412565026e-05,
+      "loss": 0.5263,
+      "step": 15865
+    },
+    {
+      "epoch": 20.30848,
+      "grad_norm": 1.1082432270050049,
+      "learning_rate": 1.8283313325330133e-05,
+      "loss": 0.477,
+      "step": 15866
+    },
+    {
+      "epoch": 20.30976,
+      "grad_norm": 1.0318244695663452,
+      "learning_rate": 1.8281312525010005e-05,
+      "loss": 0.4939,
+      "step": 15867
+    },
+    {
+      "epoch": 20.31104,
+      "grad_norm": 1.0345925092697144,
+      "learning_rate": 1.8279311724689876e-05,
+      "loss": 0.4818,
+      "step": 15868
+    },
+    {
+      "epoch": 20.31232,
+      "grad_norm": 1.1669683456420898,
+      "learning_rate": 1.8277310924369748e-05,
+      "loss": 0.5096,
+      "step": 15869
+    },
+    {
+      "epoch": 20.3136,
+      "grad_norm": 1.1364049911499023,
+      "learning_rate": 1.827531012404962e-05,
+      "loss": 0.4886,
+      "step": 15870
+    },
+    {
+      "epoch": 20.31488,
+      "grad_norm": 1.096252202987671,
+      "learning_rate": 1.8273309323729492e-05,
+      "loss": 0.4738,
+      "step": 15871
+    },
+    {
+      "epoch": 20.31616,
+      "grad_norm": 1.098500370979309,
+      "learning_rate": 1.8271308523409364e-05,
+      "loss": 0.5058,
+      "step": 15872
+    },
+    {
+      "epoch": 20.31744,
+      "grad_norm": 1.0478065013885498,
+      "learning_rate": 1.826930772308924e-05,
+      "loss": 0.4372,
+      "step": 15873
+    },
+    {
+      "epoch": 20.31872,
+      "grad_norm": 1.1461261510849,
+      "learning_rate": 1.8267306922769108e-05,
+      "loss": 0.5234,
+      "step": 15874
+    },
+    {
+      "epoch": 20.32,
+      "grad_norm": 1.1147645711898804,
+      "learning_rate": 1.826530612244898e-05,
+      "loss": 0.4923,
+      "step": 15875
+    },
+    {
+      "epoch": 20.32128,
+      "grad_norm": 1.076523780822754,
+      "learning_rate": 1.826330532212885e-05,
+      "loss": 0.4899,
+      "step": 15876
+    },
+    {
+      "epoch": 20.32256,
+      "grad_norm": 1.062248706817627,
+      "learning_rate": 1.8261304521808726e-05,
+      "loss": 0.4891,
+      "step": 15877
+    },
+    {
+      "epoch": 20.32384,
+      "grad_norm": 1.0742130279541016,
+      "learning_rate": 1.8259303721488595e-05,
+      "loss": 0.5267,
+      "step": 15878
+    },
+    {
+      "epoch": 20.32512,
+      "grad_norm": 1.046400547027588,
+      "learning_rate": 1.8257302921168467e-05,
+      "loss": 0.4743,
+      "step": 15879
+    },
+    {
+      "epoch": 20.3264,
+      "grad_norm": 1.1239981651306152,
+      "learning_rate": 1.8255302120848342e-05,
+      "loss": 0.4765,
+      "step": 15880
+    },
+    {
+      "epoch": 20.32768,
+      "grad_norm": 1.074862003326416,
+      "learning_rate": 1.8253301320528214e-05,
+      "loss": 0.4618,
+      "step": 15881
+    },
+    {
+      "epoch": 20.32896,
+      "grad_norm": 1.0675029754638672,
+      "learning_rate": 1.8251300520208082e-05,
+      "loss": 0.501,
+      "step": 15882
+    },
+    {
+      "epoch": 20.33024,
+      "grad_norm": 1.0445536375045776,
+      "learning_rate": 1.8249299719887954e-05,
+      "loss": 0.5028,
+      "step": 15883
+    },
+    {
+      "epoch": 20.33152,
+      "grad_norm": 1.1353448629379272,
+      "learning_rate": 1.824729891956783e-05,
+      "loss": 0.4808,
+      "step": 15884
+    },
+    {
+      "epoch": 20.3328,
+      "grad_norm": 1.0861884355545044,
+      "learning_rate": 1.82452981192477e-05,
+      "loss": 0.4717,
+      "step": 15885
+    },
+    {
+      "epoch": 20.33408,
+      "grad_norm": 1.0767872333526611,
+      "learning_rate": 1.824329731892757e-05,
+      "loss": 0.4831,
+      "step": 15886
+    },
+    {
+      "epoch": 20.33536,
+      "grad_norm": 1.1319317817687988,
+      "learning_rate": 1.824129651860744e-05,
+      "loss": 0.5856,
+      "step": 15887
+    },
+    {
+      "epoch": 20.33664,
+      "grad_norm": 1.1070101261138916,
+      "learning_rate": 1.8239295718287317e-05,
+      "loss": 0.4876,
+      "step": 15888
+    },
+    {
+      "epoch": 20.33792,
+      "grad_norm": 1.0813465118408203,
+      "learning_rate": 1.823729491796719e-05,
+      "loss": 0.4761,
+      "step": 15889
+    },
+    {
+      "epoch": 20.3392,
+      "grad_norm": 1.1044111251831055,
+      "learning_rate": 1.8235294117647057e-05,
+      "loss": 0.5044,
+      "step": 15890
+    },
+    {
+      "epoch": 20.34048,
+      "grad_norm": 1.0086759328842163,
+      "learning_rate": 1.8233293317326932e-05,
+      "loss": 0.5094,
+      "step": 15891
+    },
+    {
+      "epoch": 20.34176,
+      "grad_norm": 1.0328330993652344,
+      "learning_rate": 1.8231292517006804e-05,
+      "loss": 0.4408,
+      "step": 15892
+    },
+    {
+      "epoch": 20.34304,
+      "grad_norm": 1.073428988456726,
+      "learning_rate": 1.8229291716686676e-05,
+      "loss": 0.4879,
+      "step": 15893
+    },
+    {
+      "epoch": 20.34432,
+      "grad_norm": 1.0707123279571533,
+      "learning_rate": 1.8227290916366545e-05,
+      "loss": 0.4873,
+      "step": 15894
+    },
+    {
+      "epoch": 20.3456,
+      "grad_norm": 0.9961058497428894,
+      "learning_rate": 1.822529011604642e-05,
+      "loss": 0.4728,
+      "step": 15895
+    },
+    {
+      "epoch": 20.34688,
+      "grad_norm": 1.1604456901550293,
+      "learning_rate": 1.8223289315726292e-05,
+      "loss": 0.4744,
+      "step": 15896
+    },
+    {
+      "epoch": 20.34816,
+      "grad_norm": 1.1048893928527832,
+      "learning_rate": 1.8221288515406164e-05,
+      "loss": 0.4744,
+      "step": 15897
+    },
+    {
+      "epoch": 20.34944,
+      "grad_norm": 1.0889872312545776,
+      "learning_rate": 1.8219287715086035e-05,
+      "loss": 0.4792,
+      "step": 15898
+    },
+    {
+      "epoch": 20.35072,
+      "grad_norm": 1.1309471130371094,
+      "learning_rate": 1.8217286914765907e-05,
+      "loss": 0.5117,
+      "step": 15899
+    },
+    {
+      "epoch": 20.352,
+      "grad_norm": 1.1387243270874023,
+      "learning_rate": 1.821528611444578e-05,
+      "loss": 0.5127,
+      "step": 15900
+    },
+    {
+      "epoch": 20.35328,
+      "grad_norm": 1.1357046365737915,
+      "learning_rate": 1.821328531412565e-05,
+      "loss": 0.5016,
+      "step": 15901
+    },
+    {
+      "epoch": 20.35456,
+      "grad_norm": 1.1048427820205688,
+      "learning_rate": 1.8211284513805523e-05,
+      "loss": 0.4902,
+      "step": 15902
+    },
+    {
+      "epoch": 20.35584,
+      "grad_norm": 1.0928595066070557,
+      "learning_rate": 1.8209283713485395e-05,
+      "loss": 0.4721,
+      "step": 15903
+    },
+    {
+      "epoch": 20.35712,
+      "grad_norm": 1.0246015787124634,
+      "learning_rate": 1.8207282913165267e-05,
+      "loss": 0.4882,
+      "step": 15904
+    },
+    {
+      "epoch": 20.3584,
+      "grad_norm": 1.0948463678359985,
+      "learning_rate": 1.820528211284514e-05,
+      "loss": 0.4531,
+      "step": 15905
+    },
+    {
+      "epoch": 20.35968,
+      "grad_norm": 1.0775697231292725,
+      "learning_rate": 1.820328131252501e-05,
+      "loss": 0.4587,
+      "step": 15906
+    },
+    {
+      "epoch": 20.36096,
+      "grad_norm": 1.0711140632629395,
+      "learning_rate": 1.8201280512204882e-05,
+      "loss": 0.4533,
+      "step": 15907
+    },
+    {
+      "epoch": 20.36224,
+      "grad_norm": 1.0861189365386963,
+      "learning_rate": 1.8199279711884754e-05,
+      "loss": 0.4906,
+      "step": 15908
+    },
+    {
+      "epoch": 20.36352,
+      "grad_norm": 1.1456248760223389,
+      "learning_rate": 1.8197278911564626e-05,
+      "loss": 0.4666,
+      "step": 15909
+    },
+    {
+      "epoch": 20.3648,
+      "grad_norm": 1.126597285270691,
+      "learning_rate": 1.8195278111244498e-05,
+      "loss": 0.5546,
+      "step": 15910
+    },
+    {
+      "epoch": 20.36608,
+      "grad_norm": 1.1376280784606934,
+      "learning_rate": 1.819327731092437e-05,
+      "loss": 0.497,
+      "step": 15911
+    },
+    {
+      "epoch": 20.36736,
+      "grad_norm": 1.084020972251892,
+      "learning_rate": 1.8191276510604245e-05,
+      "loss": 0.4715,
+      "step": 15912
+    },
+    {
+      "epoch": 20.36864,
+      "grad_norm": 1.0613341331481934,
+      "learning_rate": 1.8189275710284113e-05,
+      "loss": 0.4906,
+      "step": 15913
+    },
+    {
+      "epoch": 20.36992,
+      "grad_norm": 1.0506778955459595,
+      "learning_rate": 1.8187274909963985e-05,
+      "loss": 0.4739,
+      "step": 15914
+    },
+    {
+      "epoch": 20.3712,
+      "grad_norm": 1.0626603364944458,
+      "learning_rate": 1.8185274109643857e-05,
+      "loss": 0.4892,
+      "step": 15915
+    },
+    {
+      "epoch": 20.37248,
+      "grad_norm": 1.02902090549469,
+      "learning_rate": 1.8183273309323732e-05,
+      "loss": 0.5032,
+      "step": 15916
+    },
+    {
+      "epoch": 20.37376,
+      "grad_norm": 1.0012882947921753,
+      "learning_rate": 1.81812725090036e-05,
+      "loss": 0.4414,
+      "step": 15917
+    },
+    {
+      "epoch": 20.37504,
+      "grad_norm": 1.0998730659484863,
+      "learning_rate": 1.8179271708683473e-05,
+      "loss": 0.5046,
+      "step": 15918
+    },
+    {
+      "epoch": 20.37632,
+      "grad_norm": 1.138268232345581,
+      "learning_rate": 1.8177270908363348e-05,
+      "loss": 0.5413,
+      "step": 15919
+    },
+    {
+      "epoch": 20.3776,
+      "grad_norm": 1.1356502771377563,
+      "learning_rate": 1.817527010804322e-05,
+      "loss": 0.5459,
+      "step": 15920
+    },
+    {
+      "epoch": 20.37888,
+      "grad_norm": 1.0949009656906128,
+      "learning_rate": 1.8173269307723088e-05,
+      "loss": 0.481,
+      "step": 15921
+    },
+    {
+      "epoch": 20.38016,
+      "grad_norm": 1.0819274187088013,
+      "learning_rate": 1.817126850740296e-05,
+      "loss": 0.4975,
+      "step": 15922
+    },
+    {
+      "epoch": 20.38144,
+      "grad_norm": 1.084596037864685,
+      "learning_rate": 1.8169267707082835e-05,
+      "loss": 0.5315,
+      "step": 15923
+    },
+    {
+      "epoch": 20.38272,
+      "grad_norm": 1.1176440715789795,
+      "learning_rate": 1.8167266906762707e-05,
+      "loss": 0.4893,
+      "step": 15924
+    },
+    {
+      "epoch": 20.384,
+      "grad_norm": 1.1294463872909546,
+      "learning_rate": 1.8165266106442576e-05,
+      "loss": 0.4869,
+      "step": 15925
+    },
+    {
+      "epoch": 20.38528,
+      "grad_norm": 1.177349328994751,
+      "learning_rate": 1.816326530612245e-05,
+      "loss": 0.476,
+      "step": 15926
+    },
+    {
+      "epoch": 20.38656,
+      "grad_norm": 1.15972101688385,
+      "learning_rate": 1.8161264505802323e-05,
+      "loss": 0.5329,
+      "step": 15927
+    },
+    {
+      "epoch": 20.38784,
+      "grad_norm": 1.0567160844802856,
+      "learning_rate": 1.8159263705482195e-05,
+      "loss": 0.4679,
+      "step": 15928
+    },
+    {
+      "epoch": 20.38912,
+      "grad_norm": 1.0844935178756714,
+      "learning_rate": 1.8157262905162063e-05,
+      "loss": 0.4588,
+      "step": 15929
+    },
+    {
+      "epoch": 20.3904,
+      "grad_norm": 1.057117223739624,
+      "learning_rate": 1.8155262104841938e-05,
+      "loss": 0.4974,
+      "step": 15930
+    },
+    {
+      "epoch": 20.39168,
+      "grad_norm": 1.1074366569519043,
+      "learning_rate": 1.815326130452181e-05,
+      "loss": 0.5375,
+      "step": 15931
+    },
+    {
+      "epoch": 20.39296,
+      "grad_norm": 1.081174612045288,
+      "learning_rate": 1.8151260504201682e-05,
+      "loss": 0.5087,
+      "step": 15932
+    },
+    {
+      "epoch": 20.39424,
+      "grad_norm": 1.090725064277649,
+      "learning_rate": 1.8149259703881554e-05,
+      "loss": 0.5162,
+      "step": 15933
+    },
+    {
+      "epoch": 20.39552,
+      "grad_norm": 1.1822116374969482,
+      "learning_rate": 1.8147258903561426e-05,
+      "loss": 0.4763,
+      "step": 15934
+    },
+    {
+      "epoch": 20.3968,
+      "grad_norm": 1.0998727083206177,
+      "learning_rate": 1.8145258103241298e-05,
+      "loss": 0.481,
+      "step": 15935
+    },
+    {
+      "epoch": 20.39808,
+      "grad_norm": 1.1375815868377686,
+      "learning_rate": 1.814325730292117e-05,
+      "loss": 0.5338,
+      "step": 15936
+    },
+    {
+      "epoch": 20.39936,
+      "grad_norm": 1.1332123279571533,
+      "learning_rate": 1.814125650260104e-05,
+      "loss": 0.5234,
+      "step": 15937
+    },
+    {
+      "epoch": 20.40064,
+      "grad_norm": 1.1074390411376953,
+      "learning_rate": 1.8139255702280913e-05,
+      "loss": 0.4858,
+      "step": 15938
+    },
+    {
+      "epoch": 20.40192,
+      "grad_norm": 1.059393286705017,
+      "learning_rate": 1.8137254901960785e-05,
+      "loss": 0.4912,
+      "step": 15939
+    },
+    {
+      "epoch": 20.4032,
+      "grad_norm": 1.132832646369934,
+      "learning_rate": 1.8135254101640657e-05,
+      "loss": 0.5001,
+      "step": 15940
+    },
+    {
+      "epoch": 20.40448,
+      "grad_norm": 1.0979952812194824,
+      "learning_rate": 1.813325330132053e-05,
+      "loss": 0.5107,
+      "step": 15941
+    },
+    {
+      "epoch": 20.40576,
+      "grad_norm": 1.1360996961593628,
+      "learning_rate": 1.81312525010004e-05,
+      "loss": 0.4633,
+      "step": 15942
+    },
+    {
+      "epoch": 20.40704,
+      "grad_norm": 1.0994714498519897,
+      "learning_rate": 1.8129251700680272e-05,
+      "loss": 0.4984,
+      "step": 15943
+    },
+    {
+      "epoch": 20.40832,
+      "grad_norm": 1.0364211797714233,
+      "learning_rate": 1.8127250900360144e-05,
+      "loss": 0.4605,
+      "step": 15944
+    },
+    {
+      "epoch": 20.4096,
+      "grad_norm": 1.0742835998535156,
+      "learning_rate": 1.8125250100040016e-05,
+      "loss": 0.4562,
+      "step": 15945
+    },
+    {
+      "epoch": 20.41088,
+      "grad_norm": 1.151170015335083,
+      "learning_rate": 1.8123249299719888e-05,
+      "loss": 0.4731,
+      "step": 15946
+    },
+    {
+      "epoch": 20.41216,
+      "grad_norm": 1.179980993270874,
+      "learning_rate": 1.812124849939976e-05,
+      "loss": 0.4855,
+      "step": 15947
+    },
+    {
+      "epoch": 20.41344,
+      "grad_norm": 1.1587342023849487,
+      "learning_rate": 1.8119247699079632e-05,
+      "loss": 0.5534,
+      "step": 15948
+    },
+    {
+      "epoch": 20.41472,
+      "grad_norm": 1.099782109260559,
+      "learning_rate": 1.8117246898759504e-05,
+      "loss": 0.4801,
+      "step": 15949
+    },
+    {
+      "epoch": 20.416,
+      "grad_norm": 1.149043321609497,
+      "learning_rate": 1.8115246098439375e-05,
+      "loss": 0.5156,
+      "step": 15950
+    },
+    {
+      "epoch": 20.41728,
+      "grad_norm": 1.0656511783599854,
+      "learning_rate": 1.811324529811925e-05,
+      "loss": 0.4889,
+      "step": 15951
+    },
+    {
+      "epoch": 20.41856,
+      "grad_norm": 1.1039444208145142,
+      "learning_rate": 1.811124449779912e-05,
+      "loss": 0.4805,
+      "step": 15952
+    },
+    {
+      "epoch": 20.41984,
+      "grad_norm": 1.2415475845336914,
+      "learning_rate": 1.810924369747899e-05,
+      "loss": 0.5199,
+      "step": 15953
+    },
+    {
+      "epoch": 20.42112,
+      "grad_norm": 1.1294299364089966,
+      "learning_rate": 1.8107242897158863e-05,
+      "loss": 0.5036,
+      "step": 15954
+    },
+    {
+      "epoch": 20.4224,
+      "grad_norm": 1.0674455165863037,
+      "learning_rate": 1.8105242096838738e-05,
+      "loss": 0.5114,
+      "step": 15955
+    },
+    {
+      "epoch": 20.42368,
+      "grad_norm": 1.1650712490081787,
+      "learning_rate": 1.8103241296518607e-05,
+      "loss": 0.4882,
+      "step": 15956
+    },
+    {
+      "epoch": 20.42496,
+      "grad_norm": 1.055827260017395,
+      "learning_rate": 1.810124049619848e-05,
+      "loss": 0.4596,
+      "step": 15957
+    },
+    {
+      "epoch": 20.42624,
+      "grad_norm": 1.0579880475997925,
+      "learning_rate": 1.8099239695878354e-05,
+      "loss": 0.4844,
+      "step": 15958
+    },
+    {
+      "epoch": 20.42752,
+      "grad_norm": 1.1154268980026245,
+      "learning_rate": 1.8097238895558226e-05,
+      "loss": 0.5422,
+      "step": 15959
+    },
+    {
+      "epoch": 20.4288,
+      "grad_norm": 1.1454224586486816,
+      "learning_rate": 1.8095238095238094e-05,
+      "loss": 0.5215,
+      "step": 15960
+    },
+    {
+      "epoch": 20.43008,
+      "grad_norm": 1.1174713373184204,
+      "learning_rate": 1.8093237294917966e-05,
+      "loss": 0.5044,
+      "step": 15961
+    },
+    {
+      "epoch": 20.43136,
+      "grad_norm": 1.1929234266281128,
+      "learning_rate": 1.809123649459784e-05,
+      "loss": 0.5606,
+      "step": 15962
+    },
+    {
+      "epoch": 20.43264,
+      "grad_norm": 1.2372697591781616,
+      "learning_rate": 1.8089235694277713e-05,
+      "loss": 0.535,
+      "step": 15963
+    },
+    {
+      "epoch": 20.43392,
+      "grad_norm": 1.1086331605911255,
+      "learning_rate": 1.808723489395758e-05,
+      "loss": 0.5191,
+      "step": 15964
+    },
+    {
+      "epoch": 20.4352,
+      "grad_norm": 1.1372125148773193,
+      "learning_rate": 1.8085234093637457e-05,
+      "loss": 0.5431,
+      "step": 15965
+    },
+    {
+      "epoch": 20.43648,
+      "grad_norm": 1.093470811843872,
+      "learning_rate": 1.808323329331733e-05,
+      "loss": 0.4876,
+      "step": 15966
+    },
+    {
+      "epoch": 20.43776,
+      "grad_norm": 1.1044193506240845,
+      "learning_rate": 1.80812324929972e-05,
+      "loss": 0.4605,
+      "step": 15967
+    },
+    {
+      "epoch": 20.43904,
+      "grad_norm": 1.069149136543274,
+      "learning_rate": 1.807923169267707e-05,
+      "loss": 0.4948,
+      "step": 15968
+    },
+    {
+      "epoch": 20.44032,
+      "grad_norm": 1.0932533740997314,
+      "learning_rate": 1.8077230892356944e-05,
+      "loss": 0.4867,
+      "step": 15969
+    },
+    {
+      "epoch": 20.4416,
+      "grad_norm": 1.1302661895751953,
+      "learning_rate": 1.8075230092036816e-05,
+      "loss": 0.4979,
+      "step": 15970
+    },
+    {
+      "epoch": 20.44288,
+      "grad_norm": 1.0727626085281372,
+      "learning_rate": 1.8073229291716688e-05,
+      "loss": 0.4647,
+      "step": 15971
+    },
+    {
+      "epoch": 20.44416,
+      "grad_norm": 1.1052957773208618,
+      "learning_rate": 1.807122849139656e-05,
+      "loss": 0.5308,
+      "step": 15972
+    },
+    {
+      "epoch": 20.44544,
+      "grad_norm": 1.0423941612243652,
+      "learning_rate": 1.806922769107643e-05,
+      "loss": 0.4499,
+      "step": 15973
+    },
+    {
+      "epoch": 20.44672,
+      "grad_norm": 1.0597472190856934,
+      "learning_rate": 1.8067226890756303e-05,
+      "loss": 0.4767,
+      "step": 15974
+    },
+    {
+      "epoch": 20.448,
+      "grad_norm": 1.0862598419189453,
+      "learning_rate": 1.8065226090436175e-05,
+      "loss": 0.5064,
+      "step": 15975
+    },
+    {
+      "epoch": 20.44928,
+      "grad_norm": 1.124342441558838,
+      "learning_rate": 1.8063225290116047e-05,
+      "loss": 0.5078,
+      "step": 15976
+    },
+    {
+      "epoch": 20.45056,
+      "grad_norm": 1.1150842905044556,
+      "learning_rate": 1.806122448979592e-05,
+      "loss": 0.5201,
+      "step": 15977
+    },
+    {
+      "epoch": 20.45184,
+      "grad_norm": 1.101412296295166,
+      "learning_rate": 1.805922368947579e-05,
+      "loss": 0.5034,
+      "step": 15978
+    },
+    {
+      "epoch": 20.45312,
+      "grad_norm": 1.0810297727584839,
+      "learning_rate": 1.8057222889155666e-05,
+      "loss": 0.5046,
+      "step": 15979
+    },
+    {
+      "epoch": 20.4544,
+      "grad_norm": 1.1513960361480713,
+      "learning_rate": 1.8055222088835534e-05,
+      "loss": 0.5528,
+      "step": 15980
+    },
+    {
+      "epoch": 20.45568,
+      "grad_norm": 1.0624333620071411,
+      "learning_rate": 1.8053221288515406e-05,
+      "loss": 0.4878,
+      "step": 15981
+    },
+    {
+      "epoch": 20.45696,
+      "grad_norm": 1.1114479303359985,
+      "learning_rate": 1.8051220488195278e-05,
+      "loss": 0.5168,
+      "step": 15982
+    },
+    {
+      "epoch": 20.45824,
+      "grad_norm": 1.1178675889968872,
+      "learning_rate": 1.8049219687875153e-05,
+      "loss": 0.5037,
+      "step": 15983
+    },
+    {
+      "epoch": 20.45952,
+      "grad_norm": 1.145005226135254,
+      "learning_rate": 1.8047218887555022e-05,
+      "loss": 0.5133,
+      "step": 15984
+    },
+    {
+      "epoch": 20.4608,
+      "grad_norm": 1.016296625137329,
+      "learning_rate": 1.8045218087234894e-05,
+      "loss": 0.4644,
+      "step": 15985
+    },
+    {
+      "epoch": 20.46208,
+      "grad_norm": 1.0797942876815796,
+      "learning_rate": 1.804321728691477e-05,
+      "loss": 0.457,
+      "step": 15986
+    },
+    {
+      "epoch": 20.46336,
+      "grad_norm": 1.0780278444290161,
+      "learning_rate": 1.804121648659464e-05,
+      "loss": 0.5116,
+      "step": 15987
+    },
+    {
+      "epoch": 20.46464,
+      "grad_norm": 1.1817333698272705,
+      "learning_rate": 1.803921568627451e-05,
+      "loss": 0.4981,
+      "step": 15988
+    },
+    {
+      "epoch": 20.46592,
+      "grad_norm": 1.0998475551605225,
+      "learning_rate": 1.803721488595438e-05,
+      "loss": 0.4816,
+      "step": 15989
+    },
+    {
+      "epoch": 20.4672,
+      "grad_norm": 1.1314204931259155,
+      "learning_rate": 1.8035214085634256e-05,
+      "loss": 0.5686,
+      "step": 15990
+    },
+    {
+      "epoch": 20.46848,
+      "grad_norm": 1.0639474391937256,
+      "learning_rate": 1.803321328531413e-05,
+      "loss": 0.4855,
+      "step": 15991
+    },
+    {
+      "epoch": 20.46976,
+      "grad_norm": 0.9945549964904785,
+      "learning_rate": 1.8031212484993997e-05,
+      "loss": 0.4438,
+      "step": 15992
+    },
+    {
+      "epoch": 20.47104,
+      "grad_norm": 1.022700548171997,
+      "learning_rate": 1.8029211684673872e-05,
+      "loss": 0.47,
+      "step": 15993
+    },
+    {
+      "epoch": 20.47232,
+      "grad_norm": 1.0951484441757202,
+      "learning_rate": 1.8027210884353744e-05,
+      "loss": 0.5013,
+      "step": 15994
+    },
+    {
+      "epoch": 20.4736,
+      "grad_norm": 1.067020058631897,
+      "learning_rate": 1.8025210084033616e-05,
+      "loss": 0.5355,
+      "step": 15995
+    },
+    {
+      "epoch": 20.47488,
+      "grad_norm": 1.0288721323013306,
+      "learning_rate": 1.8023209283713484e-05,
+      "loss": 0.4643,
+      "step": 15996
+    },
+    {
+      "epoch": 20.47616,
+      "grad_norm": 1.031306505203247,
+      "learning_rate": 1.802120848339336e-05,
+      "loss": 0.4628,
+      "step": 15997
+    },
+    {
+      "epoch": 20.47744,
+      "grad_norm": 1.183764100074768,
+      "learning_rate": 1.801920768307323e-05,
+      "loss": 0.5168,
+      "step": 15998
+    },
+    {
+      "epoch": 20.47872,
+      "grad_norm": 0.9885333180427551,
+      "learning_rate": 1.8017206882753103e-05,
+      "loss": 0.4841,
+      "step": 15999
+    },
+    {
+      "epoch": 20.48,
+      "grad_norm": 1.01718008518219,
+      "learning_rate": 1.801520608243297e-05,
+      "loss": 0.5065,
+      "step": 16000
+    },
+    {
+      "epoch": 20.48128,
+      "grad_norm": 1.058000922203064,
+      "learning_rate": 1.8013205282112847e-05,
+      "loss": 0.4702,
+      "step": 16001
+    },
+    {
+      "epoch": 20.48256,
+      "grad_norm": 1.0549604892730713,
+      "learning_rate": 1.801120448179272e-05,
+      "loss": 0.4522,
+      "step": 16002
+    },
+    {
+      "epoch": 20.48384,
+      "grad_norm": 1.2326083183288574,
+      "learning_rate": 1.800920368147259e-05,
+      "loss": 0.5391,
+      "step": 16003
+    },
+    {
+      "epoch": 20.48512,
+      "grad_norm": 1.0707967281341553,
+      "learning_rate": 1.8007202881152462e-05,
+      "loss": 0.4603,
+      "step": 16004
+    },
+    {
+      "epoch": 20.4864,
+      "grad_norm": 1.141228199005127,
+      "learning_rate": 1.8005202080832334e-05,
+      "loss": 0.5162,
+      "step": 16005
+    },
+    {
+      "epoch": 20.48768,
+      "grad_norm": 1.074081540107727,
+      "learning_rate": 1.8003201280512206e-05,
+      "loss": 0.4736,
+      "step": 16006
+    },
+    {
+      "epoch": 20.48896,
+      "grad_norm": 1.0647474527359009,
+      "learning_rate": 1.8001200480192078e-05,
+      "loss": 0.4778,
+      "step": 16007
+    },
+    {
+      "epoch": 20.49024,
+      "grad_norm": 1.1163427829742432,
+      "learning_rate": 1.799919967987195e-05,
+      "loss": 0.5002,
+      "step": 16008
+    },
+    {
+      "epoch": 20.49152,
+      "grad_norm": 1.0871421098709106,
+      "learning_rate": 1.7997198879551822e-05,
+      "loss": 0.485,
+      "step": 16009
+    },
+    {
+      "epoch": 20.4928,
+      "grad_norm": 1.0807521343231201,
+      "learning_rate": 1.7995198079231694e-05,
+      "loss": 0.4932,
+      "step": 16010
+    },
+    {
+      "epoch": 20.49408,
+      "grad_norm": 1.119227409362793,
+      "learning_rate": 1.7993197278911565e-05,
+      "loss": 0.5201,
+      "step": 16011
+    },
+    {
+      "epoch": 20.49536,
+      "grad_norm": 1.0895061492919922,
+      "learning_rate": 1.7991196478591437e-05,
+      "loss": 0.495,
+      "step": 16012
+    },
+    {
+      "epoch": 20.49664,
+      "grad_norm": 1.0847043991088867,
+      "learning_rate": 1.798919567827131e-05,
+      "loss": 0.484,
+      "step": 16013
+    },
+    {
+      "epoch": 20.49792,
+      "grad_norm": 1.0497854948043823,
+      "learning_rate": 1.798719487795118e-05,
+      "loss": 0.4332,
+      "step": 16014
+    },
+    {
+      "epoch": 20.4992,
+      "grad_norm": 1.1417521238327026,
+      "learning_rate": 1.7985194077631053e-05,
+      "loss": 0.518,
+      "step": 16015
+    },
+    {
+      "epoch": 20.50048,
+      "grad_norm": 1.1823201179504395,
+      "learning_rate": 1.7983193277310925e-05,
+      "loss": 0.4732,
+      "step": 16016
+    },
+    {
+      "epoch": 20.50176,
+      "grad_norm": 1.0755385160446167,
+      "learning_rate": 1.7981192476990797e-05,
+      "loss": 0.4858,
+      "step": 16017
+    },
+    {
+      "epoch": 20.50304,
+      "grad_norm": 1.1058756113052368,
+      "learning_rate": 1.7979191676670672e-05,
+      "loss": 0.4979,
+      "step": 16018
+    },
+    {
+      "epoch": 20.50432,
+      "grad_norm": 1.109181523323059,
+      "learning_rate": 1.797719087635054e-05,
+      "loss": 0.5062,
+      "step": 16019
+    },
+    {
+      "epoch": 20.5056,
+      "grad_norm": 1.0815677642822266,
+      "learning_rate": 1.7975190076030412e-05,
+      "loss": 0.5043,
+      "step": 16020
+    },
+    {
+      "epoch": 20.50688,
+      "grad_norm": 1.1085656881332397,
+      "learning_rate": 1.7973189275710284e-05,
+      "loss": 0.4851,
+      "step": 16021
+    },
+    {
+      "epoch": 20.50816,
+      "grad_norm": 1.077195644378662,
+      "learning_rate": 1.797118847539016e-05,
+      "loss": 0.4885,
+      "step": 16022
+    },
+    {
+      "epoch": 20.50944,
+      "grad_norm": 0.9872754812240601,
+      "learning_rate": 1.7969187675070028e-05,
+      "loss": 0.4926,
+      "step": 16023
+    },
+    {
+      "epoch": 20.51072,
+      "grad_norm": 1.0659494400024414,
+      "learning_rate": 1.79671868747499e-05,
+      "loss": 0.447,
+      "step": 16024
+    },
+    {
+      "epoch": 20.512,
+      "grad_norm": 1.0598735809326172,
+      "learning_rate": 1.7965186074429775e-05,
+      "loss": 0.487,
+      "step": 16025
+    },
+    {
+      "epoch": 20.51328,
+      "grad_norm": 1.0480681657791138,
+      "learning_rate": 1.7963185274109647e-05,
+      "loss": 0.4795,
+      "step": 16026
+    },
+    {
+      "epoch": 20.51456,
+      "grad_norm": 1.0445165634155273,
+      "learning_rate": 1.7961184473789515e-05,
+      "loss": 0.4792,
+      "step": 16027
+    },
+    {
+      "epoch": 20.51584,
+      "grad_norm": 1.084743857383728,
+      "learning_rate": 1.7959183673469387e-05,
+      "loss": 0.5358,
+      "step": 16028
+    },
+    {
+      "epoch": 20.51712,
+      "grad_norm": 1.127386212348938,
+      "learning_rate": 1.7957182873149262e-05,
+      "loss": 0.514,
+      "step": 16029
+    },
+    {
+      "epoch": 20.5184,
+      "grad_norm": 1.1147735118865967,
+      "learning_rate": 1.7955182072829134e-05,
+      "loss": 0.53,
+      "step": 16030
+    },
+    {
+      "epoch": 20.51968,
+      "grad_norm": 1.0256705284118652,
+      "learning_rate": 1.7953181272509003e-05,
+      "loss": 0.4541,
+      "step": 16031
+    },
+    {
+      "epoch": 20.52096,
+      "grad_norm": 1.1226087808609009,
+      "learning_rate": 1.7951180472188878e-05,
+      "loss": 0.5,
+      "step": 16032
+    },
+    {
+      "epoch": 20.52224,
+      "grad_norm": 1.1469954252243042,
+      "learning_rate": 1.794917967186875e-05,
+      "loss": 0.5799,
+      "step": 16033
+    },
+    {
+      "epoch": 20.52352,
+      "grad_norm": 1.1341044902801514,
+      "learning_rate": 1.794717887154862e-05,
+      "loss": 0.547,
+      "step": 16034
+    },
+    {
+      "epoch": 20.5248,
+      "grad_norm": 1.123470425605774,
+      "learning_rate": 1.794517807122849e-05,
+      "loss": 0.4843,
+      "step": 16035
+    },
+    {
+      "epoch": 20.52608,
+      "grad_norm": 1.1044238805770874,
+      "learning_rate": 1.7943177270908365e-05,
+      "loss": 0.5044,
+      "step": 16036
+    },
+    {
+      "epoch": 20.52736,
+      "grad_norm": 1.021741271018982,
+      "learning_rate": 1.7941176470588237e-05,
+      "loss": 0.4652,
+      "step": 16037
+    },
+    {
+      "epoch": 20.52864,
+      "grad_norm": 1.1132144927978516,
+      "learning_rate": 1.793917567026811e-05,
+      "loss": 0.486,
+      "step": 16038
+    },
+    {
+      "epoch": 20.52992,
+      "grad_norm": 1.0857055187225342,
+      "learning_rate": 1.793717486994798e-05,
+      "loss": 0.5323,
+      "step": 16039
+    },
+    {
+      "epoch": 20.5312,
+      "grad_norm": 1.1283453702926636,
+      "learning_rate": 1.7935174069627853e-05,
+      "loss": 0.4848,
+      "step": 16040
+    },
+    {
+      "epoch": 20.53248,
+      "grad_norm": 1.175715684890747,
+      "learning_rate": 1.7933173269307725e-05,
+      "loss": 0.5741,
+      "step": 16041
+    },
+    {
+      "epoch": 20.53376,
+      "grad_norm": 1.1087825298309326,
+      "learning_rate": 1.7931172468987596e-05,
+      "loss": 0.4799,
+      "step": 16042
+    },
+    {
+      "epoch": 20.53504,
+      "grad_norm": 1.1053529977798462,
+      "learning_rate": 1.7929171668667468e-05,
+      "loss": 0.5016,
+      "step": 16043
+    },
+    {
+      "epoch": 20.53632,
+      "grad_norm": 1.1228110790252686,
+      "learning_rate": 1.792717086834734e-05,
+      "loss": 0.5231,
+      "step": 16044
+    },
+    {
+      "epoch": 20.5376,
+      "grad_norm": 1.127894639968872,
+      "learning_rate": 1.7925170068027212e-05,
+      "loss": 0.5069,
+      "step": 16045
+    },
+    {
+      "epoch": 20.53888,
+      "grad_norm": 1.0892380475997925,
+      "learning_rate": 1.7923169267707084e-05,
+      "loss": 0.5249,
+      "step": 16046
+    },
+    {
+      "epoch": 20.54016,
+      "grad_norm": 1.0959818363189697,
+      "learning_rate": 1.7921168467386956e-05,
+      "loss": 0.511,
+      "step": 16047
+    },
+    {
+      "epoch": 20.54144,
+      "grad_norm": 1.0303337574005127,
+      "learning_rate": 1.7919167667066828e-05,
+      "loss": 0.4818,
+      "step": 16048
+    },
+    {
+      "epoch": 20.54272,
+      "grad_norm": 1.0872350931167603,
+      "learning_rate": 1.79171668667467e-05,
+      "loss": 0.4872,
+      "step": 16049
+    },
+    {
+      "epoch": 20.544,
+      "grad_norm": 1.132896065711975,
+      "learning_rate": 1.791516606642657e-05,
+      "loss": 0.5216,
+      "step": 16050
+    },
+    {
+      "epoch": 20.545279999999998,
+      "grad_norm": 1.0700019598007202,
+      "learning_rate": 1.7913165266106443e-05,
+      "loss": 0.4918,
+      "step": 16051
+    },
+    {
+      "epoch": 20.54656,
+      "grad_norm": 1.0675747394561768,
+      "learning_rate": 1.7911164465786315e-05,
+      "loss": 0.4883,
+      "step": 16052
+    },
+    {
+      "epoch": 20.54784,
+      "grad_norm": 1.1021859645843506,
+      "learning_rate": 1.7909163665466187e-05,
+      "loss": 0.4742,
+      "step": 16053
+    },
+    {
+      "epoch": 20.54912,
+      "grad_norm": 1.0696967840194702,
+      "learning_rate": 1.790716286514606e-05,
+      "loss": 0.4873,
+      "step": 16054
+    },
+    {
+      "epoch": 20.5504,
+      "grad_norm": 1.0992937088012695,
+      "learning_rate": 1.790516206482593e-05,
+      "loss": 0.5003,
+      "step": 16055
+    },
+    {
+      "epoch": 20.55168,
+      "grad_norm": 1.0972262620925903,
+      "learning_rate": 1.7903161264505802e-05,
+      "loss": 0.562,
+      "step": 16056
+    },
+    {
+      "epoch": 20.55296,
+      "grad_norm": 1.1256171464920044,
+      "learning_rate": 1.7901160464185678e-05,
+      "loss": 0.4714,
+      "step": 16057
+    },
+    {
+      "epoch": 20.55424,
+      "grad_norm": 1.1551443338394165,
+      "learning_rate": 1.7899159663865546e-05,
+      "loss": 0.5156,
+      "step": 16058
+    },
+    {
+      "epoch": 20.55552,
+      "grad_norm": 1.0866180658340454,
+      "learning_rate": 1.7897158863545418e-05,
+      "loss": 0.4726,
+      "step": 16059
+    },
+    {
+      "epoch": 20.5568,
+      "grad_norm": 1.1982043981552124,
+      "learning_rate": 1.789515806322529e-05,
+      "loss": 0.5541,
+      "step": 16060
+    },
+    {
+      "epoch": 20.55808,
+      "grad_norm": 1.107646107673645,
+      "learning_rate": 1.7893157262905165e-05,
+      "loss": 0.5068,
+      "step": 16061
+    },
+    {
+      "epoch": 20.55936,
+      "grad_norm": 1.1011683940887451,
+      "learning_rate": 1.7891156462585034e-05,
+      "loss": 0.5374,
+      "step": 16062
+    },
+    {
+      "epoch": 20.56064,
+      "grad_norm": 1.1166385412216187,
+      "learning_rate": 1.7889155662264905e-05,
+      "loss": 0.4849,
+      "step": 16063
+    },
+    {
+      "epoch": 20.56192,
+      "grad_norm": 1.1044502258300781,
+      "learning_rate": 1.788715486194478e-05,
+      "loss": 0.4947,
+      "step": 16064
+    },
+    {
+      "epoch": 20.5632,
+      "grad_norm": 1.126716136932373,
+      "learning_rate": 1.7885154061624652e-05,
+      "loss": 0.5107,
+      "step": 16065
+    },
+    {
+      "epoch": 20.56448,
+      "grad_norm": 1.119673252105713,
+      "learning_rate": 1.788315326130452e-05,
+      "loss": 0.5369,
+      "step": 16066
+    },
+    {
+      "epoch": 20.56576,
+      "grad_norm": 1.074339747428894,
+      "learning_rate": 1.7881152460984393e-05,
+      "loss": 0.4935,
+      "step": 16067
+    },
+    {
+      "epoch": 20.56704,
+      "grad_norm": 1.1229872703552246,
+      "learning_rate": 1.7879151660664268e-05,
+      "loss": 0.5122,
+      "step": 16068
+    },
+    {
+      "epoch": 20.56832,
+      "grad_norm": 1.1117804050445557,
+      "learning_rate": 1.787715086034414e-05,
+      "loss": 0.4864,
+      "step": 16069
+    },
+    {
+      "epoch": 20.5696,
+      "grad_norm": 1.0915405750274658,
+      "learning_rate": 1.787515006002401e-05,
+      "loss": 0.5351,
+      "step": 16070
+    },
+    {
+      "epoch": 20.57088,
+      "grad_norm": 1.1257494688034058,
+      "learning_rate": 1.7873149259703884e-05,
+      "loss": 0.4874,
+      "step": 16071
+    },
+    {
+      "epoch": 20.57216,
+      "grad_norm": 1.1443331241607666,
+      "learning_rate": 1.7871148459383755e-05,
+      "loss": 0.4771,
+      "step": 16072
+    },
+    {
+      "epoch": 20.57344,
+      "grad_norm": 1.1253244876861572,
+      "learning_rate": 1.7869147659063627e-05,
+      "loss": 0.492,
+      "step": 16073
+    },
+    {
+      "epoch": 20.57472,
+      "grad_norm": 1.1300221681594849,
+      "learning_rate": 1.7867146858743496e-05,
+      "loss": 0.5306,
+      "step": 16074
+    },
+    {
+      "epoch": 20.576,
+      "grad_norm": 1.1364995241165161,
+      "learning_rate": 1.786514605842337e-05,
+      "loss": 0.5069,
+      "step": 16075
+    },
+    {
+      "epoch": 20.577280000000002,
+      "grad_norm": 1.0387669801712036,
+      "learning_rate": 1.7863145258103243e-05,
+      "loss": 0.4726,
+      "step": 16076
+    },
+    {
+      "epoch": 20.57856,
+      "grad_norm": 1.049291968345642,
+      "learning_rate": 1.7861144457783115e-05,
+      "loss": 0.4852,
+      "step": 16077
+    },
+    {
+      "epoch": 20.57984,
+      "grad_norm": 1.0444282293319702,
+      "learning_rate": 1.7859143657462987e-05,
+      "loss": 0.4718,
+      "step": 16078
+    },
+    {
+      "epoch": 20.58112,
+      "grad_norm": 1.095975399017334,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.5015,
+      "step": 16079
+    },
+    {
+      "epoch": 20.5824,
+      "grad_norm": 1.104600191116333,
+      "learning_rate": 1.785514205682273e-05,
+      "loss": 0.5515,
+      "step": 16080
+    },
+    {
+      "epoch": 20.58368,
+      "grad_norm": 1.0863333940505981,
+      "learning_rate": 1.7853141256502602e-05,
+      "loss": 0.453,
+      "step": 16081
+    },
+    {
+      "epoch": 20.58496,
+      "grad_norm": 1.1413263082504272,
+      "learning_rate": 1.7851140456182474e-05,
+      "loss": 0.5646,
+      "step": 16082
+    },
+    {
+      "epoch": 20.58624,
+      "grad_norm": 1.149473786354065,
+      "learning_rate": 1.7849139655862346e-05,
+      "loss": 0.4853,
+      "step": 16083
+    },
+    {
+      "epoch": 20.58752,
+      "grad_norm": 1.069388747215271,
+      "learning_rate": 1.7847138855542218e-05,
+      "loss": 0.4969,
+      "step": 16084
+    },
+    {
+      "epoch": 20.5888,
+      "grad_norm": 1.156481385231018,
+      "learning_rate": 1.784513805522209e-05,
+      "loss": 0.5106,
+      "step": 16085
+    },
+    {
+      "epoch": 20.59008,
+      "grad_norm": 1.0707303285598755,
+      "learning_rate": 1.784313725490196e-05,
+      "loss": 0.467,
+      "step": 16086
+    },
+    {
+      "epoch": 20.59136,
+      "grad_norm": 1.1706613302230835,
+      "learning_rate": 1.7841136454581833e-05,
+      "loss": 0.5354,
+      "step": 16087
+    },
+    {
+      "epoch": 20.59264,
+      "grad_norm": 1.0919955968856812,
+      "learning_rate": 1.7839135654261705e-05,
+      "loss": 0.4906,
+      "step": 16088
+    },
+    {
+      "epoch": 20.59392,
+      "grad_norm": 1.0173168182373047,
+      "learning_rate": 1.7837134853941577e-05,
+      "loss": 0.4586,
+      "step": 16089
+    },
+    {
+      "epoch": 20.5952,
+      "grad_norm": 1.1111178398132324,
+      "learning_rate": 1.783513405362145e-05,
+      "loss": 0.5209,
+      "step": 16090
+    },
+    {
+      "epoch": 20.59648,
+      "grad_norm": 1.0708633661270142,
+      "learning_rate": 1.783313325330132e-05,
+      "loss": 0.4752,
+      "step": 16091
+    },
+    {
+      "epoch": 20.59776,
+      "grad_norm": 1.0484986305236816,
+      "learning_rate": 1.7831132452981196e-05,
+      "loss": 0.491,
+      "step": 16092
+    },
+    {
+      "epoch": 20.59904,
+      "grad_norm": 1.205859899520874,
+      "learning_rate": 1.7829131652661064e-05,
+      "loss": 0.4858,
+      "step": 16093
+    },
+    {
+      "epoch": 20.60032,
+      "grad_norm": 1.1603915691375732,
+      "learning_rate": 1.7827130852340936e-05,
+      "loss": 0.4818,
+      "step": 16094
+    },
+    {
+      "epoch": 20.6016,
+      "grad_norm": 1.0573688745498657,
+      "learning_rate": 1.7825130052020808e-05,
+      "loss": 0.4992,
+      "step": 16095
+    },
+    {
+      "epoch": 20.60288,
+      "grad_norm": 1.1221954822540283,
+      "learning_rate": 1.7823129251700683e-05,
+      "loss": 0.4625,
+      "step": 16096
+    },
+    {
+      "epoch": 20.60416,
+      "grad_norm": 1.1069375276565552,
+      "learning_rate": 1.7821128451380552e-05,
+      "loss": 0.502,
+      "step": 16097
+    },
+    {
+      "epoch": 20.60544,
+      "grad_norm": 1.0597623586654663,
+      "learning_rate": 1.7819127651060424e-05,
+      "loss": 0.4757,
+      "step": 16098
+    },
+    {
+      "epoch": 20.60672,
+      "grad_norm": 1.1148048639297485,
+      "learning_rate": 1.78171268507403e-05,
+      "loss": 0.521,
+      "step": 16099
+    },
+    {
+      "epoch": 20.608,
+      "grad_norm": 1.1164958477020264,
+      "learning_rate": 1.781512605042017e-05,
+      "loss": 0.4964,
+      "step": 16100
+    },
+    {
+      "epoch": 20.60928,
+      "grad_norm": 1.063112735748291,
+      "learning_rate": 1.781312525010004e-05,
+      "loss": 0.51,
+      "step": 16101
+    },
+    {
+      "epoch": 20.61056,
+      "grad_norm": 1.1305928230285645,
+      "learning_rate": 1.781112444977991e-05,
+      "loss": 0.5106,
+      "step": 16102
+    },
+    {
+      "epoch": 20.61184,
+      "grad_norm": 1.0931042432785034,
+      "learning_rate": 1.7809123649459786e-05,
+      "loss": 0.482,
+      "step": 16103
+    },
+    {
+      "epoch": 20.61312,
+      "grad_norm": 1.0625600814819336,
+      "learning_rate": 1.7807122849139658e-05,
+      "loss": 0.4919,
+      "step": 16104
+    },
+    {
+      "epoch": 20.6144,
+      "grad_norm": 1.0991266965866089,
+      "learning_rate": 1.7805122048819527e-05,
+      "loss": 0.481,
+      "step": 16105
+    },
+    {
+      "epoch": 20.61568,
+      "grad_norm": 1.0887682437896729,
+      "learning_rate": 1.7803121248499402e-05,
+      "loss": 0.5264,
+      "step": 16106
+    },
+    {
+      "epoch": 20.61696,
+      "grad_norm": 1.150978446006775,
+      "learning_rate": 1.7801120448179274e-05,
+      "loss": 0.5575,
+      "step": 16107
+    },
+    {
+      "epoch": 20.61824,
+      "grad_norm": 1.1687663793563843,
+      "learning_rate": 1.7799119647859146e-05,
+      "loss": 0.4825,
+      "step": 16108
+    },
+    {
+      "epoch": 20.61952,
+      "grad_norm": 1.1047236919403076,
+      "learning_rate": 1.7797118847539014e-05,
+      "loss": 0.513,
+      "step": 16109
+    },
+    {
+      "epoch": 20.6208,
+      "grad_norm": 1.1841906309127808,
+      "learning_rate": 1.779511804721889e-05,
+      "loss": 0.448,
+      "step": 16110
+    },
+    {
+      "epoch": 20.62208,
+      "grad_norm": 1.0986437797546387,
+      "learning_rate": 1.779311724689876e-05,
+      "loss": 0.5195,
+      "step": 16111
+    },
+    {
+      "epoch": 20.62336,
+      "grad_norm": 1.110182762145996,
+      "learning_rate": 1.7791116446578633e-05,
+      "loss": 0.5159,
+      "step": 16112
+    },
+    {
+      "epoch": 20.62464,
+      "grad_norm": 1.0949082374572754,
+      "learning_rate": 1.77891156462585e-05,
+      "loss": 0.5092,
+      "step": 16113
+    },
+    {
+      "epoch": 20.62592,
+      "grad_norm": 1.1522455215454102,
+      "learning_rate": 1.7787114845938377e-05,
+      "loss": 0.5514,
+      "step": 16114
+    },
+    {
+      "epoch": 20.6272,
+      "grad_norm": 1.0845671892166138,
+      "learning_rate": 1.778511404561825e-05,
+      "loss": 0.4713,
+      "step": 16115
+    },
+    {
+      "epoch": 20.62848,
+      "grad_norm": 1.1189041137695312,
+      "learning_rate": 1.778311324529812e-05,
+      "loss": 0.5218,
+      "step": 16116
+    },
+    {
+      "epoch": 20.62976,
+      "grad_norm": 1.1408662796020508,
+      "learning_rate": 1.7781112444977992e-05,
+      "loss": 0.531,
+      "step": 16117
+    },
+    {
+      "epoch": 20.63104,
+      "grad_norm": 1.071902871131897,
+      "learning_rate": 1.7779111644657864e-05,
+      "loss": 0.47,
+      "step": 16118
+    },
+    {
+      "epoch": 20.63232,
+      "grad_norm": 1.0264390707015991,
+      "learning_rate": 1.7777110844337736e-05,
+      "loss": 0.4506,
+      "step": 16119
+    },
+    {
+      "epoch": 20.6336,
+      "grad_norm": 1.073315143585205,
+      "learning_rate": 1.7775110044017608e-05,
+      "loss": 0.4636,
+      "step": 16120
+    },
+    {
+      "epoch": 20.63488,
+      "grad_norm": 1.1272789239883423,
+      "learning_rate": 1.777310924369748e-05,
+      "loss": 0.5256,
+      "step": 16121
+    },
+    {
+      "epoch": 20.63616,
+      "grad_norm": 1.1192529201507568,
+      "learning_rate": 1.7771108443377352e-05,
+      "loss": 0.5349,
+      "step": 16122
+    },
+    {
+      "epoch": 20.63744,
+      "grad_norm": 1.0640686750411987,
+      "learning_rate": 1.7769107643057224e-05,
+      "loss": 0.4645,
+      "step": 16123
+    },
+    {
+      "epoch": 20.63872,
+      "grad_norm": 1.095218539237976,
+      "learning_rate": 1.7767106842737095e-05,
+      "loss": 0.5333,
+      "step": 16124
+    },
+    {
+      "epoch": 20.64,
+      "grad_norm": 1.1425113677978516,
+      "learning_rate": 1.7765106042416967e-05,
+      "loss": 0.5289,
+      "step": 16125
+    },
+    {
+      "epoch": 20.64128,
+      "grad_norm": 1.095801830291748,
+      "learning_rate": 1.776310524209684e-05,
+      "loss": 0.496,
+      "step": 16126
+    },
+    {
+      "epoch": 20.64256,
+      "grad_norm": 1.1665611267089844,
+      "learning_rate": 1.776110444177671e-05,
+      "loss": 0.5176,
+      "step": 16127
+    },
+    {
+      "epoch": 20.64384,
+      "grad_norm": 1.109303593635559,
+      "learning_rate": 1.7759103641456583e-05,
+      "loss": 0.5234,
+      "step": 16128
+    },
+    {
+      "epoch": 20.64512,
+      "grad_norm": 1.0539172887802124,
+      "learning_rate": 1.7757102841136455e-05,
+      "loss": 0.4982,
+      "step": 16129
+    },
+    {
+      "epoch": 20.6464,
+      "grad_norm": 1.068564772605896,
+      "learning_rate": 1.7755102040816327e-05,
+      "loss": 0.5223,
+      "step": 16130
+    },
+    {
+      "epoch": 20.64768,
+      "grad_norm": 1.0728812217712402,
+      "learning_rate": 1.7753101240496202e-05,
+      "loss": 0.4921,
+      "step": 16131
+    },
+    {
+      "epoch": 20.64896,
+      "grad_norm": 1.1256548166275024,
+      "learning_rate": 1.775110044017607e-05,
+      "loss": 0.5624,
+      "step": 16132
+    },
+    {
+      "epoch": 20.65024,
+      "grad_norm": 1.0447193384170532,
+      "learning_rate": 1.7749099639855942e-05,
+      "loss": 0.5047,
+      "step": 16133
+    },
+    {
+      "epoch": 20.65152,
+      "grad_norm": 1.0693058967590332,
+      "learning_rate": 1.7747098839535814e-05,
+      "loss": 0.4991,
+      "step": 16134
+    },
+    {
+      "epoch": 20.6528,
+      "grad_norm": 1.0980228185653687,
+      "learning_rate": 1.774509803921569e-05,
+      "loss": 0.5151,
+      "step": 16135
+    },
+    {
+      "epoch": 20.65408,
+      "grad_norm": 1.066780924797058,
+      "learning_rate": 1.7743097238895558e-05,
+      "loss": 0.4867,
+      "step": 16136
+    },
+    {
+      "epoch": 20.65536,
+      "grad_norm": 1.225781798362732,
+      "learning_rate": 1.774109643857543e-05,
+      "loss": 0.5333,
+      "step": 16137
+    },
+    {
+      "epoch": 20.65664,
+      "grad_norm": 1.0326478481292725,
+      "learning_rate": 1.7739095638255305e-05,
+      "loss": 0.4637,
+      "step": 16138
+    },
+    {
+      "epoch": 20.65792,
+      "grad_norm": 1.1340293884277344,
+      "learning_rate": 1.7737094837935177e-05,
+      "loss": 0.5045,
+      "step": 16139
+    },
+    {
+      "epoch": 20.6592,
+      "grad_norm": 1.1338059902191162,
+      "learning_rate": 1.7735094037615045e-05,
+      "loss": 0.5238,
+      "step": 16140
+    },
+    {
+      "epoch": 20.66048,
+      "grad_norm": 1.1640844345092773,
+      "learning_rate": 1.7733093237294917e-05,
+      "loss": 0.5039,
+      "step": 16141
+    },
+    {
+      "epoch": 20.66176,
+      "grad_norm": 1.0547772645950317,
+      "learning_rate": 1.7731092436974792e-05,
+      "loss": 0.4398,
+      "step": 16142
+    },
+    {
+      "epoch": 20.66304,
+      "grad_norm": 1.1220600605010986,
+      "learning_rate": 1.7729091636654664e-05,
+      "loss": 0.5248,
+      "step": 16143
+    },
+    {
+      "epoch": 20.66432,
+      "grad_norm": 1.0587539672851562,
+      "learning_rate": 1.7727090836334533e-05,
+      "loss": 0.4715,
+      "step": 16144
+    },
+    {
+      "epoch": 20.6656,
+      "grad_norm": 1.1665568351745605,
+      "learning_rate": 1.7725090036014408e-05,
+      "loss": 0.5386,
+      "step": 16145
+    },
+    {
+      "epoch": 20.66688,
+      "grad_norm": 1.0843427181243896,
+      "learning_rate": 1.772308923569428e-05,
+      "loss": 0.5055,
+      "step": 16146
+    },
+    {
+      "epoch": 20.66816,
+      "grad_norm": 1.0697423219680786,
+      "learning_rate": 1.772108843537415e-05,
+      "loss": 0.5219,
+      "step": 16147
+    },
+    {
+      "epoch": 20.66944,
+      "grad_norm": 1.1367213726043701,
+      "learning_rate": 1.771908763505402e-05,
+      "loss": 0.5201,
+      "step": 16148
+    },
+    {
+      "epoch": 20.67072,
+      "grad_norm": 1.1785510778427124,
+      "learning_rate": 1.7717086834733895e-05,
+      "loss": 0.4896,
+      "step": 16149
+    },
+    {
+      "epoch": 20.672,
+      "grad_norm": 1.1276311874389648,
+      "learning_rate": 1.7715086034413767e-05,
+      "loss": 0.4954,
+      "step": 16150
+    },
+    {
+      "epoch": 20.67328,
+      "grad_norm": 1.1062501668930054,
+      "learning_rate": 1.771308523409364e-05,
+      "loss": 0.4945,
+      "step": 16151
+    },
+    {
+      "epoch": 20.67456,
+      "grad_norm": 1.0566152334213257,
+      "learning_rate": 1.771108443377351e-05,
+      "loss": 0.471,
+      "step": 16152
+    },
+    {
+      "epoch": 20.67584,
+      "grad_norm": 1.0588206052780151,
+      "learning_rate": 1.7709083633453383e-05,
+      "loss": 0.5359,
+      "step": 16153
+    },
+    {
+      "epoch": 20.67712,
+      "grad_norm": 1.1501905918121338,
+      "learning_rate": 1.7707082833133255e-05,
+      "loss": 0.4935,
+      "step": 16154
+    },
+    {
+      "epoch": 20.6784,
+      "grad_norm": 1.0157536268234253,
+      "learning_rate": 1.7705082032813126e-05,
+      "loss": 0.4498,
+      "step": 16155
+    },
+    {
+      "epoch": 20.67968,
+      "grad_norm": 1.1221078634262085,
+      "learning_rate": 1.7703081232492998e-05,
+      "loss": 0.4926,
+      "step": 16156
+    },
+    {
+      "epoch": 20.68096,
+      "grad_norm": 1.0495266914367676,
+      "learning_rate": 1.770108043217287e-05,
+      "loss": 0.471,
+      "step": 16157
+    },
+    {
+      "epoch": 20.68224,
+      "grad_norm": 1.1005525588989258,
+      "learning_rate": 1.7699079631852742e-05,
+      "loss": 0.499,
+      "step": 16158
+    },
+    {
+      "epoch": 20.68352,
+      "grad_norm": 1.0935386419296265,
+      "learning_rate": 1.7697078831532614e-05,
+      "loss": 0.4897,
+      "step": 16159
+    },
+    {
+      "epoch": 20.6848,
+      "grad_norm": 1.0938061475753784,
+      "learning_rate": 1.7695078031212486e-05,
+      "loss": 0.4632,
+      "step": 16160
+    },
+    {
+      "epoch": 20.68608,
+      "grad_norm": 1.104549765586853,
+      "learning_rate": 1.7693077230892358e-05,
+      "loss": 0.4732,
+      "step": 16161
+    },
+    {
+      "epoch": 20.687359999999998,
+      "grad_norm": 1.0983625650405884,
+      "learning_rate": 1.769107643057223e-05,
+      "loss": 0.4978,
+      "step": 16162
+    },
+    {
+      "epoch": 20.68864,
+      "grad_norm": 1.1382209062576294,
+      "learning_rate": 1.76890756302521e-05,
+      "loss": 0.5318,
+      "step": 16163
+    },
+    {
+      "epoch": 20.68992,
+      "grad_norm": 1.1396292448043823,
+      "learning_rate": 1.7687074829931973e-05,
+      "loss": 0.5042,
+      "step": 16164
+    },
+    {
+      "epoch": 20.6912,
+      "grad_norm": 1.1234673261642456,
+      "learning_rate": 1.7685074029611845e-05,
+      "loss": 0.5071,
+      "step": 16165
+    },
+    {
+      "epoch": 20.69248,
+      "grad_norm": 1.0681381225585938,
+      "learning_rate": 1.7683073229291717e-05,
+      "loss": 0.5042,
+      "step": 16166
+    },
+    {
+      "epoch": 20.69376,
+      "grad_norm": 1.0953648090362549,
+      "learning_rate": 1.768107242897159e-05,
+      "loss": 0.4874,
+      "step": 16167
+    },
+    {
+      "epoch": 20.69504,
+      "grad_norm": 1.1817059516906738,
+      "learning_rate": 1.767907162865146e-05,
+      "loss": 0.5278,
+      "step": 16168
+    },
+    {
+      "epoch": 20.69632,
+      "grad_norm": 1.130669116973877,
+      "learning_rate": 1.7677070828331332e-05,
+      "loss": 0.5081,
+      "step": 16169
+    },
+    {
+      "epoch": 20.6976,
+      "grad_norm": 1.0833494663238525,
+      "learning_rate": 1.7675070028011208e-05,
+      "loss": 0.4866,
+      "step": 16170
+    },
+    {
+      "epoch": 20.69888,
+      "grad_norm": 1.0703330039978027,
+      "learning_rate": 1.7673069227691076e-05,
+      "loss": 0.4909,
+      "step": 16171
+    },
+    {
+      "epoch": 20.70016,
+      "grad_norm": 1.13280189037323,
+      "learning_rate": 1.7671068427370948e-05,
+      "loss": 0.5435,
+      "step": 16172
+    },
+    {
+      "epoch": 20.70144,
+      "grad_norm": 1.023118495941162,
+      "learning_rate": 1.766906762705082e-05,
+      "loss": 0.4848,
+      "step": 16173
+    },
+    {
+      "epoch": 20.70272,
+      "grad_norm": 1.0794446468353271,
+      "learning_rate": 1.7667066826730695e-05,
+      "loss": 0.5403,
+      "step": 16174
+    },
+    {
+      "epoch": 20.704,
+      "grad_norm": 1.007285237312317,
+      "learning_rate": 1.7665066026410564e-05,
+      "loss": 0.4658,
+      "step": 16175
+    },
+    {
+      "epoch": 20.70528,
+      "grad_norm": 1.0927997827529907,
+      "learning_rate": 1.7663065226090435e-05,
+      "loss": 0.5081,
+      "step": 16176
+    },
+    {
+      "epoch": 20.70656,
+      "grad_norm": 1.104650855064392,
+      "learning_rate": 1.766106442577031e-05,
+      "loss": 0.4623,
+      "step": 16177
+    },
+    {
+      "epoch": 20.70784,
+      "grad_norm": 1.0667095184326172,
+      "learning_rate": 1.7659063625450182e-05,
+      "loss": 0.4841,
+      "step": 16178
+    },
+    {
+      "epoch": 20.70912,
+      "grad_norm": 1.0377103090286255,
+      "learning_rate": 1.765706282513005e-05,
+      "loss": 0.4809,
+      "step": 16179
+    },
+    {
+      "epoch": 20.7104,
+      "grad_norm": 1.1214765310287476,
+      "learning_rate": 1.7655062024809923e-05,
+      "loss": 0.4913,
+      "step": 16180
+    },
+    {
+      "epoch": 20.71168,
+      "grad_norm": 1.079932689666748,
+      "learning_rate": 1.7653061224489798e-05,
+      "loss": 0.4694,
+      "step": 16181
+    },
+    {
+      "epoch": 20.71296,
+      "grad_norm": 1.179184913635254,
+      "learning_rate": 1.765106042416967e-05,
+      "loss": 0.4864,
+      "step": 16182
+    },
+    {
+      "epoch": 20.71424,
+      "grad_norm": 1.0123862028121948,
+      "learning_rate": 1.764905962384954e-05,
+      "loss": 0.4597,
+      "step": 16183
+    },
+    {
+      "epoch": 20.71552,
+      "grad_norm": 1.1281754970550537,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.4937,
+      "step": 16184
+    },
+    {
+      "epoch": 20.7168,
+      "grad_norm": 1.0788217782974243,
+      "learning_rate": 1.7645058023209285e-05,
+      "loss": 0.4898,
+      "step": 16185
+    },
+    {
+      "epoch": 20.71808,
+      "grad_norm": 1.1888701915740967,
+      "learning_rate": 1.7643057222889157e-05,
+      "loss": 0.4891,
+      "step": 16186
+    },
+    {
+      "epoch": 20.71936,
+      "grad_norm": 1.1022487878799438,
+      "learning_rate": 1.7641056422569026e-05,
+      "loss": 0.5248,
+      "step": 16187
+    },
+    {
+      "epoch": 20.72064,
+      "grad_norm": 1.0947836637496948,
+      "learning_rate": 1.76390556222489e-05,
+      "loss": 0.4841,
+      "step": 16188
+    },
+    {
+      "epoch": 20.72192,
+      "grad_norm": 1.101966142654419,
+      "learning_rate": 1.7637054821928773e-05,
+      "loss": 0.5178,
+      "step": 16189
+    },
+    {
+      "epoch": 20.7232,
+      "grad_norm": 1.065540075302124,
+      "learning_rate": 1.7635054021608645e-05,
+      "loss": 0.452,
+      "step": 16190
+    },
+    {
+      "epoch": 20.72448,
+      "grad_norm": 1.024093508720398,
+      "learning_rate": 1.7633053221288517e-05,
+      "loss": 0.4519,
+      "step": 16191
+    },
+    {
+      "epoch": 20.72576,
+      "grad_norm": 1.0961694717407227,
+      "learning_rate": 1.763105242096839e-05,
+      "loss": 0.5014,
+      "step": 16192
+    },
+    {
+      "epoch": 20.72704,
+      "grad_norm": 1.0572614669799805,
+      "learning_rate": 1.762905162064826e-05,
+      "loss": 0.4712,
+      "step": 16193
+    },
+    {
+      "epoch": 20.72832,
+      "grad_norm": 1.2695916891098022,
+      "learning_rate": 1.7627050820328132e-05,
+      "loss": 0.5001,
+      "step": 16194
+    },
+    {
+      "epoch": 20.7296,
+      "grad_norm": 1.0590322017669678,
+      "learning_rate": 1.7625050020008004e-05,
+      "loss": 0.4618,
+      "step": 16195
+    },
+    {
+      "epoch": 20.73088,
+      "grad_norm": 1.0804553031921387,
+      "learning_rate": 1.7623049219687876e-05,
+      "loss": 0.4603,
+      "step": 16196
+    },
+    {
+      "epoch": 20.73216,
+      "grad_norm": 1.037230372428894,
+      "learning_rate": 1.7621048419367748e-05,
+      "loss": 0.4547,
+      "step": 16197
+    },
+    {
+      "epoch": 20.73344,
+      "grad_norm": 1.1001992225646973,
+      "learning_rate": 1.761904761904762e-05,
+      "loss": 0.4539,
+      "step": 16198
+    },
+    {
+      "epoch": 20.73472,
+      "grad_norm": 1.0365502834320068,
+      "learning_rate": 1.761704681872749e-05,
+      "loss": 0.4636,
+      "step": 16199
+    },
+    {
+      "epoch": 20.736,
+      "grad_norm": 1.0428606271743774,
+      "learning_rate": 1.7615046018407363e-05,
+      "loss": 0.4721,
+      "step": 16200
+    },
+    {
+      "epoch": 20.73728,
+      "grad_norm": 1.1415473222732544,
+      "learning_rate": 1.7613045218087235e-05,
+      "loss": 0.4861,
+      "step": 16201
+    },
+    {
+      "epoch": 20.73856,
+      "grad_norm": 1.120957612991333,
+      "learning_rate": 1.7611044417767107e-05,
+      "loss": 0.4992,
+      "step": 16202
+    },
+    {
+      "epoch": 20.73984,
+      "grad_norm": 1.0169925689697266,
+      "learning_rate": 1.760904361744698e-05,
+      "loss": 0.4415,
+      "step": 16203
+    },
+    {
+      "epoch": 20.74112,
+      "grad_norm": 1.1539958715438843,
+      "learning_rate": 1.760704281712685e-05,
+      "loss": 0.539,
+      "step": 16204
+    },
+    {
+      "epoch": 20.7424,
+      "grad_norm": 1.0903569459915161,
+      "learning_rate": 1.7605042016806726e-05,
+      "loss": 0.4769,
+      "step": 16205
+    },
+    {
+      "epoch": 20.74368,
+      "grad_norm": 1.1013376712799072,
+      "learning_rate": 1.7603041216486594e-05,
+      "loss": 0.4809,
+      "step": 16206
+    },
+    {
+      "epoch": 20.74496,
+      "grad_norm": 1.1020969152450562,
+      "learning_rate": 1.7601040416166466e-05,
+      "loss": 0.4534,
+      "step": 16207
+    },
+    {
+      "epoch": 20.74624,
+      "grad_norm": 1.0721830129623413,
+      "learning_rate": 1.7599039615846338e-05,
+      "loss": 0.4651,
+      "step": 16208
+    },
+    {
+      "epoch": 20.74752,
+      "grad_norm": 1.0913660526275635,
+      "learning_rate": 1.7597038815526213e-05,
+      "loss": 0.482,
+      "step": 16209
+    },
+    {
+      "epoch": 20.7488,
+      "grad_norm": 1.0793043375015259,
+      "learning_rate": 1.7595038015206082e-05,
+      "loss": 0.4536,
+      "step": 16210
+    },
+    {
+      "epoch": 20.75008,
+      "grad_norm": 1.0972751379013062,
+      "learning_rate": 1.7593037214885954e-05,
+      "loss": 0.5117,
+      "step": 16211
+    },
+    {
+      "epoch": 20.75136,
+      "grad_norm": 1.1126095056533813,
+      "learning_rate": 1.759103641456583e-05,
+      "loss": 0.5348,
+      "step": 16212
+    },
+    {
+      "epoch": 20.75264,
+      "grad_norm": 1.1272165775299072,
+      "learning_rate": 1.75890356142457e-05,
+      "loss": 0.515,
+      "step": 16213
+    },
+    {
+      "epoch": 20.75392,
+      "grad_norm": 1.1290925741195679,
+      "learning_rate": 1.758703481392557e-05,
+      "loss": 0.5332,
+      "step": 16214
+    },
+    {
+      "epoch": 20.7552,
+      "grad_norm": 1.0821077823638916,
+      "learning_rate": 1.758503401360544e-05,
+      "loss": 0.5224,
+      "step": 16215
+    },
+    {
+      "epoch": 20.75648,
+      "grad_norm": 1.0680049657821655,
+      "learning_rate": 1.7583033213285316e-05,
+      "loss": 0.5149,
+      "step": 16216
+    },
+    {
+      "epoch": 20.75776,
+      "grad_norm": 1.059745192527771,
+      "learning_rate": 1.7581032412965188e-05,
+      "loss": 0.5058,
+      "step": 16217
+    },
+    {
+      "epoch": 20.75904,
+      "grad_norm": 1.0059226751327515,
+      "learning_rate": 1.7579031612645057e-05,
+      "loss": 0.5011,
+      "step": 16218
+    },
+    {
+      "epoch": 20.76032,
+      "grad_norm": 1.1066895723342896,
+      "learning_rate": 1.757703081232493e-05,
+      "loss": 0.5139,
+      "step": 16219
+    },
+    {
+      "epoch": 20.7616,
+      "grad_norm": 1.0385751724243164,
+      "learning_rate": 1.7575030012004804e-05,
+      "loss": 0.4992,
+      "step": 16220
+    },
+    {
+      "epoch": 20.76288,
+      "grad_norm": 1.0588712692260742,
+      "learning_rate": 1.7573029211684676e-05,
+      "loss": 0.4881,
+      "step": 16221
+    },
+    {
+      "epoch": 20.76416,
+      "grad_norm": 1.1335150003433228,
+      "learning_rate": 1.7571028411364544e-05,
+      "loss": 0.5135,
+      "step": 16222
+    },
+    {
+      "epoch": 20.76544,
+      "grad_norm": 1.06964111328125,
+      "learning_rate": 1.756902761104442e-05,
+      "loss": 0.474,
+      "step": 16223
+    },
+    {
+      "epoch": 20.76672,
+      "grad_norm": 1.0755014419555664,
+      "learning_rate": 1.756702681072429e-05,
+      "loss": 0.5125,
+      "step": 16224
+    },
+    {
+      "epoch": 20.768,
+      "grad_norm": 1.129836082458496,
+      "learning_rate": 1.7565026010404163e-05,
+      "loss": 0.4916,
+      "step": 16225
+    },
+    {
+      "epoch": 20.76928,
+      "grad_norm": 1.106661319732666,
+      "learning_rate": 1.756302521008403e-05,
+      "loss": 0.5316,
+      "step": 16226
+    },
+    {
+      "epoch": 20.77056,
+      "grad_norm": 1.123422384262085,
+      "learning_rate": 1.7561024409763907e-05,
+      "loss": 0.5241,
+      "step": 16227
+    },
+    {
+      "epoch": 20.77184,
+      "grad_norm": 1.0029109716415405,
+      "learning_rate": 1.755902360944378e-05,
+      "loss": 0.4284,
+      "step": 16228
+    },
+    {
+      "epoch": 20.77312,
+      "grad_norm": 1.0901793241500854,
+      "learning_rate": 1.755702280912365e-05,
+      "loss": 0.4894,
+      "step": 16229
+    },
+    {
+      "epoch": 20.7744,
+      "grad_norm": 1.003308653831482,
+      "learning_rate": 1.7555022008803522e-05,
+      "loss": 0.4551,
+      "step": 16230
+    },
+    {
+      "epoch": 20.77568,
+      "grad_norm": 1.1484678983688354,
+      "learning_rate": 1.7553021208483394e-05,
+      "loss": 0.4894,
+      "step": 16231
+    },
+    {
+      "epoch": 20.77696,
+      "grad_norm": 1.0901198387145996,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 0.4932,
+      "step": 16232
+    },
+    {
+      "epoch": 20.77824,
+      "grad_norm": 1.1623092889785767,
+      "learning_rate": 1.7549019607843138e-05,
+      "loss": 0.5136,
+      "step": 16233
+    },
+    {
+      "epoch": 20.77952,
+      "grad_norm": 1.0638892650604248,
+      "learning_rate": 1.754701880752301e-05,
+      "loss": 0.4726,
+      "step": 16234
+    },
+    {
+      "epoch": 20.7808,
+      "grad_norm": 1.0341891050338745,
+      "learning_rate": 1.754501800720288e-05,
+      "loss": 0.4597,
+      "step": 16235
+    },
+    {
+      "epoch": 20.78208,
+      "grad_norm": 1.0782972574234009,
+      "learning_rate": 1.7543017206882754e-05,
+      "loss": 0.5077,
+      "step": 16236
+    },
+    {
+      "epoch": 20.78336,
+      "grad_norm": 1.1353981494903564,
+      "learning_rate": 1.7541016406562625e-05,
+      "loss": 0.518,
+      "step": 16237
+    },
+    {
+      "epoch": 20.78464,
+      "grad_norm": 1.0707870721817017,
+      "learning_rate": 1.7539015606242497e-05,
+      "loss": 0.5071,
+      "step": 16238
+    },
+    {
+      "epoch": 20.78592,
+      "grad_norm": 1.0623449087142944,
+      "learning_rate": 1.753701480592237e-05,
+      "loss": 0.4881,
+      "step": 16239
+    },
+    {
+      "epoch": 20.7872,
+      "grad_norm": 1.0724384784698486,
+      "learning_rate": 1.753501400560224e-05,
+      "loss": 0.5197,
+      "step": 16240
+    },
+    {
+      "epoch": 20.78848,
+      "grad_norm": 1.1770350933074951,
+      "learning_rate": 1.7533013205282113e-05,
+      "loss": 0.5326,
+      "step": 16241
+    },
+    {
+      "epoch": 20.78976,
+      "grad_norm": 1.1036503314971924,
+      "learning_rate": 1.7531012404961985e-05,
+      "loss": 0.5241,
+      "step": 16242
+    },
+    {
+      "epoch": 20.79104,
+      "grad_norm": 1.0659743547439575,
+      "learning_rate": 1.7529011604641857e-05,
+      "loss": 0.5047,
+      "step": 16243
+    },
+    {
+      "epoch": 20.79232,
+      "grad_norm": 1.0868451595306396,
+      "learning_rate": 1.7527010804321732e-05,
+      "loss": 0.4819,
+      "step": 16244
+    },
+    {
+      "epoch": 20.7936,
+      "grad_norm": 1.1210837364196777,
+      "learning_rate": 1.75250100040016e-05,
+      "loss": 0.5214,
+      "step": 16245
+    },
+    {
+      "epoch": 20.79488,
+      "grad_norm": 1.065202236175537,
+      "learning_rate": 1.7523009203681472e-05,
+      "loss": 0.4913,
+      "step": 16246
+    },
+    {
+      "epoch": 20.79616,
+      "grad_norm": 1.1045325994491577,
+      "learning_rate": 1.7521008403361344e-05,
+      "loss": 0.493,
+      "step": 16247
+    },
+    {
+      "epoch": 20.79744,
+      "grad_norm": 1.029994249343872,
+      "learning_rate": 1.751900760304122e-05,
+      "loss": 0.4934,
+      "step": 16248
+    },
+    {
+      "epoch": 20.79872,
+      "grad_norm": 1.1114221811294556,
+      "learning_rate": 1.7517006802721088e-05,
+      "loss": 0.5183,
+      "step": 16249
+    },
+    {
+      "epoch": 20.8,
+      "grad_norm": 1.0240123271942139,
+      "learning_rate": 1.751500600240096e-05,
+      "loss": 0.4572,
+      "step": 16250
+    },
+    {
+      "epoch": 20.80128,
+      "grad_norm": 1.0750398635864258,
+      "learning_rate": 1.7513005202080835e-05,
+      "loss": 0.4603,
+      "step": 16251
+    },
+    {
+      "epoch": 20.80256,
+      "grad_norm": 1.1186068058013916,
+      "learning_rate": 1.7511004401760707e-05,
+      "loss": 0.4698,
+      "step": 16252
+    },
+    {
+      "epoch": 20.80384,
+      "grad_norm": 1.1130450963974,
+      "learning_rate": 1.7509003601440575e-05,
+      "loss": 0.5152,
+      "step": 16253
+    },
+    {
+      "epoch": 20.80512,
+      "grad_norm": 1.104473352432251,
+      "learning_rate": 1.7507002801120447e-05,
+      "loss": 0.5139,
+      "step": 16254
+    },
+    {
+      "epoch": 20.8064,
+      "grad_norm": 1.1033751964569092,
+      "learning_rate": 1.7505002000800322e-05,
+      "loss": 0.4785,
+      "step": 16255
+    },
+    {
+      "epoch": 20.80768,
+      "grad_norm": 1.0060745477676392,
+      "learning_rate": 1.7503001200480194e-05,
+      "loss": 0.4259,
+      "step": 16256
+    },
+    {
+      "epoch": 20.80896,
+      "grad_norm": 1.1276196241378784,
+      "learning_rate": 1.7501000400160063e-05,
+      "loss": 0.5602,
+      "step": 16257
+    },
+    {
+      "epoch": 20.81024,
+      "grad_norm": 1.1201550960540771,
+      "learning_rate": 1.7498999599839938e-05,
+      "loss": 0.5034,
+      "step": 16258
+    },
+    {
+      "epoch": 20.81152,
+      "grad_norm": 1.0943266153335571,
+      "learning_rate": 1.749699879951981e-05,
+      "loss": 0.509,
+      "step": 16259
+    },
+    {
+      "epoch": 20.8128,
+      "grad_norm": 1.0522860288619995,
+      "learning_rate": 1.749499799919968e-05,
+      "loss": 0.4813,
+      "step": 16260
+    },
+    {
+      "epoch": 20.81408,
+      "grad_norm": 1.0484342575073242,
+      "learning_rate": 1.749299719887955e-05,
+      "loss": 0.4743,
+      "step": 16261
+    },
+    {
+      "epoch": 20.81536,
+      "grad_norm": 1.0669806003570557,
+      "learning_rate": 1.7490996398559425e-05,
+      "loss": 0.4943,
+      "step": 16262
+    },
+    {
+      "epoch": 20.81664,
+      "grad_norm": 1.063136339187622,
+      "learning_rate": 1.7488995598239297e-05,
+      "loss": 0.4856,
+      "step": 16263
+    },
+    {
+      "epoch": 20.81792,
+      "grad_norm": 1.091721534729004,
+      "learning_rate": 1.748699479791917e-05,
+      "loss": 0.481,
+      "step": 16264
+    },
+    {
+      "epoch": 20.8192,
+      "grad_norm": 1.0925869941711426,
+      "learning_rate": 1.748499399759904e-05,
+      "loss": 0.5033,
+      "step": 16265
+    },
+    {
+      "epoch": 20.82048,
+      "grad_norm": 1.1635428667068481,
+      "learning_rate": 1.7482993197278913e-05,
+      "loss": 0.5144,
+      "step": 16266
+    },
+    {
+      "epoch": 20.82176,
+      "grad_norm": 1.0611239671707153,
+      "learning_rate": 1.7480992396958784e-05,
+      "loss": 0.4716,
+      "step": 16267
+    },
+    {
+      "epoch": 20.82304,
+      "grad_norm": 1.1461703777313232,
+      "learning_rate": 1.7478991596638656e-05,
+      "loss": 0.5118,
+      "step": 16268
+    },
+    {
+      "epoch": 20.82432,
+      "grad_norm": 1.0631418228149414,
+      "learning_rate": 1.7476990796318528e-05,
+      "loss": 0.4622,
+      "step": 16269
+    },
+    {
+      "epoch": 20.8256,
+      "grad_norm": 1.0580074787139893,
+      "learning_rate": 1.74749899959984e-05,
+      "loss": 0.5003,
+      "step": 16270
+    },
+    {
+      "epoch": 20.82688,
+      "grad_norm": 1.1611815690994263,
+      "learning_rate": 1.7472989195678272e-05,
+      "loss": 0.5646,
+      "step": 16271
+    },
+    {
+      "epoch": 20.82816,
+      "grad_norm": 1.1409611701965332,
+      "learning_rate": 1.7470988395358144e-05,
+      "loss": 0.4652,
+      "step": 16272
+    },
+    {
+      "epoch": 20.829439999999998,
+      "grad_norm": 1.1084599494934082,
+      "learning_rate": 1.7468987595038016e-05,
+      "loss": 0.5027,
+      "step": 16273
+    },
+    {
+      "epoch": 20.83072,
+      "grad_norm": 1.1274982690811157,
+      "learning_rate": 1.7466986794717887e-05,
+      "loss": 0.486,
+      "step": 16274
+    },
+    {
+      "epoch": 20.832,
+      "grad_norm": 1.1236320734024048,
+      "learning_rate": 1.746498599439776e-05,
+      "loss": 0.5091,
+      "step": 16275
+    },
+    {
+      "epoch": 20.83328,
+      "grad_norm": 1.1362968683242798,
+      "learning_rate": 1.746298519407763e-05,
+      "loss": 0.4902,
+      "step": 16276
+    },
+    {
+      "epoch": 20.83456,
+      "grad_norm": 1.092063069343567,
+      "learning_rate": 1.7460984393757503e-05,
+      "loss": 0.5271,
+      "step": 16277
+    },
+    {
+      "epoch": 20.83584,
+      "grad_norm": 1.1088082790374756,
+      "learning_rate": 1.7458983593437375e-05,
+      "loss": 0.4834,
+      "step": 16278
+    },
+    {
+      "epoch": 20.83712,
+      "grad_norm": 1.081045150756836,
+      "learning_rate": 1.7456982793117247e-05,
+      "loss": 0.4928,
+      "step": 16279
+    },
+    {
+      "epoch": 20.8384,
+      "grad_norm": 1.1172006130218506,
+      "learning_rate": 1.745498199279712e-05,
+      "loss": 0.5346,
+      "step": 16280
+    },
+    {
+      "epoch": 20.83968,
+      "grad_norm": 1.0738575458526611,
+      "learning_rate": 1.745298119247699e-05,
+      "loss": 0.4861,
+      "step": 16281
+    },
+    {
+      "epoch": 20.84096,
+      "grad_norm": 1.0271522998809814,
+      "learning_rate": 1.7450980392156862e-05,
+      "loss": 0.5072,
+      "step": 16282
+    },
+    {
+      "epoch": 20.84224,
+      "grad_norm": 1.1056361198425293,
+      "learning_rate": 1.7448979591836738e-05,
+      "loss": 0.5387,
+      "step": 16283
+    },
+    {
+      "epoch": 20.84352,
+      "grad_norm": 1.07204270362854,
+      "learning_rate": 1.7446978791516606e-05,
+      "loss": 0.4802,
+      "step": 16284
+    },
+    {
+      "epoch": 20.8448,
+      "grad_norm": 1.0948388576507568,
+      "learning_rate": 1.7444977991196478e-05,
+      "loss": 0.4963,
+      "step": 16285
+    },
+    {
+      "epoch": 20.84608,
+      "grad_norm": 1.1299151182174683,
+      "learning_rate": 1.744297719087635e-05,
+      "loss": 0.5061,
+      "step": 16286
+    },
+    {
+      "epoch": 20.84736,
+      "grad_norm": 1.0689219236373901,
+      "learning_rate": 1.7440976390556225e-05,
+      "loss": 0.4855,
+      "step": 16287
+    },
+    {
+      "epoch": 20.84864,
+      "grad_norm": 1.1455039978027344,
+      "learning_rate": 1.7438975590236093e-05,
+      "loss": 0.5598,
+      "step": 16288
+    },
+    {
+      "epoch": 20.84992,
+      "grad_norm": 1.1063296794891357,
+      "learning_rate": 1.7436974789915965e-05,
+      "loss": 0.4986,
+      "step": 16289
+    },
+    {
+      "epoch": 20.8512,
+      "grad_norm": 1.1937898397445679,
+      "learning_rate": 1.743497398959584e-05,
+      "loss": 0.513,
+      "step": 16290
+    },
+    {
+      "epoch": 20.85248,
+      "grad_norm": 1.005609154701233,
+      "learning_rate": 1.7432973189275712e-05,
+      "loss": 0.4549,
+      "step": 16291
+    },
+    {
+      "epoch": 20.85376,
+      "grad_norm": 1.0839899778366089,
+      "learning_rate": 1.743097238895558e-05,
+      "loss": 0.4952,
+      "step": 16292
+    },
+    {
+      "epoch": 20.85504,
+      "grad_norm": 1.1058770418167114,
+      "learning_rate": 1.7428971588635453e-05,
+      "loss": 0.4729,
+      "step": 16293
+    },
+    {
+      "epoch": 20.85632,
+      "grad_norm": 1.0722662210464478,
+      "learning_rate": 1.7426970788315328e-05,
+      "loss": 0.4774,
+      "step": 16294
+    },
+    {
+      "epoch": 20.8576,
+      "grad_norm": 1.0828754901885986,
+      "learning_rate": 1.74249699879952e-05,
+      "loss": 0.5036,
+      "step": 16295
+    },
+    {
+      "epoch": 20.85888,
+      "grad_norm": 1.0864346027374268,
+      "learning_rate": 1.742296918767507e-05,
+      "loss": 0.5187,
+      "step": 16296
+    },
+    {
+      "epoch": 20.86016,
+      "grad_norm": 1.0293843746185303,
+      "learning_rate": 1.7420968387354944e-05,
+      "loss": 0.4526,
+      "step": 16297
+    },
+    {
+      "epoch": 20.86144,
+      "grad_norm": 1.0849385261535645,
+      "learning_rate": 1.7418967587034815e-05,
+      "loss": 0.4959,
+      "step": 16298
+    },
+    {
+      "epoch": 20.86272,
+      "grad_norm": 1.0619242191314697,
+      "learning_rate": 1.7416966786714687e-05,
+      "loss": 0.504,
+      "step": 16299
+    },
+    {
+      "epoch": 20.864,
+      "grad_norm": 1.0861916542053223,
+      "learning_rate": 1.7414965986394556e-05,
+      "loss": 0.4619,
+      "step": 16300
+    },
+    {
+      "epoch": 20.86528,
+      "grad_norm": 1.0871127843856812,
+      "learning_rate": 1.741296518607443e-05,
+      "loss": 0.4972,
+      "step": 16301
+    },
+    {
+      "epoch": 20.86656,
+      "grad_norm": 1.1219013929367065,
+      "learning_rate": 1.7410964385754303e-05,
+      "loss": 0.4946,
+      "step": 16302
+    },
+    {
+      "epoch": 20.86784,
+      "grad_norm": 1.0501264333724976,
+      "learning_rate": 1.7408963585434175e-05,
+      "loss": 0.5039,
+      "step": 16303
+    },
+    {
+      "epoch": 20.86912,
+      "grad_norm": 1.1329530477523804,
+      "learning_rate": 1.7406962785114047e-05,
+      "loss": 0.5356,
+      "step": 16304
+    },
+    {
+      "epoch": 20.8704,
+      "grad_norm": 1.0762802362442017,
+      "learning_rate": 1.740496198479392e-05,
+      "loss": 0.5171,
+      "step": 16305
+    },
+    {
+      "epoch": 20.87168,
+      "grad_norm": 1.0678480863571167,
+      "learning_rate": 1.740296118447379e-05,
+      "loss": 0.4828,
+      "step": 16306
+    },
+    {
+      "epoch": 20.87296,
+      "grad_norm": 1.1031467914581299,
+      "learning_rate": 1.7400960384153662e-05,
+      "loss": 0.4672,
+      "step": 16307
+    },
+    {
+      "epoch": 20.87424,
+      "grad_norm": 1.0675780773162842,
+      "learning_rate": 1.7398959583833534e-05,
+      "loss": 0.5382,
+      "step": 16308
+    },
+    {
+      "epoch": 20.87552,
+      "grad_norm": 1.0954948663711548,
+      "learning_rate": 1.7396958783513406e-05,
+      "loss": 0.5027,
+      "step": 16309
+    },
+    {
+      "epoch": 20.8768,
+      "grad_norm": 1.0903716087341309,
+      "learning_rate": 1.7394957983193278e-05,
+      "loss": 0.4796,
+      "step": 16310
+    },
+    {
+      "epoch": 20.87808,
+      "grad_norm": 1.147133708000183,
+      "learning_rate": 1.7392957182873153e-05,
+      "loss": 0.5004,
+      "step": 16311
+    },
+    {
+      "epoch": 20.87936,
+      "grad_norm": 1.1145248413085938,
+      "learning_rate": 1.739095638255302e-05,
+      "loss": 0.5235,
+      "step": 16312
+    },
+    {
+      "epoch": 20.88064,
+      "grad_norm": 1.0918245315551758,
+      "learning_rate": 1.7388955582232893e-05,
+      "loss": 0.4825,
+      "step": 16313
+    },
+    {
+      "epoch": 20.88192,
+      "grad_norm": 1.1098552942276,
+      "learning_rate": 1.7386954781912765e-05,
+      "loss": 0.4949,
+      "step": 16314
+    },
+    {
+      "epoch": 20.8832,
+      "grad_norm": 1.1097396612167358,
+      "learning_rate": 1.738495398159264e-05,
+      "loss": 0.49,
+      "step": 16315
+    },
+    {
+      "epoch": 20.88448,
+      "grad_norm": 1.1164050102233887,
+      "learning_rate": 1.738295318127251e-05,
+      "loss": 0.5204,
+      "step": 16316
+    },
+    {
+      "epoch": 20.88576,
+      "grad_norm": 1.1462451219558716,
+      "learning_rate": 1.738095238095238e-05,
+      "loss": 0.5449,
+      "step": 16317
+    },
+    {
+      "epoch": 20.88704,
+      "grad_norm": 1.052255392074585,
+      "learning_rate": 1.7378951580632256e-05,
+      "loss": 0.481,
+      "step": 16318
+    },
+    {
+      "epoch": 20.88832,
+      "grad_norm": 1.0906624794006348,
+      "learning_rate": 1.7376950780312128e-05,
+      "loss": 0.4848,
+      "step": 16319
+    },
+    {
+      "epoch": 20.8896,
+      "grad_norm": 1.0164730548858643,
+      "learning_rate": 1.7374949979991996e-05,
+      "loss": 0.4646,
+      "step": 16320
+    },
+    {
+      "epoch": 20.89088,
+      "grad_norm": 1.1174423694610596,
+      "learning_rate": 1.7372949179671868e-05,
+      "loss": 0.5213,
+      "step": 16321
+    },
+    {
+      "epoch": 20.89216,
+      "grad_norm": 1.0460309982299805,
+      "learning_rate": 1.7370948379351743e-05,
+      "loss": 0.5021,
+      "step": 16322
+    },
+    {
+      "epoch": 20.89344,
+      "grad_norm": 1.0785523653030396,
+      "learning_rate": 1.7368947579031615e-05,
+      "loss": 0.4775,
+      "step": 16323
+    },
+    {
+      "epoch": 20.89472,
+      "grad_norm": 1.173263430595398,
+      "learning_rate": 1.7366946778711484e-05,
+      "loss": 0.5126,
+      "step": 16324
+    },
+    {
+      "epoch": 20.896,
+      "grad_norm": 1.0594722032546997,
+      "learning_rate": 1.736494597839136e-05,
+      "loss": 0.4804,
+      "step": 16325
+    },
+    {
+      "epoch": 20.89728,
+      "grad_norm": 1.1119956970214844,
+      "learning_rate": 1.736294517807123e-05,
+      "loss": 0.4952,
+      "step": 16326
+    },
+    {
+      "epoch": 20.89856,
+      "grad_norm": 1.0069000720977783,
+      "learning_rate": 1.7360944377751103e-05,
+      "loss": 0.4821,
+      "step": 16327
+    },
+    {
+      "epoch": 20.89984,
+      "grad_norm": 1.118168830871582,
+      "learning_rate": 1.735894357743097e-05,
+      "loss": 0.513,
+      "step": 16328
+    },
+    {
+      "epoch": 20.90112,
+      "grad_norm": 1.0356378555297852,
+      "learning_rate": 1.7356942777110846e-05,
+      "loss": 0.435,
+      "step": 16329
+    },
+    {
+      "epoch": 20.9024,
+      "grad_norm": 1.0259177684783936,
+      "learning_rate": 1.7354941976790718e-05,
+      "loss": 0.4688,
+      "step": 16330
+    },
+    {
+      "epoch": 20.90368,
+      "grad_norm": 1.057941198348999,
+      "learning_rate": 1.735294117647059e-05,
+      "loss": 0.4662,
+      "step": 16331
+    },
+    {
+      "epoch": 20.90496,
+      "grad_norm": 1.1319576501846313,
+      "learning_rate": 1.735094037615046e-05,
+      "loss": 0.5064,
+      "step": 16332
+    },
+    {
+      "epoch": 20.90624,
+      "grad_norm": 1.1116087436676025,
+      "learning_rate": 1.7348939575830334e-05,
+      "loss": 0.5034,
+      "step": 16333
+    },
+    {
+      "epoch": 20.90752,
+      "grad_norm": 1.1020821332931519,
+      "learning_rate": 1.7346938775510206e-05,
+      "loss": 0.509,
+      "step": 16334
+    },
+    {
+      "epoch": 20.9088,
+      "grad_norm": 1.024755835533142,
+      "learning_rate": 1.7344937975190078e-05,
+      "loss": 0.4732,
+      "step": 16335
+    },
+    {
+      "epoch": 20.91008,
+      "grad_norm": 1.1245591640472412,
+      "learning_rate": 1.734293717486995e-05,
+      "loss": 0.5083,
+      "step": 16336
+    },
+    {
+      "epoch": 20.91136,
+      "grad_norm": 1.0561660528182983,
+      "learning_rate": 1.734093637454982e-05,
+      "loss": 0.4551,
+      "step": 16337
+    },
+    {
+      "epoch": 20.91264,
+      "grad_norm": 1.017407774925232,
+      "learning_rate": 1.7338935574229693e-05,
+      "loss": 0.4703,
+      "step": 16338
+    },
+    {
+      "epoch": 20.91392,
+      "grad_norm": 1.145080804824829,
+      "learning_rate": 1.7336934773909565e-05,
+      "loss": 0.5039,
+      "step": 16339
+    },
+    {
+      "epoch": 20.9152,
+      "grad_norm": 1.074860692024231,
+      "learning_rate": 1.7334933973589437e-05,
+      "loss": 0.4906,
+      "step": 16340
+    },
+    {
+      "epoch": 20.91648,
+      "grad_norm": 1.1369560956954956,
+      "learning_rate": 1.733293317326931e-05,
+      "loss": 0.5045,
+      "step": 16341
+    },
+    {
+      "epoch": 20.91776,
+      "grad_norm": 1.0858222246170044,
+      "learning_rate": 1.733093237294918e-05,
+      "loss": 0.4916,
+      "step": 16342
+    },
+    {
+      "epoch": 20.91904,
+      "grad_norm": 1.1016911268234253,
+      "learning_rate": 1.7328931572629052e-05,
+      "loss": 0.521,
+      "step": 16343
+    },
+    {
+      "epoch": 20.92032,
+      "grad_norm": 1.0700989961624146,
+      "learning_rate": 1.7326930772308924e-05,
+      "loss": 0.4737,
+      "step": 16344
+    },
+    {
+      "epoch": 20.9216,
+      "grad_norm": 1.066236972808838,
+      "learning_rate": 1.7324929971988796e-05,
+      "loss": 0.4873,
+      "step": 16345
+    },
+    {
+      "epoch": 20.92288,
+      "grad_norm": 1.1762385368347168,
+      "learning_rate": 1.7322929171668668e-05,
+      "loss": 0.5469,
+      "step": 16346
+    },
+    {
+      "epoch": 20.92416,
+      "grad_norm": 1.086787462234497,
+      "learning_rate": 1.732092837134854e-05,
+      "loss": 0.5013,
+      "step": 16347
+    },
+    {
+      "epoch": 20.925440000000002,
+      "grad_norm": 1.0452648401260376,
+      "learning_rate": 1.731892757102841e-05,
+      "loss": 0.4687,
+      "step": 16348
+    },
+    {
+      "epoch": 20.92672,
+      "grad_norm": 1.1296664476394653,
+      "learning_rate": 1.7316926770708284e-05,
+      "loss": 0.5306,
+      "step": 16349
+    },
+    {
+      "epoch": 20.928,
+      "grad_norm": 1.0765169858932495,
+      "learning_rate": 1.731492597038816e-05,
+      "loss": 0.4865,
+      "step": 16350
+    },
+    {
+      "epoch": 20.92928,
+      "grad_norm": 1.122309684753418,
+      "learning_rate": 1.7312925170068027e-05,
+      "loss": 0.4692,
+      "step": 16351
+    },
+    {
+      "epoch": 20.93056,
+      "grad_norm": 1.1600003242492676,
+      "learning_rate": 1.73109243697479e-05,
+      "loss": 0.545,
+      "step": 16352
+    },
+    {
+      "epoch": 20.93184,
+      "grad_norm": 1.06452214717865,
+      "learning_rate": 1.730892356942777e-05,
+      "loss": 0.4528,
+      "step": 16353
+    },
+    {
+      "epoch": 20.93312,
+      "grad_norm": 1.0116653442382812,
+      "learning_rate": 1.7306922769107646e-05,
+      "loss": 0.444,
+      "step": 16354
+    },
+    {
+      "epoch": 20.9344,
+      "grad_norm": 1.0115692615509033,
+      "learning_rate": 1.7304921968787515e-05,
+      "loss": 0.4983,
+      "step": 16355
+    },
+    {
+      "epoch": 20.93568,
+      "grad_norm": 1.1115832328796387,
+      "learning_rate": 1.7302921168467387e-05,
+      "loss": 0.4955,
+      "step": 16356
+    },
+    {
+      "epoch": 20.93696,
+      "grad_norm": 1.0773097276687622,
+      "learning_rate": 1.7300920368147262e-05,
+      "loss": 0.4978,
+      "step": 16357
+    },
+    {
+      "epoch": 20.93824,
+      "grad_norm": 1.102440595626831,
+      "learning_rate": 1.7298919567827134e-05,
+      "loss": 0.5104,
+      "step": 16358
+    },
+    {
+      "epoch": 20.93952,
+      "grad_norm": 1.0837904214859009,
+      "learning_rate": 1.7296918767507002e-05,
+      "loss": 0.4988,
+      "step": 16359
+    },
+    {
+      "epoch": 20.9408,
+      "grad_norm": 1.091955542564392,
+      "learning_rate": 1.7294917967186874e-05,
+      "loss": 0.5182,
+      "step": 16360
+    },
+    {
+      "epoch": 20.94208,
+      "grad_norm": 1.13055419921875,
+      "learning_rate": 1.729291716686675e-05,
+      "loss": 0.5187,
+      "step": 16361
+    },
+    {
+      "epoch": 20.94336,
+      "grad_norm": 1.1152286529541016,
+      "learning_rate": 1.729091636654662e-05,
+      "loss": 0.498,
+      "step": 16362
+    },
+    {
+      "epoch": 20.94464,
+      "grad_norm": 1.0456947088241577,
+      "learning_rate": 1.728891556622649e-05,
+      "loss": 0.5033,
+      "step": 16363
+    },
+    {
+      "epoch": 20.94592,
+      "grad_norm": 1.0851963758468628,
+      "learning_rate": 1.7286914765906365e-05,
+      "loss": 0.5229,
+      "step": 16364
+    },
+    {
+      "epoch": 20.9472,
+      "grad_norm": 1.03732168674469,
+      "learning_rate": 1.7284913965586237e-05,
+      "loss": 0.4809,
+      "step": 16365
+    },
+    {
+      "epoch": 20.94848,
+      "grad_norm": 1.0764819383621216,
+      "learning_rate": 1.728291316526611e-05,
+      "loss": 0.4754,
+      "step": 16366
+    },
+    {
+      "epoch": 20.94976,
+      "grad_norm": 1.1544440984725952,
+      "learning_rate": 1.7280912364945977e-05,
+      "loss": 0.5008,
+      "step": 16367
+    },
+    {
+      "epoch": 20.95104,
+      "grad_norm": 1.1239932775497437,
+      "learning_rate": 1.7278911564625852e-05,
+      "loss": 0.4808,
+      "step": 16368
+    },
+    {
+      "epoch": 20.95232,
+      "grad_norm": 1.1171079874038696,
+      "learning_rate": 1.7276910764305724e-05,
+      "loss": 0.4859,
+      "step": 16369
+    },
+    {
+      "epoch": 20.9536,
+      "grad_norm": 1.1345586776733398,
+      "learning_rate": 1.7274909963985596e-05,
+      "loss": 0.4935,
+      "step": 16370
+    },
+    {
+      "epoch": 20.95488,
+      "grad_norm": 1.1230216026306152,
+      "learning_rate": 1.7272909163665468e-05,
+      "loss": 0.5274,
+      "step": 16371
+    },
+    {
+      "epoch": 20.95616,
+      "grad_norm": 1.2072840929031372,
+      "learning_rate": 1.727090836334534e-05,
+      "loss": 0.5538,
+      "step": 16372
+    },
+    {
+      "epoch": 20.95744,
+      "grad_norm": 1.1466865539550781,
+      "learning_rate": 1.726890756302521e-05,
+      "loss": 0.5492,
+      "step": 16373
+    },
+    {
+      "epoch": 20.95872,
+      "grad_norm": 1.017502784729004,
+      "learning_rate": 1.7266906762705083e-05,
+      "loss": 0.4899,
+      "step": 16374
+    },
+    {
+      "epoch": 20.96,
+      "grad_norm": 1.1090508699417114,
+      "learning_rate": 1.7264905962384955e-05,
+      "loss": 0.4935,
+      "step": 16375
+    },
+    {
+      "epoch": 20.96128,
+      "grad_norm": 1.078791618347168,
+      "learning_rate": 1.7262905162064827e-05,
+      "loss": 0.4829,
+      "step": 16376
+    },
+    {
+      "epoch": 20.96256,
+      "grad_norm": 1.1684792041778564,
+      "learning_rate": 1.72609043617447e-05,
+      "loss": 0.5098,
+      "step": 16377
+    },
+    {
+      "epoch": 20.96384,
+      "grad_norm": 1.1406118869781494,
+      "learning_rate": 1.725890356142457e-05,
+      "loss": 0.5024,
+      "step": 16378
+    },
+    {
+      "epoch": 20.96512,
+      "grad_norm": 1.0845208168029785,
+      "learning_rate": 1.7256902761104443e-05,
+      "loss": 0.4965,
+      "step": 16379
+    },
+    {
+      "epoch": 20.9664,
+      "grad_norm": 1.04744291305542,
+      "learning_rate": 1.7254901960784314e-05,
+      "loss": 0.4817,
+      "step": 16380
+    },
+    {
+      "epoch": 20.96768,
+      "grad_norm": 1.1002355813980103,
+      "learning_rate": 1.7252901160464186e-05,
+      "loss": 0.4886,
+      "step": 16381
+    },
+    {
+      "epoch": 20.96896,
+      "grad_norm": 1.0374670028686523,
+      "learning_rate": 1.7250900360144058e-05,
+      "loss": 0.4639,
+      "step": 16382
+    },
+    {
+      "epoch": 20.97024,
+      "grad_norm": 1.0553470849990845,
+      "learning_rate": 1.724889955982393e-05,
+      "loss": 0.4705,
+      "step": 16383
+    },
+    {
+      "epoch": 20.97152,
+      "grad_norm": 1.1559804677963257,
+      "learning_rate": 1.7246898759503802e-05,
+      "loss": 0.5046,
+      "step": 16384
+    },
+    {
+      "epoch": 20.9728,
+      "grad_norm": 1.1546366214752197,
+      "learning_rate": 1.7244897959183677e-05,
+      "loss": 0.5199,
+      "step": 16385
+    },
+    {
+      "epoch": 20.97408,
+      "grad_norm": 1.0644662380218506,
+      "learning_rate": 1.7242897158863546e-05,
+      "loss": 0.4589,
+      "step": 16386
+    },
+    {
+      "epoch": 20.97536,
+      "grad_norm": 1.0808895826339722,
+      "learning_rate": 1.7240896358543417e-05,
+      "loss": 0.5159,
+      "step": 16387
+    },
+    {
+      "epoch": 20.97664,
+      "grad_norm": 1.1637095212936401,
+      "learning_rate": 1.723889555822329e-05,
+      "loss": 0.5412,
+      "step": 16388
+    },
+    {
+      "epoch": 20.97792,
+      "grad_norm": 1.0618444681167603,
+      "learning_rate": 1.7236894757903165e-05,
+      "loss": 0.5048,
+      "step": 16389
+    },
+    {
+      "epoch": 20.9792,
+      "grad_norm": 1.1485847234725952,
+      "learning_rate": 1.7234893957583033e-05,
+      "loss": 0.4985,
+      "step": 16390
+    },
+    {
+      "epoch": 20.98048,
+      "grad_norm": 1.1214594841003418,
+      "learning_rate": 1.7232893157262905e-05,
+      "loss": 0.4636,
+      "step": 16391
+    },
+    {
+      "epoch": 20.98176,
+      "grad_norm": 1.1146374940872192,
+      "learning_rate": 1.7230892356942777e-05,
+      "loss": 0.5329,
+      "step": 16392
+    },
+    {
+      "epoch": 20.98304,
+      "grad_norm": 1.092591404914856,
+      "learning_rate": 1.7228891556622652e-05,
+      "loss": 0.5354,
+      "step": 16393
+    },
+    {
+      "epoch": 20.98432,
+      "grad_norm": 1.1593587398529053,
+      "learning_rate": 1.722689075630252e-05,
+      "loss": 0.4776,
+      "step": 16394
+    },
+    {
+      "epoch": 20.9856,
+      "grad_norm": 1.0387076139450073,
+      "learning_rate": 1.7224889955982392e-05,
+      "loss": 0.4753,
+      "step": 16395
+    },
+    {
+      "epoch": 20.98688,
+      "grad_norm": 1.0778547525405884,
+      "learning_rate": 1.7222889155662268e-05,
+      "loss": 0.4997,
+      "step": 16396
+    },
+    {
+      "epoch": 20.98816,
+      "grad_norm": 1.0431793928146362,
+      "learning_rate": 1.722088835534214e-05,
+      "loss": 0.4463,
+      "step": 16397
+    },
+    {
+      "epoch": 20.98944,
+      "grad_norm": 1.0882987976074219,
+      "learning_rate": 1.7218887555022008e-05,
+      "loss": 0.4593,
+      "step": 16398
+    },
+    {
+      "epoch": 20.99072,
+      "grad_norm": 1.1702048778533936,
+      "learning_rate": 1.721688675470188e-05,
+      "loss": 0.5429,
+      "step": 16399
+    },
+    {
+      "epoch": 20.992,
+      "grad_norm": 1.0901694297790527,
+      "learning_rate": 1.7214885954381755e-05,
+      "loss": 0.5139,
+      "step": 16400
+    },
+    {
+      "epoch": 20.99328,
+      "grad_norm": 1.1674569845199585,
+      "learning_rate": 1.7212885154061627e-05,
+      "loss": 0.5385,
+      "step": 16401
+    },
+    {
+      "epoch": 20.99456,
+      "grad_norm": 1.1179885864257812,
+      "learning_rate": 1.7210884353741495e-05,
+      "loss": 0.4636,
+      "step": 16402
+    },
+    {
+      "epoch": 20.99584,
+      "grad_norm": 1.1010475158691406,
+      "learning_rate": 1.720888355342137e-05,
+      "loss": 0.5093,
+      "step": 16403
+    },
+    {
+      "epoch": 20.99712,
+      "grad_norm": 1.157423734664917,
+      "learning_rate": 1.7206882753101242e-05,
+      "loss": 0.5069,
+      "step": 16404
+    },
+    {
+      "epoch": 20.9984,
+      "grad_norm": 1.0208972692489624,
+      "learning_rate": 1.7204881952781114e-05,
+      "loss": 0.4783,
+      "step": 16405
+    },
+    {
+      "epoch": 20.99968,
+      "grad_norm": 1.0992622375488281,
+      "learning_rate": 1.7202881152460983e-05,
+      "loss": 0.4992,
+      "step": 16406
+    },
+    {
+      "epoch": 21.00096,
+      "grad_norm": 2.2934412956237793,
+      "learning_rate": 1.7200880352140858e-05,
+      "loss": 0.8021,
+      "step": 16407
+    },
+    {
+      "epoch": 21.00224,
+      "grad_norm": 1.0393415689468384,
+      "learning_rate": 1.719887955182073e-05,
+      "loss": 0.4699,
+      "step": 16408
+    },
+    {
+      "epoch": 21.00352,
+      "grad_norm": 1.0242365598678589,
+      "learning_rate": 1.7196878751500602e-05,
+      "loss": 0.4762,
+      "step": 16409
+    },
+    {
+      "epoch": 21.0048,
+      "grad_norm": 1.0307321548461914,
+      "learning_rate": 1.7194877951180474e-05,
+      "loss": 0.4812,
+      "step": 16410
+    },
+    {
+      "epoch": 21.00608,
+      "grad_norm": 1.07740318775177,
+      "learning_rate": 1.7192877150860345e-05,
+      "loss": 0.4985,
+      "step": 16411
+    },
+    {
+      "epoch": 21.00736,
+      "grad_norm": 1.0476981401443481,
+      "learning_rate": 1.7190876350540217e-05,
+      "loss": 0.4478,
+      "step": 16412
+    },
+    {
+      "epoch": 21.00864,
+      "grad_norm": 1.041973352432251,
+      "learning_rate": 1.718887555022009e-05,
+      "loss": 0.4452,
+      "step": 16413
+    },
+    {
+      "epoch": 21.00992,
+      "grad_norm": 1.1095020771026611,
+      "learning_rate": 1.718687474989996e-05,
+      "loss": 0.5099,
+      "step": 16414
+    },
+    {
+      "epoch": 21.0112,
+      "grad_norm": 1.123745083808899,
+      "learning_rate": 1.7184873949579833e-05,
+      "loss": 0.4954,
+      "step": 16415
+    },
+    {
+      "epoch": 21.01248,
+      "grad_norm": 1.1033985614776611,
+      "learning_rate": 1.7182873149259705e-05,
+      "loss": 0.4863,
+      "step": 16416
+    },
+    {
+      "epoch": 21.01376,
+      "grad_norm": 1.1041131019592285,
+      "learning_rate": 1.7180872348939577e-05,
+      "loss": 0.4556,
+      "step": 16417
+    },
+    {
+      "epoch": 21.01504,
+      "grad_norm": 1.0607541799545288,
+      "learning_rate": 1.717887154861945e-05,
+      "loss": 0.4738,
+      "step": 16418
+    },
+    {
+      "epoch": 21.01632,
+      "grad_norm": 1.0398081541061401,
+      "learning_rate": 1.717687074829932e-05,
+      "loss": 0.4566,
+      "step": 16419
+    },
+    {
+      "epoch": 21.0176,
+      "grad_norm": 1.0724793672561646,
+      "learning_rate": 1.7174869947979192e-05,
+      "loss": 0.4586,
+      "step": 16420
+    },
+    {
+      "epoch": 21.01888,
+      "grad_norm": 1.109832525253296,
+      "learning_rate": 1.7172869147659064e-05,
+      "loss": 0.4661,
+      "step": 16421
+    },
+    {
+      "epoch": 21.02016,
+      "grad_norm": 1.0665154457092285,
+      "learning_rate": 1.7170868347338936e-05,
+      "loss": 0.4561,
+      "step": 16422
+    },
+    {
+      "epoch": 21.02144,
+      "grad_norm": 1.094376564025879,
+      "learning_rate": 1.7168867547018808e-05,
+      "loss": 0.4649,
+      "step": 16423
+    },
+    {
+      "epoch": 21.02272,
+      "grad_norm": 1.061568021774292,
+      "learning_rate": 1.7166866746698683e-05,
+      "loss": 0.4897,
+      "step": 16424
+    },
+    {
+      "epoch": 21.024,
+      "grad_norm": 1.0756134986877441,
+      "learning_rate": 1.716486594637855e-05,
+      "loss": 0.4575,
+      "step": 16425
+    },
+    {
+      "epoch": 21.02528,
+      "grad_norm": 1.0889450311660767,
+      "learning_rate": 1.7162865146058423e-05,
+      "loss": 0.5075,
+      "step": 16426
+    },
+    {
+      "epoch": 21.02656,
+      "grad_norm": 1.0604745149612427,
+      "learning_rate": 1.7160864345738295e-05,
+      "loss": 0.4671,
+      "step": 16427
+    },
+    {
+      "epoch": 21.02784,
+      "grad_norm": 1.0580859184265137,
+      "learning_rate": 1.715886354541817e-05,
+      "loss": 0.4691,
+      "step": 16428
+    },
+    {
+      "epoch": 21.02912,
+      "grad_norm": 1.1256431341171265,
+      "learning_rate": 1.715686274509804e-05,
+      "loss": 0.5068,
+      "step": 16429
+    },
+    {
+      "epoch": 21.0304,
+      "grad_norm": 1.1371206045150757,
+      "learning_rate": 1.715486194477791e-05,
+      "loss": 0.484,
+      "step": 16430
+    },
+    {
+      "epoch": 21.03168,
+      "grad_norm": 1.045066237449646,
+      "learning_rate": 1.7152861144457786e-05,
+      "loss": 0.4497,
+      "step": 16431
+    },
+    {
+      "epoch": 21.03296,
+      "grad_norm": 1.1001946926116943,
+      "learning_rate": 1.7150860344137658e-05,
+      "loss": 0.4488,
+      "step": 16432
+    },
+    {
+      "epoch": 21.03424,
+      "grad_norm": 1.087947964668274,
+      "learning_rate": 1.7148859543817526e-05,
+      "loss": 0.475,
+      "step": 16433
+    },
+    {
+      "epoch": 21.03552,
+      "grad_norm": 1.0743104219436646,
+      "learning_rate": 1.7146858743497398e-05,
+      "loss": 0.4484,
+      "step": 16434
+    },
+    {
+      "epoch": 21.0368,
+      "grad_norm": 1.0668704509735107,
+      "learning_rate": 1.7144857943177273e-05,
+      "loss": 0.4431,
+      "step": 16435
+    },
+    {
+      "epoch": 21.03808,
+      "grad_norm": 1.1153063774108887,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": 0.4967,
+      "step": 16436
+    },
+    {
+      "epoch": 21.03936,
+      "grad_norm": 1.122804045677185,
+      "learning_rate": 1.7140856342537014e-05,
+      "loss": 0.4632,
+      "step": 16437
+    },
+    {
+      "epoch": 21.04064,
+      "grad_norm": 1.1385273933410645,
+      "learning_rate": 1.713885554221689e-05,
+      "loss": 0.4943,
+      "step": 16438
+    },
+    {
+      "epoch": 21.04192,
+      "grad_norm": 1.0751020908355713,
+      "learning_rate": 1.713685474189676e-05,
+      "loss": 0.4541,
+      "step": 16439
+    },
+    {
+      "epoch": 21.0432,
+      "grad_norm": 1.0615102052688599,
+      "learning_rate": 1.7134853941576633e-05,
+      "loss": 0.48,
+      "step": 16440
+    },
+    {
+      "epoch": 21.04448,
+      "grad_norm": 1.1338032484054565,
+      "learning_rate": 1.71328531412565e-05,
+      "loss": 0.4953,
+      "step": 16441
+    },
+    {
+      "epoch": 21.04576,
+      "grad_norm": 1.0364360809326172,
+      "learning_rate": 1.7130852340936376e-05,
+      "loss": 0.4364,
+      "step": 16442
+    },
+    {
+      "epoch": 21.04704,
+      "grad_norm": 1.0834779739379883,
+      "learning_rate": 1.7128851540616248e-05,
+      "loss": 0.5213,
+      "step": 16443
+    },
+    {
+      "epoch": 21.04832,
+      "grad_norm": 1.0926162004470825,
+      "learning_rate": 1.712685074029612e-05,
+      "loss": 0.498,
+      "step": 16444
+    },
+    {
+      "epoch": 21.0496,
+      "grad_norm": 1.0313501358032227,
+      "learning_rate": 1.712484993997599e-05,
+      "loss": 0.4867,
+      "step": 16445
+    },
+    {
+      "epoch": 21.05088,
+      "grad_norm": 1.0741840600967407,
+      "learning_rate": 1.7122849139655864e-05,
+      "loss": 0.5016,
+      "step": 16446
+    },
+    {
+      "epoch": 21.05216,
+      "grad_norm": 1.1211599111557007,
+      "learning_rate": 1.7120848339335736e-05,
+      "loss": 0.5157,
+      "step": 16447
+    },
+    {
+      "epoch": 21.05344,
+      "grad_norm": 1.0967553853988647,
+      "learning_rate": 1.7118847539015608e-05,
+      "loss": 0.4737,
+      "step": 16448
+    },
+    {
+      "epoch": 21.05472,
+      "grad_norm": 1.204904317855835,
+      "learning_rate": 1.711684673869548e-05,
+      "loss": 0.5274,
+      "step": 16449
+    },
+    {
+      "epoch": 21.056,
+      "grad_norm": 1.0651218891143799,
+      "learning_rate": 1.711484593837535e-05,
+      "loss": 0.4312,
+      "step": 16450
+    },
+    {
+      "epoch": 21.05728,
+      "grad_norm": 1.0928947925567627,
+      "learning_rate": 1.7112845138055223e-05,
+      "loss": 0.4495,
+      "step": 16451
+    },
+    {
+      "epoch": 21.05856,
+      "grad_norm": 1.0684500932693481,
+      "learning_rate": 1.7110844337735095e-05,
+      "loss": 0.4851,
+      "step": 16452
+    },
+    {
+      "epoch": 21.05984,
+      "grad_norm": 1.095968246459961,
+      "learning_rate": 1.7108843537414967e-05,
+      "loss": 0.452,
+      "step": 16453
+    },
+    {
+      "epoch": 21.06112,
+      "grad_norm": 1.140784502029419,
+      "learning_rate": 1.710684273709484e-05,
+      "loss": 0.5047,
+      "step": 16454
+    },
+    {
+      "epoch": 21.0624,
+      "grad_norm": 1.1561800241470337,
+      "learning_rate": 1.710484193677471e-05,
+      "loss": 0.5386,
+      "step": 16455
+    },
+    {
+      "epoch": 21.06368,
+      "grad_norm": 1.110007643699646,
+      "learning_rate": 1.7102841136454582e-05,
+      "loss": 0.4753,
+      "step": 16456
+    },
+    {
+      "epoch": 21.06496,
+      "grad_norm": 1.0784857273101807,
+      "learning_rate": 1.7100840336134454e-05,
+      "loss": 0.4374,
+      "step": 16457
+    },
+    {
+      "epoch": 21.06624,
+      "grad_norm": 1.0909050703048706,
+      "learning_rate": 1.7098839535814326e-05,
+      "loss": 0.4375,
+      "step": 16458
+    },
+    {
+      "epoch": 21.06752,
+      "grad_norm": 1.12933349609375,
+      "learning_rate": 1.7096838735494198e-05,
+      "loss": 0.4876,
+      "step": 16459
+    },
+    {
+      "epoch": 21.0688,
+      "grad_norm": 1.1255989074707031,
+      "learning_rate": 1.709483793517407e-05,
+      "loss": 0.4721,
+      "step": 16460
+    },
+    {
+      "epoch": 21.07008,
+      "grad_norm": 1.059699535369873,
+      "learning_rate": 1.709283713485394e-05,
+      "loss": 0.4565,
+      "step": 16461
+    },
+    {
+      "epoch": 21.07136,
+      "grad_norm": 1.0751200914382935,
+      "learning_rate": 1.7090836334533814e-05,
+      "loss": 0.464,
+      "step": 16462
+    },
+    {
+      "epoch": 21.07264,
+      "grad_norm": 1.1244473457336426,
+      "learning_rate": 1.708883553421369e-05,
+      "loss": 0.4842,
+      "step": 16463
+    },
+    {
+      "epoch": 21.07392,
+      "grad_norm": 1.1796867847442627,
+      "learning_rate": 1.7086834733893557e-05,
+      "loss": 0.5435,
+      "step": 16464
+    },
+    {
+      "epoch": 21.0752,
+      "grad_norm": 1.0969107151031494,
+      "learning_rate": 1.708483393357343e-05,
+      "loss": 0.479,
+      "step": 16465
+    },
+    {
+      "epoch": 21.07648,
+      "grad_norm": 1.1163842678070068,
+      "learning_rate": 1.70828331332533e-05,
+      "loss": 0.4498,
+      "step": 16466
+    },
+    {
+      "epoch": 21.07776,
+      "grad_norm": 1.1781340837478638,
+      "learning_rate": 1.7080832332933176e-05,
+      "loss": 0.5003,
+      "step": 16467
+    },
+    {
+      "epoch": 21.07904,
+      "grad_norm": 1.1196157932281494,
+      "learning_rate": 1.7078831532613045e-05,
+      "loss": 0.5118,
+      "step": 16468
+    },
+    {
+      "epoch": 21.08032,
+      "grad_norm": 1.1168984174728394,
+      "learning_rate": 1.7076830732292917e-05,
+      "loss": 0.4505,
+      "step": 16469
+    },
+    {
+      "epoch": 21.0816,
+      "grad_norm": 1.1065702438354492,
+      "learning_rate": 1.7074829931972792e-05,
+      "loss": 0.5232,
+      "step": 16470
+    },
+    {
+      "epoch": 21.08288,
+      "grad_norm": 1.0598688125610352,
+      "learning_rate": 1.7072829131652664e-05,
+      "loss": 0.4615,
+      "step": 16471
+    },
+    {
+      "epoch": 21.08416,
+      "grad_norm": 1.0742090940475464,
+      "learning_rate": 1.7070828331332532e-05,
+      "loss": 0.4675,
+      "step": 16472
+    },
+    {
+      "epoch": 21.08544,
+      "grad_norm": 1.0973572731018066,
+      "learning_rate": 1.7068827531012404e-05,
+      "loss": 0.4869,
+      "step": 16473
+    },
+    {
+      "epoch": 21.08672,
+      "grad_norm": 1.0886359214782715,
+      "learning_rate": 1.706682673069228e-05,
+      "loss": 0.4684,
+      "step": 16474
+    },
+    {
+      "epoch": 21.088,
+      "grad_norm": 1.0763150453567505,
+      "learning_rate": 1.706482593037215e-05,
+      "loss": 0.4558,
+      "step": 16475
+    },
+    {
+      "epoch": 21.08928,
+      "grad_norm": 1.0928971767425537,
+      "learning_rate": 1.706282513005202e-05,
+      "loss": 0.4765,
+      "step": 16476
+    },
+    {
+      "epoch": 21.09056,
+      "grad_norm": 1.1339545249938965,
+      "learning_rate": 1.7060824329731895e-05,
+      "loss": 0.5047,
+      "step": 16477
+    },
+    {
+      "epoch": 21.09184,
+      "grad_norm": 1.0725475549697876,
+      "learning_rate": 1.7058823529411767e-05,
+      "loss": 0.4662,
+      "step": 16478
+    },
+    {
+      "epoch": 21.09312,
+      "grad_norm": 1.1563411951065063,
+      "learning_rate": 1.705682272909164e-05,
+      "loss": 0.4984,
+      "step": 16479
+    },
+    {
+      "epoch": 21.0944,
+      "grad_norm": 1.0735126733779907,
+      "learning_rate": 1.7054821928771507e-05,
+      "loss": 0.4125,
+      "step": 16480
+    },
+    {
+      "epoch": 21.09568,
+      "grad_norm": 1.1152347326278687,
+      "learning_rate": 1.7052821128451382e-05,
+      "loss": 0.505,
+      "step": 16481
+    },
+    {
+      "epoch": 21.09696,
+      "grad_norm": 1.0870592594146729,
+      "learning_rate": 1.7050820328131254e-05,
+      "loss": 0.5002,
+      "step": 16482
+    },
+    {
+      "epoch": 21.09824,
+      "grad_norm": 1.116709589958191,
+      "learning_rate": 1.7048819527811126e-05,
+      "loss": 0.462,
+      "step": 16483
+    },
+    {
+      "epoch": 21.09952,
+      "grad_norm": 1.1316735744476318,
+      "learning_rate": 1.7046818727490998e-05,
+      "loss": 0.5289,
+      "step": 16484
+    },
+    {
+      "epoch": 21.1008,
+      "grad_norm": 1.127074122428894,
+      "learning_rate": 1.704481792717087e-05,
+      "loss": 0.4713,
+      "step": 16485
+    },
+    {
+      "epoch": 21.10208,
+      "grad_norm": 1.1193740367889404,
+      "learning_rate": 1.704281712685074e-05,
+      "loss": 0.4759,
+      "step": 16486
+    },
+    {
+      "epoch": 21.10336,
+      "grad_norm": 1.0757756233215332,
+      "learning_rate": 1.7040816326530613e-05,
+      "loss": 0.5199,
+      "step": 16487
+    },
+    {
+      "epoch": 21.10464,
+      "grad_norm": 1.0991090536117554,
+      "learning_rate": 1.7038815526210485e-05,
+      "loss": 0.523,
+      "step": 16488
+    },
+    {
+      "epoch": 21.10592,
+      "grad_norm": 1.0792702436447144,
+      "learning_rate": 1.7036814725890357e-05,
+      "loss": 0.4515,
+      "step": 16489
+    },
+    {
+      "epoch": 21.1072,
+      "grad_norm": 1.1250958442687988,
+      "learning_rate": 1.703481392557023e-05,
+      "loss": 0.4843,
+      "step": 16490
+    },
+    {
+      "epoch": 21.10848,
+      "grad_norm": 1.1384851932525635,
+      "learning_rate": 1.70328131252501e-05,
+      "loss": 0.523,
+      "step": 16491
+    },
+    {
+      "epoch": 21.10976,
+      "grad_norm": 1.0551738739013672,
+      "learning_rate": 1.7030812324929973e-05,
+      "loss": 0.454,
+      "step": 16492
+    },
+    {
+      "epoch": 21.11104,
+      "grad_norm": 1.0307825803756714,
+      "learning_rate": 1.7028811524609844e-05,
+      "loss": 0.4157,
+      "step": 16493
+    },
+    {
+      "epoch": 21.11232,
+      "grad_norm": 1.1250662803649902,
+      "learning_rate": 1.7026810724289716e-05,
+      "loss": 0.5015,
+      "step": 16494
+    },
+    {
+      "epoch": 21.1136,
+      "grad_norm": 1.0940858125686646,
+      "learning_rate": 1.7024809923969588e-05,
+      "loss": 0.4605,
+      "step": 16495
+    },
+    {
+      "epoch": 21.11488,
+      "grad_norm": 1.0869758129119873,
+      "learning_rate": 1.702280912364946e-05,
+      "loss": 0.544,
+      "step": 16496
+    },
+    {
+      "epoch": 21.11616,
+      "grad_norm": 1.0667223930358887,
+      "learning_rate": 1.7020808323329332e-05,
+      "loss": 0.4369,
+      "step": 16497
+    },
+    {
+      "epoch": 21.11744,
+      "grad_norm": 1.0950666666030884,
+      "learning_rate": 1.7018807523009204e-05,
+      "loss": 0.5017,
+      "step": 16498
+    },
+    {
+      "epoch": 21.11872,
+      "grad_norm": 1.0515073537826538,
+      "learning_rate": 1.7016806722689076e-05,
+      "loss": 0.4371,
+      "step": 16499
+    },
+    {
+      "epoch": 21.12,
+      "grad_norm": 1.133811354637146,
+      "learning_rate": 1.7014805922368947e-05,
+      "loss": 0.5154,
+      "step": 16500
+    },
+    {
+      "epoch": 21.12128,
+      "grad_norm": 1.1043317317962646,
+      "learning_rate": 1.701280512204882e-05,
+      "loss": 0.4841,
+      "step": 16501
+    },
+    {
+      "epoch": 21.12256,
+      "grad_norm": 1.1179126501083374,
+      "learning_rate": 1.7010804321728695e-05,
+      "loss": 0.5082,
+      "step": 16502
+    },
+    {
+      "epoch": 21.12384,
+      "grad_norm": 1.0235763788223267,
+      "learning_rate": 1.7008803521408563e-05,
+      "loss": 0.4659,
+      "step": 16503
+    },
+    {
+      "epoch": 21.12512,
+      "grad_norm": 1.1011422872543335,
+      "learning_rate": 1.7006802721088435e-05,
+      "loss": 0.4946,
+      "step": 16504
+    },
+    {
+      "epoch": 21.1264,
+      "grad_norm": 1.1294410228729248,
+      "learning_rate": 1.7004801920768307e-05,
+      "loss": 0.4528,
+      "step": 16505
+    },
+    {
+      "epoch": 21.12768,
+      "grad_norm": 1.088690996170044,
+      "learning_rate": 1.7002801120448182e-05,
+      "loss": 0.4638,
+      "step": 16506
+    },
+    {
+      "epoch": 21.12896,
+      "grad_norm": 1.1901614665985107,
+      "learning_rate": 1.700080032012805e-05,
+      "loss": 0.4745,
+      "step": 16507
+    },
+    {
+      "epoch": 21.13024,
+      "grad_norm": 1.177269458770752,
+      "learning_rate": 1.6998799519807922e-05,
+      "loss": 0.4941,
+      "step": 16508
+    },
+    {
+      "epoch": 21.13152,
+      "grad_norm": 1.117848515510559,
+      "learning_rate": 1.6996798719487798e-05,
+      "loss": 0.4575,
+      "step": 16509
+    },
+    {
+      "epoch": 21.1328,
+      "grad_norm": 1.1764754056930542,
+      "learning_rate": 1.699479791916767e-05,
+      "loss": 0.4967,
+      "step": 16510
+    },
+    {
+      "epoch": 21.13408,
+      "grad_norm": 1.094519853591919,
+      "learning_rate": 1.6992797118847538e-05,
+      "loss": 0.4677,
+      "step": 16511
+    },
+    {
+      "epoch": 21.13536,
+      "grad_norm": 1.1102967262268066,
+      "learning_rate": 1.699079631852741e-05,
+      "loss": 0.4894,
+      "step": 16512
+    },
+    {
+      "epoch": 21.13664,
+      "grad_norm": 1.0887529850006104,
+      "learning_rate": 1.6988795518207285e-05,
+      "loss": 0.4896,
+      "step": 16513
+    },
+    {
+      "epoch": 21.13792,
+      "grad_norm": 1.1254242658615112,
+      "learning_rate": 1.6986794717887157e-05,
+      "loss": 0.4719,
+      "step": 16514
+    },
+    {
+      "epoch": 21.1392,
+      "grad_norm": 1.1468827724456787,
+      "learning_rate": 1.6984793917567025e-05,
+      "loss": 0.4629,
+      "step": 16515
+    },
+    {
+      "epoch": 21.14048,
+      "grad_norm": 1.1172226667404175,
+      "learning_rate": 1.69827931172469e-05,
+      "loss": 0.4621,
+      "step": 16516
+    },
+    {
+      "epoch": 21.14176,
+      "grad_norm": 1.1248224973678589,
+      "learning_rate": 1.6980792316926772e-05,
+      "loss": 0.4801,
+      "step": 16517
+    },
+    {
+      "epoch": 21.14304,
+      "grad_norm": 1.124162197113037,
+      "learning_rate": 1.6978791516606644e-05,
+      "loss": 0.4939,
+      "step": 16518
+    },
+    {
+      "epoch": 21.14432,
+      "grad_norm": 1.0810459852218628,
+      "learning_rate": 1.6976790716286513e-05,
+      "loss": 0.48,
+      "step": 16519
+    },
+    {
+      "epoch": 21.1456,
+      "grad_norm": 1.0889114141464233,
+      "learning_rate": 1.6974789915966388e-05,
+      "loss": 0.4854,
+      "step": 16520
+    },
+    {
+      "epoch": 21.14688,
+      "grad_norm": 1.0843100547790527,
+      "learning_rate": 1.697278911564626e-05,
+      "loss": 0.4766,
+      "step": 16521
+    },
+    {
+      "epoch": 21.14816,
+      "grad_norm": 1.0566636323928833,
+      "learning_rate": 1.697078831532613e-05,
+      "loss": 0.4388,
+      "step": 16522
+    },
+    {
+      "epoch": 21.14944,
+      "grad_norm": 1.102766513824463,
+      "learning_rate": 1.6968787515006004e-05,
+      "loss": 0.4835,
+      "step": 16523
+    },
+    {
+      "epoch": 21.15072,
+      "grad_norm": 1.0825378894805908,
+      "learning_rate": 1.6966786714685875e-05,
+      "loss": 0.4682,
+      "step": 16524
+    },
+    {
+      "epoch": 21.152,
+      "grad_norm": 1.1190485954284668,
+      "learning_rate": 1.6964785914365747e-05,
+      "loss": 0.4853,
+      "step": 16525
+    },
+    {
+      "epoch": 21.15328,
+      "grad_norm": 1.0945488214492798,
+      "learning_rate": 1.696278511404562e-05,
+      "loss": 0.5039,
+      "step": 16526
+    },
+    {
+      "epoch": 21.15456,
+      "grad_norm": 1.0824321508407593,
+      "learning_rate": 1.696078431372549e-05,
+      "loss": 0.4791,
+      "step": 16527
+    },
+    {
+      "epoch": 21.15584,
+      "grad_norm": 1.0835765600204468,
+      "learning_rate": 1.6958783513405363e-05,
+      "loss": 0.4846,
+      "step": 16528
+    },
+    {
+      "epoch": 21.15712,
+      "grad_norm": 1.0822176933288574,
+      "learning_rate": 1.6956782713085235e-05,
+      "loss": 0.4914,
+      "step": 16529
+    },
+    {
+      "epoch": 21.1584,
+      "grad_norm": 1.0641273260116577,
+      "learning_rate": 1.6954781912765107e-05,
+      "loss": 0.4497,
+      "step": 16530
+    },
+    {
+      "epoch": 21.15968,
+      "grad_norm": 1.1108111143112183,
+      "learning_rate": 1.695278111244498e-05,
+      "loss": 0.4962,
+      "step": 16531
+    },
+    {
+      "epoch": 21.16096,
+      "grad_norm": 1.0912580490112305,
+      "learning_rate": 1.695078031212485e-05,
+      "loss": 0.4981,
+      "step": 16532
+    },
+    {
+      "epoch": 21.16224,
+      "grad_norm": 1.1157331466674805,
+      "learning_rate": 1.6948779511804722e-05,
+      "loss": 0.4682,
+      "step": 16533
+    },
+    {
+      "epoch": 21.16352,
+      "grad_norm": 1.1047836542129517,
+      "learning_rate": 1.6946778711484594e-05,
+      "loss": 0.479,
+      "step": 16534
+    },
+    {
+      "epoch": 21.1648,
+      "grad_norm": 1.1271506547927856,
+      "learning_rate": 1.6944777911164466e-05,
+      "loss": 0.5429,
+      "step": 16535
+    },
+    {
+      "epoch": 21.16608,
+      "grad_norm": 1.075492024421692,
+      "learning_rate": 1.6942777110844338e-05,
+      "loss": 0.4732,
+      "step": 16536
+    },
+    {
+      "epoch": 21.16736,
+      "grad_norm": 1.1787910461425781,
+      "learning_rate": 1.6940776310524213e-05,
+      "loss": 0.5362,
+      "step": 16537
+    },
+    {
+      "epoch": 21.16864,
+      "grad_norm": 1.0927817821502686,
+      "learning_rate": 1.693877551020408e-05,
+      "loss": 0.4621,
+      "step": 16538
+    },
+    {
+      "epoch": 21.16992,
+      "grad_norm": 1.1615148782730103,
+      "learning_rate": 1.6936774709883953e-05,
+      "loss": 0.5075,
+      "step": 16539
+    },
+    {
+      "epoch": 21.1712,
+      "grad_norm": 1.115074634552002,
+      "learning_rate": 1.6934773909563825e-05,
+      "loss": 0.4464,
+      "step": 16540
+    },
+    {
+      "epoch": 21.17248,
+      "grad_norm": 1.1260422468185425,
+      "learning_rate": 1.69327731092437e-05,
+      "loss": 0.4683,
+      "step": 16541
+    },
+    {
+      "epoch": 21.17376,
+      "grad_norm": 1.0940788984298706,
+      "learning_rate": 1.693077230892357e-05,
+      "loss": 0.5185,
+      "step": 16542
+    },
+    {
+      "epoch": 21.17504,
+      "grad_norm": 1.095339059829712,
+      "learning_rate": 1.692877150860344e-05,
+      "loss": 0.4741,
+      "step": 16543
+    },
+    {
+      "epoch": 21.17632,
+      "grad_norm": 1.0709000825881958,
+      "learning_rate": 1.6926770708283316e-05,
+      "loss": 0.4933,
+      "step": 16544
+    },
+    {
+      "epoch": 21.1776,
+      "grad_norm": 1.1118566989898682,
+      "learning_rate": 1.6924769907963188e-05,
+      "loss": 0.5025,
+      "step": 16545
+    },
+    {
+      "epoch": 21.17888,
+      "grad_norm": 1.1093112230300903,
+      "learning_rate": 1.6922769107643056e-05,
+      "loss": 0.5049,
+      "step": 16546
+    },
+    {
+      "epoch": 21.18016,
+      "grad_norm": 1.0397313833236694,
+      "learning_rate": 1.6920768307322928e-05,
+      "loss": 0.4659,
+      "step": 16547
+    },
+    {
+      "epoch": 21.18144,
+      "grad_norm": 1.1054112911224365,
+      "learning_rate": 1.6918767507002803e-05,
+      "loss": 0.4411,
+      "step": 16548
+    },
+    {
+      "epoch": 21.18272,
+      "grad_norm": 1.1328340768814087,
+      "learning_rate": 1.6916766706682675e-05,
+      "loss": 0.4802,
+      "step": 16549
+    },
+    {
+      "epoch": 21.184,
+      "grad_norm": 1.0890549421310425,
+      "learning_rate": 1.6914765906362544e-05,
+      "loss": 0.4922,
+      "step": 16550
+    },
+    {
+      "epoch": 21.18528,
+      "grad_norm": 1.1533173322677612,
+      "learning_rate": 1.691276510604242e-05,
+      "loss": 0.515,
+      "step": 16551
+    },
+    {
+      "epoch": 21.18656,
+      "grad_norm": 1.0926942825317383,
+      "learning_rate": 1.691076430572229e-05,
+      "loss": 0.4696,
+      "step": 16552
+    },
+    {
+      "epoch": 21.18784,
+      "grad_norm": 1.0669362545013428,
+      "learning_rate": 1.6908763505402163e-05,
+      "loss": 0.4982,
+      "step": 16553
+    },
+    {
+      "epoch": 21.18912,
+      "grad_norm": 1.0951491594314575,
+      "learning_rate": 1.690676270508203e-05,
+      "loss": 0.4715,
+      "step": 16554
+    },
+    {
+      "epoch": 21.1904,
+      "grad_norm": 1.072893500328064,
+      "learning_rate": 1.6904761904761906e-05,
+      "loss": 0.4721,
+      "step": 16555
+    },
+    {
+      "epoch": 21.19168,
+      "grad_norm": 1.0877991914749146,
+      "learning_rate": 1.6902761104441778e-05,
+      "loss": 0.4584,
+      "step": 16556
+    },
+    {
+      "epoch": 21.19296,
+      "grad_norm": 1.12388014793396,
+      "learning_rate": 1.690076030412165e-05,
+      "loss": 0.4523,
+      "step": 16557
+    },
+    {
+      "epoch": 21.19424,
+      "grad_norm": 1.1295223236083984,
+      "learning_rate": 1.689875950380152e-05,
+      "loss": 0.4936,
+      "step": 16558
+    },
+    {
+      "epoch": 21.19552,
+      "grad_norm": 1.1079462766647339,
+      "learning_rate": 1.6896758703481394e-05,
+      "loss": 0.4463,
+      "step": 16559
+    },
+    {
+      "epoch": 21.1968,
+      "grad_norm": 1.1547305583953857,
+      "learning_rate": 1.6894757903161266e-05,
+      "loss": 0.4691,
+      "step": 16560
+    },
+    {
+      "epoch": 21.19808,
+      "grad_norm": 1.1566081047058105,
+      "learning_rate": 1.6892757102841137e-05,
+      "loss": 0.5163,
+      "step": 16561
+    },
+    {
+      "epoch": 21.19936,
+      "grad_norm": 1.1385910511016846,
+      "learning_rate": 1.689075630252101e-05,
+      "loss": 0.4941,
+      "step": 16562
+    },
+    {
+      "epoch": 21.20064,
+      "grad_norm": 1.1214505434036255,
+      "learning_rate": 1.688875550220088e-05,
+      "loss": 0.5138,
+      "step": 16563
+    },
+    {
+      "epoch": 21.20192,
+      "grad_norm": 1.089640498161316,
+      "learning_rate": 1.6886754701880753e-05,
+      "loss": 0.4491,
+      "step": 16564
+    },
+    {
+      "epoch": 21.2032,
+      "grad_norm": 1.0873479843139648,
+      "learning_rate": 1.6884753901560625e-05,
+      "loss": 0.4711,
+      "step": 16565
+    },
+    {
+      "epoch": 21.20448,
+      "grad_norm": 1.0695767402648926,
+      "learning_rate": 1.6882753101240497e-05,
+      "loss": 0.4575,
+      "step": 16566
+    },
+    {
+      "epoch": 21.20576,
+      "grad_norm": 1.1109639406204224,
+      "learning_rate": 1.688075230092037e-05,
+      "loss": 0.5093,
+      "step": 16567
+    },
+    {
+      "epoch": 21.20704,
+      "grad_norm": 1.10574471950531,
+      "learning_rate": 1.687875150060024e-05,
+      "loss": 0.5033,
+      "step": 16568
+    },
+    {
+      "epoch": 21.20832,
+      "grad_norm": 1.0415626764297485,
+      "learning_rate": 1.6876750700280112e-05,
+      "loss": 0.4781,
+      "step": 16569
+    },
+    {
+      "epoch": 21.209600000000002,
+      "grad_norm": 1.1042823791503906,
+      "learning_rate": 1.6874749899959984e-05,
+      "loss": 0.4813,
+      "step": 16570
+    },
+    {
+      "epoch": 21.21088,
+      "grad_norm": 1.1033847332000732,
+      "learning_rate": 1.6872749099639856e-05,
+      "loss": 0.4547,
+      "step": 16571
+    },
+    {
+      "epoch": 21.21216,
+      "grad_norm": 1.0810890197753906,
+      "learning_rate": 1.6870748299319728e-05,
+      "loss": 0.5008,
+      "step": 16572
+    },
+    {
+      "epoch": 21.21344,
+      "grad_norm": 1.111965537071228,
+      "learning_rate": 1.68687474989996e-05,
+      "loss": 0.4542,
+      "step": 16573
+    },
+    {
+      "epoch": 21.21472,
+      "grad_norm": 1.168083906173706,
+      "learning_rate": 1.686674669867947e-05,
+      "loss": 0.4931,
+      "step": 16574
+    },
+    {
+      "epoch": 21.216,
+      "grad_norm": 1.1118204593658447,
+      "learning_rate": 1.6864745898359343e-05,
+      "loss": 0.4682,
+      "step": 16575
+    },
+    {
+      "epoch": 21.21728,
+      "grad_norm": 1.1220096349716187,
+      "learning_rate": 1.686274509803922e-05,
+      "loss": 0.5033,
+      "step": 16576
+    },
+    {
+      "epoch": 21.21856,
+      "grad_norm": 1.061000943183899,
+      "learning_rate": 1.6860744297719087e-05,
+      "loss": 0.5032,
+      "step": 16577
+    },
+    {
+      "epoch": 21.21984,
+      "grad_norm": 1.1358115673065186,
+      "learning_rate": 1.685874349739896e-05,
+      "loss": 0.4291,
+      "step": 16578
+    },
+    {
+      "epoch": 21.22112,
+      "grad_norm": 1.0819488763809204,
+      "learning_rate": 1.685674269707883e-05,
+      "loss": 0.4413,
+      "step": 16579
+    },
+    {
+      "epoch": 21.2224,
+      "grad_norm": 1.0310840606689453,
+      "learning_rate": 1.6854741896758706e-05,
+      "loss": 0.436,
+      "step": 16580
+    },
+    {
+      "epoch": 21.22368,
+      "grad_norm": 1.0708487033843994,
+      "learning_rate": 1.6852741096438575e-05,
+      "loss": 0.4718,
+      "step": 16581
+    },
+    {
+      "epoch": 21.22496,
+      "grad_norm": 1.1317819356918335,
+      "learning_rate": 1.6850740296118446e-05,
+      "loss": 0.528,
+      "step": 16582
+    },
+    {
+      "epoch": 21.22624,
+      "grad_norm": 1.1294492483139038,
+      "learning_rate": 1.6848739495798322e-05,
+      "loss": 0.4881,
+      "step": 16583
+    },
+    {
+      "epoch": 21.22752,
+      "grad_norm": 1.1252028942108154,
+      "learning_rate": 1.6846738695478194e-05,
+      "loss": 0.4865,
+      "step": 16584
+    },
+    {
+      "epoch": 21.2288,
+      "grad_norm": 1.187008261680603,
+      "learning_rate": 1.6844737895158062e-05,
+      "loss": 0.5291,
+      "step": 16585
+    },
+    {
+      "epoch": 21.23008,
+      "grad_norm": 1.1691592931747437,
+      "learning_rate": 1.6842737094837934e-05,
+      "loss": 0.5374,
+      "step": 16586
+    },
+    {
+      "epoch": 21.23136,
+      "grad_norm": 1.116336464881897,
+      "learning_rate": 1.684073629451781e-05,
+      "loss": 0.5126,
+      "step": 16587
+    },
+    {
+      "epoch": 21.23264,
+      "grad_norm": 1.038357138633728,
+      "learning_rate": 1.683873549419768e-05,
+      "loss": 0.4979,
+      "step": 16588
+    },
+    {
+      "epoch": 21.23392,
+      "grad_norm": 1.0832995176315308,
+      "learning_rate": 1.683673469387755e-05,
+      "loss": 0.4649,
+      "step": 16589
+    },
+    {
+      "epoch": 21.2352,
+      "grad_norm": 1.0173653364181519,
+      "learning_rate": 1.6834733893557425e-05,
+      "loss": 0.4228,
+      "step": 16590
+    },
+    {
+      "epoch": 21.23648,
+      "grad_norm": 1.0729964971542358,
+      "learning_rate": 1.6832733093237297e-05,
+      "loss": 0.4692,
+      "step": 16591
+    },
+    {
+      "epoch": 21.23776,
+      "grad_norm": 1.036226749420166,
+      "learning_rate": 1.683073229291717e-05,
+      "loss": 0.4212,
+      "step": 16592
+    },
+    {
+      "epoch": 21.23904,
+      "grad_norm": 1.1111071109771729,
+      "learning_rate": 1.6828731492597037e-05,
+      "loss": 0.5403,
+      "step": 16593
+    },
+    {
+      "epoch": 21.24032,
+      "grad_norm": 1.0485975742340088,
+      "learning_rate": 1.6826730692276912e-05,
+      "loss": 0.4712,
+      "step": 16594
+    },
+    {
+      "epoch": 21.2416,
+      "grad_norm": 0.9971339106559753,
+      "learning_rate": 1.6824729891956784e-05,
+      "loss": 0.3999,
+      "step": 16595
+    },
+    {
+      "epoch": 21.24288,
+      "grad_norm": 1.1354920864105225,
+      "learning_rate": 1.6822729091636656e-05,
+      "loss": 0.533,
+      "step": 16596
+    },
+    {
+      "epoch": 21.24416,
+      "grad_norm": 1.1078267097473145,
+      "learning_rate": 1.6820728291316528e-05,
+      "loss": 0.4979,
+      "step": 16597
+    },
+    {
+      "epoch": 21.24544,
+      "grad_norm": 1.1035330295562744,
+      "learning_rate": 1.68187274909964e-05,
+      "loss": 0.4989,
+      "step": 16598
+    },
+    {
+      "epoch": 21.24672,
+      "grad_norm": 1.091853380203247,
+      "learning_rate": 1.681672669067627e-05,
+      "loss": 0.4654,
+      "step": 16599
+    },
+    {
+      "epoch": 21.248,
+      "grad_norm": 1.0949825048446655,
+      "learning_rate": 1.6814725890356143e-05,
+      "loss": 0.4446,
+      "step": 16600
+    },
+    {
+      "epoch": 21.24928,
+      "grad_norm": 1.165686011314392,
+      "learning_rate": 1.6812725090036015e-05,
+      "loss": 0.5106,
+      "step": 16601
+    },
+    {
+      "epoch": 21.25056,
+      "grad_norm": 1.0916712284088135,
+      "learning_rate": 1.6810724289715887e-05,
+      "loss": 0.4743,
+      "step": 16602
+    },
+    {
+      "epoch": 21.25184,
+      "grad_norm": 1.0554473400115967,
+      "learning_rate": 1.680872348939576e-05,
+      "loss": 0.459,
+      "step": 16603
+    },
+    {
+      "epoch": 21.25312,
+      "grad_norm": 1.1290833950042725,
+      "learning_rate": 1.6806722689075634e-05,
+      "loss": 0.5159,
+      "step": 16604
+    },
+    {
+      "epoch": 21.2544,
+      "grad_norm": 1.066178560256958,
+      "learning_rate": 1.6804721888755503e-05,
+      "loss": 0.4695,
+      "step": 16605
+    },
+    {
+      "epoch": 21.25568,
+      "grad_norm": 1.0555775165557861,
+      "learning_rate": 1.6802721088435374e-05,
+      "loss": 0.4638,
+      "step": 16606
+    },
+    {
+      "epoch": 21.25696,
+      "grad_norm": 1.1663188934326172,
+      "learning_rate": 1.6800720288115246e-05,
+      "loss": 0.5057,
+      "step": 16607
+    },
+    {
+      "epoch": 21.25824,
+      "grad_norm": 1.0777236223220825,
+      "learning_rate": 1.679871948779512e-05,
+      "loss": 0.457,
+      "step": 16608
+    },
+    {
+      "epoch": 21.25952,
+      "grad_norm": 1.185142993927002,
+      "learning_rate": 1.679671868747499e-05,
+      "loss": 0.4742,
+      "step": 16609
+    },
+    {
+      "epoch": 21.2608,
+      "grad_norm": 1.0730408430099487,
+      "learning_rate": 1.6794717887154862e-05,
+      "loss": 0.5001,
+      "step": 16610
+    },
+    {
+      "epoch": 21.26208,
+      "grad_norm": 1.0346087217330933,
+      "learning_rate": 1.6792717086834734e-05,
+      "loss": 0.4242,
+      "step": 16611
+    },
+    {
+      "epoch": 21.26336,
+      "grad_norm": 1.0942175388336182,
+      "learning_rate": 1.6790716286514606e-05,
+      "loss": 0.4677,
+      "step": 16612
+    },
+    {
+      "epoch": 21.26464,
+      "grad_norm": 1.15269935131073,
+      "learning_rate": 1.6788715486194477e-05,
+      "loss": 0.5306,
+      "step": 16613
+    },
+    {
+      "epoch": 21.26592,
+      "grad_norm": 1.0454847812652588,
+      "learning_rate": 1.678671468587435e-05,
+      "loss": 0.4561,
+      "step": 16614
+    },
+    {
+      "epoch": 21.2672,
+      "grad_norm": 1.16476309299469,
+      "learning_rate": 1.6784713885554225e-05,
+      "loss": 0.53,
+      "step": 16615
+    },
+    {
+      "epoch": 21.26848,
+      "grad_norm": 1.286267638206482,
+      "learning_rate": 1.6782713085234093e-05,
+      "loss": 0.5394,
+      "step": 16616
+    },
+    {
+      "epoch": 21.26976,
+      "grad_norm": 1.114345908164978,
+      "learning_rate": 1.6780712284913965e-05,
+      "loss": 0.4361,
+      "step": 16617
+    },
+    {
+      "epoch": 21.27104,
+      "grad_norm": 1.149522304534912,
+      "learning_rate": 1.6778711484593837e-05,
+      "loss": 0.4684,
+      "step": 16618
+    },
+    {
+      "epoch": 21.27232,
+      "grad_norm": 1.143465280532837,
+      "learning_rate": 1.6776710684273712e-05,
+      "loss": 0.5119,
+      "step": 16619
+    },
+    {
+      "epoch": 21.2736,
+      "grad_norm": 1.0972861051559448,
+      "learning_rate": 1.677470988395358e-05,
+      "loss": 0.482,
+      "step": 16620
+    },
+    {
+      "epoch": 21.27488,
+      "grad_norm": 1.0499613285064697,
+      "learning_rate": 1.6772709083633452e-05,
+      "loss": 0.465,
+      "step": 16621
+    },
+    {
+      "epoch": 21.27616,
+      "grad_norm": 1.0492812395095825,
+      "learning_rate": 1.6770708283313328e-05,
+      "loss": 0.4625,
+      "step": 16622
+    },
+    {
+      "epoch": 21.27744,
+      "grad_norm": 1.0873419046401978,
+      "learning_rate": 1.67687074829932e-05,
+      "loss": 0.4993,
+      "step": 16623
+    },
+    {
+      "epoch": 21.27872,
+      "grad_norm": 1.1490446329116821,
+      "learning_rate": 1.6766706682673068e-05,
+      "loss": 0.4968,
+      "step": 16624
+    },
+    {
+      "epoch": 21.28,
+      "grad_norm": 1.1244101524353027,
+      "learning_rate": 1.676470588235294e-05,
+      "loss": 0.5022,
+      "step": 16625
+    },
+    {
+      "epoch": 21.28128,
+      "grad_norm": 1.0936533212661743,
+      "learning_rate": 1.6762705082032815e-05,
+      "loss": 0.4537,
+      "step": 16626
+    },
+    {
+      "epoch": 21.28256,
+      "grad_norm": 1.1399465799331665,
+      "learning_rate": 1.6760704281712687e-05,
+      "loss": 0.4794,
+      "step": 16627
+    },
+    {
+      "epoch": 21.28384,
+      "grad_norm": 1.14096999168396,
+      "learning_rate": 1.6758703481392555e-05,
+      "loss": 0.4995,
+      "step": 16628
+    },
+    {
+      "epoch": 21.28512,
+      "grad_norm": 1.1549474000930786,
+      "learning_rate": 1.675670268107243e-05,
+      "loss": 0.4909,
+      "step": 16629
+    },
+    {
+      "epoch": 21.2864,
+      "grad_norm": 1.088848352432251,
+      "learning_rate": 1.6754701880752302e-05,
+      "loss": 0.4626,
+      "step": 16630
+    },
+    {
+      "epoch": 21.28768,
+      "grad_norm": 1.114532709121704,
+      "learning_rate": 1.6752701080432174e-05,
+      "loss": 0.4492,
+      "step": 16631
+    },
+    {
+      "epoch": 21.28896,
+      "grad_norm": 1.1168134212493896,
+      "learning_rate": 1.6750700280112043e-05,
+      "loss": 0.505,
+      "step": 16632
+    },
+    {
+      "epoch": 21.29024,
+      "grad_norm": 1.1612701416015625,
+      "learning_rate": 1.6748699479791918e-05,
+      "loss": 0.519,
+      "step": 16633
+    },
+    {
+      "epoch": 21.29152,
+      "grad_norm": 1.0742192268371582,
+      "learning_rate": 1.674669867947179e-05,
+      "loss": 0.4603,
+      "step": 16634
+    },
+    {
+      "epoch": 21.2928,
+      "grad_norm": 1.1029329299926758,
+      "learning_rate": 1.674469787915166e-05,
+      "loss": 0.4851,
+      "step": 16635
+    },
+    {
+      "epoch": 21.29408,
+      "grad_norm": 1.106626033782959,
+      "learning_rate": 1.6742697078831534e-05,
+      "loss": 0.5193,
+      "step": 16636
+    },
+    {
+      "epoch": 21.29536,
+      "grad_norm": 1.117716908454895,
+      "learning_rate": 1.6740696278511405e-05,
+      "loss": 0.5412,
+      "step": 16637
+    },
+    {
+      "epoch": 21.29664,
+      "grad_norm": 1.0421884059906006,
+      "learning_rate": 1.6738695478191277e-05,
+      "loss": 0.446,
+      "step": 16638
+    },
+    {
+      "epoch": 21.29792,
+      "grad_norm": 1.0915087461471558,
+      "learning_rate": 1.673669467787115e-05,
+      "loss": 0.5246,
+      "step": 16639
+    },
+    {
+      "epoch": 21.2992,
+      "grad_norm": 1.1888045072555542,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 0.4996,
+      "step": 16640
+    },
+    {
+      "epoch": 21.30048,
+      "grad_norm": 1.0700676441192627,
+      "learning_rate": 1.6732693077230893e-05,
+      "loss": 0.4903,
+      "step": 16641
+    },
+    {
+      "epoch": 21.30176,
+      "grad_norm": 1.1567354202270508,
+      "learning_rate": 1.6730692276910765e-05,
+      "loss": 0.525,
+      "step": 16642
+    },
+    {
+      "epoch": 21.30304,
+      "grad_norm": 1.1310677528381348,
+      "learning_rate": 1.672869147659064e-05,
+      "loss": 0.5157,
+      "step": 16643
+    },
+    {
+      "epoch": 21.30432,
+      "grad_norm": 1.101754903793335,
+      "learning_rate": 1.672669067627051e-05,
+      "loss": 0.4957,
+      "step": 16644
+    },
+    {
+      "epoch": 21.3056,
+      "grad_norm": 1.1254335641860962,
+      "learning_rate": 1.672468987595038e-05,
+      "loss": 0.4955,
+      "step": 16645
+    },
+    {
+      "epoch": 21.30688,
+      "grad_norm": 1.1134346723556519,
+      "learning_rate": 1.6722689075630252e-05,
+      "loss": 0.4501,
+      "step": 16646
+    },
+    {
+      "epoch": 21.30816,
+      "grad_norm": 1.1738286018371582,
+      "learning_rate": 1.6720688275310127e-05,
+      "loss": 0.4818,
+      "step": 16647
+    },
+    {
+      "epoch": 21.30944,
+      "grad_norm": 1.0866729021072388,
+      "learning_rate": 1.6718687474989996e-05,
+      "loss": 0.4542,
+      "step": 16648
+    },
+    {
+      "epoch": 21.31072,
+      "grad_norm": 1.1546677350997925,
+      "learning_rate": 1.6716686674669868e-05,
+      "loss": 0.4684,
+      "step": 16649
+    },
+    {
+      "epoch": 21.312,
+      "grad_norm": 1.077710747718811,
+      "learning_rate": 1.6714685874349743e-05,
+      "loss": 0.4649,
+      "step": 16650
+    },
+    {
+      "epoch": 21.31328,
+      "grad_norm": 1.1036674976348877,
+      "learning_rate": 1.6712685074029615e-05,
+      "loss": 0.498,
+      "step": 16651
+    },
+    {
+      "epoch": 21.31456,
+      "grad_norm": 1.1394011974334717,
+      "learning_rate": 1.6710684273709483e-05,
+      "loss": 0.4669,
+      "step": 16652
+    },
+    {
+      "epoch": 21.31584,
+      "grad_norm": 1.1064188480377197,
+      "learning_rate": 1.6708683473389355e-05,
+      "loss": 0.4772,
+      "step": 16653
+    },
+    {
+      "epoch": 21.31712,
+      "grad_norm": 1.1202365159988403,
+      "learning_rate": 1.670668267306923e-05,
+      "loss": 0.4958,
+      "step": 16654
+    },
+    {
+      "epoch": 21.3184,
+      "grad_norm": 1.156451940536499,
+      "learning_rate": 1.6704681872749102e-05,
+      "loss": 0.5121,
+      "step": 16655
+    },
+    {
+      "epoch": 21.31968,
+      "grad_norm": 1.1051130294799805,
+      "learning_rate": 1.670268107242897e-05,
+      "loss": 0.5077,
+      "step": 16656
+    },
+    {
+      "epoch": 21.32096,
+      "grad_norm": 1.0539714097976685,
+      "learning_rate": 1.6700680272108846e-05,
+      "loss": 0.4521,
+      "step": 16657
+    },
+    {
+      "epoch": 21.32224,
+      "grad_norm": 1.1256710290908813,
+      "learning_rate": 1.6698679471788718e-05,
+      "loss": 0.4814,
+      "step": 16658
+    },
+    {
+      "epoch": 21.32352,
+      "grad_norm": 1.1350125074386597,
+      "learning_rate": 1.669667867146859e-05,
+      "loss": 0.509,
+      "step": 16659
+    },
+    {
+      "epoch": 21.3248,
+      "grad_norm": 0.9891694784164429,
+      "learning_rate": 1.6694677871148458e-05,
+      "loss": 0.4096,
+      "step": 16660
+    },
+    {
+      "epoch": 21.32608,
+      "grad_norm": 1.1129761934280396,
+      "learning_rate": 1.6692677070828333e-05,
+      "loss": 0.4794,
+      "step": 16661
+    },
+    {
+      "epoch": 21.32736,
+      "grad_norm": 1.0066285133361816,
+      "learning_rate": 1.6690676270508205e-05,
+      "loss": 0.4442,
+      "step": 16662
+    },
+    {
+      "epoch": 21.32864,
+      "grad_norm": 1.067672848701477,
+      "learning_rate": 1.6688675470188077e-05,
+      "loss": 0.4584,
+      "step": 16663
+    },
+    {
+      "epoch": 21.32992,
+      "grad_norm": 1.1201221942901611,
+      "learning_rate": 1.668667466986795e-05,
+      "loss": 0.4726,
+      "step": 16664
+    },
+    {
+      "epoch": 21.3312,
+      "grad_norm": 1.104931116104126,
+      "learning_rate": 1.668467386954782e-05,
+      "loss": 0.4799,
+      "step": 16665
+    },
+    {
+      "epoch": 21.33248,
+      "grad_norm": 1.0619887113571167,
+      "learning_rate": 1.6682673069227693e-05,
+      "loss": 0.4639,
+      "step": 16666
+    },
+    {
+      "epoch": 21.33376,
+      "grad_norm": 1.1233429908752441,
+      "learning_rate": 1.6680672268907564e-05,
+      "loss": 0.4904,
+      "step": 16667
+    },
+    {
+      "epoch": 21.33504,
+      "grad_norm": 1.137324333190918,
+      "learning_rate": 1.6678671468587436e-05,
+      "loss": 0.4927,
+      "step": 16668
+    },
+    {
+      "epoch": 21.33632,
+      "grad_norm": 1.0907856225967407,
+      "learning_rate": 1.6676670668267308e-05,
+      "loss": 0.4698,
+      "step": 16669
+    },
+    {
+      "epoch": 21.3376,
+      "grad_norm": 1.1330994367599487,
+      "learning_rate": 1.667466986794718e-05,
+      "loss": 0.5419,
+      "step": 16670
+    },
+    {
+      "epoch": 21.33888,
+      "grad_norm": 1.1534572839736938,
+      "learning_rate": 1.6672669067627052e-05,
+      "loss": 0.4913,
+      "step": 16671
+    },
+    {
+      "epoch": 21.34016,
+      "grad_norm": 1.070062279701233,
+      "learning_rate": 1.6670668267306924e-05,
+      "loss": 0.4565,
+      "step": 16672
+    },
+    {
+      "epoch": 21.34144,
+      "grad_norm": 1.0346976518630981,
+      "learning_rate": 1.6668667466986796e-05,
+      "loss": 0.4802,
+      "step": 16673
+    },
+    {
+      "epoch": 21.34272,
+      "grad_norm": 1.0610109567642212,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.4705,
+      "step": 16674
+    },
+    {
+      "epoch": 21.344,
+      "grad_norm": 1.103748083114624,
+      "learning_rate": 1.666466586634654e-05,
+      "loss": 0.4516,
+      "step": 16675
+    },
+    {
+      "epoch": 21.34528,
+      "grad_norm": 1.1500375270843506,
+      "learning_rate": 1.666266506602641e-05,
+      "loss": 0.5045,
+      "step": 16676
+    },
+    {
+      "epoch": 21.34656,
+      "grad_norm": 1.0690054893493652,
+      "learning_rate": 1.6660664265706283e-05,
+      "loss": 0.4846,
+      "step": 16677
+    },
+    {
+      "epoch": 21.34784,
+      "grad_norm": 1.1877983808517456,
+      "learning_rate": 1.6658663465386155e-05,
+      "loss": 0.5534,
+      "step": 16678
+    },
+    {
+      "epoch": 21.34912,
+      "grad_norm": 1.092886209487915,
+      "learning_rate": 1.6656662665066027e-05,
+      "loss": 0.4751,
+      "step": 16679
+    },
+    {
+      "epoch": 21.3504,
+      "grad_norm": 1.1173007488250732,
+      "learning_rate": 1.66546618647459e-05,
+      "loss": 0.5033,
+      "step": 16680
+    },
+    {
+      "epoch": 21.35168,
+      "grad_norm": 1.1595876216888428,
+      "learning_rate": 1.665266106442577e-05,
+      "loss": 0.5458,
+      "step": 16681
+    },
+    {
+      "epoch": 21.35296,
+      "grad_norm": 1.059984803199768,
+      "learning_rate": 1.6650660264105646e-05,
+      "loss": 0.4982,
+      "step": 16682
+    },
+    {
+      "epoch": 21.35424,
+      "grad_norm": 1.0998493432998657,
+      "learning_rate": 1.6648659463785514e-05,
+      "loss": 0.5241,
+      "step": 16683
+    },
+    {
+      "epoch": 21.35552,
+      "grad_norm": 1.1519036293029785,
+      "learning_rate": 1.6646658663465386e-05,
+      "loss": 0.5199,
+      "step": 16684
+    },
+    {
+      "epoch": 21.3568,
+      "grad_norm": 1.141817331314087,
+      "learning_rate": 1.6644657863145258e-05,
+      "loss": 0.5133,
+      "step": 16685
+    },
+    {
+      "epoch": 21.35808,
+      "grad_norm": 1.0783271789550781,
+      "learning_rate": 1.6642657062825133e-05,
+      "loss": 0.4728,
+      "step": 16686
+    },
+    {
+      "epoch": 21.35936,
+      "grad_norm": 1.1302711963653564,
+      "learning_rate": 1.6640656262505e-05,
+      "loss": 0.5232,
+      "step": 16687
+    },
+    {
+      "epoch": 21.36064,
+      "grad_norm": 1.1916934251785278,
+      "learning_rate": 1.6638655462184873e-05,
+      "loss": 0.5078,
+      "step": 16688
+    },
+    {
+      "epoch": 21.36192,
+      "grad_norm": 1.1410053968429565,
+      "learning_rate": 1.663665466186475e-05,
+      "loss": 0.4956,
+      "step": 16689
+    },
+    {
+      "epoch": 21.3632,
+      "grad_norm": 1.1169198751449585,
+      "learning_rate": 1.663465386154462e-05,
+      "loss": 0.4981,
+      "step": 16690
+    },
+    {
+      "epoch": 21.36448,
+      "grad_norm": 1.0799909830093384,
+      "learning_rate": 1.663265306122449e-05,
+      "loss": 0.4978,
+      "step": 16691
+    },
+    {
+      "epoch": 21.36576,
+      "grad_norm": 1.1033358573913574,
+      "learning_rate": 1.663065226090436e-05,
+      "loss": 0.4631,
+      "step": 16692
+    },
+    {
+      "epoch": 21.36704,
+      "grad_norm": 1.146831750869751,
+      "learning_rate": 1.6628651460584236e-05,
+      "loss": 0.4868,
+      "step": 16693
+    },
+    {
+      "epoch": 21.36832,
+      "grad_norm": 1.1381210088729858,
+      "learning_rate": 1.6626650660264108e-05,
+      "loss": 0.4916,
+      "step": 16694
+    },
+    {
+      "epoch": 21.3696,
+      "grad_norm": 1.1156655550003052,
+      "learning_rate": 1.6624649859943976e-05,
+      "loss": 0.4718,
+      "step": 16695
+    },
+    {
+      "epoch": 21.37088,
+      "grad_norm": 1.1188489198684692,
+      "learning_rate": 1.6622649059623852e-05,
+      "loss": 0.481,
+      "step": 16696
+    },
+    {
+      "epoch": 21.37216,
+      "grad_norm": 1.0974246263504028,
+      "learning_rate": 1.6620648259303724e-05,
+      "loss": 0.4989,
+      "step": 16697
+    },
+    {
+      "epoch": 21.37344,
+      "grad_norm": 1.2472598552703857,
+      "learning_rate": 1.6618647458983595e-05,
+      "loss": 0.621,
+      "step": 16698
+    },
+    {
+      "epoch": 21.37472,
+      "grad_norm": 1.074589490890503,
+      "learning_rate": 1.6616646658663464e-05,
+      "loss": 0.4802,
+      "step": 16699
+    },
+    {
+      "epoch": 21.376,
+      "grad_norm": 1.0904700756072998,
+      "learning_rate": 1.661464585834334e-05,
+      "loss": 0.4855,
+      "step": 16700
+    },
+    {
+      "epoch": 21.37728,
+      "grad_norm": 1.093536376953125,
+      "learning_rate": 1.661264505802321e-05,
+      "loss": 0.4616,
+      "step": 16701
+    },
+    {
+      "epoch": 21.37856,
+      "grad_norm": 1.0414737462997437,
+      "learning_rate": 1.6610644257703083e-05,
+      "loss": 0.4223,
+      "step": 16702
+    },
+    {
+      "epoch": 21.37984,
+      "grad_norm": 1.1210976839065552,
+      "learning_rate": 1.6608643457382955e-05,
+      "loss": 0.5263,
+      "step": 16703
+    },
+    {
+      "epoch": 21.38112,
+      "grad_norm": 1.0790395736694336,
+      "learning_rate": 1.6606642657062827e-05,
+      "loss": 0.4368,
+      "step": 16704
+    },
+    {
+      "epoch": 21.3824,
+      "grad_norm": 1.146754503250122,
+      "learning_rate": 1.66046418567427e-05,
+      "loss": 0.501,
+      "step": 16705
+    },
+    {
+      "epoch": 21.38368,
+      "grad_norm": 1.1394240856170654,
+      "learning_rate": 1.660264105642257e-05,
+      "loss": 0.53,
+      "step": 16706
+    },
+    {
+      "epoch": 21.38496,
+      "grad_norm": 1.1505063772201538,
+      "learning_rate": 1.6600640256102442e-05,
+      "loss": 0.4901,
+      "step": 16707
+    },
+    {
+      "epoch": 21.38624,
+      "grad_norm": 1.1096645593643188,
+      "learning_rate": 1.6598639455782314e-05,
+      "loss": 0.4851,
+      "step": 16708
+    },
+    {
+      "epoch": 21.38752,
+      "grad_norm": 1.1275689601898193,
+      "learning_rate": 1.6596638655462186e-05,
+      "loss": 0.5176,
+      "step": 16709
+    },
+    {
+      "epoch": 21.3888,
+      "grad_norm": 1.1523791551589966,
+      "learning_rate": 1.6594637855142058e-05,
+      "loss": 0.5758,
+      "step": 16710
+    },
+    {
+      "epoch": 21.39008,
+      "grad_norm": 1.1662170886993408,
+      "learning_rate": 1.659263705482193e-05,
+      "loss": 0.5227,
+      "step": 16711
+    },
+    {
+      "epoch": 21.39136,
+      "grad_norm": 1.178915023803711,
+      "learning_rate": 1.65906362545018e-05,
+      "loss": 0.4966,
+      "step": 16712
+    },
+    {
+      "epoch": 21.39264,
+      "grad_norm": 1.153525471687317,
+      "learning_rate": 1.6588635454181673e-05,
+      "loss": 0.509,
+      "step": 16713
+    },
+    {
+      "epoch": 21.39392,
+      "grad_norm": 1.0840404033660889,
+      "learning_rate": 1.6586634653861545e-05,
+      "loss": 0.45,
+      "step": 16714
+    },
+    {
+      "epoch": 21.3952,
+      "grad_norm": 1.0108500719070435,
+      "learning_rate": 1.6584633853541417e-05,
+      "loss": 0.4983,
+      "step": 16715
+    },
+    {
+      "epoch": 21.39648,
+      "grad_norm": 1.0287790298461914,
+      "learning_rate": 1.658263305322129e-05,
+      "loss": 0.447,
+      "step": 16716
+    },
+    {
+      "epoch": 21.39776,
+      "grad_norm": 1.1262785196304321,
+      "learning_rate": 1.6580632252901164e-05,
+      "loss": 0.5199,
+      "step": 16717
+    },
+    {
+      "epoch": 21.39904,
+      "grad_norm": 1.1663216352462769,
+      "learning_rate": 1.6578631452581033e-05,
+      "loss": 0.5389,
+      "step": 16718
+    },
+    {
+      "epoch": 21.40032,
+      "grad_norm": 1.050490379333496,
+      "learning_rate": 1.6576630652260904e-05,
+      "loss": 0.4721,
+      "step": 16719
+    },
+    {
+      "epoch": 21.4016,
+      "grad_norm": 1.0672640800476074,
+      "learning_rate": 1.6574629851940776e-05,
+      "loss": 0.4858,
+      "step": 16720
+    },
+    {
+      "epoch": 21.40288,
+      "grad_norm": 1.0625916719436646,
+      "learning_rate": 1.657262905162065e-05,
+      "loss": 0.4438,
+      "step": 16721
+    },
+    {
+      "epoch": 21.40416,
+      "grad_norm": 1.169717788696289,
+      "learning_rate": 1.657062825130052e-05,
+      "loss": 0.4998,
+      "step": 16722
+    },
+    {
+      "epoch": 21.40544,
+      "grad_norm": 1.1702299118041992,
+      "learning_rate": 1.6568627450980392e-05,
+      "loss": 0.4903,
+      "step": 16723
+    },
+    {
+      "epoch": 21.40672,
+      "grad_norm": 1.062159538269043,
+      "learning_rate": 1.6566626650660264e-05,
+      "loss": 0.4464,
+      "step": 16724
+    },
+    {
+      "epoch": 21.408,
+      "grad_norm": 1.1465506553649902,
+      "learning_rate": 1.656462585034014e-05,
+      "loss": 0.511,
+      "step": 16725
+    },
+    {
+      "epoch": 21.40928,
+      "grad_norm": 1.079226016998291,
+      "learning_rate": 1.6562625050020007e-05,
+      "loss": 0.4868,
+      "step": 16726
+    },
+    {
+      "epoch": 21.41056,
+      "grad_norm": 1.0724486112594604,
+      "learning_rate": 1.656062424969988e-05,
+      "loss": 0.4862,
+      "step": 16727
+    },
+    {
+      "epoch": 21.41184,
+      "grad_norm": 1.142055869102478,
+      "learning_rate": 1.6558623449379755e-05,
+      "loss": 0.5397,
+      "step": 16728
+    },
+    {
+      "epoch": 21.41312,
+      "grad_norm": 1.0549299716949463,
+      "learning_rate": 1.6556622649059626e-05,
+      "loss": 0.44,
+      "step": 16729
+    },
+    {
+      "epoch": 21.4144,
+      "grad_norm": 1.177176833152771,
+      "learning_rate": 1.6554621848739495e-05,
+      "loss": 0.5188,
+      "step": 16730
+    },
+    {
+      "epoch": 21.41568,
+      "grad_norm": 1.174822449684143,
+      "learning_rate": 1.6552621048419367e-05,
+      "loss": 0.502,
+      "step": 16731
+    },
+    {
+      "epoch": 21.41696,
+      "grad_norm": 1.1037957668304443,
+      "learning_rate": 1.6550620248099242e-05,
+      "loss": 0.486,
+      "step": 16732
+    },
+    {
+      "epoch": 21.41824,
+      "grad_norm": 1.1217528581619263,
+      "learning_rate": 1.6548619447779114e-05,
+      "loss": 0.4969,
+      "step": 16733
+    },
+    {
+      "epoch": 21.41952,
+      "grad_norm": 1.146602749824524,
+      "learning_rate": 1.6546618647458982e-05,
+      "loss": 0.4549,
+      "step": 16734
+    },
+    {
+      "epoch": 21.4208,
+      "grad_norm": 1.062936782836914,
+      "learning_rate": 1.6544617847138858e-05,
+      "loss": 0.467,
+      "step": 16735
+    },
+    {
+      "epoch": 21.42208,
+      "grad_norm": 1.1040713787078857,
+      "learning_rate": 1.654261704681873e-05,
+      "loss": 0.4766,
+      "step": 16736
+    },
+    {
+      "epoch": 21.42336,
+      "grad_norm": 1.055647373199463,
+      "learning_rate": 1.65406162464986e-05,
+      "loss": 0.4662,
+      "step": 16737
+    },
+    {
+      "epoch": 21.42464,
+      "grad_norm": 1.1249991655349731,
+      "learning_rate": 1.653861544617847e-05,
+      "loss": 0.5188,
+      "step": 16738
+    },
+    {
+      "epoch": 21.42592,
+      "grad_norm": 1.0827454328536987,
+      "learning_rate": 1.6536614645858345e-05,
+      "loss": 0.4846,
+      "step": 16739
+    },
+    {
+      "epoch": 21.4272,
+      "grad_norm": 1.1406104564666748,
+      "learning_rate": 1.6534613845538217e-05,
+      "loss": 0.5539,
+      "step": 16740
+    },
+    {
+      "epoch": 21.42848,
+      "grad_norm": 1.1139311790466309,
+      "learning_rate": 1.653261304521809e-05,
+      "loss": 0.4687,
+      "step": 16741
+    },
+    {
+      "epoch": 21.42976,
+      "grad_norm": 1.1412529945373535,
+      "learning_rate": 1.653061224489796e-05,
+      "loss": 0.5136,
+      "step": 16742
+    },
+    {
+      "epoch": 21.43104,
+      "grad_norm": 1.1076371669769287,
+      "learning_rate": 1.6528611444577832e-05,
+      "loss": 0.4687,
+      "step": 16743
+    },
+    {
+      "epoch": 21.43232,
+      "grad_norm": 1.0816609859466553,
+      "learning_rate": 1.6526610644257704e-05,
+      "loss": 0.4853,
+      "step": 16744
+    },
+    {
+      "epoch": 21.4336,
+      "grad_norm": 1.0708239078521729,
+      "learning_rate": 1.6524609843937576e-05,
+      "loss": 0.4779,
+      "step": 16745
+    },
+    {
+      "epoch": 21.43488,
+      "grad_norm": 1.1649770736694336,
+      "learning_rate": 1.6522609043617448e-05,
+      "loss": 0.498,
+      "step": 16746
+    },
+    {
+      "epoch": 21.43616,
+      "grad_norm": 1.1269488334655762,
+      "learning_rate": 1.652060824329732e-05,
+      "loss": 0.4976,
+      "step": 16747
+    },
+    {
+      "epoch": 21.43744,
+      "grad_norm": 1.0875318050384521,
+      "learning_rate": 1.651860744297719e-05,
+      "loss": 0.514,
+      "step": 16748
+    },
+    {
+      "epoch": 21.43872,
+      "grad_norm": 1.0458475351333618,
+      "learning_rate": 1.6516606642657063e-05,
+      "loss": 0.4691,
+      "step": 16749
+    },
+    {
+      "epoch": 21.44,
+      "grad_norm": 1.1326485872268677,
+      "learning_rate": 1.6514605842336935e-05,
+      "loss": 0.4746,
+      "step": 16750
+    },
+    {
+      "epoch": 21.44128,
+      "grad_norm": 1.032772183418274,
+      "learning_rate": 1.6512605042016807e-05,
+      "loss": 0.4662,
+      "step": 16751
+    },
+    {
+      "epoch": 21.44256,
+      "grad_norm": 1.1388492584228516,
+      "learning_rate": 1.651060424169668e-05,
+      "loss": 0.49,
+      "step": 16752
+    },
+    {
+      "epoch": 21.44384,
+      "grad_norm": 1.1491458415985107,
+      "learning_rate": 1.650860344137655e-05,
+      "loss": 0.486,
+      "step": 16753
+    },
+    {
+      "epoch": 21.44512,
+      "grad_norm": 1.0875811576843262,
+      "learning_rate": 1.6506602641056423e-05,
+      "loss": 0.4755,
+      "step": 16754
+    },
+    {
+      "epoch": 21.4464,
+      "grad_norm": 1.056416630744934,
+      "learning_rate": 1.6504601840736295e-05,
+      "loss": 0.4494,
+      "step": 16755
+    },
+    {
+      "epoch": 21.44768,
+      "grad_norm": 1.0903964042663574,
+      "learning_rate": 1.650260104041617e-05,
+      "loss": 0.5011,
+      "step": 16756
+    },
+    {
+      "epoch": 21.44896,
+      "grad_norm": 1.0801723003387451,
+      "learning_rate": 1.650060024009604e-05,
+      "loss": 0.492,
+      "step": 16757
+    },
+    {
+      "epoch": 21.45024,
+      "grad_norm": 1.1389845609664917,
+      "learning_rate": 1.649859943977591e-05,
+      "loss": 0.51,
+      "step": 16758
+    },
+    {
+      "epoch": 21.45152,
+      "grad_norm": 1.1262261867523193,
+      "learning_rate": 1.6496598639455782e-05,
+      "loss": 0.5022,
+      "step": 16759
+    },
+    {
+      "epoch": 21.4528,
+      "grad_norm": 1.123065710067749,
+      "learning_rate": 1.6494597839135657e-05,
+      "loss": 0.5005,
+      "step": 16760
+    },
+    {
+      "epoch": 21.45408,
+      "grad_norm": 1.099016785621643,
+      "learning_rate": 1.6492597038815526e-05,
+      "loss": 0.5023,
+      "step": 16761
+    },
+    {
+      "epoch": 21.45536,
+      "grad_norm": 1.1234606504440308,
+      "learning_rate": 1.6490596238495398e-05,
+      "loss": 0.4718,
+      "step": 16762
+    },
+    {
+      "epoch": 21.45664,
+      "grad_norm": 1.1033679246902466,
+      "learning_rate": 1.6488595438175273e-05,
+      "loss": 0.4908,
+      "step": 16763
+    },
+    {
+      "epoch": 21.45792,
+      "grad_norm": 1.0978657007217407,
+      "learning_rate": 1.6486594637855145e-05,
+      "loss": 0.5028,
+      "step": 16764
+    },
+    {
+      "epoch": 21.4592,
+      "grad_norm": 1.0921522378921509,
+      "learning_rate": 1.6484593837535013e-05,
+      "loss": 0.4639,
+      "step": 16765
+    },
+    {
+      "epoch": 21.46048,
+      "grad_norm": 1.0589871406555176,
+      "learning_rate": 1.6482593037214885e-05,
+      "loss": 0.4662,
+      "step": 16766
+    },
+    {
+      "epoch": 21.46176,
+      "grad_norm": 1.120967149734497,
+      "learning_rate": 1.648059223689476e-05,
+      "loss": 0.4901,
+      "step": 16767
+    },
+    {
+      "epoch": 21.46304,
+      "grad_norm": 1.1276214122772217,
+      "learning_rate": 1.6478591436574632e-05,
+      "loss": 0.4807,
+      "step": 16768
+    },
+    {
+      "epoch": 21.46432,
+      "grad_norm": 1.0669405460357666,
+      "learning_rate": 1.64765906362545e-05,
+      "loss": 0.4683,
+      "step": 16769
+    },
+    {
+      "epoch": 21.4656,
+      "grad_norm": 1.0860059261322021,
+      "learning_rate": 1.6474589835934376e-05,
+      "loss": 0.5079,
+      "step": 16770
+    },
+    {
+      "epoch": 21.46688,
+      "grad_norm": 1.073190689086914,
+      "learning_rate": 1.6472589035614248e-05,
+      "loss": 0.4723,
+      "step": 16771
+    },
+    {
+      "epoch": 21.46816,
+      "grad_norm": 1.1541026830673218,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 0.5063,
+      "step": 16772
+    },
+    {
+      "epoch": 21.46944,
+      "grad_norm": 1.0719937086105347,
+      "learning_rate": 1.6468587434973988e-05,
+      "loss": 0.4436,
+      "step": 16773
+    },
+    {
+      "epoch": 21.47072,
+      "grad_norm": 1.048323631286621,
+      "learning_rate": 1.6466586634653863e-05,
+      "loss": 0.439,
+      "step": 16774
+    },
+    {
+      "epoch": 21.472,
+      "grad_norm": 1.0535809993743896,
+      "learning_rate": 1.6464585834333735e-05,
+      "loss": 0.4863,
+      "step": 16775
+    },
+    {
+      "epoch": 21.47328,
+      "grad_norm": 1.0624300241470337,
+      "learning_rate": 1.6462585034013607e-05,
+      "loss": 0.5091,
+      "step": 16776
+    },
+    {
+      "epoch": 21.47456,
+      "grad_norm": 1.1307200193405151,
+      "learning_rate": 1.6460584233693475e-05,
+      "loss": 0.4733,
+      "step": 16777
+    },
+    {
+      "epoch": 21.47584,
+      "grad_norm": 1.1632894277572632,
+      "learning_rate": 1.645858343337335e-05,
+      "loss": 0.4962,
+      "step": 16778
+    },
+    {
+      "epoch": 21.47712,
+      "grad_norm": 1.1236180067062378,
+      "learning_rate": 1.6456582633053223e-05,
+      "loss": 0.4846,
+      "step": 16779
+    },
+    {
+      "epoch": 21.4784,
+      "grad_norm": 1.083125114440918,
+      "learning_rate": 1.6454581832733094e-05,
+      "loss": 0.4867,
+      "step": 16780
+    },
+    {
+      "epoch": 21.47968,
+      "grad_norm": 1.0312892198562622,
+      "learning_rate": 1.6452581032412966e-05,
+      "loss": 0.4648,
+      "step": 16781
+    },
+    {
+      "epoch": 21.48096,
+      "grad_norm": 1.0965737104415894,
+      "learning_rate": 1.6450580232092838e-05,
+      "loss": 0.4893,
+      "step": 16782
+    },
+    {
+      "epoch": 21.48224,
+      "grad_norm": 1.157957673072815,
+      "learning_rate": 1.644857943177271e-05,
+      "loss": 0.5055,
+      "step": 16783
+    },
+    {
+      "epoch": 21.48352,
+      "grad_norm": 1.0796724557876587,
+      "learning_rate": 1.6446578631452582e-05,
+      "loss": 0.4762,
+      "step": 16784
+    },
+    {
+      "epoch": 21.4848,
+      "grad_norm": 1.1314095258712769,
+      "learning_rate": 1.6444577831132454e-05,
+      "loss": 0.49,
+      "step": 16785
+    },
+    {
+      "epoch": 21.48608,
+      "grad_norm": 1.070644736289978,
+      "learning_rate": 1.6442577030812326e-05,
+      "loss": 0.4916,
+      "step": 16786
+    },
+    {
+      "epoch": 21.48736,
+      "grad_norm": 1.101872205734253,
+      "learning_rate": 1.6440576230492197e-05,
+      "loss": 0.4969,
+      "step": 16787
+    },
+    {
+      "epoch": 21.48864,
+      "grad_norm": 1.0998544692993164,
+      "learning_rate": 1.643857543017207e-05,
+      "loss": 0.4433,
+      "step": 16788
+    },
+    {
+      "epoch": 21.48992,
+      "grad_norm": 1.1133356094360352,
+      "learning_rate": 1.643657462985194e-05,
+      "loss": 0.4838,
+      "step": 16789
+    },
+    {
+      "epoch": 21.4912,
+      "grad_norm": 1.0400280952453613,
+      "learning_rate": 1.6434573829531813e-05,
+      "loss": 0.4781,
+      "step": 16790
+    },
+    {
+      "epoch": 21.49248,
+      "grad_norm": 1.1187825202941895,
+      "learning_rate": 1.6432573029211685e-05,
+      "loss": 0.4502,
+      "step": 16791
+    },
+    {
+      "epoch": 21.49376,
+      "grad_norm": 1.09933340549469,
+      "learning_rate": 1.6430572228891557e-05,
+      "loss": 0.4945,
+      "step": 16792
+    },
+    {
+      "epoch": 21.49504,
+      "grad_norm": 1.0932108163833618,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 0.4643,
+      "step": 16793
+    },
+    {
+      "epoch": 21.49632,
+      "grad_norm": 1.1463568210601807,
+      "learning_rate": 1.64265706282513e-05,
+      "loss": 0.5326,
+      "step": 16794
+    },
+    {
+      "epoch": 21.4976,
+      "grad_norm": 1.0978707075119019,
+      "learning_rate": 1.6424569827931176e-05,
+      "loss": 0.5083,
+      "step": 16795
+    },
+    {
+      "epoch": 21.49888,
+      "grad_norm": 1.1023026704788208,
+      "learning_rate": 1.6422569027611044e-05,
+      "loss": 0.4388,
+      "step": 16796
+    },
+    {
+      "epoch": 21.50016,
+      "grad_norm": 1.119019865989685,
+      "learning_rate": 1.6420568227290916e-05,
+      "loss": 0.4728,
+      "step": 16797
+    },
+    {
+      "epoch": 21.50144,
+      "grad_norm": 1.1367695331573486,
+      "learning_rate": 1.6418567426970788e-05,
+      "loss": 0.4819,
+      "step": 16798
+    },
+    {
+      "epoch": 21.50272,
+      "grad_norm": 1.096197485923767,
+      "learning_rate": 1.6416566626650663e-05,
+      "loss": 0.459,
+      "step": 16799
+    },
+    {
+      "epoch": 21.504,
+      "grad_norm": 1.1106514930725098,
+      "learning_rate": 1.641456582633053e-05,
+      "loss": 0.4894,
+      "step": 16800
+    },
+    {
+      "epoch": 21.50528,
+      "grad_norm": 1.0969781875610352,
+      "learning_rate": 1.6412565026010403e-05,
+      "loss": 0.5059,
+      "step": 16801
+    },
+    {
+      "epoch": 21.50656,
+      "grad_norm": 1.158473014831543,
+      "learning_rate": 1.641056422569028e-05,
+      "loss": 0.4895,
+      "step": 16802
+    },
+    {
+      "epoch": 21.50784,
+      "grad_norm": 1.147180199623108,
+      "learning_rate": 1.640856342537015e-05,
+      "loss": 0.5291,
+      "step": 16803
+    },
+    {
+      "epoch": 21.50912,
+      "grad_norm": 1.0704094171524048,
+      "learning_rate": 1.640656262505002e-05,
+      "loss": 0.4578,
+      "step": 16804
+    },
+    {
+      "epoch": 21.5104,
+      "grad_norm": 1.1559474468231201,
+      "learning_rate": 1.640456182472989e-05,
+      "loss": 0.5084,
+      "step": 16805
+    },
+    {
+      "epoch": 21.51168,
+      "grad_norm": 1.14545738697052,
+      "learning_rate": 1.6402561024409766e-05,
+      "loss": 0.5353,
+      "step": 16806
+    },
+    {
+      "epoch": 21.51296,
+      "grad_norm": 1.1325570344924927,
+      "learning_rate": 1.6400560224089638e-05,
+      "loss": 0.4731,
+      "step": 16807
+    },
+    {
+      "epoch": 21.51424,
+      "grad_norm": 1.1258713006973267,
+      "learning_rate": 1.6398559423769506e-05,
+      "loss": 0.4792,
+      "step": 16808
+    },
+    {
+      "epoch": 21.51552,
+      "grad_norm": 1.1287716627120972,
+      "learning_rate": 1.639655862344938e-05,
+      "loss": 0.4661,
+      "step": 16809
+    },
+    {
+      "epoch": 21.5168,
+      "grad_norm": 1.1289795637130737,
+      "learning_rate": 1.6394557823129254e-05,
+      "loss": 0.4774,
+      "step": 16810
+    },
+    {
+      "epoch": 21.51808,
+      "grad_norm": 1.1608408689498901,
+      "learning_rate": 1.6392557022809125e-05,
+      "loss": 0.4932,
+      "step": 16811
+    },
+    {
+      "epoch": 21.51936,
+      "grad_norm": 1.1531096696853638,
+      "learning_rate": 1.6390556222488994e-05,
+      "loss": 0.5028,
+      "step": 16812
+    },
+    {
+      "epoch": 21.52064,
+      "grad_norm": 1.1289916038513184,
+      "learning_rate": 1.638855542216887e-05,
+      "loss": 0.4894,
+      "step": 16813
+    },
+    {
+      "epoch": 21.52192,
+      "grad_norm": 1.131188988685608,
+      "learning_rate": 1.638655462184874e-05,
+      "loss": 0.5029,
+      "step": 16814
+    },
+    {
+      "epoch": 21.5232,
+      "grad_norm": 1.0534837245941162,
+      "learning_rate": 1.6384553821528613e-05,
+      "loss": 0.4684,
+      "step": 16815
+    },
+    {
+      "epoch": 21.52448,
+      "grad_norm": 1.1278927326202393,
+      "learning_rate": 1.6382553021208485e-05,
+      "loss": 0.5104,
+      "step": 16816
+    },
+    {
+      "epoch": 21.52576,
+      "grad_norm": 1.177605152130127,
+      "learning_rate": 1.6380552220888357e-05,
+      "loss": 0.4751,
+      "step": 16817
+    },
+    {
+      "epoch": 21.52704,
+      "grad_norm": 1.1548734903335571,
+      "learning_rate": 1.637855142056823e-05,
+      "loss": 0.502,
+      "step": 16818
+    },
+    {
+      "epoch": 21.52832,
+      "grad_norm": 1.145601749420166,
+      "learning_rate": 1.63765506202481e-05,
+      "loss": 0.5197,
+      "step": 16819
+    },
+    {
+      "epoch": 21.5296,
+      "grad_norm": 1.058839201927185,
+      "learning_rate": 1.6374549819927972e-05,
+      "loss": 0.4424,
+      "step": 16820
+    },
+    {
+      "epoch": 21.53088,
+      "grad_norm": 1.1241230964660645,
+      "learning_rate": 1.6372549019607844e-05,
+      "loss": 0.4831,
+      "step": 16821
+    },
+    {
+      "epoch": 21.53216,
+      "grad_norm": 1.130538821220398,
+      "learning_rate": 1.6370548219287716e-05,
+      "loss": 0.4936,
+      "step": 16822
+    },
+    {
+      "epoch": 21.53344,
+      "grad_norm": 1.1460503339767456,
+      "learning_rate": 1.6368547418967588e-05,
+      "loss": 0.4879,
+      "step": 16823
+    },
+    {
+      "epoch": 21.53472,
+      "grad_norm": 1.1377558708190918,
+      "learning_rate": 1.636654661864746e-05,
+      "loss": 0.5494,
+      "step": 16824
+    },
+    {
+      "epoch": 21.536,
+      "grad_norm": 1.1145014762878418,
+      "learning_rate": 1.636454581832733e-05,
+      "loss": 0.5045,
+      "step": 16825
+    },
+    {
+      "epoch": 21.53728,
+      "grad_norm": 1.0829366445541382,
+      "learning_rate": 1.6362545018007203e-05,
+      "loss": 0.4957,
+      "step": 16826
+    },
+    {
+      "epoch": 21.53856,
+      "grad_norm": 1.0551543235778809,
+      "learning_rate": 1.6360544217687075e-05,
+      "loss": 0.4979,
+      "step": 16827
+    },
+    {
+      "epoch": 21.53984,
+      "grad_norm": 1.0742313861846924,
+      "learning_rate": 1.6358543417366947e-05,
+      "loss": 0.5055,
+      "step": 16828
+    },
+    {
+      "epoch": 21.54112,
+      "grad_norm": 1.1106947660446167,
+      "learning_rate": 1.635654261704682e-05,
+      "loss": 0.5044,
+      "step": 16829
+    },
+    {
+      "epoch": 21.5424,
+      "grad_norm": 1.0783088207244873,
+      "learning_rate": 1.6354541816726694e-05,
+      "loss": 0.5078,
+      "step": 16830
+    },
+    {
+      "epoch": 21.54368,
+      "grad_norm": 1.1662565469741821,
+      "learning_rate": 1.6352541016406563e-05,
+      "loss": 0.5252,
+      "step": 16831
+    },
+    {
+      "epoch": 21.54496,
+      "grad_norm": 1.0940940380096436,
+      "learning_rate": 1.6350540216086434e-05,
+      "loss": 0.4984,
+      "step": 16832
+    },
+    {
+      "epoch": 21.54624,
+      "grad_norm": 1.0466233491897583,
+      "learning_rate": 1.6348539415766306e-05,
+      "loss": 0.498,
+      "step": 16833
+    },
+    {
+      "epoch": 21.54752,
+      "grad_norm": 1.097116470336914,
+      "learning_rate": 1.634653861544618e-05,
+      "loss": 0.4793,
+      "step": 16834
+    },
+    {
+      "epoch": 21.5488,
+      "grad_norm": 1.19340980052948,
+      "learning_rate": 1.634453781512605e-05,
+      "loss": 0.5476,
+      "step": 16835
+    },
+    {
+      "epoch": 21.55008,
+      "grad_norm": 1.0786470174789429,
+      "learning_rate": 1.6342537014805922e-05,
+      "loss": 0.4835,
+      "step": 16836
+    },
+    {
+      "epoch": 21.55136,
+      "grad_norm": 1.1250004768371582,
+      "learning_rate": 1.6340536214485794e-05,
+      "loss": 0.4834,
+      "step": 16837
+    },
+    {
+      "epoch": 21.55264,
+      "grad_norm": 1.1461224555969238,
+      "learning_rate": 1.633853541416567e-05,
+      "loss": 0.5397,
+      "step": 16838
+    },
+    {
+      "epoch": 21.55392,
+      "grad_norm": 1.153669834136963,
+      "learning_rate": 1.6336534613845537e-05,
+      "loss": 0.5246,
+      "step": 16839
+    },
+    {
+      "epoch": 21.5552,
+      "grad_norm": 1.1360396146774292,
+      "learning_rate": 1.633453381352541e-05,
+      "loss": 0.4887,
+      "step": 16840
+    },
+    {
+      "epoch": 21.55648,
+      "grad_norm": 1.1669892072677612,
+      "learning_rate": 1.6332533013205284e-05,
+      "loss": 0.5377,
+      "step": 16841
+    },
+    {
+      "epoch": 21.557760000000002,
+      "grad_norm": 1.094748854637146,
+      "learning_rate": 1.6330532212885156e-05,
+      "loss": 0.4918,
+      "step": 16842
+    },
+    {
+      "epoch": 21.55904,
+      "grad_norm": 1.1243646144866943,
+      "learning_rate": 1.6328531412565025e-05,
+      "loss": 0.4567,
+      "step": 16843
+    },
+    {
+      "epoch": 21.56032,
+      "grad_norm": 1.0712041854858398,
+      "learning_rate": 1.6326530612244897e-05,
+      "loss": 0.4915,
+      "step": 16844
+    },
+    {
+      "epoch": 21.5616,
+      "grad_norm": 1.2099251747131348,
+      "learning_rate": 1.6324529811924772e-05,
+      "loss": 0.5528,
+      "step": 16845
+    },
+    {
+      "epoch": 21.56288,
+      "grad_norm": 1.102573037147522,
+      "learning_rate": 1.6322529011604644e-05,
+      "loss": 0.4806,
+      "step": 16846
+    },
+    {
+      "epoch": 21.56416,
+      "grad_norm": 1.1252408027648926,
+      "learning_rate": 1.6320528211284512e-05,
+      "loss": 0.48,
+      "step": 16847
+    },
+    {
+      "epoch": 21.56544,
+      "grad_norm": 1.112750768661499,
+      "learning_rate": 1.6318527410964387e-05,
+      "loss": 0.5078,
+      "step": 16848
+    },
+    {
+      "epoch": 21.56672,
+      "grad_norm": 1.0940823554992676,
+      "learning_rate": 1.631652661064426e-05,
+      "loss": 0.499,
+      "step": 16849
+    },
+    {
+      "epoch": 21.568,
+      "grad_norm": 1.04874849319458,
+      "learning_rate": 1.631452581032413e-05,
+      "loss": 0.4649,
+      "step": 16850
+    },
+    {
+      "epoch": 21.56928,
+      "grad_norm": 1.0783095359802246,
+      "learning_rate": 1.6312525010004e-05,
+      "loss": 0.5148,
+      "step": 16851
+    },
+    {
+      "epoch": 21.57056,
+      "grad_norm": 1.1552876234054565,
+      "learning_rate": 1.6310524209683875e-05,
+      "loss": 0.5087,
+      "step": 16852
+    },
+    {
+      "epoch": 21.57184,
+      "grad_norm": 1.1094108819961548,
+      "learning_rate": 1.6308523409363747e-05,
+      "loss": 0.4625,
+      "step": 16853
+    },
+    {
+      "epoch": 21.57312,
+      "grad_norm": 1.1274285316467285,
+      "learning_rate": 1.630652260904362e-05,
+      "loss": 0.5366,
+      "step": 16854
+    },
+    {
+      "epoch": 21.5744,
+      "grad_norm": 1.1100611686706543,
+      "learning_rate": 1.630452180872349e-05,
+      "loss": 0.5105,
+      "step": 16855
+    },
+    {
+      "epoch": 21.57568,
+      "grad_norm": 1.108077883720398,
+      "learning_rate": 1.6302521008403362e-05,
+      "loss": 0.4801,
+      "step": 16856
+    },
+    {
+      "epoch": 21.57696,
+      "grad_norm": 1.1188905239105225,
+      "learning_rate": 1.6300520208083234e-05,
+      "loss": 0.4432,
+      "step": 16857
+    },
+    {
+      "epoch": 21.57824,
+      "grad_norm": 1.2001028060913086,
+      "learning_rate": 1.6298519407763106e-05,
+      "loss": 0.5044,
+      "step": 16858
+    },
+    {
+      "epoch": 21.57952,
+      "grad_norm": 1.151809573173523,
+      "learning_rate": 1.6296518607442978e-05,
+      "loss": 0.4696,
+      "step": 16859
+    },
+    {
+      "epoch": 21.5808,
+      "grad_norm": 1.1924430131912231,
+      "learning_rate": 1.629451780712285e-05,
+      "loss": 0.5198,
+      "step": 16860
+    },
+    {
+      "epoch": 21.58208,
+      "grad_norm": 1.1216545104980469,
+      "learning_rate": 1.629251700680272e-05,
+      "loss": 0.4543,
+      "step": 16861
+    },
+    {
+      "epoch": 21.58336,
+      "grad_norm": 1.150597333908081,
+      "learning_rate": 1.6290516206482593e-05,
+      "loss": 0.5262,
+      "step": 16862
+    },
+    {
+      "epoch": 21.58464,
+      "grad_norm": 1.1352801322937012,
+      "learning_rate": 1.6288515406162465e-05,
+      "loss": 0.4916,
+      "step": 16863
+    },
+    {
+      "epoch": 21.58592,
+      "grad_norm": 1.0616695880889893,
+      "learning_rate": 1.6286514605842337e-05,
+      "loss": 0.4861,
+      "step": 16864
+    },
+    {
+      "epoch": 21.5872,
+      "grad_norm": 1.079972743988037,
+      "learning_rate": 1.628451380552221e-05,
+      "loss": 0.4737,
+      "step": 16865
+    },
+    {
+      "epoch": 21.58848,
+      "grad_norm": 1.0440356731414795,
+      "learning_rate": 1.628251300520208e-05,
+      "loss": 0.499,
+      "step": 16866
+    },
+    {
+      "epoch": 21.58976,
+      "grad_norm": 1.1359690427780151,
+      "learning_rate": 1.6280512204881953e-05,
+      "loss": 0.5054,
+      "step": 16867
+    },
+    {
+      "epoch": 21.59104,
+      "grad_norm": 1.1668392419815063,
+      "learning_rate": 1.6278511404561825e-05,
+      "loss": 0.5426,
+      "step": 16868
+    },
+    {
+      "epoch": 21.59232,
+      "grad_norm": 1.2051527500152588,
+      "learning_rate": 1.62765106042417e-05,
+      "loss": 0.4973,
+      "step": 16869
+    },
+    {
+      "epoch": 21.5936,
+      "grad_norm": 1.0257412195205688,
+      "learning_rate": 1.627450980392157e-05,
+      "loss": 0.4355,
+      "step": 16870
+    },
+    {
+      "epoch": 21.59488,
+      "grad_norm": 1.0660890340805054,
+      "learning_rate": 1.627250900360144e-05,
+      "loss": 0.4721,
+      "step": 16871
+    },
+    {
+      "epoch": 21.59616,
+      "grad_norm": 1.1143311262130737,
+      "learning_rate": 1.6270508203281312e-05,
+      "loss": 0.4979,
+      "step": 16872
+    },
+    {
+      "epoch": 21.59744,
+      "grad_norm": 1.1307166814804077,
+      "learning_rate": 1.6268507402961187e-05,
+      "loss": 0.516,
+      "step": 16873
+    },
+    {
+      "epoch": 21.59872,
+      "grad_norm": 1.0710972547531128,
+      "learning_rate": 1.6266506602641056e-05,
+      "loss": 0.4866,
+      "step": 16874
+    },
+    {
+      "epoch": 21.6,
+      "grad_norm": 1.1194899082183838,
+      "learning_rate": 1.6264505802320928e-05,
+      "loss": 0.4978,
+      "step": 16875
+    },
+    {
+      "epoch": 21.60128,
+      "grad_norm": 1.1724941730499268,
+      "learning_rate": 1.6262505002000803e-05,
+      "loss": 0.5422,
+      "step": 16876
+    },
+    {
+      "epoch": 21.60256,
+      "grad_norm": 1.1926066875457764,
+      "learning_rate": 1.6260504201680675e-05,
+      "loss": 0.5132,
+      "step": 16877
+    },
+    {
+      "epoch": 21.60384,
+      "grad_norm": 1.1342614889144897,
+      "learning_rate": 1.6258503401360543e-05,
+      "loss": 0.4642,
+      "step": 16878
+    },
+    {
+      "epoch": 21.60512,
+      "grad_norm": 1.0815118551254272,
+      "learning_rate": 1.6256502601040415e-05,
+      "loss": 0.4648,
+      "step": 16879
+    },
+    {
+      "epoch": 21.6064,
+      "grad_norm": 1.073204517364502,
+      "learning_rate": 1.625450180072029e-05,
+      "loss": 0.4942,
+      "step": 16880
+    },
+    {
+      "epoch": 21.60768,
+      "grad_norm": 1.1120681762695312,
+      "learning_rate": 1.6252501000400162e-05,
+      "loss": 0.5087,
+      "step": 16881
+    },
+    {
+      "epoch": 21.60896,
+      "grad_norm": 1.085679292678833,
+      "learning_rate": 1.625050020008003e-05,
+      "loss": 0.4773,
+      "step": 16882
+    },
+    {
+      "epoch": 21.61024,
+      "grad_norm": 1.0857542753219604,
+      "learning_rate": 1.6248499399759906e-05,
+      "loss": 0.4978,
+      "step": 16883
+    },
+    {
+      "epoch": 21.61152,
+      "grad_norm": 1.0734546184539795,
+      "learning_rate": 1.6246498599439778e-05,
+      "loss": 0.4487,
+      "step": 16884
+    },
+    {
+      "epoch": 21.6128,
+      "grad_norm": 1.1066250801086426,
+      "learning_rate": 1.624449779911965e-05,
+      "loss": 0.5063,
+      "step": 16885
+    },
+    {
+      "epoch": 21.61408,
+      "grad_norm": 1.0745172500610352,
+      "learning_rate": 1.6242496998799518e-05,
+      "loss": 0.4654,
+      "step": 16886
+    },
+    {
+      "epoch": 21.61536,
+      "grad_norm": 1.1120097637176514,
+      "learning_rate": 1.6240496198479393e-05,
+      "loss": 0.4833,
+      "step": 16887
+    },
+    {
+      "epoch": 21.61664,
+      "grad_norm": 1.0820595026016235,
+      "learning_rate": 1.6238495398159265e-05,
+      "loss": 0.4938,
+      "step": 16888
+    },
+    {
+      "epoch": 21.61792,
+      "grad_norm": 1.0889438390731812,
+      "learning_rate": 1.6236494597839137e-05,
+      "loss": 0.4847,
+      "step": 16889
+    },
+    {
+      "epoch": 21.6192,
+      "grad_norm": 1.1894570589065552,
+      "learning_rate": 1.6234493797519005e-05,
+      "loss": 0.5127,
+      "step": 16890
+    },
+    {
+      "epoch": 21.62048,
+      "grad_norm": 1.123768925666809,
+      "learning_rate": 1.623249299719888e-05,
+      "loss": 0.4727,
+      "step": 16891
+    },
+    {
+      "epoch": 21.62176,
+      "grad_norm": 1.0649784803390503,
+      "learning_rate": 1.6230492196878753e-05,
+      "loss": 0.4568,
+      "step": 16892
+    },
+    {
+      "epoch": 21.62304,
+      "grad_norm": 1.1826279163360596,
+      "learning_rate": 1.6228491396558624e-05,
+      "loss": 0.4925,
+      "step": 16893
+    },
+    {
+      "epoch": 21.62432,
+      "grad_norm": 1.1905529499053955,
+      "learning_rate": 1.6226490596238496e-05,
+      "loss": 0.495,
+      "step": 16894
+    },
+    {
+      "epoch": 21.6256,
+      "grad_norm": 1.206997275352478,
+      "learning_rate": 1.6224489795918368e-05,
+      "loss": 0.4867,
+      "step": 16895
+    },
+    {
+      "epoch": 21.62688,
+      "grad_norm": 1.1951205730438232,
+      "learning_rate": 1.622248899559824e-05,
+      "loss": 0.5134,
+      "step": 16896
+    },
+    {
+      "epoch": 21.62816,
+      "grad_norm": 1.1452414989471436,
+      "learning_rate": 1.6220488195278112e-05,
+      "loss": 0.4927,
+      "step": 16897
+    },
+    {
+      "epoch": 21.62944,
+      "grad_norm": 1.1808103322982788,
+      "learning_rate": 1.6218487394957984e-05,
+      "loss": 0.5282,
+      "step": 16898
+    },
+    {
+      "epoch": 21.63072,
+      "grad_norm": 1.0460494756698608,
+      "learning_rate": 1.6216486594637856e-05,
+      "loss": 0.4496,
+      "step": 16899
+    },
+    {
+      "epoch": 21.632,
+      "grad_norm": 1.1484510898590088,
+      "learning_rate": 1.6214485794317727e-05,
+      "loss": 0.4895,
+      "step": 16900
+    },
+    {
+      "epoch": 21.63328,
+      "grad_norm": 1.153800129890442,
+      "learning_rate": 1.62124849939976e-05,
+      "loss": 0.5,
+      "step": 16901
+    },
+    {
+      "epoch": 21.63456,
+      "grad_norm": 1.1488093137741089,
+      "learning_rate": 1.621048419367747e-05,
+      "loss": 0.5194,
+      "step": 16902
+    },
+    {
+      "epoch": 21.63584,
+      "grad_norm": 1.0916675329208374,
+      "learning_rate": 1.6208483393357343e-05,
+      "loss": 0.4648,
+      "step": 16903
+    },
+    {
+      "epoch": 21.63712,
+      "grad_norm": 1.0833429098129272,
+      "learning_rate": 1.6206482593037215e-05,
+      "loss": 0.4548,
+      "step": 16904
+    },
+    {
+      "epoch": 21.6384,
+      "grad_norm": 1.1267287731170654,
+      "learning_rate": 1.6204481792717087e-05,
+      "loss": 0.5134,
+      "step": 16905
+    },
+    {
+      "epoch": 21.63968,
+      "grad_norm": 1.2103078365325928,
+      "learning_rate": 1.620248099239696e-05,
+      "loss": 0.5598,
+      "step": 16906
+    },
+    {
+      "epoch": 21.64096,
+      "grad_norm": 1.0767333507537842,
+      "learning_rate": 1.620048019207683e-05,
+      "loss": 0.4599,
+      "step": 16907
+    },
+    {
+      "epoch": 21.64224,
+      "grad_norm": 1.12871515750885,
+      "learning_rate": 1.6198479391756706e-05,
+      "loss": 0.5094,
+      "step": 16908
+    },
+    {
+      "epoch": 21.64352,
+      "grad_norm": 1.0893502235412598,
+      "learning_rate": 1.6196478591436574e-05,
+      "loss": 0.4617,
+      "step": 16909
+    },
+    {
+      "epoch": 21.6448,
+      "grad_norm": 1.1467949151992798,
+      "learning_rate": 1.6194477791116446e-05,
+      "loss": 0.4717,
+      "step": 16910
+    },
+    {
+      "epoch": 21.64608,
+      "grad_norm": 1.151532530784607,
+      "learning_rate": 1.6192476990796318e-05,
+      "loss": 0.4958,
+      "step": 16911
+    },
+    {
+      "epoch": 21.64736,
+      "grad_norm": 1.151410698890686,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.5203,
+      "step": 16912
+    },
+    {
+      "epoch": 21.64864,
+      "grad_norm": 1.044360876083374,
+      "learning_rate": 1.618847539015606e-05,
+      "loss": 0.4547,
+      "step": 16913
+    },
+    {
+      "epoch": 21.64992,
+      "grad_norm": 1.1051994562149048,
+      "learning_rate": 1.6186474589835933e-05,
+      "loss": 0.5285,
+      "step": 16914
+    },
+    {
+      "epoch": 21.6512,
+      "grad_norm": 1.0989066362380981,
+      "learning_rate": 1.618447378951581e-05,
+      "loss": 0.4682,
+      "step": 16915
+    },
+    {
+      "epoch": 21.65248,
+      "grad_norm": 1.1165852546691895,
+      "learning_rate": 1.618247298919568e-05,
+      "loss": 0.5312,
+      "step": 16916
+    },
+    {
+      "epoch": 21.65376,
+      "grad_norm": 1.1108812093734741,
+      "learning_rate": 1.618047218887555e-05,
+      "loss": 0.5191,
+      "step": 16917
+    },
+    {
+      "epoch": 21.65504,
+      "grad_norm": 1.1192561388015747,
+      "learning_rate": 1.617847138855542e-05,
+      "loss": 0.4527,
+      "step": 16918
+    },
+    {
+      "epoch": 21.65632,
+      "grad_norm": 1.137689232826233,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 0.4527,
+      "step": 16919
+    },
+    {
+      "epoch": 21.6576,
+      "grad_norm": 1.1603665351867676,
+      "learning_rate": 1.6174469787915168e-05,
+      "loss": 0.5427,
+      "step": 16920
+    },
+    {
+      "epoch": 21.65888,
+      "grad_norm": 1.0487416982650757,
+      "learning_rate": 1.6172468987595036e-05,
+      "loss": 0.4543,
+      "step": 16921
+    },
+    {
+      "epoch": 21.66016,
+      "grad_norm": 1.0956952571868896,
+      "learning_rate": 1.617046818727491e-05,
+      "loss": 0.4952,
+      "step": 16922
+    },
+    {
+      "epoch": 21.66144,
+      "grad_norm": 1.0673739910125732,
+      "learning_rate": 1.6168467386954784e-05,
+      "loss": 0.4378,
+      "step": 16923
+    },
+    {
+      "epoch": 21.66272,
+      "grad_norm": 1.1492853164672852,
+      "learning_rate": 1.6166466586634655e-05,
+      "loss": 0.5302,
+      "step": 16924
+    },
+    {
+      "epoch": 21.664,
+      "grad_norm": 1.058276653289795,
+      "learning_rate": 1.6164465786314524e-05,
+      "loss": 0.4846,
+      "step": 16925
+    },
+    {
+      "epoch": 21.66528,
+      "grad_norm": 1.1349074840545654,
+      "learning_rate": 1.61624649859944e-05,
+      "loss": 0.5074,
+      "step": 16926
+    },
+    {
+      "epoch": 21.66656,
+      "grad_norm": 1.1578072309494019,
+      "learning_rate": 1.616046418567427e-05,
+      "loss": 0.5537,
+      "step": 16927
+    },
+    {
+      "epoch": 21.667839999999998,
+      "grad_norm": 1.1258023977279663,
+      "learning_rate": 1.6158463385354143e-05,
+      "loss": 0.5015,
+      "step": 16928
+    },
+    {
+      "epoch": 21.66912,
+      "grad_norm": 1.1199811697006226,
+      "learning_rate": 1.6156462585034015e-05,
+      "loss": 0.5215,
+      "step": 16929
+    },
+    {
+      "epoch": 21.6704,
+      "grad_norm": 1.0602338314056396,
+      "learning_rate": 1.6154461784713887e-05,
+      "loss": 0.4738,
+      "step": 16930
+    },
+    {
+      "epoch": 21.67168,
+      "grad_norm": 1.1654423475265503,
+      "learning_rate": 1.615246098439376e-05,
+      "loss": 0.5589,
+      "step": 16931
+    },
+    {
+      "epoch": 21.67296,
+      "grad_norm": 1.081931710243225,
+      "learning_rate": 1.615046018407363e-05,
+      "loss": 0.4466,
+      "step": 16932
+    },
+    {
+      "epoch": 21.67424,
+      "grad_norm": 1.1227881908416748,
+      "learning_rate": 1.6148459383753502e-05,
+      "loss": 0.5191,
+      "step": 16933
+    },
+    {
+      "epoch": 21.67552,
+      "grad_norm": 1.1291007995605469,
+      "learning_rate": 1.6146458583433374e-05,
+      "loss": 0.5003,
+      "step": 16934
+    },
+    {
+      "epoch": 21.6768,
+      "grad_norm": 1.0886553525924683,
+      "learning_rate": 1.6144457783113246e-05,
+      "loss": 0.4575,
+      "step": 16935
+    },
+    {
+      "epoch": 21.67808,
+      "grad_norm": 1.2142690420150757,
+      "learning_rate": 1.614245698279312e-05,
+      "loss": 0.5728,
+      "step": 16936
+    },
+    {
+      "epoch": 21.67936,
+      "grad_norm": 1.0875095129013062,
+      "learning_rate": 1.614045618247299e-05,
+      "loss": 0.4926,
+      "step": 16937
+    },
+    {
+      "epoch": 21.68064,
+      "grad_norm": 1.092563509941101,
+      "learning_rate": 1.613845538215286e-05,
+      "loss": 0.4254,
+      "step": 16938
+    },
+    {
+      "epoch": 21.68192,
+      "grad_norm": 1.0964988470077515,
+      "learning_rate": 1.6136454581832733e-05,
+      "loss": 0.4502,
+      "step": 16939
+    },
+    {
+      "epoch": 21.6832,
+      "grad_norm": 1.0522211790084839,
+      "learning_rate": 1.613445378151261e-05,
+      "loss": 0.4743,
+      "step": 16940
+    },
+    {
+      "epoch": 21.68448,
+      "grad_norm": 1.0800302028656006,
+      "learning_rate": 1.6132452981192477e-05,
+      "loss": 0.4934,
+      "step": 16941
+    },
+    {
+      "epoch": 21.68576,
+      "grad_norm": 1.1527082920074463,
+      "learning_rate": 1.613045218087235e-05,
+      "loss": 0.5126,
+      "step": 16942
+    },
+    {
+      "epoch": 21.68704,
+      "grad_norm": 1.0669933557510376,
+      "learning_rate": 1.6128451380552224e-05,
+      "loss": 0.4885,
+      "step": 16943
+    },
+    {
+      "epoch": 21.68832,
+      "grad_norm": 1.1759413480758667,
+      "learning_rate": 1.6126450580232096e-05,
+      "loss": 0.5232,
+      "step": 16944
+    },
+    {
+      "epoch": 21.6896,
+      "grad_norm": 1.2183908224105835,
+      "learning_rate": 1.6124449779911964e-05,
+      "loss": 0.5372,
+      "step": 16945
+    },
+    {
+      "epoch": 21.69088,
+      "grad_norm": 1.1375067234039307,
+      "learning_rate": 1.6122448979591836e-05,
+      "loss": 0.5206,
+      "step": 16946
+    },
+    {
+      "epoch": 21.69216,
+      "grad_norm": 1.2400755882263184,
+      "learning_rate": 1.612044817927171e-05,
+      "loss": 0.5103,
+      "step": 16947
+    },
+    {
+      "epoch": 21.69344,
+      "grad_norm": 1.172790288925171,
+      "learning_rate": 1.6118447378951583e-05,
+      "loss": 0.4674,
+      "step": 16948
+    },
+    {
+      "epoch": 21.69472,
+      "grad_norm": 1.0851976871490479,
+      "learning_rate": 1.6116446578631452e-05,
+      "loss": 0.4606,
+      "step": 16949
+    },
+    {
+      "epoch": 21.696,
+      "grad_norm": 1.1035497188568115,
+      "learning_rate": 1.6114445778311324e-05,
+      "loss": 0.4504,
+      "step": 16950
+    },
+    {
+      "epoch": 21.69728,
+      "grad_norm": 1.1531620025634766,
+      "learning_rate": 1.61124449779912e-05,
+      "loss": 0.5273,
+      "step": 16951
+    },
+    {
+      "epoch": 21.69856,
+      "grad_norm": 1.0832749605178833,
+      "learning_rate": 1.611044417767107e-05,
+      "loss": 0.4824,
+      "step": 16952
+    },
+    {
+      "epoch": 21.699840000000002,
+      "grad_norm": 1.1482528448104858,
+      "learning_rate": 1.610844337735094e-05,
+      "loss": 0.5408,
+      "step": 16953
+    },
+    {
+      "epoch": 21.70112,
+      "grad_norm": 1.0921155214309692,
+      "learning_rate": 1.6106442577030814e-05,
+      "loss": 0.4788,
+      "step": 16954
+    },
+    {
+      "epoch": 21.7024,
+      "grad_norm": 1.0282621383666992,
+      "learning_rate": 1.6104441776710686e-05,
+      "loss": 0.422,
+      "step": 16955
+    },
+    {
+      "epoch": 21.70368,
+      "grad_norm": 1.1054280996322632,
+      "learning_rate": 1.6102440976390558e-05,
+      "loss": 0.4924,
+      "step": 16956
+    },
+    {
+      "epoch": 21.70496,
+      "grad_norm": 1.1694878339767456,
+      "learning_rate": 1.6100440176070427e-05,
+      "loss": 0.5017,
+      "step": 16957
+    },
+    {
+      "epoch": 21.70624,
+      "grad_norm": 1.0920751094818115,
+      "learning_rate": 1.6098439375750302e-05,
+      "loss": 0.5053,
+      "step": 16958
+    },
+    {
+      "epoch": 21.70752,
+      "grad_norm": 1.0558879375457764,
+      "learning_rate": 1.6096438575430174e-05,
+      "loss": 0.4535,
+      "step": 16959
+    },
+    {
+      "epoch": 21.7088,
+      "grad_norm": 1.084283709526062,
+      "learning_rate": 1.6094437775110046e-05,
+      "loss": 0.4792,
+      "step": 16960
+    },
+    {
+      "epoch": 21.71008,
+      "grad_norm": 1.0939773321151733,
+      "learning_rate": 1.6092436974789917e-05,
+      "loss": 0.466,
+      "step": 16961
+    },
+    {
+      "epoch": 21.71136,
+      "grad_norm": 1.151667594909668,
+      "learning_rate": 1.609043617446979e-05,
+      "loss": 0.5377,
+      "step": 16962
+    },
+    {
+      "epoch": 21.71264,
+      "grad_norm": 1.173499345779419,
+      "learning_rate": 1.608843537414966e-05,
+      "loss": 0.5031,
+      "step": 16963
+    },
+    {
+      "epoch": 21.71392,
+      "grad_norm": 1.102555751800537,
+      "learning_rate": 1.6086434573829533e-05,
+      "loss": 0.457,
+      "step": 16964
+    },
+    {
+      "epoch": 21.7152,
+      "grad_norm": 1.1618655920028687,
+      "learning_rate": 1.6084433773509405e-05,
+      "loss": 0.513,
+      "step": 16965
+    },
+    {
+      "epoch": 21.71648,
+      "grad_norm": 1.0982638597488403,
+      "learning_rate": 1.6082432973189277e-05,
+      "loss": 0.4711,
+      "step": 16966
+    },
+    {
+      "epoch": 21.71776,
+      "grad_norm": 1.179809808731079,
+      "learning_rate": 1.608043217286915e-05,
+      "loss": 0.5369,
+      "step": 16967
+    },
+    {
+      "epoch": 21.71904,
+      "grad_norm": 1.1830394268035889,
+      "learning_rate": 1.607843137254902e-05,
+      "loss": 0.5041,
+      "step": 16968
+    },
+    {
+      "epoch": 21.72032,
+      "grad_norm": 1.1552484035491943,
+      "learning_rate": 1.6076430572228892e-05,
+      "loss": 0.5089,
+      "step": 16969
+    },
+    {
+      "epoch": 21.7216,
+      "grad_norm": 1.0561984777450562,
+      "learning_rate": 1.6074429771908764e-05,
+      "loss": 0.4571,
+      "step": 16970
+    },
+    {
+      "epoch": 21.72288,
+      "grad_norm": 1.2352055311203003,
+      "learning_rate": 1.6072428971588636e-05,
+      "loss": 0.5468,
+      "step": 16971
+    },
+    {
+      "epoch": 21.72416,
+      "grad_norm": 1.0787138938903809,
+      "learning_rate": 1.6070428171268508e-05,
+      "loss": 0.4551,
+      "step": 16972
+    },
+    {
+      "epoch": 21.72544,
+      "grad_norm": 1.0198187828063965,
+      "learning_rate": 1.606842737094838e-05,
+      "loss": 0.4421,
+      "step": 16973
+    },
+    {
+      "epoch": 21.72672,
+      "grad_norm": 1.1738202571868896,
+      "learning_rate": 1.606642657062825e-05,
+      "loss": 0.5602,
+      "step": 16974
+    },
+    {
+      "epoch": 21.728,
+      "grad_norm": 1.08427095413208,
+      "learning_rate": 1.6064425770308127e-05,
+      "loss": 0.5051,
+      "step": 16975
+    },
+    {
+      "epoch": 21.72928,
+      "grad_norm": 1.0601693391799927,
+      "learning_rate": 1.6062424969987995e-05,
+      "loss": 0.4876,
+      "step": 16976
+    },
+    {
+      "epoch": 21.73056,
+      "grad_norm": 1.016387701034546,
+      "learning_rate": 1.6060424169667867e-05,
+      "loss": 0.4689,
+      "step": 16977
+    },
+    {
+      "epoch": 21.73184,
+      "grad_norm": 1.1087530851364136,
+      "learning_rate": 1.605842336934774e-05,
+      "loss": 0.5092,
+      "step": 16978
+    },
+    {
+      "epoch": 21.73312,
+      "grad_norm": 1.0978829860687256,
+      "learning_rate": 1.6056422569027614e-05,
+      "loss": 0.4665,
+      "step": 16979
+    },
+    {
+      "epoch": 21.7344,
+      "grad_norm": 1.1613261699676514,
+      "learning_rate": 1.6054421768707483e-05,
+      "loss": 0.5061,
+      "step": 16980
+    },
+    {
+      "epoch": 21.73568,
+      "grad_norm": 1.1227811574935913,
+      "learning_rate": 1.6052420968387355e-05,
+      "loss": 0.5022,
+      "step": 16981
+    },
+    {
+      "epoch": 21.73696,
+      "grad_norm": 1.1268730163574219,
+      "learning_rate": 1.605042016806723e-05,
+      "loss": 0.4788,
+      "step": 16982
+    },
+    {
+      "epoch": 21.73824,
+      "grad_norm": 1.1465482711791992,
+      "learning_rate": 1.6048419367747102e-05,
+      "loss": 0.4893,
+      "step": 16983
+    },
+    {
+      "epoch": 21.73952,
+      "grad_norm": 1.1241455078125,
+      "learning_rate": 1.604641856742697e-05,
+      "loss": 0.5033,
+      "step": 16984
+    },
+    {
+      "epoch": 21.7408,
+      "grad_norm": 1.1142101287841797,
+      "learning_rate": 1.6044417767106842e-05,
+      "loss": 0.5004,
+      "step": 16985
+    },
+    {
+      "epoch": 21.74208,
+      "grad_norm": 1.125550389289856,
+      "learning_rate": 1.6042416966786717e-05,
+      "loss": 0.5053,
+      "step": 16986
+    },
+    {
+      "epoch": 21.74336,
+      "grad_norm": 1.0979461669921875,
+      "learning_rate": 1.604041616646659e-05,
+      "loss": 0.514,
+      "step": 16987
+    },
+    {
+      "epoch": 21.74464,
+      "grad_norm": 1.0835514068603516,
+      "learning_rate": 1.6038415366146458e-05,
+      "loss": 0.4502,
+      "step": 16988
+    },
+    {
+      "epoch": 21.74592,
+      "grad_norm": 1.1370912790298462,
+      "learning_rate": 1.6036414565826333e-05,
+      "loss": 0.5205,
+      "step": 16989
+    },
+    {
+      "epoch": 21.7472,
+      "grad_norm": 1.076702356338501,
+      "learning_rate": 1.6034413765506205e-05,
+      "loss": 0.4769,
+      "step": 16990
+    },
+    {
+      "epoch": 21.74848,
+      "grad_norm": 1.1314976215362549,
+      "learning_rate": 1.6032412965186077e-05,
+      "loss": 0.473,
+      "step": 16991
+    },
+    {
+      "epoch": 21.74976,
+      "grad_norm": 1.1105777025222778,
+      "learning_rate": 1.6030412164865945e-05,
+      "loss": 0.5232,
+      "step": 16992
+    },
+    {
+      "epoch": 21.75104,
+      "grad_norm": 1.0831820964813232,
+      "learning_rate": 1.602841136454582e-05,
+      "loss": 0.4784,
+      "step": 16993
+    },
+    {
+      "epoch": 21.75232,
+      "grad_norm": 1.2100034952163696,
+      "learning_rate": 1.6026410564225692e-05,
+      "loss": 0.4977,
+      "step": 16994
+    },
+    {
+      "epoch": 21.7536,
+      "grad_norm": 1.1598615646362305,
+      "learning_rate": 1.6024409763905564e-05,
+      "loss": 0.5115,
+      "step": 16995
+    },
+    {
+      "epoch": 21.75488,
+      "grad_norm": 1.0910362005233765,
+      "learning_rate": 1.6022408963585436e-05,
+      "loss": 0.5047,
+      "step": 16996
+    },
+    {
+      "epoch": 21.75616,
+      "grad_norm": 1.12237548828125,
+      "learning_rate": 1.6020408163265308e-05,
+      "loss": 0.4748,
+      "step": 16997
+    },
+    {
+      "epoch": 21.75744,
+      "grad_norm": 1.175667643547058,
+      "learning_rate": 1.601840736294518e-05,
+      "loss": 0.5035,
+      "step": 16998
+    },
+    {
+      "epoch": 21.75872,
+      "grad_norm": 1.1076536178588867,
+      "learning_rate": 1.601640656262505e-05,
+      "loss": 0.5181,
+      "step": 16999
+    },
+    {
+      "epoch": 21.76,
+      "grad_norm": 1.105377197265625,
+      "learning_rate": 1.6014405762304923e-05,
+      "loss": 0.4939,
+      "step": 17000
+    },
+    {
+      "epoch": 21.76128,
+      "grad_norm": 1.052820086479187,
+      "learning_rate": 1.6012404961984795e-05,
+      "loss": 0.4644,
+      "step": 17001
+    },
+    {
+      "epoch": 21.76256,
+      "grad_norm": 1.0277769565582275,
+      "learning_rate": 1.6010404161664667e-05,
+      "loss": 0.4438,
+      "step": 17002
+    },
+    {
+      "epoch": 21.76384,
+      "grad_norm": 1.1701537370681763,
+      "learning_rate": 1.600840336134454e-05,
+      "loss": 0.5511,
+      "step": 17003
+    },
+    {
+      "epoch": 21.76512,
+      "grad_norm": 1.1238141059875488,
+      "learning_rate": 1.600640256102441e-05,
+      "loss": 0.5343,
+      "step": 17004
+    },
+    {
+      "epoch": 21.7664,
+      "grad_norm": 1.1096354722976685,
+      "learning_rate": 1.6004401760704283e-05,
+      "loss": 0.576,
+      "step": 17005
+    },
+    {
+      "epoch": 21.76768,
+      "grad_norm": 1.1454472541809082,
+      "learning_rate": 1.6002400960384154e-05,
+      "loss": 0.5276,
+      "step": 17006
+    },
+    {
+      "epoch": 21.76896,
+      "grad_norm": 1.095155954360962,
+      "learning_rate": 1.6000400160064026e-05,
+      "loss": 0.4968,
+      "step": 17007
+    },
+    {
+      "epoch": 21.77024,
+      "grad_norm": 1.0994304418563843,
+      "learning_rate": 1.5998399359743898e-05,
+      "loss": 0.4959,
+      "step": 17008
+    },
+    {
+      "epoch": 21.77152,
+      "grad_norm": 1.117372989654541,
+      "learning_rate": 1.599639855942377e-05,
+      "loss": 0.5205,
+      "step": 17009
+    },
+    {
+      "epoch": 21.7728,
+      "grad_norm": 1.0571292638778687,
+      "learning_rate": 1.5994397759103642e-05,
+      "loss": 0.4693,
+      "step": 17010
+    },
+    {
+      "epoch": 21.77408,
+      "grad_norm": 1.0867185592651367,
+      "learning_rate": 1.5992396958783514e-05,
+      "loss": 0.4867,
+      "step": 17011
+    },
+    {
+      "epoch": 21.77536,
+      "grad_norm": 1.0914740562438965,
+      "learning_rate": 1.5990396158463386e-05,
+      "loss": 0.4593,
+      "step": 17012
+    },
+    {
+      "epoch": 21.77664,
+      "grad_norm": 1.1249064207077026,
+      "learning_rate": 1.5988395358143257e-05,
+      "loss": 0.4779,
+      "step": 17013
+    },
+    {
+      "epoch": 21.77792,
+      "grad_norm": 1.1130808591842651,
+      "learning_rate": 1.5986394557823133e-05,
+      "loss": 0.4954,
+      "step": 17014
+    },
+    {
+      "epoch": 21.7792,
+      "grad_norm": 1.0818630456924438,
+      "learning_rate": 1.5984393757503e-05,
+      "loss": 0.4842,
+      "step": 17015
+    },
+    {
+      "epoch": 21.78048,
+      "grad_norm": 1.1350693702697754,
+      "learning_rate": 1.5982392957182873e-05,
+      "loss": 0.4944,
+      "step": 17016
+    },
+    {
+      "epoch": 21.78176,
+      "grad_norm": 1.1261863708496094,
+      "learning_rate": 1.5980392156862745e-05,
+      "loss": 0.4952,
+      "step": 17017
+    },
+    {
+      "epoch": 21.78304,
+      "grad_norm": 1.1032289266586304,
+      "learning_rate": 1.597839135654262e-05,
+      "loss": 0.505,
+      "step": 17018
+    },
+    {
+      "epoch": 21.78432,
+      "grad_norm": 1.1199618577957153,
+      "learning_rate": 1.597639055622249e-05,
+      "loss": 0.4496,
+      "step": 17019
+    },
+    {
+      "epoch": 21.7856,
+      "grad_norm": 1.105373740196228,
+      "learning_rate": 1.597438975590236e-05,
+      "loss": 0.5072,
+      "step": 17020
+    },
+    {
+      "epoch": 21.78688,
+      "grad_norm": 1.0840733051300049,
+      "learning_rate": 1.5972388955582236e-05,
+      "loss": 0.5159,
+      "step": 17021
+    },
+    {
+      "epoch": 21.78816,
+      "grad_norm": 1.0092965364456177,
+      "learning_rate": 1.5970388155262107e-05,
+      "loss": 0.4299,
+      "step": 17022
+    },
+    {
+      "epoch": 21.78944,
+      "grad_norm": 1.0409196615219116,
+      "learning_rate": 1.5968387354941976e-05,
+      "loss": 0.4441,
+      "step": 17023
+    },
+    {
+      "epoch": 21.79072,
+      "grad_norm": 1.1237512826919556,
+      "learning_rate": 1.5966386554621848e-05,
+      "loss": 0.5225,
+      "step": 17024
+    },
+    {
+      "epoch": 21.792,
+      "grad_norm": 1.1367634534835815,
+      "learning_rate": 1.5964385754301723e-05,
+      "loss": 0.4826,
+      "step": 17025
+    },
+    {
+      "epoch": 21.79328,
+      "grad_norm": 1.146195650100708,
+      "learning_rate": 1.5962384953981595e-05,
+      "loss": 0.5016,
+      "step": 17026
+    },
+    {
+      "epoch": 21.79456,
+      "grad_norm": 1.0935816764831543,
+      "learning_rate": 1.5960384153661463e-05,
+      "loss": 0.4626,
+      "step": 17027
+    },
+    {
+      "epoch": 21.79584,
+      "grad_norm": 1.169556975364685,
+      "learning_rate": 1.595838335334134e-05,
+      "loss": 0.5184,
+      "step": 17028
+    },
+    {
+      "epoch": 21.79712,
+      "grad_norm": 1.1683391332626343,
+      "learning_rate": 1.595638255302121e-05,
+      "loss": 0.5,
+      "step": 17029
+    },
+    {
+      "epoch": 21.7984,
+      "grad_norm": 1.1834429502487183,
+      "learning_rate": 1.5954381752701082e-05,
+      "loss": 0.5633,
+      "step": 17030
+    },
+    {
+      "epoch": 21.79968,
+      "grad_norm": 1.1966886520385742,
+      "learning_rate": 1.595238095238095e-05,
+      "loss": 0.5166,
+      "step": 17031
+    },
+    {
+      "epoch": 21.80096,
+      "grad_norm": 1.1592307090759277,
+      "learning_rate": 1.5950380152060826e-05,
+      "loss": 0.5007,
+      "step": 17032
+    },
+    {
+      "epoch": 21.80224,
+      "grad_norm": 1.0894125699996948,
+      "learning_rate": 1.5948379351740698e-05,
+      "loss": 0.4405,
+      "step": 17033
+    },
+    {
+      "epoch": 21.80352,
+      "grad_norm": 1.1028647422790527,
+      "learning_rate": 1.594637855142057e-05,
+      "loss": 0.4736,
+      "step": 17034
+    },
+    {
+      "epoch": 21.8048,
+      "grad_norm": 1.01607084274292,
+      "learning_rate": 1.594437775110044e-05,
+      "loss": 0.4253,
+      "step": 17035
+    },
+    {
+      "epoch": 21.80608,
+      "grad_norm": 1.1347633600234985,
+      "learning_rate": 1.5942376950780313e-05,
+      "loss": 0.5113,
+      "step": 17036
+    },
+    {
+      "epoch": 21.80736,
+      "grad_norm": 1.1249445676803589,
+      "learning_rate": 1.5940376150460185e-05,
+      "loss": 0.4994,
+      "step": 17037
+    },
+    {
+      "epoch": 21.80864,
+      "grad_norm": 1.192696452140808,
+      "learning_rate": 1.5938375350140057e-05,
+      "loss": 0.5374,
+      "step": 17038
+    },
+    {
+      "epoch": 21.809919999999998,
+      "grad_norm": 1.092063307762146,
+      "learning_rate": 1.593637454981993e-05,
+      "loss": 0.4376,
+      "step": 17039
+    },
+    {
+      "epoch": 21.8112,
+      "grad_norm": 1.0697550773620605,
+      "learning_rate": 1.59343737494998e-05,
+      "loss": 0.4639,
+      "step": 17040
+    },
+    {
+      "epoch": 21.81248,
+      "grad_norm": 1.066964864730835,
+      "learning_rate": 1.5932372949179673e-05,
+      "loss": 0.4451,
+      "step": 17041
+    },
+    {
+      "epoch": 21.81376,
+      "grad_norm": 1.099763035774231,
+      "learning_rate": 1.5930372148859545e-05,
+      "loss": 0.4918,
+      "step": 17042
+    },
+    {
+      "epoch": 21.81504,
+      "grad_norm": 1.1521753072738647,
+      "learning_rate": 1.5928371348539416e-05,
+      "loss": 0.5071,
+      "step": 17043
+    },
+    {
+      "epoch": 21.81632,
+      "grad_norm": 1.0961254835128784,
+      "learning_rate": 1.592637054821929e-05,
+      "loss": 0.496,
+      "step": 17044
+    },
+    {
+      "epoch": 21.8176,
+      "grad_norm": 1.0640747547149658,
+      "learning_rate": 1.592436974789916e-05,
+      "loss": 0.4542,
+      "step": 17045
+    },
+    {
+      "epoch": 21.81888,
+      "grad_norm": 1.2090388536453247,
+      "learning_rate": 1.5922368947579032e-05,
+      "loss": 0.5355,
+      "step": 17046
+    },
+    {
+      "epoch": 21.82016,
+      "grad_norm": 1.0608830451965332,
+      "learning_rate": 1.5920368147258904e-05,
+      "loss": 0.4686,
+      "step": 17047
+    },
+    {
+      "epoch": 21.82144,
+      "grad_norm": 1.1006875038146973,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 0.5012,
+      "step": 17048
+    },
+    {
+      "epoch": 21.82272,
+      "grad_norm": 1.132010817527771,
+      "learning_rate": 1.591636654661865e-05,
+      "loss": 0.5102,
+      "step": 17049
+    },
+    {
+      "epoch": 21.824,
+      "grad_norm": 1.0770601034164429,
+      "learning_rate": 1.591436574629852e-05,
+      "loss": 0.4736,
+      "step": 17050
+    },
+    {
+      "epoch": 21.82528,
+      "grad_norm": 1.1132420301437378,
+      "learning_rate": 1.591236494597839e-05,
+      "loss": 0.512,
+      "step": 17051
+    },
+    {
+      "epoch": 21.82656,
+      "grad_norm": 1.1608811616897583,
+      "learning_rate": 1.5910364145658263e-05,
+      "loss": 0.4529,
+      "step": 17052
+    },
+    {
+      "epoch": 21.82784,
+      "grad_norm": 1.1335759162902832,
+      "learning_rate": 1.590836334533814e-05,
+      "loss": 0.501,
+      "step": 17053
+    },
+    {
+      "epoch": 21.82912,
+      "grad_norm": 1.0787115097045898,
+      "learning_rate": 1.5906362545018007e-05,
+      "loss": 0.478,
+      "step": 17054
+    },
+    {
+      "epoch": 21.8304,
+      "grad_norm": 1.0791386365890503,
+      "learning_rate": 1.590436174469788e-05,
+      "loss": 0.4745,
+      "step": 17055
+    },
+    {
+      "epoch": 21.83168,
+      "grad_norm": 1.0508241653442383,
+      "learning_rate": 1.590236094437775e-05,
+      "loss": 0.4423,
+      "step": 17056
+    },
+    {
+      "epoch": 21.83296,
+      "grad_norm": 1.061378836631775,
+      "learning_rate": 1.5900360144057626e-05,
+      "loss": 0.4637,
+      "step": 17057
+    },
+    {
+      "epoch": 21.83424,
+      "grad_norm": 1.1394723653793335,
+      "learning_rate": 1.5898359343737494e-05,
+      "loss": 0.4634,
+      "step": 17058
+    },
+    {
+      "epoch": 21.83552,
+      "grad_norm": 1.1414905786514282,
+      "learning_rate": 1.5896358543417366e-05,
+      "loss": 0.4838,
+      "step": 17059
+    },
+    {
+      "epoch": 21.8368,
+      "grad_norm": 1.127100944519043,
+      "learning_rate": 1.589435774309724e-05,
+      "loss": 0.5292,
+      "step": 17060
+    },
+    {
+      "epoch": 21.83808,
+      "grad_norm": 1.1181409358978271,
+      "learning_rate": 1.5892356942777113e-05,
+      "loss": 0.5023,
+      "step": 17061
+    },
+    {
+      "epoch": 21.83936,
+      "grad_norm": 1.1021265983581543,
+      "learning_rate": 1.5890356142456982e-05,
+      "loss": 0.4538,
+      "step": 17062
+    },
+    {
+      "epoch": 21.84064,
+      "grad_norm": 1.042770504951477,
+      "learning_rate": 1.5888355342136854e-05,
+      "loss": 0.4804,
+      "step": 17063
+    },
+    {
+      "epoch": 21.841920000000002,
+      "grad_norm": 1.1434727907180786,
+      "learning_rate": 1.588635454181673e-05,
+      "loss": 0.5127,
+      "step": 17064
+    },
+    {
+      "epoch": 21.8432,
+      "grad_norm": 1.111856460571289,
+      "learning_rate": 1.58843537414966e-05,
+      "loss": 0.5108,
+      "step": 17065
+    },
+    {
+      "epoch": 21.84448,
+      "grad_norm": 1.0827891826629639,
+      "learning_rate": 1.588235294117647e-05,
+      "loss": 0.4725,
+      "step": 17066
+    },
+    {
+      "epoch": 21.84576,
+      "grad_norm": 1.1117992401123047,
+      "learning_rate": 1.5880352140856344e-05,
+      "loss": 0.474,
+      "step": 17067
+    },
+    {
+      "epoch": 21.84704,
+      "grad_norm": 1.1064029932022095,
+      "learning_rate": 1.5878351340536216e-05,
+      "loss": 0.4948,
+      "step": 17068
+    },
+    {
+      "epoch": 21.84832,
+      "grad_norm": 1.0790382623672485,
+      "learning_rate": 1.5876350540216088e-05,
+      "loss": 0.5079,
+      "step": 17069
+    },
+    {
+      "epoch": 21.8496,
+      "grad_norm": 1.1318869590759277,
+      "learning_rate": 1.5874349739895957e-05,
+      "loss": 0.5018,
+      "step": 17070
+    },
+    {
+      "epoch": 21.85088,
+      "grad_norm": 1.1790701150894165,
+      "learning_rate": 1.5872348939575832e-05,
+      "loss": 0.5168,
+      "step": 17071
+    },
+    {
+      "epoch": 21.85216,
+      "grad_norm": 1.1890596151351929,
+      "learning_rate": 1.5870348139255704e-05,
+      "loss": 0.5259,
+      "step": 17072
+    },
+    {
+      "epoch": 21.85344,
+      "grad_norm": 1.1023216247558594,
+      "learning_rate": 1.5868347338935576e-05,
+      "loss": 0.5027,
+      "step": 17073
+    },
+    {
+      "epoch": 21.85472,
+      "grad_norm": 1.2085516452789307,
+      "learning_rate": 1.5866346538615447e-05,
+      "loss": 0.5627,
+      "step": 17074
+    },
+    {
+      "epoch": 21.856,
+      "grad_norm": 1.0791484117507935,
+      "learning_rate": 1.586434573829532e-05,
+      "loss": 0.5077,
+      "step": 17075
+    },
+    {
+      "epoch": 21.85728,
+      "grad_norm": 1.1096769571304321,
+      "learning_rate": 1.586234493797519e-05,
+      "loss": 0.4913,
+      "step": 17076
+    },
+    {
+      "epoch": 21.85856,
+      "grad_norm": 1.1315524578094482,
+      "learning_rate": 1.5860344137655063e-05,
+      "loss": 0.4864,
+      "step": 17077
+    },
+    {
+      "epoch": 21.85984,
+      "grad_norm": 1.126545786857605,
+      "learning_rate": 1.5858343337334935e-05,
+      "loss": 0.4946,
+      "step": 17078
+    },
+    {
+      "epoch": 21.86112,
+      "grad_norm": 1.0645451545715332,
+      "learning_rate": 1.5856342537014807e-05,
+      "loss": 0.4946,
+      "step": 17079
+    },
+    {
+      "epoch": 21.8624,
+      "grad_norm": 1.1454330682754517,
+      "learning_rate": 1.585434173669468e-05,
+      "loss": 0.4943,
+      "step": 17080
+    },
+    {
+      "epoch": 21.86368,
+      "grad_norm": 1.156087040901184,
+      "learning_rate": 1.585234093637455e-05,
+      "loss": 0.493,
+      "step": 17081
+    },
+    {
+      "epoch": 21.86496,
+      "grad_norm": 1.093726634979248,
+      "learning_rate": 1.5850340136054422e-05,
+      "loss": 0.4829,
+      "step": 17082
+    },
+    {
+      "epoch": 21.86624,
+      "grad_norm": 1.1023703813552856,
+      "learning_rate": 1.5848339335734294e-05,
+      "loss": 0.5139,
+      "step": 17083
+    },
+    {
+      "epoch": 21.86752,
+      "grad_norm": 1.0602704286575317,
+      "learning_rate": 1.5846338535414166e-05,
+      "loss": 0.4675,
+      "step": 17084
+    },
+    {
+      "epoch": 21.8688,
+      "grad_norm": 1.1112686395645142,
+      "learning_rate": 1.5844337735094038e-05,
+      "loss": 0.4874,
+      "step": 17085
+    },
+    {
+      "epoch": 21.87008,
+      "grad_norm": 1.07467520236969,
+      "learning_rate": 1.584233693477391e-05,
+      "loss": 0.4669,
+      "step": 17086
+    },
+    {
+      "epoch": 21.87136,
+      "grad_norm": 1.1608061790466309,
+      "learning_rate": 1.584033613445378e-05,
+      "loss": 0.5248,
+      "step": 17087
+    },
+    {
+      "epoch": 21.87264,
+      "grad_norm": 1.1059867143630981,
+      "learning_rate": 1.5838335334133657e-05,
+      "loss": 0.5052,
+      "step": 17088
+    },
+    {
+      "epoch": 21.87392,
+      "grad_norm": 1.120044231414795,
+      "learning_rate": 1.5836334533813525e-05,
+      "loss": 0.4999,
+      "step": 17089
+    },
+    {
+      "epoch": 21.8752,
+      "grad_norm": 1.1224958896636963,
+      "learning_rate": 1.5834333733493397e-05,
+      "loss": 0.5023,
+      "step": 17090
+    },
+    {
+      "epoch": 21.87648,
+      "grad_norm": 1.1289787292480469,
+      "learning_rate": 1.583233293317327e-05,
+      "loss": 0.5192,
+      "step": 17091
+    },
+    {
+      "epoch": 21.87776,
+      "grad_norm": 1.0791914463043213,
+      "learning_rate": 1.5830332132853144e-05,
+      "loss": 0.4934,
+      "step": 17092
+    },
+    {
+      "epoch": 21.87904,
+      "grad_norm": 1.1275081634521484,
+      "learning_rate": 1.5828331332533013e-05,
+      "loss": 0.5088,
+      "step": 17093
+    },
+    {
+      "epoch": 21.88032,
+      "grad_norm": 1.1475367546081543,
+      "learning_rate": 1.5826330532212885e-05,
+      "loss": 0.5101,
+      "step": 17094
+    },
+    {
+      "epoch": 21.8816,
+      "grad_norm": 1.1276719570159912,
+      "learning_rate": 1.582432973189276e-05,
+      "loss": 0.5062,
+      "step": 17095
+    },
+    {
+      "epoch": 21.88288,
+      "grad_norm": 1.0605647563934326,
+      "learning_rate": 1.582232893157263e-05,
+      "loss": 0.465,
+      "step": 17096
+    },
+    {
+      "epoch": 21.88416,
+      "grad_norm": 1.1123402118682861,
+      "learning_rate": 1.58203281312525e-05,
+      "loss": 0.5084,
+      "step": 17097
+    },
+    {
+      "epoch": 21.88544,
+      "grad_norm": 1.1372113227844238,
+      "learning_rate": 1.5818327330932372e-05,
+      "loss": 0.5107,
+      "step": 17098
+    },
+    {
+      "epoch": 21.88672,
+      "grad_norm": 1.104132890701294,
+      "learning_rate": 1.5816326530612247e-05,
+      "loss": 0.4793,
+      "step": 17099
+    },
+    {
+      "epoch": 21.888,
+      "grad_norm": 1.0645356178283691,
+      "learning_rate": 1.581432573029212e-05,
+      "loss": 0.4723,
+      "step": 17100
+    },
+    {
+      "epoch": 21.88928,
+      "grad_norm": 1.1151045560836792,
+      "learning_rate": 1.5812324929971988e-05,
+      "loss": 0.4642,
+      "step": 17101
+    },
+    {
+      "epoch": 21.89056,
+      "grad_norm": 1.1927440166473389,
+      "learning_rate": 1.5810324129651863e-05,
+      "loss": 0.5306,
+      "step": 17102
+    },
+    {
+      "epoch": 21.89184,
+      "grad_norm": 1.148725986480713,
+      "learning_rate": 1.5808323329331735e-05,
+      "loss": 0.5302,
+      "step": 17103
+    },
+    {
+      "epoch": 21.89312,
+      "grad_norm": 1.1260452270507812,
+      "learning_rate": 1.5806322529011607e-05,
+      "loss": 0.5168,
+      "step": 17104
+    },
+    {
+      "epoch": 21.8944,
+      "grad_norm": 1.1221392154693604,
+      "learning_rate": 1.5804321728691475e-05,
+      "loss": 0.5141,
+      "step": 17105
+    },
+    {
+      "epoch": 21.89568,
+      "grad_norm": 1.172025442123413,
+      "learning_rate": 1.580232092837135e-05,
+      "loss": 0.5186,
+      "step": 17106
+    },
+    {
+      "epoch": 21.89696,
+      "grad_norm": 1.1051148176193237,
+      "learning_rate": 1.5800320128051222e-05,
+      "loss": 0.4797,
+      "step": 17107
+    },
+    {
+      "epoch": 21.89824,
+      "grad_norm": 1.0419796705245972,
+      "learning_rate": 1.5798319327731094e-05,
+      "loss": 0.4483,
+      "step": 17108
+    },
+    {
+      "epoch": 21.89952,
+      "grad_norm": 1.0614566802978516,
+      "learning_rate": 1.5796318527410966e-05,
+      "loss": 0.4695,
+      "step": 17109
+    },
+    {
+      "epoch": 21.9008,
+      "grad_norm": 1.111011266708374,
+      "learning_rate": 1.5794317727090838e-05,
+      "loss": 0.4986,
+      "step": 17110
+    },
+    {
+      "epoch": 21.90208,
+      "grad_norm": 1.1020454168319702,
+      "learning_rate": 1.579231692677071e-05,
+      "loss": 0.5104,
+      "step": 17111
+    },
+    {
+      "epoch": 21.90336,
+      "grad_norm": 1.1246579885482788,
+      "learning_rate": 1.579031612645058e-05,
+      "loss": 0.5063,
+      "step": 17112
+    },
+    {
+      "epoch": 21.90464,
+      "grad_norm": 1.0644440650939941,
+      "learning_rate": 1.5788315326130453e-05,
+      "loss": 0.4876,
+      "step": 17113
+    },
+    {
+      "epoch": 21.90592,
+      "grad_norm": 1.0820109844207764,
+      "learning_rate": 1.5786314525810325e-05,
+      "loss": 0.5087,
+      "step": 17114
+    },
+    {
+      "epoch": 21.9072,
+      "grad_norm": 1.0998029708862305,
+      "learning_rate": 1.5784313725490197e-05,
+      "loss": 0.4956,
+      "step": 17115
+    },
+    {
+      "epoch": 21.90848,
+      "grad_norm": 1.1597471237182617,
+      "learning_rate": 1.578231292517007e-05,
+      "loss": 0.5432,
+      "step": 17116
+    },
+    {
+      "epoch": 21.90976,
+      "grad_norm": 1.1102598905563354,
+      "learning_rate": 1.578031212484994e-05,
+      "loss": 0.528,
+      "step": 17117
+    },
+    {
+      "epoch": 21.91104,
+      "grad_norm": 1.1272238492965698,
+      "learning_rate": 1.5778311324529813e-05,
+      "loss": 0.4563,
+      "step": 17118
+    },
+    {
+      "epoch": 21.91232,
+      "grad_norm": 1.0742859840393066,
+      "learning_rate": 1.5776310524209684e-05,
+      "loss": 0.4609,
+      "step": 17119
+    },
+    {
+      "epoch": 21.9136,
+      "grad_norm": 1.1245981454849243,
+      "learning_rate": 1.5774309723889556e-05,
+      "loss": 0.5284,
+      "step": 17120
+    },
+    {
+      "epoch": 21.91488,
+      "grad_norm": 1.1140938997268677,
+      "learning_rate": 1.5772308923569428e-05,
+      "loss": 0.4731,
+      "step": 17121
+    },
+    {
+      "epoch": 21.91616,
+      "grad_norm": 1.1234639883041382,
+      "learning_rate": 1.57703081232493e-05,
+      "loss": 0.4937,
+      "step": 17122
+    },
+    {
+      "epoch": 21.91744,
+      "grad_norm": 1.1589488983154297,
+      "learning_rate": 1.5768307322929172e-05,
+      "loss": 0.5186,
+      "step": 17123
+    },
+    {
+      "epoch": 21.91872,
+      "grad_norm": 1.1331297159194946,
+      "learning_rate": 1.5766306522609044e-05,
+      "loss": 0.4893,
+      "step": 17124
+    },
+    {
+      "epoch": 21.92,
+      "grad_norm": 1.1356956958770752,
+      "learning_rate": 1.5764305722288916e-05,
+      "loss": 0.4674,
+      "step": 17125
+    },
+    {
+      "epoch": 21.92128,
+      "grad_norm": 1.0896925926208496,
+      "learning_rate": 1.5762304921968787e-05,
+      "loss": 0.462,
+      "step": 17126
+    },
+    {
+      "epoch": 21.92256,
+      "grad_norm": 1.1289901733398438,
+      "learning_rate": 1.5760304121648663e-05,
+      "loss": 0.5332,
+      "step": 17127
+    },
+    {
+      "epoch": 21.92384,
+      "grad_norm": 1.0392855405807495,
+      "learning_rate": 1.575830332132853e-05,
+      "loss": 0.4433,
+      "step": 17128
+    },
+    {
+      "epoch": 21.92512,
+      "grad_norm": 1.1507699489593506,
+      "learning_rate": 1.5756302521008403e-05,
+      "loss": 0.5408,
+      "step": 17129
+    },
+    {
+      "epoch": 21.9264,
+      "grad_norm": 1.0881316661834717,
+      "learning_rate": 1.5754301720688275e-05,
+      "loss": 0.5205,
+      "step": 17130
+    },
+    {
+      "epoch": 21.92768,
+      "grad_norm": 1.1345676183700562,
+      "learning_rate": 1.575230092036815e-05,
+      "loss": 0.4812,
+      "step": 17131
+    },
+    {
+      "epoch": 21.92896,
+      "grad_norm": 1.1051517724990845,
+      "learning_rate": 1.575030012004802e-05,
+      "loss": 0.5129,
+      "step": 17132
+    },
+    {
+      "epoch": 21.93024,
+      "grad_norm": 1.1058467626571655,
+      "learning_rate": 1.574829931972789e-05,
+      "loss": 0.4401,
+      "step": 17133
+    },
+    {
+      "epoch": 21.93152,
+      "grad_norm": 1.1275383234024048,
+      "learning_rate": 1.5746298519407766e-05,
+      "loss": 0.5036,
+      "step": 17134
+    },
+    {
+      "epoch": 21.9328,
+      "grad_norm": 1.1110889911651611,
+      "learning_rate": 1.5744297719087637e-05,
+      "loss": 0.4844,
+      "step": 17135
+    },
+    {
+      "epoch": 21.93408,
+      "grad_norm": 1.1405858993530273,
+      "learning_rate": 1.5742296918767506e-05,
+      "loss": 0.5322,
+      "step": 17136
+    },
+    {
+      "epoch": 21.93536,
+      "grad_norm": 1.1357598304748535,
+      "learning_rate": 1.5740296118447378e-05,
+      "loss": 0.4834,
+      "step": 17137
+    },
+    {
+      "epoch": 21.93664,
+      "grad_norm": 1.1561405658721924,
+      "learning_rate": 1.5738295318127253e-05,
+      "loss": 0.5107,
+      "step": 17138
+    },
+    {
+      "epoch": 21.93792,
+      "grad_norm": 1.1202070713043213,
+      "learning_rate": 1.5736294517807125e-05,
+      "loss": 0.5176,
+      "step": 17139
+    },
+    {
+      "epoch": 21.9392,
+      "grad_norm": 1.141701579093933,
+      "learning_rate": 1.5734293717486993e-05,
+      "loss": 0.5304,
+      "step": 17140
+    },
+    {
+      "epoch": 21.94048,
+      "grad_norm": 1.0925021171569824,
+      "learning_rate": 1.573229291716687e-05,
+      "loss": 0.4811,
+      "step": 17141
+    },
+    {
+      "epoch": 21.94176,
+      "grad_norm": 1.147238850593567,
+      "learning_rate": 1.573029211684674e-05,
+      "loss": 0.5093,
+      "step": 17142
+    },
+    {
+      "epoch": 21.94304,
+      "grad_norm": 1.1315157413482666,
+      "learning_rate": 1.5728291316526612e-05,
+      "loss": 0.4744,
+      "step": 17143
+    },
+    {
+      "epoch": 21.94432,
+      "grad_norm": 1.1635738611221313,
+      "learning_rate": 1.572629051620648e-05,
+      "loss": 0.5085,
+      "step": 17144
+    },
+    {
+      "epoch": 21.9456,
+      "grad_norm": 1.1979985237121582,
+      "learning_rate": 1.5724289715886356e-05,
+      "loss": 0.5047,
+      "step": 17145
+    },
+    {
+      "epoch": 21.94688,
+      "grad_norm": 1.1460307836532593,
+      "learning_rate": 1.5722288915566228e-05,
+      "loss": 0.487,
+      "step": 17146
+    },
+    {
+      "epoch": 21.94816,
+      "grad_norm": 1.087184190750122,
+      "learning_rate": 1.57202881152461e-05,
+      "loss": 0.524,
+      "step": 17147
+    },
+    {
+      "epoch": 21.94944,
+      "grad_norm": 1.1512465476989746,
+      "learning_rate": 1.571828731492597e-05,
+      "loss": 0.4737,
+      "step": 17148
+    },
+    {
+      "epoch": 21.95072,
+      "grad_norm": 1.1650768518447876,
+      "learning_rate": 1.5716286514605843e-05,
+      "loss": 0.5084,
+      "step": 17149
+    },
+    {
+      "epoch": 21.951999999999998,
+      "grad_norm": 1.0953246355056763,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.5255,
+      "step": 17150
+    },
+    {
+      "epoch": 21.95328,
+      "grad_norm": 1.0524414777755737,
+      "learning_rate": 1.5712284913965587e-05,
+      "loss": 0.485,
+      "step": 17151
+    },
+    {
+      "epoch": 21.95456,
+      "grad_norm": 1.0316898822784424,
+      "learning_rate": 1.571028411364546e-05,
+      "loss": 0.4774,
+      "step": 17152
+    },
+    {
+      "epoch": 21.95584,
+      "grad_norm": 1.0071184635162354,
+      "learning_rate": 1.570828331332533e-05,
+      "loss": 0.4582,
+      "step": 17153
+    },
+    {
+      "epoch": 21.95712,
+      "grad_norm": 1.1721525192260742,
+      "learning_rate": 1.5706282513005203e-05,
+      "loss": 0.5407,
+      "step": 17154
+    },
+    {
+      "epoch": 21.9584,
+      "grad_norm": 1.0725557804107666,
+      "learning_rate": 1.5704281712685075e-05,
+      "loss": 0.4728,
+      "step": 17155
+    },
+    {
+      "epoch": 21.95968,
+      "grad_norm": 1.0535809993743896,
+      "learning_rate": 1.5702280912364946e-05,
+      "loss": 0.455,
+      "step": 17156
+    },
+    {
+      "epoch": 21.96096,
+      "grad_norm": 1.1127283573150635,
+      "learning_rate": 1.570028011204482e-05,
+      "loss": 0.5051,
+      "step": 17157
+    },
+    {
+      "epoch": 21.96224,
+      "grad_norm": 1.1096875667572021,
+      "learning_rate": 1.569827931172469e-05,
+      "loss": 0.4694,
+      "step": 17158
+    },
+    {
+      "epoch": 21.96352,
+      "grad_norm": 1.105588436126709,
+      "learning_rate": 1.5696278511404562e-05,
+      "loss": 0.4654,
+      "step": 17159
+    },
+    {
+      "epoch": 21.9648,
+      "grad_norm": 1.1018520593643188,
+      "learning_rate": 1.5694277711084434e-05,
+      "loss": 0.4964,
+      "step": 17160
+    },
+    {
+      "epoch": 21.96608,
+      "grad_norm": 1.1417279243469238,
+      "learning_rate": 1.5692276910764306e-05,
+      "loss": 0.5166,
+      "step": 17161
+    },
+    {
+      "epoch": 21.96736,
+      "grad_norm": 1.0698052644729614,
+      "learning_rate": 1.569027611044418e-05,
+      "loss": 0.5072,
+      "step": 17162
+    },
+    {
+      "epoch": 21.96864,
+      "grad_norm": 1.1088814735412598,
+      "learning_rate": 1.568827531012405e-05,
+      "loss": 0.5034,
+      "step": 17163
+    },
+    {
+      "epoch": 21.96992,
+      "grad_norm": 1.0650029182434082,
+      "learning_rate": 1.568627450980392e-05,
+      "loss": 0.4262,
+      "step": 17164
+    },
+    {
+      "epoch": 21.9712,
+      "grad_norm": 1.0559251308441162,
+      "learning_rate": 1.5684273709483793e-05,
+      "loss": 0.4375,
+      "step": 17165
+    },
+    {
+      "epoch": 21.97248,
+      "grad_norm": 1.107661485671997,
+      "learning_rate": 1.568227290916367e-05,
+      "loss": 0.4978,
+      "step": 17166
+    },
+    {
+      "epoch": 21.97376,
+      "grad_norm": 1.109234094619751,
+      "learning_rate": 1.5680272108843537e-05,
+      "loss": 0.5112,
+      "step": 17167
+    },
+    {
+      "epoch": 21.97504,
+      "grad_norm": 1.1099008321762085,
+      "learning_rate": 1.567827130852341e-05,
+      "loss": 0.4588,
+      "step": 17168
+    },
+    {
+      "epoch": 21.97632,
+      "grad_norm": 1.1135979890823364,
+      "learning_rate": 1.567627050820328e-05,
+      "loss": 0.4801,
+      "step": 17169
+    },
+    {
+      "epoch": 21.9776,
+      "grad_norm": 1.1482584476470947,
+      "learning_rate": 1.5674269707883156e-05,
+      "loss": 0.4637,
+      "step": 17170
+    },
+    {
+      "epoch": 21.97888,
+      "grad_norm": 1.0247658491134644,
+      "learning_rate": 1.5672268907563024e-05,
+      "loss": 0.4485,
+      "step": 17171
+    },
+    {
+      "epoch": 21.98016,
+      "grad_norm": 1.0352832078933716,
+      "learning_rate": 1.5670268107242896e-05,
+      "loss": 0.4452,
+      "step": 17172
+    },
+    {
+      "epoch": 21.98144,
+      "grad_norm": 1.1640503406524658,
+      "learning_rate": 1.566826730692277e-05,
+      "loss": 0.5008,
+      "step": 17173
+    },
+    {
+      "epoch": 21.98272,
+      "grad_norm": 1.086647629737854,
+      "learning_rate": 1.5666266506602643e-05,
+      "loss": 0.4714,
+      "step": 17174
+    },
+    {
+      "epoch": 21.984,
+      "grad_norm": 1.1195262670516968,
+      "learning_rate": 1.5664265706282512e-05,
+      "loss": 0.5151,
+      "step": 17175
+    },
+    {
+      "epoch": 21.98528,
+      "grad_norm": 1.1734910011291504,
+      "learning_rate": 1.5662264905962384e-05,
+      "loss": 0.5262,
+      "step": 17176
+    },
+    {
+      "epoch": 21.98656,
+      "grad_norm": 1.1472947597503662,
+      "learning_rate": 1.566026410564226e-05,
+      "loss": 0.4849,
+      "step": 17177
+    },
+    {
+      "epoch": 21.98784,
+      "grad_norm": 1.114270806312561,
+      "learning_rate": 1.565826330532213e-05,
+      "loss": 0.5082,
+      "step": 17178
+    },
+    {
+      "epoch": 21.98912,
+      "grad_norm": 1.0618884563446045,
+      "learning_rate": 1.5656262505002e-05,
+      "loss": 0.4387,
+      "step": 17179
+    },
+    {
+      "epoch": 21.9904,
+      "grad_norm": 1.098699927330017,
+      "learning_rate": 1.5654261704681874e-05,
+      "loss": 0.5168,
+      "step": 17180
+    },
+    {
+      "epoch": 21.99168,
+      "grad_norm": 1.1803953647613525,
+      "learning_rate": 1.5652260904361746e-05,
+      "loss": 0.5424,
+      "step": 17181
+    },
+    {
+      "epoch": 21.99296,
+      "grad_norm": 1.1238080263137817,
+      "learning_rate": 1.5650260104041618e-05,
+      "loss": 0.4951,
+      "step": 17182
+    },
+    {
+      "epoch": 21.99424,
+      "grad_norm": 1.0754061937332153,
+      "learning_rate": 1.5648259303721487e-05,
+      "loss": 0.4725,
+      "step": 17183
+    },
+    {
+      "epoch": 21.99552,
+      "grad_norm": 1.0654562711715698,
+      "learning_rate": 1.5646258503401362e-05,
+      "loss": 0.486,
+      "step": 17184
+    },
+    {
+      "epoch": 21.9968,
+      "grad_norm": 1.1009260416030884,
+      "learning_rate": 1.5644257703081234e-05,
+      "loss": 0.5057,
+      "step": 17185
+    },
+    {
+      "epoch": 21.99808,
+      "grad_norm": 1.091380000114441,
+      "learning_rate": 1.5642256902761106e-05,
+      "loss": 0.4666,
+      "step": 17186
+    },
+    {
+      "epoch": 21.99936,
+      "grad_norm": 1.0999908447265625,
+      "learning_rate": 1.5640256102440977e-05,
+      "loss": 0.4635,
+      "step": 17187
+    },
+    {
+      "epoch": 22.00064,
+      "grad_norm": Infinity,
+      "learning_rate": 1.5640256102440977e-05,
+      "loss": 0.957,
+      "step": 17188
+    },
+    {
+      "epoch": 22.00192,
+      "grad_norm": 1.0367313623428345,
+      "learning_rate": 1.563825530212085e-05,
+      "loss": 0.4574,
+      "step": 17189
+    },
+    {
+      "epoch": 22.0032,
+      "grad_norm": 1.0084816217422485,
+      "learning_rate": 1.563625450180072e-05,
+      "loss": 0.4657,
+      "step": 17190
+    },
+    {
+      "epoch": 22.00448,
+      "grad_norm": 1.096815824508667,
+      "learning_rate": 1.5634253701480593e-05,
+      "loss": 0.4837,
+      "step": 17191
+    },
+    {
+      "epoch": 22.00576,
+      "grad_norm": 1.0867575407028198,
+      "learning_rate": 1.5632252901160465e-05,
+      "loss": 0.4679,
+      "step": 17192
+    },
+    {
+      "epoch": 22.00704,
+      "grad_norm": 1.0765024423599243,
+      "learning_rate": 1.5630252100840337e-05,
+      "loss": 0.4796,
+      "step": 17193
+    },
+    {
+      "epoch": 22.00832,
+      "grad_norm": 1.0676796436309814,
+      "learning_rate": 1.562825130052021e-05,
+      "loss": 0.4513,
+      "step": 17194
+    },
+    {
+      "epoch": 22.0096,
+      "grad_norm": 1.0853677988052368,
+      "learning_rate": 1.562625050020008e-05,
+      "loss": 0.4669,
+      "step": 17195
+    },
+    {
+      "epoch": 22.01088,
+      "grad_norm": 1.1179803609848022,
+      "learning_rate": 1.5624249699879952e-05,
+      "loss": 0.4737,
+      "step": 17196
+    },
+    {
+      "epoch": 22.01216,
+      "grad_norm": 1.1370209455490112,
+      "learning_rate": 1.5622248899559824e-05,
+      "loss": 0.4873,
+      "step": 17197
+    },
+    {
+      "epoch": 22.01344,
+      "grad_norm": 1.0936213731765747,
+      "learning_rate": 1.5620248099239696e-05,
+      "loss": 0.4791,
+      "step": 17198
+    },
+    {
+      "epoch": 22.01472,
+      "grad_norm": 1.1142305135726929,
+      "learning_rate": 1.5618247298919568e-05,
+      "loss": 0.4956,
+      "step": 17199
+    },
+    {
+      "epoch": 22.016,
+      "grad_norm": 1.0316848754882812,
+      "learning_rate": 1.561624649859944e-05,
+      "loss": 0.3966,
+      "step": 17200
+    },
+    {
+      "epoch": 22.01728,
+      "grad_norm": 1.1402348279953003,
+      "learning_rate": 1.561424569827931e-05,
+      "loss": 0.4791,
+      "step": 17201
+    },
+    {
+      "epoch": 22.01856,
+      "grad_norm": 1.1665503978729248,
+      "learning_rate": 1.5612244897959187e-05,
+      "loss": 0.5207,
+      "step": 17202
+    },
+    {
+      "epoch": 22.01984,
+      "grad_norm": 1.1318738460540771,
+      "learning_rate": 1.5610244097639055e-05,
+      "loss": 0.4978,
+      "step": 17203
+    },
+    {
+      "epoch": 22.02112,
+      "grad_norm": 1.0995848178863525,
+      "learning_rate": 1.5608243297318927e-05,
+      "loss": 0.4897,
+      "step": 17204
+    },
+    {
+      "epoch": 22.0224,
+      "grad_norm": 1.0910522937774658,
+      "learning_rate": 1.56062424969988e-05,
+      "loss": 0.4392,
+      "step": 17205
+    },
+    {
+      "epoch": 22.02368,
+      "grad_norm": 1.0444594621658325,
+      "learning_rate": 1.5604241696678674e-05,
+      "loss": 0.428,
+      "step": 17206
+    },
+    {
+      "epoch": 22.02496,
+      "grad_norm": 1.1223368644714355,
+      "learning_rate": 1.5602240896358543e-05,
+      "loss": 0.4473,
+      "step": 17207
+    },
+    {
+      "epoch": 22.02624,
+      "grad_norm": 1.035774827003479,
+      "learning_rate": 1.5600240096038415e-05,
+      "loss": 0.4201,
+      "step": 17208
+    },
+    {
+      "epoch": 22.02752,
+      "grad_norm": 1.155383586883545,
+      "learning_rate": 1.559823929571829e-05,
+      "loss": 0.4817,
+      "step": 17209
+    },
+    {
+      "epoch": 22.0288,
+      "grad_norm": 1.180153727531433,
+      "learning_rate": 1.559623849539816e-05,
+      "loss": 0.5142,
+      "step": 17210
+    },
+    {
+      "epoch": 22.03008,
+      "grad_norm": 1.0396239757537842,
+      "learning_rate": 1.559423769507803e-05,
+      "loss": 0.4348,
+      "step": 17211
+    },
+    {
+      "epoch": 22.03136,
+      "grad_norm": 1.1152554750442505,
+      "learning_rate": 1.5592236894757902e-05,
+      "loss": 0.5252,
+      "step": 17212
+    },
+    {
+      "epoch": 22.03264,
+      "grad_norm": 1.0972347259521484,
+      "learning_rate": 1.5590236094437777e-05,
+      "loss": 0.4627,
+      "step": 17213
+    },
+    {
+      "epoch": 22.03392,
+      "grad_norm": 1.1150223016738892,
+      "learning_rate": 1.558823529411765e-05,
+      "loss": 0.4778,
+      "step": 17214
+    },
+    {
+      "epoch": 22.0352,
+      "grad_norm": 1.1251050233840942,
+      "learning_rate": 1.5586234493797518e-05,
+      "loss": 0.5056,
+      "step": 17215
+    },
+    {
+      "epoch": 22.03648,
+      "grad_norm": 1.1292921304702759,
+      "learning_rate": 1.5584233693477393e-05,
+      "loss": 0.4896,
+      "step": 17216
+    },
+    {
+      "epoch": 22.03776,
+      "grad_norm": 1.169345736503601,
+      "learning_rate": 1.5582232893157265e-05,
+      "loss": 0.4869,
+      "step": 17217
+    },
+    {
+      "epoch": 22.03904,
+      "grad_norm": 1.10861074924469,
+      "learning_rate": 1.5580232092837137e-05,
+      "loss": 0.4686,
+      "step": 17218
+    },
+    {
+      "epoch": 22.04032,
+      "grad_norm": 1.080115556716919,
+      "learning_rate": 1.5578231292517005e-05,
+      "loss": 0.4709,
+      "step": 17219
+    },
+    {
+      "epoch": 22.0416,
+      "grad_norm": 1.0956518650054932,
+      "learning_rate": 1.557623049219688e-05,
+      "loss": 0.4785,
+      "step": 17220
+    },
+    {
+      "epoch": 22.04288,
+      "grad_norm": 1.1570404767990112,
+      "learning_rate": 1.5574229691876752e-05,
+      "loss": 0.539,
+      "step": 17221
+    },
+    {
+      "epoch": 22.04416,
+      "grad_norm": 1.1641499996185303,
+      "learning_rate": 1.5572228891556624e-05,
+      "loss": 0.5051,
+      "step": 17222
+    },
+    {
+      "epoch": 22.04544,
+      "grad_norm": 1.096850872039795,
+      "learning_rate": 1.5570228091236496e-05,
+      "loss": 0.4491,
+      "step": 17223
+    },
+    {
+      "epoch": 22.04672,
+      "grad_norm": 1.103550910949707,
+      "learning_rate": 1.5568227290916368e-05,
+      "loss": 0.5001,
+      "step": 17224
+    },
+    {
+      "epoch": 22.048,
+      "grad_norm": 1.1057217121124268,
+      "learning_rate": 1.556622649059624e-05,
+      "loss": 0.4451,
+      "step": 17225
+    },
+    {
+      "epoch": 22.04928,
+      "grad_norm": 1.1657497882843018,
+      "learning_rate": 1.556422569027611e-05,
+      "loss": 0.5181,
+      "step": 17226
+    },
+    {
+      "epoch": 22.05056,
+      "grad_norm": 1.1382044553756714,
+      "learning_rate": 1.5562224889955983e-05,
+      "loss": 0.4842,
+      "step": 17227
+    },
+    {
+      "epoch": 22.05184,
+      "grad_norm": 1.1846833229064941,
+      "learning_rate": 1.5560224089635855e-05,
+      "loss": 0.5218,
+      "step": 17228
+    },
+    {
+      "epoch": 22.05312,
+      "grad_norm": 1.137743592262268,
+      "learning_rate": 1.5558223289315727e-05,
+      "loss": 0.4667,
+      "step": 17229
+    },
+    {
+      "epoch": 22.0544,
+      "grad_norm": 1.1043981313705444,
+      "learning_rate": 1.55562224889956e-05,
+      "loss": 0.4647,
+      "step": 17230
+    },
+    {
+      "epoch": 22.05568,
+      "grad_norm": 1.121191143989563,
+      "learning_rate": 1.555422168867547e-05,
+      "loss": 0.4709,
+      "step": 17231
+    },
+    {
+      "epoch": 22.05696,
+      "grad_norm": 1.0988284349441528,
+      "learning_rate": 1.5552220888355343e-05,
+      "loss": 0.4861,
+      "step": 17232
+    },
+    {
+      "epoch": 22.05824,
+      "grad_norm": 1.153282880783081,
+      "learning_rate": 1.5550220088035214e-05,
+      "loss": 0.4873,
+      "step": 17233
+    },
+    {
+      "epoch": 22.05952,
+      "grad_norm": 1.1652368307113647,
+      "learning_rate": 1.5548219287715086e-05,
+      "loss": 0.4896,
+      "step": 17234
+    },
+    {
+      "epoch": 22.0608,
+      "grad_norm": 1.1135512590408325,
+      "learning_rate": 1.5546218487394958e-05,
+      "loss": 0.4406,
+      "step": 17235
+    },
+    {
+      "epoch": 22.06208,
+      "grad_norm": 1.166584849357605,
+      "learning_rate": 1.554421768707483e-05,
+      "loss": 0.507,
+      "step": 17236
+    },
+    {
+      "epoch": 22.06336,
+      "grad_norm": 1.1483736038208008,
+      "learning_rate": 1.5542216886754702e-05,
+      "loss": 0.5168,
+      "step": 17237
+    },
+    {
+      "epoch": 22.06464,
+      "grad_norm": 1.1366267204284668,
+      "learning_rate": 1.5540216086434574e-05,
+      "loss": 0.5365,
+      "step": 17238
+    },
+    {
+      "epoch": 22.06592,
+      "grad_norm": 1.1409132480621338,
+      "learning_rate": 1.5538215286114446e-05,
+      "loss": 0.4238,
+      "step": 17239
+    },
+    {
+      "epoch": 22.0672,
+      "grad_norm": 1.1683720350265503,
+      "learning_rate": 1.5536214485794317e-05,
+      "loss": 0.5112,
+      "step": 17240
+    },
+    {
+      "epoch": 22.06848,
+      "grad_norm": 1.0807946920394897,
+      "learning_rate": 1.5534213685474193e-05,
+      "loss": 0.4384,
+      "step": 17241
+    },
+    {
+      "epoch": 22.06976,
+      "grad_norm": 1.2099087238311768,
+      "learning_rate": 1.553221288515406e-05,
+      "loss": 0.523,
+      "step": 17242
+    },
+    {
+      "epoch": 22.07104,
+      "grad_norm": 1.1151604652404785,
+      "learning_rate": 1.5530212084833933e-05,
+      "loss": 0.4653,
+      "step": 17243
+    },
+    {
+      "epoch": 22.07232,
+      "grad_norm": 1.0875353813171387,
+      "learning_rate": 1.5528211284513805e-05,
+      "loss": 0.4273,
+      "step": 17244
+    },
+    {
+      "epoch": 22.0736,
+      "grad_norm": 1.2180532217025757,
+      "learning_rate": 1.552621048419368e-05,
+      "loss": 0.5473,
+      "step": 17245
+    },
+    {
+      "epoch": 22.07488,
+      "grad_norm": 1.1339551210403442,
+      "learning_rate": 1.552420968387355e-05,
+      "loss": 0.4595,
+      "step": 17246
+    },
+    {
+      "epoch": 22.07616,
+      "grad_norm": 1.1073286533355713,
+      "learning_rate": 1.552220888355342e-05,
+      "loss": 0.4464,
+      "step": 17247
+    },
+    {
+      "epoch": 22.07744,
+      "grad_norm": 1.136877417564392,
+      "learning_rate": 1.5520208083233296e-05,
+      "loss": 0.494,
+      "step": 17248
+    },
+    {
+      "epoch": 22.07872,
+      "grad_norm": 1.1355657577514648,
+      "learning_rate": 1.5518207282913167e-05,
+      "loss": 0.4883,
+      "step": 17249
+    },
+    {
+      "epoch": 22.08,
+      "grad_norm": 1.1080483198165894,
+      "learning_rate": 1.5516206482593036e-05,
+      "loss": 0.4671,
+      "step": 17250
+    },
+    {
+      "epoch": 22.08128,
+      "grad_norm": 1.1329385042190552,
+      "learning_rate": 1.5514205682272908e-05,
+      "loss": 0.536,
+      "step": 17251
+    },
+    {
+      "epoch": 22.08256,
+      "grad_norm": 1.0859447717666626,
+      "learning_rate": 1.5512204881952783e-05,
+      "loss": 0.4559,
+      "step": 17252
+    },
+    {
+      "epoch": 22.08384,
+      "grad_norm": 1.134740948677063,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 0.4539,
+      "step": 17253
+    },
+    {
+      "epoch": 22.08512,
+      "grad_norm": 1.1605035066604614,
+      "learning_rate": 1.5508203281312523e-05,
+      "loss": 0.4898,
+      "step": 17254
+    },
+    {
+      "epoch": 22.0864,
+      "grad_norm": 1.0824528932571411,
+      "learning_rate": 1.55062024809924e-05,
+      "loss": 0.4827,
+      "step": 17255
+    },
+    {
+      "epoch": 22.08768,
+      "grad_norm": 1.0842760801315308,
+      "learning_rate": 1.550420168067227e-05,
+      "loss": 0.4643,
+      "step": 17256
+    },
+    {
+      "epoch": 22.08896,
+      "grad_norm": 1.152551531791687,
+      "learning_rate": 1.5502200880352142e-05,
+      "loss": 0.5292,
+      "step": 17257
+    },
+    {
+      "epoch": 22.09024,
+      "grad_norm": 1.086955189704895,
+      "learning_rate": 1.550020008003201e-05,
+      "loss": 0.4563,
+      "step": 17258
+    },
+    {
+      "epoch": 22.09152,
+      "grad_norm": 1.1958043575286865,
+      "learning_rate": 1.5498199279711886e-05,
+      "loss": 0.5174,
+      "step": 17259
+    },
+    {
+      "epoch": 22.0928,
+      "grad_norm": 1.111706256866455,
+      "learning_rate": 1.5496198479391758e-05,
+      "loss": 0.4607,
+      "step": 17260
+    },
+    {
+      "epoch": 22.09408,
+      "grad_norm": 1.1034555435180664,
+      "learning_rate": 1.549419767907163e-05,
+      "loss": 0.4702,
+      "step": 17261
+    },
+    {
+      "epoch": 22.09536,
+      "grad_norm": 1.1045020818710327,
+      "learning_rate": 1.54921968787515e-05,
+      "loss": 0.4442,
+      "step": 17262
+    },
+    {
+      "epoch": 22.09664,
+      "grad_norm": 1.1338021755218506,
+      "learning_rate": 1.5490196078431373e-05,
+      "loss": 0.4941,
+      "step": 17263
+    },
+    {
+      "epoch": 22.09792,
+      "grad_norm": 1.1852363348007202,
+      "learning_rate": 1.5488195278111245e-05,
+      "loss": 0.5476,
+      "step": 17264
+    },
+    {
+      "epoch": 22.0992,
+      "grad_norm": 1.109511137008667,
+      "learning_rate": 1.5486194477791117e-05,
+      "loss": 0.4669,
+      "step": 17265
+    },
+    {
+      "epoch": 22.10048,
+      "grad_norm": 1.1228883266448975,
+      "learning_rate": 1.548419367747099e-05,
+      "loss": 0.4729,
+      "step": 17266
+    },
+    {
+      "epoch": 22.10176,
+      "grad_norm": 1.1240496635437012,
+      "learning_rate": 1.548219287715086e-05,
+      "loss": 0.5041,
+      "step": 17267
+    },
+    {
+      "epoch": 22.10304,
+      "grad_norm": 1.1427851915359497,
+      "learning_rate": 1.5480192076830733e-05,
+      "loss": 0.5104,
+      "step": 17268
+    },
+    {
+      "epoch": 22.10432,
+      "grad_norm": 1.1009935140609741,
+      "learning_rate": 1.5478191276510608e-05,
+      "loss": 0.4835,
+      "step": 17269
+    },
+    {
+      "epoch": 22.1056,
+      "grad_norm": 1.2154606580734253,
+      "learning_rate": 1.5476190476190476e-05,
+      "loss": 0.5254,
+      "step": 17270
+    },
+    {
+      "epoch": 22.10688,
+      "grad_norm": 1.1332104206085205,
+      "learning_rate": 1.5474189675870348e-05,
+      "loss": 0.475,
+      "step": 17271
+    },
+    {
+      "epoch": 22.10816,
+      "grad_norm": 1.156253457069397,
+      "learning_rate": 1.547218887555022e-05,
+      "loss": 0.4589,
+      "step": 17272
+    },
+    {
+      "epoch": 22.10944,
+      "grad_norm": 1.0867305994033813,
+      "learning_rate": 1.5470188075230095e-05,
+      "loss": 0.4505,
+      "step": 17273
+    },
+    {
+      "epoch": 22.11072,
+      "grad_norm": 1.1919612884521484,
+      "learning_rate": 1.5468187274909964e-05,
+      "loss": 0.4891,
+      "step": 17274
+    },
+    {
+      "epoch": 22.112,
+      "grad_norm": 1.1696399450302124,
+      "learning_rate": 1.5466186474589836e-05,
+      "loss": 0.5135,
+      "step": 17275
+    },
+    {
+      "epoch": 22.11328,
+      "grad_norm": 1.1965242624282837,
+      "learning_rate": 1.546418567426971e-05,
+      "loss": 0.5167,
+      "step": 17276
+    },
+    {
+      "epoch": 22.11456,
+      "grad_norm": 1.1492869853973389,
+      "learning_rate": 1.5462184873949583e-05,
+      "loss": 0.5027,
+      "step": 17277
+    },
+    {
+      "epoch": 22.11584,
+      "grad_norm": 1.1493687629699707,
+      "learning_rate": 1.546018407362945e-05,
+      "loss": 0.4764,
+      "step": 17278
+    },
+    {
+      "epoch": 22.11712,
+      "grad_norm": 1.1883156299591064,
+      "learning_rate": 1.5458183273309323e-05,
+      "loss": 0.4979,
+      "step": 17279
+    },
+    {
+      "epoch": 22.1184,
+      "grad_norm": 1.198495864868164,
+      "learning_rate": 1.54561824729892e-05,
+      "loss": 0.5273,
+      "step": 17280
+    },
+    {
+      "epoch": 22.11968,
+      "grad_norm": 1.1346237659454346,
+      "learning_rate": 1.545418167266907e-05,
+      "loss": 0.4574,
+      "step": 17281
+    },
+    {
+      "epoch": 22.12096,
+      "grad_norm": 1.1258900165557861,
+      "learning_rate": 1.545218087234894e-05,
+      "loss": 0.4931,
+      "step": 17282
+    },
+    {
+      "epoch": 22.12224,
+      "grad_norm": 1.1091734170913696,
+      "learning_rate": 1.545018007202881e-05,
+      "loss": 0.4793,
+      "step": 17283
+    },
+    {
+      "epoch": 22.12352,
+      "grad_norm": 1.0940433740615845,
+      "learning_rate": 1.5448179271708686e-05,
+      "loss": 0.4683,
+      "step": 17284
+    },
+    {
+      "epoch": 22.1248,
+      "grad_norm": 1.120511531829834,
+      "learning_rate": 1.5446178471388558e-05,
+      "loss": 0.4579,
+      "step": 17285
+    },
+    {
+      "epoch": 22.12608,
+      "grad_norm": 1.1699022054672241,
+      "learning_rate": 1.5444177671068426e-05,
+      "loss": 0.4941,
+      "step": 17286
+    },
+    {
+      "epoch": 22.12736,
+      "grad_norm": 1.0578786134719849,
+      "learning_rate": 1.54421768707483e-05,
+      "loss": 0.4419,
+      "step": 17287
+    },
+    {
+      "epoch": 22.12864,
+      "grad_norm": 1.104691982269287,
+      "learning_rate": 1.5440176070428173e-05,
+      "loss": 0.4954,
+      "step": 17288
+    },
+    {
+      "epoch": 22.12992,
+      "grad_norm": 1.1554813385009766,
+      "learning_rate": 1.5438175270108045e-05,
+      "loss": 0.5006,
+      "step": 17289
+    },
+    {
+      "epoch": 22.1312,
+      "grad_norm": 1.1532070636749268,
+      "learning_rate": 1.5436174469787914e-05,
+      "loss": 0.4952,
+      "step": 17290
+    },
+    {
+      "epoch": 22.13248,
+      "grad_norm": 1.0986944437026978,
+      "learning_rate": 1.543417366946779e-05,
+      "loss": 0.4291,
+      "step": 17291
+    },
+    {
+      "epoch": 22.13376,
+      "grad_norm": 1.1358639001846313,
+      "learning_rate": 1.543217286914766e-05,
+      "loss": 0.4807,
+      "step": 17292
+    },
+    {
+      "epoch": 22.13504,
+      "grad_norm": 1.113168716430664,
+      "learning_rate": 1.5430172068827533e-05,
+      "loss": 0.4816,
+      "step": 17293
+    },
+    {
+      "epoch": 22.13632,
+      "grad_norm": 1.1050689220428467,
+      "learning_rate": 1.5428171268507404e-05,
+      "loss": 0.4404,
+      "step": 17294
+    },
+    {
+      "epoch": 22.1376,
+      "grad_norm": 1.091439962387085,
+      "learning_rate": 1.5426170468187276e-05,
+      "loss": 0.5036,
+      "step": 17295
+    },
+    {
+      "epoch": 22.13888,
+      "grad_norm": 1.1166634559631348,
+      "learning_rate": 1.5424169667867148e-05,
+      "loss": 0.5347,
+      "step": 17296
+    },
+    {
+      "epoch": 22.14016,
+      "grad_norm": 1.1514959335327148,
+      "learning_rate": 1.542216886754702e-05,
+      "loss": 0.494,
+      "step": 17297
+    },
+    {
+      "epoch": 22.14144,
+      "grad_norm": 1.070124864578247,
+      "learning_rate": 1.5420168067226892e-05,
+      "loss": 0.4533,
+      "step": 17298
+    },
+    {
+      "epoch": 22.14272,
+      "grad_norm": 1.0914841890335083,
+      "learning_rate": 1.5418167266906764e-05,
+      "loss": 0.509,
+      "step": 17299
+    },
+    {
+      "epoch": 22.144,
+      "grad_norm": 1.1017416715621948,
+      "learning_rate": 1.5416166466586636e-05,
+      "loss": 0.4744,
+      "step": 17300
+    },
+    {
+      "epoch": 22.14528,
+      "grad_norm": 1.1203486919403076,
+      "learning_rate": 1.5414165666266507e-05,
+      "loss": 0.4563,
+      "step": 17301
+    },
+    {
+      "epoch": 22.14656,
+      "grad_norm": 1.1463284492492676,
+      "learning_rate": 1.541216486594638e-05,
+      "loss": 0.4892,
+      "step": 17302
+    },
+    {
+      "epoch": 22.14784,
+      "grad_norm": 1.1561495065689087,
+      "learning_rate": 1.541016406562625e-05,
+      "loss": 0.4916,
+      "step": 17303
+    },
+    {
+      "epoch": 22.14912,
+      "grad_norm": 1.1047677993774414,
+      "learning_rate": 1.5408163265306123e-05,
+      "loss": 0.4706,
+      "step": 17304
+    },
+    {
+      "epoch": 22.1504,
+      "grad_norm": 1.1294938325881958,
+      "learning_rate": 1.5406162464985995e-05,
+      "loss": 0.4652,
+      "step": 17305
+    },
+    {
+      "epoch": 22.15168,
+      "grad_norm": 1.166673183441162,
+      "learning_rate": 1.5404161664665867e-05,
+      "loss": 0.478,
+      "step": 17306
+    },
+    {
+      "epoch": 22.15296,
+      "grad_norm": 1.1008706092834473,
+      "learning_rate": 1.540216086434574e-05,
+      "loss": 0.4525,
+      "step": 17307
+    },
+    {
+      "epoch": 22.15424,
+      "grad_norm": 1.1509610414505005,
+      "learning_rate": 1.5400160064025614e-05,
+      "loss": 0.5268,
+      "step": 17308
+    },
+    {
+      "epoch": 22.15552,
+      "grad_norm": 1.1664763689041138,
+      "learning_rate": 1.5398159263705482e-05,
+      "loss": 0.5145,
+      "step": 17309
+    },
+    {
+      "epoch": 22.1568,
+      "grad_norm": 1.1705671548843384,
+      "learning_rate": 1.5396158463385354e-05,
+      "loss": 0.4866,
+      "step": 17310
+    },
+    {
+      "epoch": 22.158079999999998,
+      "grad_norm": 1.0866199731826782,
+      "learning_rate": 1.5394157663065226e-05,
+      "loss": 0.4587,
+      "step": 17311
+    },
+    {
+      "epoch": 22.15936,
+      "grad_norm": 1.0654592514038086,
+      "learning_rate": 1.53921568627451e-05,
+      "loss": 0.4489,
+      "step": 17312
+    },
+    {
+      "epoch": 22.16064,
+      "grad_norm": 1.1411796808242798,
+      "learning_rate": 1.539015606242497e-05,
+      "loss": 0.5293,
+      "step": 17313
+    },
+    {
+      "epoch": 22.16192,
+      "grad_norm": 1.1217213869094849,
+      "learning_rate": 1.538815526210484e-05,
+      "loss": 0.4792,
+      "step": 17314
+    },
+    {
+      "epoch": 22.1632,
+      "grad_norm": 1.133726954460144,
+      "learning_rate": 1.5386154461784717e-05,
+      "loss": 0.4706,
+      "step": 17315
+    },
+    {
+      "epoch": 22.16448,
+      "grad_norm": 1.0698765516281128,
+      "learning_rate": 1.538415366146459e-05,
+      "loss": 0.4544,
+      "step": 17316
+    },
+    {
+      "epoch": 22.16576,
+      "grad_norm": 1.112635850906372,
+      "learning_rate": 1.5382152861144457e-05,
+      "loss": 0.4675,
+      "step": 17317
+    },
+    {
+      "epoch": 22.16704,
+      "grad_norm": 1.0733355283737183,
+      "learning_rate": 1.538015206082433e-05,
+      "loss": 0.4737,
+      "step": 17318
+    },
+    {
+      "epoch": 22.16832,
+      "grad_norm": 1.1422197818756104,
+      "learning_rate": 1.5378151260504204e-05,
+      "loss": 0.4691,
+      "step": 17319
+    },
+    {
+      "epoch": 22.1696,
+      "grad_norm": 1.1713340282440186,
+      "learning_rate": 1.5376150460184076e-05,
+      "loss": 0.4922,
+      "step": 17320
+    },
+    {
+      "epoch": 22.17088,
+      "grad_norm": 1.114890217781067,
+      "learning_rate": 1.5374149659863945e-05,
+      "loss": 0.4539,
+      "step": 17321
+    },
+    {
+      "epoch": 22.17216,
+      "grad_norm": 1.1028982400894165,
+      "learning_rate": 1.537214885954382e-05,
+      "loss": 0.4329,
+      "step": 17322
+    },
+    {
+      "epoch": 22.17344,
+      "grad_norm": 1.1502256393432617,
+      "learning_rate": 1.537014805922369e-05,
+      "loss": 0.486,
+      "step": 17323
+    },
+    {
+      "epoch": 22.17472,
+      "grad_norm": 1.1716766357421875,
+      "learning_rate": 1.5368147258903563e-05,
+      "loss": 0.5172,
+      "step": 17324
+    },
+    {
+      "epoch": 22.176,
+      "grad_norm": 1.1179568767547607,
+      "learning_rate": 1.5366146458583432e-05,
+      "loss": 0.481,
+      "step": 17325
+    },
+    {
+      "epoch": 22.17728,
+      "grad_norm": 1.0973176956176758,
+      "learning_rate": 1.5364145658263307e-05,
+      "loss": 0.4588,
+      "step": 17326
+    },
+    {
+      "epoch": 22.17856,
+      "grad_norm": 1.1206368207931519,
+      "learning_rate": 1.536214485794318e-05,
+      "loss": 0.4673,
+      "step": 17327
+    },
+    {
+      "epoch": 22.17984,
+      "grad_norm": 1.128866195678711,
+      "learning_rate": 1.536014405762305e-05,
+      "loss": 0.4766,
+      "step": 17328
+    },
+    {
+      "epoch": 22.18112,
+      "grad_norm": 1.0996205806732178,
+      "learning_rate": 1.5358143257302923e-05,
+      "loss": 0.4679,
+      "step": 17329
+    },
+    {
+      "epoch": 22.1824,
+      "grad_norm": 1.087931752204895,
+      "learning_rate": 1.5356142456982795e-05,
+      "loss": 0.4724,
+      "step": 17330
+    },
+    {
+      "epoch": 22.18368,
+      "grad_norm": 1.1762522459030151,
+      "learning_rate": 1.5354141656662666e-05,
+      "loss": 0.5389,
+      "step": 17331
+    },
+    {
+      "epoch": 22.18496,
+      "grad_norm": 1.0654993057250977,
+      "learning_rate": 1.535214085634254e-05,
+      "loss": 0.4851,
+      "step": 17332
+    },
+    {
+      "epoch": 22.18624,
+      "grad_norm": 1.0894544124603271,
+      "learning_rate": 1.535014005602241e-05,
+      "loss": 0.5082,
+      "step": 17333
+    },
+    {
+      "epoch": 22.18752,
+      "grad_norm": 1.0549466609954834,
+      "learning_rate": 1.5348139255702282e-05,
+      "loss": 0.4361,
+      "step": 17334
+    },
+    {
+      "epoch": 22.1888,
+      "grad_norm": 1.16023588180542,
+      "learning_rate": 1.5346138455382154e-05,
+      "loss": 0.5082,
+      "step": 17335
+    },
+    {
+      "epoch": 22.19008,
+      "grad_norm": 1.1312403678894043,
+      "learning_rate": 1.5344137655062026e-05,
+      "loss": 0.4884,
+      "step": 17336
+    },
+    {
+      "epoch": 22.19136,
+      "grad_norm": 1.1789931058883667,
+      "learning_rate": 1.5342136854741898e-05,
+      "loss": 0.4722,
+      "step": 17337
+    },
+    {
+      "epoch": 22.19264,
+      "grad_norm": 1.1645036935806274,
+      "learning_rate": 1.534013605442177e-05,
+      "loss": 0.5172,
+      "step": 17338
+    },
+    {
+      "epoch": 22.19392,
+      "grad_norm": 1.106508731842041,
+      "learning_rate": 1.533813525410164e-05,
+      "loss": 0.466,
+      "step": 17339
+    },
+    {
+      "epoch": 22.1952,
+      "grad_norm": 1.1362861394882202,
+      "learning_rate": 1.5336134453781513e-05,
+      "loss": 0.4865,
+      "step": 17340
+    },
+    {
+      "epoch": 22.19648,
+      "grad_norm": 1.10444974899292,
+      "learning_rate": 1.5334133653461385e-05,
+      "loss": 0.4501,
+      "step": 17341
+    },
+    {
+      "epoch": 22.19776,
+      "grad_norm": 1.1265275478363037,
+      "learning_rate": 1.5332132853141257e-05,
+      "loss": 0.4673,
+      "step": 17342
+    },
+    {
+      "epoch": 22.19904,
+      "grad_norm": 1.1572017669677734,
+      "learning_rate": 1.533013205282113e-05,
+      "loss": 0.4611,
+      "step": 17343
+    },
+    {
+      "epoch": 22.20032,
+      "grad_norm": 1.2215920686721802,
+      "learning_rate": 1.5328131252501e-05,
+      "loss": 0.5035,
+      "step": 17344
+    },
+    {
+      "epoch": 22.2016,
+      "grad_norm": 1.2130608558654785,
+      "learning_rate": 1.5326130452180872e-05,
+      "loss": 0.4942,
+      "step": 17345
+    },
+    {
+      "epoch": 22.20288,
+      "grad_norm": 1.133594036102295,
+      "learning_rate": 1.5324129651860744e-05,
+      "loss": 0.4936,
+      "step": 17346
+    },
+    {
+      "epoch": 22.20416,
+      "grad_norm": 1.1092051267623901,
+      "learning_rate": 1.532212885154062e-05,
+      "loss": 0.5127,
+      "step": 17347
+    },
+    {
+      "epoch": 22.20544,
+      "grad_norm": 1.145732045173645,
+      "learning_rate": 1.5320128051220488e-05,
+      "loss": 0.4747,
+      "step": 17348
+    },
+    {
+      "epoch": 22.20672,
+      "grad_norm": 1.0972932577133179,
+      "learning_rate": 1.531812725090036e-05,
+      "loss": 0.4997,
+      "step": 17349
+    },
+    {
+      "epoch": 22.208,
+      "grad_norm": 1.0953388214111328,
+      "learning_rate": 1.5316126450580232e-05,
+      "loss": 0.4583,
+      "step": 17350
+    },
+    {
+      "epoch": 22.20928,
+      "grad_norm": 1.086438775062561,
+      "learning_rate": 1.5314125650260107e-05,
+      "loss": 0.4578,
+      "step": 17351
+    },
+    {
+      "epoch": 22.21056,
+      "grad_norm": 1.1954177618026733,
+      "learning_rate": 1.5312124849939975e-05,
+      "loss": 0.5342,
+      "step": 17352
+    },
+    {
+      "epoch": 22.21184,
+      "grad_norm": 1.1370768547058105,
+      "learning_rate": 1.5310124049619847e-05,
+      "loss": 0.5075,
+      "step": 17353
+    },
+    {
+      "epoch": 22.21312,
+      "grad_norm": 1.1684788465499878,
+      "learning_rate": 1.5308123249299723e-05,
+      "loss": 0.4789,
+      "step": 17354
+    },
+    {
+      "epoch": 22.2144,
+      "grad_norm": 1.108345866203308,
+      "learning_rate": 1.5306122448979594e-05,
+      "loss": 0.4686,
+      "step": 17355
+    },
+    {
+      "epoch": 22.21568,
+      "grad_norm": 1.0191106796264648,
+      "learning_rate": 1.5304121648659463e-05,
+      "loss": 0.4258,
+      "step": 17356
+    },
+    {
+      "epoch": 22.21696,
+      "grad_norm": 1.1205805540084839,
+      "learning_rate": 1.5302120848339335e-05,
+      "loss": 0.4757,
+      "step": 17357
+    },
+    {
+      "epoch": 22.21824,
+      "grad_norm": 1.1553397178649902,
+      "learning_rate": 1.530012004801921e-05,
+      "loss": 0.4878,
+      "step": 17358
+    },
+    {
+      "epoch": 22.21952,
+      "grad_norm": 1.1630548238754272,
+      "learning_rate": 1.5298119247699082e-05,
+      "loss": 0.4906,
+      "step": 17359
+    },
+    {
+      "epoch": 22.2208,
+      "grad_norm": 1.0727124214172363,
+      "learning_rate": 1.529611844737895e-05,
+      "loss": 0.4877,
+      "step": 17360
+    },
+    {
+      "epoch": 22.22208,
+      "grad_norm": 1.0958082675933838,
+      "learning_rate": 1.5294117647058826e-05,
+      "loss": 0.4236,
+      "step": 17361
+    },
+    {
+      "epoch": 22.22336,
+      "grad_norm": 1.1463395357131958,
+      "learning_rate": 1.5292116846738697e-05,
+      "loss": 0.4354,
+      "step": 17362
+    },
+    {
+      "epoch": 22.22464,
+      "grad_norm": 1.125491976737976,
+      "learning_rate": 1.529011604641857e-05,
+      "loss": 0.4664,
+      "step": 17363
+    },
+    {
+      "epoch": 22.22592,
+      "grad_norm": 1.1401169300079346,
+      "learning_rate": 1.5288115246098438e-05,
+      "loss": 0.4903,
+      "step": 17364
+    },
+    {
+      "epoch": 22.2272,
+      "grad_norm": 1.1379910707473755,
+      "learning_rate": 1.5286114445778313e-05,
+      "loss": 0.4734,
+      "step": 17365
+    },
+    {
+      "epoch": 22.22848,
+      "grad_norm": 1.0853389501571655,
+      "learning_rate": 1.5284113645458185e-05,
+      "loss": 0.4546,
+      "step": 17366
+    },
+    {
+      "epoch": 22.22976,
+      "grad_norm": 1.176604151725769,
+      "learning_rate": 1.5282112845138057e-05,
+      "loss": 0.5138,
+      "step": 17367
+    },
+    {
+      "epoch": 22.23104,
+      "grad_norm": 1.2261618375778198,
+      "learning_rate": 1.528011204481793e-05,
+      "loss": 0.5195,
+      "step": 17368
+    },
+    {
+      "epoch": 22.23232,
+      "grad_norm": 1.1438370943069458,
+      "learning_rate": 1.52781112444978e-05,
+      "loss": 0.5111,
+      "step": 17369
+    },
+    {
+      "epoch": 22.2336,
+      "grad_norm": 1.0962119102478027,
+      "learning_rate": 1.5276110444177672e-05,
+      "loss": 0.4471,
+      "step": 17370
+    },
+    {
+      "epoch": 22.23488,
+      "grad_norm": 1.1545382738113403,
+      "learning_rate": 1.5274109643857544e-05,
+      "loss": 0.4741,
+      "step": 17371
+    },
+    {
+      "epoch": 22.23616,
+      "grad_norm": 1.122843861579895,
+      "learning_rate": 1.5272108843537416e-05,
+      "loss": 0.4445,
+      "step": 17372
+    },
+    {
+      "epoch": 22.23744,
+      "grad_norm": 1.114371418952942,
+      "learning_rate": 1.5270108043217288e-05,
+      "loss": 0.4685,
+      "step": 17373
+    },
+    {
+      "epoch": 22.23872,
+      "grad_norm": 1.1114271879196167,
+      "learning_rate": 1.526810724289716e-05,
+      "loss": 0.5048,
+      "step": 17374
+    },
+    {
+      "epoch": 22.24,
+      "grad_norm": 1.1175719499588013,
+      "learning_rate": 1.526610644257703e-05,
+      "loss": 0.4901,
+      "step": 17375
+    },
+    {
+      "epoch": 22.24128,
+      "grad_norm": 1.1345455646514893,
+      "learning_rate": 1.5264105642256903e-05,
+      "loss": 0.4525,
+      "step": 17376
+    },
+    {
+      "epoch": 22.24256,
+      "grad_norm": 1.138181447982788,
+      "learning_rate": 1.5262104841936775e-05,
+      "loss": 0.4898,
+      "step": 17377
+    },
+    {
+      "epoch": 22.24384,
+      "grad_norm": 1.1441845893859863,
+      "learning_rate": 1.5260104041616647e-05,
+      "loss": 0.4646,
+      "step": 17378
+    },
+    {
+      "epoch": 22.24512,
+      "grad_norm": 1.102541446685791,
+      "learning_rate": 1.525810324129652e-05,
+      "loss": 0.4854,
+      "step": 17379
+    },
+    {
+      "epoch": 22.2464,
+      "grad_norm": 1.0756680965423584,
+      "learning_rate": 1.5256102440976391e-05,
+      "loss": 0.4716,
+      "step": 17380
+    },
+    {
+      "epoch": 22.24768,
+      "grad_norm": 1.1094943284988403,
+      "learning_rate": 1.5254101640656263e-05,
+      "loss": 0.4975,
+      "step": 17381
+    },
+    {
+      "epoch": 22.24896,
+      "grad_norm": 1.1272087097167969,
+      "learning_rate": 1.5252100840336136e-05,
+      "loss": 0.4882,
+      "step": 17382
+    },
+    {
+      "epoch": 22.25024,
+      "grad_norm": 1.134270191192627,
+      "learning_rate": 1.5250100040016008e-05,
+      "loss": 0.4806,
+      "step": 17383
+    },
+    {
+      "epoch": 22.25152,
+      "grad_norm": 1.1215940713882446,
+      "learning_rate": 1.5248099239695878e-05,
+      "loss": 0.4555,
+      "step": 17384
+    },
+    {
+      "epoch": 22.2528,
+      "grad_norm": 1.0712652206420898,
+      "learning_rate": 1.524609843937575e-05,
+      "loss": 0.4371,
+      "step": 17385
+    },
+    {
+      "epoch": 22.25408,
+      "grad_norm": 1.0350494384765625,
+      "learning_rate": 1.5244097639055624e-05,
+      "loss": 0.4513,
+      "step": 17386
+    },
+    {
+      "epoch": 22.25536,
+      "grad_norm": 1.1774425506591797,
+      "learning_rate": 1.5242096838735496e-05,
+      "loss": 0.5087,
+      "step": 17387
+    },
+    {
+      "epoch": 22.25664,
+      "grad_norm": 1.137933611869812,
+      "learning_rate": 1.5240096038415366e-05,
+      "loss": 0.5026,
+      "step": 17388
+    },
+    {
+      "epoch": 22.25792,
+      "grad_norm": 1.1481642723083496,
+      "learning_rate": 1.5238095238095241e-05,
+      "loss": 0.474,
+      "step": 17389
+    },
+    {
+      "epoch": 22.2592,
+      "grad_norm": 1.067002296447754,
+      "learning_rate": 1.5236094437775111e-05,
+      "loss": 0.4291,
+      "step": 17390
+    },
+    {
+      "epoch": 22.26048,
+      "grad_norm": 1.1963670253753662,
+      "learning_rate": 1.5234093637454983e-05,
+      "loss": 0.5206,
+      "step": 17391
+    },
+    {
+      "epoch": 22.26176,
+      "grad_norm": 1.1314671039581299,
+      "learning_rate": 1.5232092837134853e-05,
+      "loss": 0.4769,
+      "step": 17392
+    },
+    {
+      "epoch": 22.26304,
+      "grad_norm": 1.0662577152252197,
+      "learning_rate": 1.5230092036814728e-05,
+      "loss": 0.4752,
+      "step": 17393
+    },
+    {
+      "epoch": 22.26432,
+      "grad_norm": 1.1250959634780884,
+      "learning_rate": 1.5228091236494599e-05,
+      "loss": 0.5012,
+      "step": 17394
+    },
+    {
+      "epoch": 22.2656,
+      "grad_norm": 1.1448113918304443,
+      "learning_rate": 1.522609043617447e-05,
+      "loss": 0.5028,
+      "step": 17395
+    },
+    {
+      "epoch": 22.26688,
+      "grad_norm": 1.1659822463989258,
+      "learning_rate": 1.522408963585434e-05,
+      "loss": 0.4969,
+      "step": 17396
+    },
+    {
+      "epoch": 22.26816,
+      "grad_norm": 1.1037496328353882,
+      "learning_rate": 1.5222088835534216e-05,
+      "loss": 0.501,
+      "step": 17397
+    },
+    {
+      "epoch": 22.26944,
+      "grad_norm": 1.1459519863128662,
+      "learning_rate": 1.5220088035214086e-05,
+      "loss": 0.4836,
+      "step": 17398
+    },
+    {
+      "epoch": 22.27072,
+      "grad_norm": 1.0809226036071777,
+      "learning_rate": 1.5218087234893958e-05,
+      "loss": 0.4574,
+      "step": 17399
+    },
+    {
+      "epoch": 22.272,
+      "grad_norm": 1.1643927097320557,
+      "learning_rate": 1.5216086434573831e-05,
+      "loss": 0.5028,
+      "step": 17400
+    },
+    {
+      "epoch": 22.27328,
+      "grad_norm": 1.097196102142334,
+      "learning_rate": 1.5214085634253703e-05,
+      "loss": 0.4355,
+      "step": 17401
+    },
+    {
+      "epoch": 22.27456,
+      "grad_norm": 1.1606805324554443,
+      "learning_rate": 1.5212084833933573e-05,
+      "loss": 0.4937,
+      "step": 17402
+    },
+    {
+      "epoch": 22.27584,
+      "grad_norm": 1.1250141859054565,
+      "learning_rate": 1.5210084033613445e-05,
+      "loss": 0.4682,
+      "step": 17403
+    },
+    {
+      "epoch": 22.27712,
+      "grad_norm": 1.1419126987457275,
+      "learning_rate": 1.5208083233293319e-05,
+      "loss": 0.4867,
+      "step": 17404
+    },
+    {
+      "epoch": 22.2784,
+      "grad_norm": 1.1214076280593872,
+      "learning_rate": 1.520608243297319e-05,
+      "loss": 0.4733,
+      "step": 17405
+    },
+    {
+      "epoch": 22.27968,
+      "grad_norm": 1.1369352340698242,
+      "learning_rate": 1.520408163265306e-05,
+      "loss": 0.4595,
+      "step": 17406
+    },
+    {
+      "epoch": 22.28096,
+      "grad_norm": 1.0881363153457642,
+      "learning_rate": 1.5202080832332934e-05,
+      "loss": 0.479,
+      "step": 17407
+    },
+    {
+      "epoch": 22.28224,
+      "grad_norm": 1.1420772075653076,
+      "learning_rate": 1.5200080032012806e-05,
+      "loss": 0.4535,
+      "step": 17408
+    },
+    {
+      "epoch": 22.28352,
+      "grad_norm": 1.1348367929458618,
+      "learning_rate": 1.5198079231692678e-05,
+      "loss": 0.4644,
+      "step": 17409
+    },
+    {
+      "epoch": 22.2848,
+      "grad_norm": 1.0882881879806519,
+      "learning_rate": 1.5196078431372548e-05,
+      "loss": 0.4883,
+      "step": 17410
+    },
+    {
+      "epoch": 22.28608,
+      "grad_norm": 1.1225090026855469,
+      "learning_rate": 1.5194077631052422e-05,
+      "loss": 0.4732,
+      "step": 17411
+    },
+    {
+      "epoch": 22.28736,
+      "grad_norm": 1.195477843284607,
+      "learning_rate": 1.5192076830732294e-05,
+      "loss": 0.5217,
+      "step": 17412
+    },
+    {
+      "epoch": 22.28864,
+      "grad_norm": 1.1594408750534058,
+      "learning_rate": 1.5190076030412166e-05,
+      "loss": 0.5314,
+      "step": 17413
+    },
+    {
+      "epoch": 22.28992,
+      "grad_norm": 1.1369751691818237,
+      "learning_rate": 1.5188075230092039e-05,
+      "loss": 0.4771,
+      "step": 17414
+    },
+    {
+      "epoch": 22.2912,
+      "grad_norm": 1.1328139305114746,
+      "learning_rate": 1.518607442977191e-05,
+      "loss": 0.5051,
+      "step": 17415
+    },
+    {
+      "epoch": 22.29248,
+      "grad_norm": 1.1395426988601685,
+      "learning_rate": 1.5184073629451781e-05,
+      "loss": 0.4516,
+      "step": 17416
+    },
+    {
+      "epoch": 22.29376,
+      "grad_norm": 1.1057137250900269,
+      "learning_rate": 1.5182072829131653e-05,
+      "loss": 0.5055,
+      "step": 17417
+    },
+    {
+      "epoch": 22.29504,
+      "grad_norm": 1.0750411748886108,
+      "learning_rate": 1.5180072028811526e-05,
+      "loss": 0.4643,
+      "step": 17418
+    },
+    {
+      "epoch": 22.29632,
+      "grad_norm": 1.134002685546875,
+      "learning_rate": 1.5178071228491397e-05,
+      "loss": 0.4959,
+      "step": 17419
+    },
+    {
+      "epoch": 22.2976,
+      "grad_norm": 1.1353780031204224,
+      "learning_rate": 1.5176070428171269e-05,
+      "loss": 0.4992,
+      "step": 17420
+    },
+    {
+      "epoch": 22.29888,
+      "grad_norm": 1.131705641746521,
+      "learning_rate": 1.5174069627851142e-05,
+      "loss": 0.4595,
+      "step": 17421
+    },
+    {
+      "epoch": 22.300159999999998,
+      "grad_norm": 1.1513729095458984,
+      "learning_rate": 1.5172068827531014e-05,
+      "loss": 0.4984,
+      "step": 17422
+    },
+    {
+      "epoch": 22.30144,
+      "grad_norm": 1.126017689704895,
+      "learning_rate": 1.5170068027210884e-05,
+      "loss": 0.4743,
+      "step": 17423
+    },
+    {
+      "epoch": 22.30272,
+      "grad_norm": 1.056380271911621,
+      "learning_rate": 1.5168067226890756e-05,
+      "loss": 0.4702,
+      "step": 17424
+    },
+    {
+      "epoch": 22.304,
+      "grad_norm": 1.079280972480774,
+      "learning_rate": 1.516606642657063e-05,
+      "loss": 0.4556,
+      "step": 17425
+    },
+    {
+      "epoch": 22.30528,
+      "grad_norm": 1.1339120864868164,
+      "learning_rate": 1.5164065626250501e-05,
+      "loss": 0.4703,
+      "step": 17426
+    },
+    {
+      "epoch": 22.30656,
+      "grad_norm": 1.1718413829803467,
+      "learning_rate": 1.5162064825930372e-05,
+      "loss": 0.4925,
+      "step": 17427
+    },
+    {
+      "epoch": 22.30784,
+      "grad_norm": 1.1251111030578613,
+      "learning_rate": 1.5160064025610247e-05,
+      "loss": 0.4401,
+      "step": 17428
+    },
+    {
+      "epoch": 22.30912,
+      "grad_norm": 1.1180959939956665,
+      "learning_rate": 1.5158063225290117e-05,
+      "loss": 0.5187,
+      "step": 17429
+    },
+    {
+      "epoch": 22.3104,
+      "grad_norm": 1.1028566360473633,
+      "learning_rate": 1.5156062424969989e-05,
+      "loss": 0.4702,
+      "step": 17430
+    },
+    {
+      "epoch": 22.31168,
+      "grad_norm": 1.1383028030395508,
+      "learning_rate": 1.5154061624649859e-05,
+      "loss": 0.4931,
+      "step": 17431
+    },
+    {
+      "epoch": 22.31296,
+      "grad_norm": 1.0866107940673828,
+      "learning_rate": 1.5152060824329734e-05,
+      "loss": 0.5136,
+      "step": 17432
+    },
+    {
+      "epoch": 22.31424,
+      "grad_norm": 1.1053640842437744,
+      "learning_rate": 1.5150060024009604e-05,
+      "loss": 0.5046,
+      "step": 17433
+    },
+    {
+      "epoch": 22.31552,
+      "grad_norm": 1.0949057340621948,
+      "learning_rate": 1.5148059223689476e-05,
+      "loss": 0.524,
+      "step": 17434
+    },
+    {
+      "epoch": 22.3168,
+      "grad_norm": 1.101889729499817,
+      "learning_rate": 1.514605842336935e-05,
+      "loss": 0.4857,
+      "step": 17435
+    },
+    {
+      "epoch": 22.31808,
+      "grad_norm": 1.0940123796463013,
+      "learning_rate": 1.5144057623049222e-05,
+      "loss": 0.4603,
+      "step": 17436
+    },
+    {
+      "epoch": 22.31936,
+      "grad_norm": 1.0854915380477905,
+      "learning_rate": 1.5142056822729092e-05,
+      "loss": 0.506,
+      "step": 17437
+    },
+    {
+      "epoch": 22.32064,
+      "grad_norm": 1.1346979141235352,
+      "learning_rate": 1.5140056022408964e-05,
+      "loss": 0.5258,
+      "step": 17438
+    },
+    {
+      "epoch": 22.32192,
+      "grad_norm": 1.0600947141647339,
+      "learning_rate": 1.5138055222088837e-05,
+      "loss": 0.4525,
+      "step": 17439
+    },
+    {
+      "epoch": 22.3232,
+      "grad_norm": 1.082271695137024,
+      "learning_rate": 1.5136054421768709e-05,
+      "loss": 0.4732,
+      "step": 17440
+    },
+    {
+      "epoch": 22.32448,
+      "grad_norm": 1.145754098892212,
+      "learning_rate": 1.513405362144858e-05,
+      "loss": 0.4547,
+      "step": 17441
+    },
+    {
+      "epoch": 22.32576,
+      "grad_norm": 1.162156581878662,
+      "learning_rate": 1.5132052821128453e-05,
+      "loss": 0.4697,
+      "step": 17442
+    },
+    {
+      "epoch": 22.32704,
+      "grad_norm": 1.1728843450546265,
+      "learning_rate": 1.5130052020808325e-05,
+      "loss": 0.4654,
+      "step": 17443
+    },
+    {
+      "epoch": 22.32832,
+      "grad_norm": 1.1649385690689087,
+      "learning_rate": 1.5128051220488196e-05,
+      "loss": 0.5223,
+      "step": 17444
+    },
+    {
+      "epoch": 22.3296,
+      "grad_norm": 1.1600626707077026,
+      "learning_rate": 1.5126050420168067e-05,
+      "loss": 0.4684,
+      "step": 17445
+    },
+    {
+      "epoch": 22.33088,
+      "grad_norm": 1.1548811197280884,
+      "learning_rate": 1.512404961984794e-05,
+      "loss": 0.4642,
+      "step": 17446
+    },
+    {
+      "epoch": 22.332160000000002,
+      "grad_norm": 1.2105642557144165,
+      "learning_rate": 1.5122048819527812e-05,
+      "loss": 0.4687,
+      "step": 17447
+    },
+    {
+      "epoch": 22.33344,
+      "grad_norm": 1.204110860824585,
+      "learning_rate": 1.5120048019207684e-05,
+      "loss": 0.4745,
+      "step": 17448
+    },
+    {
+      "epoch": 22.33472,
+      "grad_norm": 1.189799189567566,
+      "learning_rate": 1.5118047218887554e-05,
+      "loss": 0.4882,
+      "step": 17449
+    },
+    {
+      "epoch": 22.336,
+      "grad_norm": 1.0634477138519287,
+      "learning_rate": 1.5116046418567428e-05,
+      "loss": 0.4457,
+      "step": 17450
+    },
+    {
+      "epoch": 22.33728,
+      "grad_norm": 1.1369130611419678,
+      "learning_rate": 1.51140456182473e-05,
+      "loss": 0.5071,
+      "step": 17451
+    },
+    {
+      "epoch": 22.33856,
+      "grad_norm": 1.1142311096191406,
+      "learning_rate": 1.5112044817927171e-05,
+      "loss": 0.4664,
+      "step": 17452
+    },
+    {
+      "epoch": 22.33984,
+      "grad_norm": 1.1078466176986694,
+      "learning_rate": 1.5110044017607045e-05,
+      "loss": 0.4767,
+      "step": 17453
+    },
+    {
+      "epoch": 22.34112,
+      "grad_norm": 1.1485053300857544,
+      "learning_rate": 1.5108043217286915e-05,
+      "loss": 0.4765,
+      "step": 17454
+    },
+    {
+      "epoch": 22.3424,
+      "grad_norm": 1.105195164680481,
+      "learning_rate": 1.5106042416966787e-05,
+      "loss": 0.4429,
+      "step": 17455
+    },
+    {
+      "epoch": 22.34368,
+      "grad_norm": 1.1941795349121094,
+      "learning_rate": 1.5104041616646659e-05,
+      "loss": 0.5371,
+      "step": 17456
+    },
+    {
+      "epoch": 22.34496,
+      "grad_norm": 1.0616157054901123,
+      "learning_rate": 1.5102040816326532e-05,
+      "loss": 0.421,
+      "step": 17457
+    },
+    {
+      "epoch": 22.34624,
+      "grad_norm": 1.1069750785827637,
+      "learning_rate": 1.5100040016006402e-05,
+      "loss": 0.4859,
+      "step": 17458
+    },
+    {
+      "epoch": 22.34752,
+      "grad_norm": 1.1655443906784058,
+      "learning_rate": 1.5098039215686274e-05,
+      "loss": 0.519,
+      "step": 17459
+    },
+    {
+      "epoch": 22.3488,
+      "grad_norm": 1.1244604587554932,
+      "learning_rate": 1.5096038415366148e-05,
+      "loss": 0.4697,
+      "step": 17460
+    },
+    {
+      "epoch": 22.35008,
+      "grad_norm": 1.1628854274749756,
+      "learning_rate": 1.509403761504602e-05,
+      "loss": 0.5347,
+      "step": 17461
+    },
+    {
+      "epoch": 22.35136,
+      "grad_norm": 1.0745189189910889,
+      "learning_rate": 1.509203681472589e-05,
+      "loss": 0.4332,
+      "step": 17462
+    },
+    {
+      "epoch": 22.35264,
+      "grad_norm": 1.117801308631897,
+      "learning_rate": 1.5090036014405762e-05,
+      "loss": 0.439,
+      "step": 17463
+    },
+    {
+      "epoch": 22.35392,
+      "grad_norm": 1.1143686771392822,
+      "learning_rate": 1.5088035214085635e-05,
+      "loss": 0.4714,
+      "step": 17464
+    },
+    {
+      "epoch": 22.3552,
+      "grad_norm": 1.1027582883834839,
+      "learning_rate": 1.5086034413765507e-05,
+      "loss": 0.4566,
+      "step": 17465
+    },
+    {
+      "epoch": 22.35648,
+      "grad_norm": 1.114611268043518,
+      "learning_rate": 1.5084033613445377e-05,
+      "loss": 0.5007,
+      "step": 17466
+    },
+    {
+      "epoch": 22.35776,
+      "grad_norm": 1.12099027633667,
+      "learning_rate": 1.5082032813125253e-05,
+      "loss": 0.4968,
+      "step": 17467
+    },
+    {
+      "epoch": 22.35904,
+      "grad_norm": 1.1713063716888428,
+      "learning_rate": 1.5080032012805123e-05,
+      "loss": 0.5194,
+      "step": 17468
+    },
+    {
+      "epoch": 22.36032,
+      "grad_norm": 1.1638789176940918,
+      "learning_rate": 1.5078031212484995e-05,
+      "loss": 0.5699,
+      "step": 17469
+    },
+    {
+      "epoch": 22.3616,
+      "grad_norm": 1.118815541267395,
+      "learning_rate": 1.5076030412164865e-05,
+      "loss": 0.4611,
+      "step": 17470
+    },
+    {
+      "epoch": 22.36288,
+      "grad_norm": 1.087066650390625,
+      "learning_rate": 1.507402961184474e-05,
+      "loss": 0.4478,
+      "step": 17471
+    },
+    {
+      "epoch": 22.36416,
+      "grad_norm": 1.1488027572631836,
+      "learning_rate": 1.507202881152461e-05,
+      "loss": 0.4589,
+      "step": 17472
+    },
+    {
+      "epoch": 22.36544,
+      "grad_norm": 1.1603463888168335,
+      "learning_rate": 1.5070028011204482e-05,
+      "loss": 0.5202,
+      "step": 17473
+    },
+    {
+      "epoch": 22.36672,
+      "grad_norm": 1.1120957136154175,
+      "learning_rate": 1.5068027210884356e-05,
+      "loss": 0.4665,
+      "step": 17474
+    },
+    {
+      "epoch": 22.368,
+      "grad_norm": 1.0825188159942627,
+      "learning_rate": 1.5066026410564227e-05,
+      "loss": 0.4775,
+      "step": 17475
+    },
+    {
+      "epoch": 22.36928,
+      "grad_norm": 1.1528738737106323,
+      "learning_rate": 1.5064025610244098e-05,
+      "loss": 0.529,
+      "step": 17476
+    },
+    {
+      "epoch": 22.37056,
+      "grad_norm": 1.176680326461792,
+      "learning_rate": 1.506202480992397e-05,
+      "loss": 0.4932,
+      "step": 17477
+    },
+    {
+      "epoch": 22.37184,
+      "grad_norm": 1.1375823020935059,
+      "learning_rate": 1.5060024009603843e-05,
+      "loss": 0.484,
+      "step": 17478
+    },
+    {
+      "epoch": 22.37312,
+      "grad_norm": 1.109999179840088,
+      "learning_rate": 1.5058023209283715e-05,
+      "loss": 0.4834,
+      "step": 17479
+    },
+    {
+      "epoch": 22.3744,
+      "grad_norm": 1.1392533779144287,
+      "learning_rate": 1.5056022408963585e-05,
+      "loss": 0.4511,
+      "step": 17480
+    },
+    {
+      "epoch": 22.37568,
+      "grad_norm": 1.074696660041809,
+      "learning_rate": 1.5054021608643459e-05,
+      "loss": 0.4497,
+      "step": 17481
+    },
+    {
+      "epoch": 22.37696,
+      "grad_norm": 1.1152911186218262,
+      "learning_rate": 1.505202080832333e-05,
+      "loss": 0.456,
+      "step": 17482
+    },
+    {
+      "epoch": 22.37824,
+      "grad_norm": 1.1230454444885254,
+      "learning_rate": 1.5050020008003202e-05,
+      "loss": 0.4766,
+      "step": 17483
+    },
+    {
+      "epoch": 22.37952,
+      "grad_norm": 1.068898320198059,
+      "learning_rate": 1.5048019207683072e-05,
+      "loss": 0.425,
+      "step": 17484
+    },
+    {
+      "epoch": 22.3808,
+      "grad_norm": 1.1210873126983643,
+      "learning_rate": 1.5046018407362946e-05,
+      "loss": 0.5016,
+      "step": 17485
+    },
+    {
+      "epoch": 22.38208,
+      "grad_norm": 1.0686876773834229,
+      "learning_rate": 1.5044017607042818e-05,
+      "loss": 0.4212,
+      "step": 17486
+    },
+    {
+      "epoch": 22.38336,
+      "grad_norm": 1.0619049072265625,
+      "learning_rate": 1.504201680672269e-05,
+      "loss": 0.45,
+      "step": 17487
+    },
+    {
+      "epoch": 22.38464,
+      "grad_norm": 1.1653971672058105,
+      "learning_rate": 1.5040016006402563e-05,
+      "loss": 0.4943,
+      "step": 17488
+    },
+    {
+      "epoch": 22.38592,
+      "grad_norm": 1.1745988130569458,
+      "learning_rate": 1.5038015206082433e-05,
+      "loss": 0.4886,
+      "step": 17489
+    },
+    {
+      "epoch": 22.3872,
+      "grad_norm": 1.1317180395126343,
+      "learning_rate": 1.5036014405762305e-05,
+      "loss": 0.4911,
+      "step": 17490
+    },
+    {
+      "epoch": 22.38848,
+      "grad_norm": 1.1715854406356812,
+      "learning_rate": 1.5034013605442177e-05,
+      "loss": 0.4806,
+      "step": 17491
+    },
+    {
+      "epoch": 22.38976,
+      "grad_norm": 1.1153987646102905,
+      "learning_rate": 1.503201280512205e-05,
+      "loss": 0.5112,
+      "step": 17492
+    },
+    {
+      "epoch": 22.39104,
+      "grad_norm": 1.1248096227645874,
+      "learning_rate": 1.503001200480192e-05,
+      "loss": 0.4911,
+      "step": 17493
+    },
+    {
+      "epoch": 22.39232,
+      "grad_norm": 1.1459614038467407,
+      "learning_rate": 1.5028011204481793e-05,
+      "loss": 0.4958,
+      "step": 17494
+    },
+    {
+      "epoch": 22.3936,
+      "grad_norm": 1.153074860572815,
+      "learning_rate": 1.5026010404161666e-05,
+      "loss": 0.4822,
+      "step": 17495
+    },
+    {
+      "epoch": 22.39488,
+      "grad_norm": 1.1298890113830566,
+      "learning_rate": 1.5024009603841538e-05,
+      "loss": 0.4773,
+      "step": 17496
+    },
+    {
+      "epoch": 22.39616,
+      "grad_norm": 1.125186800956726,
+      "learning_rate": 1.5022008803521408e-05,
+      "loss": 0.4854,
+      "step": 17497
+    },
+    {
+      "epoch": 22.39744,
+      "grad_norm": 1.1420212984085083,
+      "learning_rate": 1.502000800320128e-05,
+      "loss": 0.5225,
+      "step": 17498
+    },
+    {
+      "epoch": 22.39872,
+      "grad_norm": 1.0702089071273804,
+      "learning_rate": 1.5018007202881154e-05,
+      "loss": 0.4647,
+      "step": 17499
+    },
+    {
+      "epoch": 22.4,
+      "grad_norm": 1.172632098197937,
+      "learning_rate": 1.5016006402561026e-05,
+      "loss": 0.5238,
+      "step": 17500
+    },
+    {
+      "epoch": 22.40128,
+      "grad_norm": 1.1643844842910767,
+      "learning_rate": 1.5014005602240896e-05,
+      "loss": 0.5163,
+      "step": 17501
+    },
+    {
+      "epoch": 22.40256,
+      "grad_norm": 1.044021725654602,
+      "learning_rate": 1.5012004801920771e-05,
+      "loss": 0.4328,
+      "step": 17502
+    },
+    {
+      "epoch": 22.40384,
+      "grad_norm": 1.0835877656936646,
+      "learning_rate": 1.5010004001600641e-05,
+      "loss": 0.4673,
+      "step": 17503
+    },
+    {
+      "epoch": 22.40512,
+      "grad_norm": 1.095995306968689,
+      "learning_rate": 1.5008003201280513e-05,
+      "loss": 0.5019,
+      "step": 17504
+    },
+    {
+      "epoch": 22.4064,
+      "grad_norm": 1.1427990198135376,
+      "learning_rate": 1.5006002400960383e-05,
+      "loss": 0.4884,
+      "step": 17505
+    },
+    {
+      "epoch": 22.40768,
+      "grad_norm": 1.1321765184402466,
+      "learning_rate": 1.5004001600640258e-05,
+      "loss": 0.4775,
+      "step": 17506
+    },
+    {
+      "epoch": 22.40896,
+      "grad_norm": 1.107414722442627,
+      "learning_rate": 1.5002000800320129e-05,
+      "loss": 0.4403,
+      "step": 17507
+    },
+    {
+      "epoch": 22.41024,
+      "grad_norm": 1.1043665409088135,
+      "learning_rate": 1.5e-05,
+      "loss": 0.4972,
+      "step": 17508
+    },
+    {
+      "epoch": 22.41152,
+      "grad_norm": 1.120118260383606,
+      "learning_rate": 1.499799919967987e-05,
+      "loss": 0.4897,
+      "step": 17509
+    },
+    {
+      "epoch": 22.4128,
+      "grad_norm": 1.11980140209198,
+      "learning_rate": 1.4995998399359746e-05,
+      "loss": 0.4861,
+      "step": 17510
+    },
+    {
+      "epoch": 22.41408,
+      "grad_norm": 1.1767655611038208,
+      "learning_rate": 1.4993997599039616e-05,
+      "loss": 0.518,
+      "step": 17511
+    },
+    {
+      "epoch": 22.41536,
+      "grad_norm": 1.1098543405532837,
+      "learning_rate": 1.4991996798719488e-05,
+      "loss": 0.472,
+      "step": 17512
+    },
+    {
+      "epoch": 22.41664,
+      "grad_norm": 1.2055127620697021,
+      "learning_rate": 1.4989995998399361e-05,
+      "loss": 0.5145,
+      "step": 17513
+    },
+    {
+      "epoch": 22.41792,
+      "grad_norm": 1.160184621810913,
+      "learning_rate": 1.4987995198079233e-05,
+      "loss": 0.5199,
+      "step": 17514
+    },
+    {
+      "epoch": 22.4192,
+      "grad_norm": 1.1769219636917114,
+      "learning_rate": 1.4985994397759103e-05,
+      "loss": 0.5057,
+      "step": 17515
+    },
+    {
+      "epoch": 22.42048,
+      "grad_norm": 1.1221567392349243,
+      "learning_rate": 1.4983993597438975e-05,
+      "loss": 0.4786,
+      "step": 17516
+    },
+    {
+      "epoch": 22.42176,
+      "grad_norm": 1.167410135269165,
+      "learning_rate": 1.4981992797118849e-05,
+      "loss": 0.4793,
+      "step": 17517
+    },
+    {
+      "epoch": 22.42304,
+      "grad_norm": 1.1696871519088745,
+      "learning_rate": 1.497999199679872e-05,
+      "loss": 0.4783,
+      "step": 17518
+    },
+    {
+      "epoch": 22.42432,
+      "grad_norm": 1.1420179605484009,
+      "learning_rate": 1.497799119647859e-05,
+      "loss": 0.47,
+      "step": 17519
+    },
+    {
+      "epoch": 22.4256,
+      "grad_norm": 1.1280018091201782,
+      "learning_rate": 1.4975990396158466e-05,
+      "loss": 0.5005,
+      "step": 17520
+    },
+    {
+      "epoch": 22.42688,
+      "grad_norm": 1.126597285270691,
+      "learning_rate": 1.4973989595838336e-05,
+      "loss": 0.4808,
+      "step": 17521
+    },
+    {
+      "epoch": 22.42816,
+      "grad_norm": 1.1016881465911865,
+      "learning_rate": 1.4971988795518208e-05,
+      "loss": 0.4692,
+      "step": 17522
+    },
+    {
+      "epoch": 22.42944,
+      "grad_norm": 1.1398890018463135,
+      "learning_rate": 1.4969987995198078e-05,
+      "loss": 0.5017,
+      "step": 17523
+    },
+    {
+      "epoch": 22.43072,
+      "grad_norm": 1.1369463205337524,
+      "learning_rate": 1.4967987194877953e-05,
+      "loss": 0.4744,
+      "step": 17524
+    },
+    {
+      "epoch": 22.432,
+      "grad_norm": 1.1468337774276733,
+      "learning_rate": 1.4965986394557824e-05,
+      "loss": 0.4802,
+      "step": 17525
+    },
+    {
+      "epoch": 22.43328,
+      "grad_norm": 1.184463620185852,
+      "learning_rate": 1.4963985594237695e-05,
+      "loss": 0.5436,
+      "step": 17526
+    },
+    {
+      "epoch": 22.43456,
+      "grad_norm": 1.0913583040237427,
+      "learning_rate": 1.4961984793917569e-05,
+      "loss": 0.4606,
+      "step": 17527
+    },
+    {
+      "epoch": 22.43584,
+      "grad_norm": 1.1201003789901733,
+      "learning_rate": 1.4959983993597441e-05,
+      "loss": 0.4782,
+      "step": 17528
+    },
+    {
+      "epoch": 22.43712,
+      "grad_norm": 1.1170686483383179,
+      "learning_rate": 1.4957983193277311e-05,
+      "loss": 0.4602,
+      "step": 17529
+    },
+    {
+      "epoch": 22.4384,
+      "grad_norm": 1.1606577634811401,
+      "learning_rate": 1.4955982392957183e-05,
+      "loss": 0.501,
+      "step": 17530
+    },
+    {
+      "epoch": 22.43968,
+      "grad_norm": 1.1875215768814087,
+      "learning_rate": 1.4953981592637056e-05,
+      "loss": 0.4986,
+      "step": 17531
+    },
+    {
+      "epoch": 22.44096,
+      "grad_norm": 1.1248286962509155,
+      "learning_rate": 1.4951980792316928e-05,
+      "loss": 0.51,
+      "step": 17532
+    },
+    {
+      "epoch": 22.44224,
+      "grad_norm": 1.106138825416565,
+      "learning_rate": 1.4949979991996798e-05,
+      "loss": 0.4864,
+      "step": 17533
+    },
+    {
+      "epoch": 22.44352,
+      "grad_norm": 1.1583833694458008,
+      "learning_rate": 1.4947979191676672e-05,
+      "loss": 0.4824,
+      "step": 17534
+    },
+    {
+      "epoch": 22.4448,
+      "grad_norm": 1.1970548629760742,
+      "learning_rate": 1.4945978391356544e-05,
+      "loss": 0.5043,
+      "step": 17535
+    },
+    {
+      "epoch": 22.44608,
+      "grad_norm": 1.1255377531051636,
+      "learning_rate": 1.4943977591036416e-05,
+      "loss": 0.4703,
+      "step": 17536
+    },
+    {
+      "epoch": 22.44736,
+      "grad_norm": 1.137693166732788,
+      "learning_rate": 1.4941976790716286e-05,
+      "loss": 0.5076,
+      "step": 17537
+    },
+    {
+      "epoch": 22.44864,
+      "grad_norm": 1.099455714225769,
+      "learning_rate": 1.493997599039616e-05,
+      "loss": 0.4662,
+      "step": 17538
+    },
+    {
+      "epoch": 22.44992,
+      "grad_norm": 1.1596556901931763,
+      "learning_rate": 1.4937975190076031e-05,
+      "loss": 0.4969,
+      "step": 17539
+    },
+    {
+      "epoch": 22.4512,
+      "grad_norm": 1.1630852222442627,
+      "learning_rate": 1.4935974389755903e-05,
+      "loss": 0.5131,
+      "step": 17540
+    },
+    {
+      "epoch": 22.45248,
+      "grad_norm": 1.1254620552062988,
+      "learning_rate": 1.4933973589435777e-05,
+      "loss": 0.5012,
+      "step": 17541
+    },
+    {
+      "epoch": 22.45376,
+      "grad_norm": 1.1461201906204224,
+      "learning_rate": 1.4931972789115647e-05,
+      "loss": 0.4934,
+      "step": 17542
+    },
+    {
+      "epoch": 22.45504,
+      "grad_norm": 1.1740776300430298,
+      "learning_rate": 1.4929971988795519e-05,
+      "loss": 0.4746,
+      "step": 17543
+    },
+    {
+      "epoch": 22.45632,
+      "grad_norm": 1.1553022861480713,
+      "learning_rate": 1.492797118847539e-05,
+      "loss": 0.485,
+      "step": 17544
+    },
+    {
+      "epoch": 22.4576,
+      "grad_norm": 1.1020857095718384,
+      "learning_rate": 1.4925970388155264e-05,
+      "loss": 0.4326,
+      "step": 17545
+    },
+    {
+      "epoch": 22.45888,
+      "grad_norm": 1.125576376914978,
+      "learning_rate": 1.4923969587835134e-05,
+      "loss": 0.4976,
+      "step": 17546
+    },
+    {
+      "epoch": 22.46016,
+      "grad_norm": 1.1211215257644653,
+      "learning_rate": 1.4921968787515006e-05,
+      "loss": 0.4824,
+      "step": 17547
+    },
+    {
+      "epoch": 22.46144,
+      "grad_norm": 1.1188104152679443,
+      "learning_rate": 1.491996798719488e-05,
+      "loss": 0.4772,
+      "step": 17548
+    },
+    {
+      "epoch": 22.46272,
+      "grad_norm": 1.074813723564148,
+      "learning_rate": 1.4917967186874752e-05,
+      "loss": 0.4494,
+      "step": 17549
+    },
+    {
+      "epoch": 22.464,
+      "grad_norm": 1.0762443542480469,
+      "learning_rate": 1.4915966386554622e-05,
+      "loss": 0.4373,
+      "step": 17550
+    },
+    {
+      "epoch": 22.46528,
+      "grad_norm": 1.1819483041763306,
+      "learning_rate": 1.4913965586234494e-05,
+      "loss": 0.5661,
+      "step": 17551
+    },
+    {
+      "epoch": 22.46656,
+      "grad_norm": 1.104282259941101,
+      "learning_rate": 1.4911964785914367e-05,
+      "loss": 0.4914,
+      "step": 17552
+    },
+    {
+      "epoch": 22.46784,
+      "grad_norm": 1.1156835556030273,
+      "learning_rate": 1.4909963985594239e-05,
+      "loss": 0.4791,
+      "step": 17553
+    },
+    {
+      "epoch": 22.46912,
+      "grad_norm": 1.198472023010254,
+      "learning_rate": 1.490796318527411e-05,
+      "loss": 0.4995,
+      "step": 17554
+    },
+    {
+      "epoch": 22.4704,
+      "grad_norm": 1.1563735008239746,
+      "learning_rate": 1.4905962384953984e-05,
+      "loss": 0.4815,
+      "step": 17555
+    },
+    {
+      "epoch": 22.47168,
+      "grad_norm": 1.1188043355941772,
+      "learning_rate": 1.4903961584633855e-05,
+      "loss": 0.4759,
+      "step": 17556
+    },
+    {
+      "epoch": 22.47296,
+      "grad_norm": 1.130955457687378,
+      "learning_rate": 1.4901960784313726e-05,
+      "loss": 0.4884,
+      "step": 17557
+    },
+    {
+      "epoch": 22.47424,
+      "grad_norm": 1.1355321407318115,
+      "learning_rate": 1.4899959983993597e-05,
+      "loss": 0.4986,
+      "step": 17558
+    },
+    {
+      "epoch": 22.47552,
+      "grad_norm": 1.1446267366409302,
+      "learning_rate": 1.4897959183673472e-05,
+      "loss": 0.4902,
+      "step": 17559
+    },
+    {
+      "epoch": 22.4768,
+      "grad_norm": 1.1643836498260498,
+      "learning_rate": 1.4895958383353342e-05,
+      "loss": 0.4702,
+      "step": 17560
+    },
+    {
+      "epoch": 22.47808,
+      "grad_norm": 1.1795804500579834,
+      "learning_rate": 1.4893957583033214e-05,
+      "loss": 0.4978,
+      "step": 17561
+    },
+    {
+      "epoch": 22.47936,
+      "grad_norm": 1.0704967975616455,
+      "learning_rate": 1.4891956782713084e-05,
+      "loss": 0.4907,
+      "step": 17562
+    },
+    {
+      "epoch": 22.48064,
+      "grad_norm": 1.0626331567764282,
+      "learning_rate": 1.488995598239296e-05,
+      "loss": 0.4522,
+      "step": 17563
+    },
+    {
+      "epoch": 22.48192,
+      "grad_norm": 1.0941236019134521,
+      "learning_rate": 1.488795518207283e-05,
+      "loss": 0.4381,
+      "step": 17564
+    },
+    {
+      "epoch": 22.4832,
+      "grad_norm": 1.100813388824463,
+      "learning_rate": 1.4885954381752701e-05,
+      "loss": 0.5071,
+      "step": 17565
+    },
+    {
+      "epoch": 22.48448,
+      "grad_norm": 1.2816932201385498,
+      "learning_rate": 1.4883953581432575e-05,
+      "loss": 0.4937,
+      "step": 17566
+    },
+    {
+      "epoch": 22.48576,
+      "grad_norm": 1.2014461755752563,
+      "learning_rate": 1.4881952781112447e-05,
+      "loss": 0.4908,
+      "step": 17567
+    },
+    {
+      "epoch": 22.48704,
+      "grad_norm": 1.0948677062988281,
+      "learning_rate": 1.4879951980792317e-05,
+      "loss": 0.4935,
+      "step": 17568
+    },
+    {
+      "epoch": 22.48832,
+      "grad_norm": 1.114828109741211,
+      "learning_rate": 1.4877951180472189e-05,
+      "loss": 0.4639,
+      "step": 17569
+    },
+    {
+      "epoch": 22.4896,
+      "grad_norm": 1.118323802947998,
+      "learning_rate": 1.4875950380152062e-05,
+      "loss": 0.5029,
+      "step": 17570
+    },
+    {
+      "epoch": 22.49088,
+      "grad_norm": 1.11355721950531,
+      "learning_rate": 1.4873949579831934e-05,
+      "loss": 0.4878,
+      "step": 17571
+    },
+    {
+      "epoch": 22.49216,
+      "grad_norm": 1.1527410745620728,
+      "learning_rate": 1.4871948779511804e-05,
+      "loss": 0.5242,
+      "step": 17572
+    },
+    {
+      "epoch": 22.49344,
+      "grad_norm": 1.1299474239349365,
+      "learning_rate": 1.4869947979191678e-05,
+      "loss": 0.4836,
+      "step": 17573
+    },
+    {
+      "epoch": 22.49472,
+      "grad_norm": 1.0945117473602295,
+      "learning_rate": 1.486794717887155e-05,
+      "loss": 0.4503,
+      "step": 17574
+    },
+    {
+      "epoch": 22.496,
+      "grad_norm": 1.0921000242233276,
+      "learning_rate": 1.4865946378551422e-05,
+      "loss": 0.5056,
+      "step": 17575
+    },
+    {
+      "epoch": 22.49728,
+      "grad_norm": 1.1385952234268188,
+      "learning_rate": 1.4863945578231292e-05,
+      "loss": 0.5017,
+      "step": 17576
+    },
+    {
+      "epoch": 22.49856,
+      "grad_norm": 1.1188937425613403,
+      "learning_rate": 1.4861944777911165e-05,
+      "loss": 0.4629,
+      "step": 17577
+    },
+    {
+      "epoch": 22.49984,
+      "grad_norm": 1.1394176483154297,
+      "learning_rate": 1.4859943977591037e-05,
+      "loss": 0.483,
+      "step": 17578
+    },
+    {
+      "epoch": 22.50112,
+      "grad_norm": 1.0991370677947998,
+      "learning_rate": 1.4857943177270909e-05,
+      "loss": 0.49,
+      "step": 17579
+    },
+    {
+      "epoch": 22.5024,
+      "grad_norm": 1.1252307891845703,
+      "learning_rate": 1.4855942376950783e-05,
+      "loss": 0.4859,
+      "step": 17580
+    },
+    {
+      "epoch": 22.50368,
+      "grad_norm": 1.1384406089782715,
+      "learning_rate": 1.4853941576630653e-05,
+      "loss": 0.4985,
+      "step": 17581
+    },
+    {
+      "epoch": 22.50496,
+      "grad_norm": 1.1817059516906738,
+      "learning_rate": 1.4851940776310525e-05,
+      "loss": 0.4953,
+      "step": 17582
+    },
+    {
+      "epoch": 22.50624,
+      "grad_norm": 1.1219290494918823,
+      "learning_rate": 1.4849939975990396e-05,
+      "loss": 0.4729,
+      "step": 17583
+    },
+    {
+      "epoch": 22.50752,
+      "grad_norm": 1.1384577751159668,
+      "learning_rate": 1.484793917567027e-05,
+      "loss": 0.5415,
+      "step": 17584
+    },
+    {
+      "epoch": 22.5088,
+      "grad_norm": 1.1150089502334595,
+      "learning_rate": 1.484593837535014e-05,
+      "loss": 0.4845,
+      "step": 17585
+    },
+    {
+      "epoch": 22.51008,
+      "grad_norm": 1.1471636295318604,
+      "learning_rate": 1.4843937575030012e-05,
+      "loss": 0.4882,
+      "step": 17586
+    },
+    {
+      "epoch": 22.51136,
+      "grad_norm": 1.1583935022354126,
+      "learning_rate": 1.4841936774709886e-05,
+      "loss": 0.5159,
+      "step": 17587
+    },
+    {
+      "epoch": 22.51264,
+      "grad_norm": 1.1290314197540283,
+      "learning_rate": 1.4839935974389757e-05,
+      "loss": 0.4779,
+      "step": 17588
+    },
+    {
+      "epoch": 22.51392,
+      "grad_norm": 1.116462230682373,
+      "learning_rate": 1.4837935174069628e-05,
+      "loss": 0.4735,
+      "step": 17589
+    },
+    {
+      "epoch": 22.5152,
+      "grad_norm": 1.1295406818389893,
+      "learning_rate": 1.48359343737495e-05,
+      "loss": 0.4985,
+      "step": 17590
+    },
+    {
+      "epoch": 22.51648,
+      "grad_norm": 1.0936286449432373,
+      "learning_rate": 1.4833933573429373e-05,
+      "loss": 0.4902,
+      "step": 17591
+    },
+    {
+      "epoch": 22.51776,
+      "grad_norm": 1.1139616966247559,
+      "learning_rate": 1.4831932773109245e-05,
+      "loss": 0.5094,
+      "step": 17592
+    },
+    {
+      "epoch": 22.51904,
+      "grad_norm": 1.0469192266464233,
+      "learning_rate": 1.4829931972789115e-05,
+      "loss": 0.4364,
+      "step": 17593
+    },
+    {
+      "epoch": 22.52032,
+      "grad_norm": 1.0828360319137573,
+      "learning_rate": 1.482793117246899e-05,
+      "loss": 0.483,
+      "step": 17594
+    },
+    {
+      "epoch": 22.5216,
+      "grad_norm": 1.1085882186889648,
+      "learning_rate": 1.482593037214886e-05,
+      "loss": 0.5008,
+      "step": 17595
+    },
+    {
+      "epoch": 22.52288,
+      "grad_norm": 1.1540615558624268,
+      "learning_rate": 1.4823929571828732e-05,
+      "loss": 0.4861,
+      "step": 17596
+    },
+    {
+      "epoch": 22.52416,
+      "grad_norm": 1.106768012046814,
+      "learning_rate": 1.4821928771508602e-05,
+      "loss": 0.4946,
+      "step": 17597
+    },
+    {
+      "epoch": 22.52544,
+      "grad_norm": 1.1145246028900146,
+      "learning_rate": 1.4819927971188478e-05,
+      "loss": 0.4877,
+      "step": 17598
+    },
+    {
+      "epoch": 22.52672,
+      "grad_norm": 1.1328237056732178,
+      "learning_rate": 1.4817927170868348e-05,
+      "loss": 0.4733,
+      "step": 17599
+    },
+    {
+      "epoch": 22.528,
+      "grad_norm": 1.1732732057571411,
+      "learning_rate": 1.481592637054822e-05,
+      "loss": 0.4773,
+      "step": 17600
+    },
+    {
+      "epoch": 22.52928,
+      "grad_norm": 1.093930721282959,
+      "learning_rate": 1.4813925570228093e-05,
+      "loss": 0.4651,
+      "step": 17601
+    },
+    {
+      "epoch": 22.53056,
+      "grad_norm": 1.0944005250930786,
+      "learning_rate": 1.4811924769907965e-05,
+      "loss": 0.4874,
+      "step": 17602
+    },
+    {
+      "epoch": 22.53184,
+      "grad_norm": 1.1620426177978516,
+      "learning_rate": 1.4809923969587835e-05,
+      "loss": 0.4759,
+      "step": 17603
+    },
+    {
+      "epoch": 22.53312,
+      "grad_norm": 1.131779670715332,
+      "learning_rate": 1.4807923169267707e-05,
+      "loss": 0.4777,
+      "step": 17604
+    },
+    {
+      "epoch": 22.5344,
+      "grad_norm": 1.1112638711929321,
+      "learning_rate": 1.480592236894758e-05,
+      "loss": 0.4639,
+      "step": 17605
+    },
+    {
+      "epoch": 22.53568,
+      "grad_norm": 1.039440393447876,
+      "learning_rate": 1.4803921568627453e-05,
+      "loss": 0.4314,
+      "step": 17606
+    },
+    {
+      "epoch": 22.53696,
+      "grad_norm": 1.0729459524154663,
+      "learning_rate": 1.4801920768307323e-05,
+      "loss": 0.4708,
+      "step": 17607
+    },
+    {
+      "epoch": 22.538240000000002,
+      "grad_norm": 1.1430833339691162,
+      "learning_rate": 1.4799919967987196e-05,
+      "loss": 0.4659,
+      "step": 17608
+    },
+    {
+      "epoch": 22.53952,
+      "grad_norm": 1.1749787330627441,
+      "learning_rate": 1.4797919167667068e-05,
+      "loss": 0.5147,
+      "step": 17609
+    },
+    {
+      "epoch": 22.5408,
+      "grad_norm": 1.1035168170928955,
+      "learning_rate": 1.479591836734694e-05,
+      "loss": 0.4272,
+      "step": 17610
+    },
+    {
+      "epoch": 22.54208,
+      "grad_norm": 1.131881594657898,
+      "learning_rate": 1.479391756702681e-05,
+      "loss": 0.4769,
+      "step": 17611
+    },
+    {
+      "epoch": 22.54336,
+      "grad_norm": 1.0516685247421265,
+      "learning_rate": 1.4791916766706684e-05,
+      "loss": 0.4525,
+      "step": 17612
+    },
+    {
+      "epoch": 22.54464,
+      "grad_norm": 1.0958613157272339,
+      "learning_rate": 1.4789915966386556e-05,
+      "loss": 0.4753,
+      "step": 17613
+    },
+    {
+      "epoch": 22.54592,
+      "grad_norm": 1.1349154710769653,
+      "learning_rate": 1.4787915166066427e-05,
+      "loss": 0.4946,
+      "step": 17614
+    },
+    {
+      "epoch": 22.5472,
+      "grad_norm": 1.055051565170288,
+      "learning_rate": 1.4785914365746298e-05,
+      "loss": 0.4331,
+      "step": 17615
+    },
+    {
+      "epoch": 22.54848,
+      "grad_norm": 1.1691598892211914,
+      "learning_rate": 1.4783913565426171e-05,
+      "loss": 0.5193,
+      "step": 17616
+    },
+    {
+      "epoch": 22.54976,
+      "grad_norm": 1.0969427824020386,
+      "learning_rate": 1.4781912765106043e-05,
+      "loss": 0.4621,
+      "step": 17617
+    },
+    {
+      "epoch": 22.55104,
+      "grad_norm": 1.1640121936798096,
+      "learning_rate": 1.4779911964785915e-05,
+      "loss": 0.4855,
+      "step": 17618
+    },
+    {
+      "epoch": 22.55232,
+      "grad_norm": 1.2234057188034058,
+      "learning_rate": 1.4777911164465788e-05,
+      "loss": 0.4956,
+      "step": 17619
+    },
+    {
+      "epoch": 22.5536,
+      "grad_norm": 1.1525243520736694,
+      "learning_rate": 1.4775910364145659e-05,
+      "loss": 0.4738,
+      "step": 17620
+    },
+    {
+      "epoch": 22.55488,
+      "grad_norm": 1.1218520402908325,
+      "learning_rate": 1.477390956382553e-05,
+      "loss": 0.4954,
+      "step": 17621
+    },
+    {
+      "epoch": 22.55616,
+      "grad_norm": 1.1429941654205322,
+      "learning_rate": 1.4771908763505402e-05,
+      "loss": 0.4926,
+      "step": 17622
+    },
+    {
+      "epoch": 22.55744,
+      "grad_norm": 1.072890043258667,
+      "learning_rate": 1.4769907963185276e-05,
+      "loss": 0.4441,
+      "step": 17623
+    },
+    {
+      "epoch": 22.55872,
+      "grad_norm": 1.085209608078003,
+      "learning_rate": 1.4767907162865146e-05,
+      "loss": 0.4751,
+      "step": 17624
+    },
+    {
+      "epoch": 22.56,
+      "grad_norm": 1.085693597793579,
+      "learning_rate": 1.4765906362545018e-05,
+      "loss": 0.4678,
+      "step": 17625
+    },
+    {
+      "epoch": 22.56128,
+      "grad_norm": 1.1705639362335205,
+      "learning_rate": 1.4763905562224891e-05,
+      "loss": 0.4686,
+      "step": 17626
+    },
+    {
+      "epoch": 22.56256,
+      "grad_norm": 1.130325198173523,
+      "learning_rate": 1.4761904761904763e-05,
+      "loss": 0.5101,
+      "step": 17627
+    },
+    {
+      "epoch": 22.56384,
+      "grad_norm": 1.1805752515792847,
+      "learning_rate": 1.4759903961584633e-05,
+      "loss": 0.5458,
+      "step": 17628
+    },
+    {
+      "epoch": 22.56512,
+      "grad_norm": 1.1645885705947876,
+      "learning_rate": 1.4757903161264505e-05,
+      "loss": 0.5055,
+      "step": 17629
+    },
+    {
+      "epoch": 22.5664,
+      "grad_norm": 1.2138627767562866,
+      "learning_rate": 1.4755902360944379e-05,
+      "loss": 0.5324,
+      "step": 17630
+    },
+    {
+      "epoch": 22.56768,
+      "grad_norm": 1.119138240814209,
+      "learning_rate": 1.475390156062425e-05,
+      "loss": 0.4681,
+      "step": 17631
+    },
+    {
+      "epoch": 22.56896,
+      "grad_norm": 1.09098482131958,
+      "learning_rate": 1.475190076030412e-05,
+      "loss": 0.4839,
+      "step": 17632
+    },
+    {
+      "epoch": 22.57024,
+      "grad_norm": 1.1301281452178955,
+      "learning_rate": 1.4749899959983996e-05,
+      "loss": 0.4629,
+      "step": 17633
+    },
+    {
+      "epoch": 22.57152,
+      "grad_norm": 1.1431349515914917,
+      "learning_rate": 1.4747899159663866e-05,
+      "loss": 0.5064,
+      "step": 17634
+    },
+    {
+      "epoch": 22.5728,
+      "grad_norm": 1.023766279220581,
+      "learning_rate": 1.4745898359343738e-05,
+      "loss": 0.4402,
+      "step": 17635
+    },
+    {
+      "epoch": 22.57408,
+      "grad_norm": 1.1107232570648193,
+      "learning_rate": 1.4743897559023608e-05,
+      "loss": 0.4486,
+      "step": 17636
+    },
+    {
+      "epoch": 22.57536,
+      "grad_norm": 1.0915488004684448,
+      "learning_rate": 1.4741896758703483e-05,
+      "loss": 0.4598,
+      "step": 17637
+    },
+    {
+      "epoch": 22.57664,
+      "grad_norm": 1.1454777717590332,
+      "learning_rate": 1.4739895958383354e-05,
+      "loss": 0.5209,
+      "step": 17638
+    },
+    {
+      "epoch": 22.57792,
+      "grad_norm": 1.0833492279052734,
+      "learning_rate": 1.4737895158063225e-05,
+      "loss": 0.5036,
+      "step": 17639
+    },
+    {
+      "epoch": 22.5792,
+      "grad_norm": 1.115124225616455,
+      "learning_rate": 1.4735894357743099e-05,
+      "loss": 0.5157,
+      "step": 17640
+    },
+    {
+      "epoch": 22.58048,
+      "grad_norm": 1.121090292930603,
+      "learning_rate": 1.4733893557422971e-05,
+      "loss": 0.5237,
+      "step": 17641
+    },
+    {
+      "epoch": 22.58176,
+      "grad_norm": 1.0920593738555908,
+      "learning_rate": 1.4731892757102841e-05,
+      "loss": 0.4775,
+      "step": 17642
+    },
+    {
+      "epoch": 22.58304,
+      "grad_norm": 1.0497217178344727,
+      "learning_rate": 1.4729891956782713e-05,
+      "loss": 0.4622,
+      "step": 17643
+    },
+    {
+      "epoch": 22.584319999999998,
+      "grad_norm": 1.1190520524978638,
+      "learning_rate": 1.4727891156462586e-05,
+      "loss": 0.4555,
+      "step": 17644
+    },
+    {
+      "epoch": 22.5856,
+      "grad_norm": 1.1475340127944946,
+      "learning_rate": 1.4725890356142458e-05,
+      "loss": 0.5076,
+      "step": 17645
+    },
+    {
+      "epoch": 22.58688,
+      "grad_norm": 1.2062084674835205,
+      "learning_rate": 1.4723889555822328e-05,
+      "loss": 0.4723,
+      "step": 17646
+    },
+    {
+      "epoch": 22.58816,
+      "grad_norm": 1.1473878622055054,
+      "learning_rate": 1.4721888755502202e-05,
+      "loss": 0.522,
+      "step": 17647
+    },
+    {
+      "epoch": 22.58944,
+      "grad_norm": 1.0924144983291626,
+      "learning_rate": 1.4719887955182074e-05,
+      "loss": 0.4567,
+      "step": 17648
+    },
+    {
+      "epoch": 22.59072,
+      "grad_norm": 1.141308069229126,
+      "learning_rate": 1.4717887154861946e-05,
+      "loss": 0.4708,
+      "step": 17649
+    },
+    {
+      "epoch": 22.592,
+      "grad_norm": 1.0392951965332031,
+      "learning_rate": 1.4715886354541816e-05,
+      "loss": 0.4772,
+      "step": 17650
+    },
+    {
+      "epoch": 22.59328,
+      "grad_norm": 1.0937608480453491,
+      "learning_rate": 1.471388555422169e-05,
+      "loss": 0.4681,
+      "step": 17651
+    },
+    {
+      "epoch": 22.59456,
+      "grad_norm": 1.1153881549835205,
+      "learning_rate": 1.4711884753901561e-05,
+      "loss": 0.5083,
+      "step": 17652
+    },
+    {
+      "epoch": 22.59584,
+      "grad_norm": 1.1007659435272217,
+      "learning_rate": 1.4709883953581433e-05,
+      "loss": 0.4735,
+      "step": 17653
+    },
+    {
+      "epoch": 22.59712,
+      "grad_norm": 1.126693844795227,
+      "learning_rate": 1.4707883153261307e-05,
+      "loss": 0.4602,
+      "step": 17654
+    },
+    {
+      "epoch": 22.5984,
+      "grad_norm": 1.103529453277588,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.4932,
+      "step": 17655
+    },
+    {
+      "epoch": 22.59968,
+      "grad_norm": 1.045106053352356,
+      "learning_rate": 1.4703881552621049e-05,
+      "loss": 0.4576,
+      "step": 17656
+    },
+    {
+      "epoch": 22.60096,
+      "grad_norm": 1.0801410675048828,
+      "learning_rate": 1.470188075230092e-05,
+      "loss": 0.4534,
+      "step": 17657
+    },
+    {
+      "epoch": 22.60224,
+      "grad_norm": 1.1122349500656128,
+      "learning_rate": 1.4699879951980794e-05,
+      "loss": 0.4549,
+      "step": 17658
+    },
+    {
+      "epoch": 22.60352,
+      "grad_norm": 1.1067173480987549,
+      "learning_rate": 1.4697879151660664e-05,
+      "loss": 0.478,
+      "step": 17659
+    },
+    {
+      "epoch": 22.6048,
+      "grad_norm": 1.1408361196517944,
+      "learning_rate": 1.4695878351340536e-05,
+      "loss": 0.4798,
+      "step": 17660
+    },
+    {
+      "epoch": 22.60608,
+      "grad_norm": 1.1362731456756592,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 0.4567,
+      "step": 17661
+    },
+    {
+      "epoch": 22.60736,
+      "grad_norm": 1.2117960453033447,
+      "learning_rate": 1.4691876750700282e-05,
+      "loss": 0.5187,
+      "step": 17662
+    },
+    {
+      "epoch": 22.60864,
+      "grad_norm": 1.155263066291809,
+      "learning_rate": 1.4689875950380152e-05,
+      "loss": 0.531,
+      "step": 17663
+    },
+    {
+      "epoch": 22.60992,
+      "grad_norm": 1.1107439994812012,
+      "learning_rate": 1.4687875150060024e-05,
+      "loss": 0.471,
+      "step": 17664
+    },
+    {
+      "epoch": 22.6112,
+      "grad_norm": 1.1272735595703125,
+      "learning_rate": 1.4685874349739897e-05,
+      "loss": 0.4807,
+      "step": 17665
+    },
+    {
+      "epoch": 22.61248,
+      "grad_norm": 1.142755150794983,
+      "learning_rate": 1.4683873549419769e-05,
+      "loss": 0.4807,
+      "step": 17666
+    },
+    {
+      "epoch": 22.61376,
+      "grad_norm": 1.160695195198059,
+      "learning_rate": 1.468187274909964e-05,
+      "loss": 0.5241,
+      "step": 17667
+    },
+    {
+      "epoch": 22.61504,
+      "grad_norm": 1.1584571599960327,
+      "learning_rate": 1.4679871948779514e-05,
+      "loss": 0.4977,
+      "step": 17668
+    },
+    {
+      "epoch": 22.61632,
+      "grad_norm": 1.08279287815094,
+      "learning_rate": 1.4677871148459385e-05,
+      "loss": 0.4404,
+      "step": 17669
+    },
+    {
+      "epoch": 22.6176,
+      "grad_norm": 1.0838192701339722,
+      "learning_rate": 1.4675870348139256e-05,
+      "loss": 0.4565,
+      "step": 17670
+    },
+    {
+      "epoch": 22.61888,
+      "grad_norm": 1.1284055709838867,
+      "learning_rate": 1.4673869547819127e-05,
+      "loss": 0.5155,
+      "step": 17671
+    },
+    {
+      "epoch": 22.62016,
+      "grad_norm": 1.0866196155548096,
+      "learning_rate": 1.4671868747499002e-05,
+      "loss": 0.4731,
+      "step": 17672
+    },
+    {
+      "epoch": 22.62144,
+      "grad_norm": 1.1500025987625122,
+      "learning_rate": 1.4669867947178872e-05,
+      "loss": 0.5287,
+      "step": 17673
+    },
+    {
+      "epoch": 22.62272,
+      "grad_norm": 1.1272227764129639,
+      "learning_rate": 1.4667867146858744e-05,
+      "loss": 0.4799,
+      "step": 17674
+    },
+    {
+      "epoch": 22.624,
+      "grad_norm": 1.1563876867294312,
+      "learning_rate": 1.4665866346538614e-05,
+      "loss": 0.4674,
+      "step": 17675
+    },
+    {
+      "epoch": 22.62528,
+      "grad_norm": 1.1425659656524658,
+      "learning_rate": 1.466386554621849e-05,
+      "loss": 0.4739,
+      "step": 17676
+    },
+    {
+      "epoch": 22.62656,
+      "grad_norm": 1.1223922967910767,
+      "learning_rate": 1.466186474589836e-05,
+      "loss": 0.4803,
+      "step": 17677
+    },
+    {
+      "epoch": 22.62784,
+      "grad_norm": 1.1119972467422485,
+      "learning_rate": 1.4659863945578231e-05,
+      "loss": 0.5031,
+      "step": 17678
+    },
+    {
+      "epoch": 22.62912,
+      "grad_norm": 1.1872590780258179,
+      "learning_rate": 1.4657863145258105e-05,
+      "loss": 0.5111,
+      "step": 17679
+    },
+    {
+      "epoch": 22.6304,
+      "grad_norm": 1.173486351966858,
+      "learning_rate": 1.4655862344937977e-05,
+      "loss": 0.5107,
+      "step": 17680
+    },
+    {
+      "epoch": 22.63168,
+      "grad_norm": 1.1515588760375977,
+      "learning_rate": 1.4653861544617847e-05,
+      "loss": 0.4845,
+      "step": 17681
+    },
+    {
+      "epoch": 22.63296,
+      "grad_norm": 1.1229499578475952,
+      "learning_rate": 1.4651860744297719e-05,
+      "loss": 0.4891,
+      "step": 17682
+    },
+    {
+      "epoch": 22.63424,
+      "grad_norm": 1.1395021677017212,
+      "learning_rate": 1.4649859943977592e-05,
+      "loss": 0.4833,
+      "step": 17683
+    },
+    {
+      "epoch": 22.63552,
+      "grad_norm": 1.1398979425430298,
+      "learning_rate": 1.4647859143657464e-05,
+      "loss": 0.4845,
+      "step": 17684
+    },
+    {
+      "epoch": 22.6368,
+      "grad_norm": 1.1859970092773438,
+      "learning_rate": 1.4645858343337334e-05,
+      "loss": 0.5003,
+      "step": 17685
+    },
+    {
+      "epoch": 22.63808,
+      "grad_norm": 1.1473771333694458,
+      "learning_rate": 1.464385754301721e-05,
+      "loss": 0.4632,
+      "step": 17686
+    },
+    {
+      "epoch": 22.63936,
+      "grad_norm": 1.087823748588562,
+      "learning_rate": 1.464185674269708e-05,
+      "loss": 0.4444,
+      "step": 17687
+    },
+    {
+      "epoch": 22.64064,
+      "grad_norm": 1.0971521139144897,
+      "learning_rate": 1.4639855942376952e-05,
+      "loss": 0.4446,
+      "step": 17688
+    },
+    {
+      "epoch": 22.64192,
+      "grad_norm": 1.1230698823928833,
+      "learning_rate": 1.4637855142056822e-05,
+      "loss": 0.45,
+      "step": 17689
+    },
+    {
+      "epoch": 22.6432,
+      "grad_norm": 1.176695704460144,
+      "learning_rate": 1.4635854341736697e-05,
+      "loss": 0.5211,
+      "step": 17690
+    },
+    {
+      "epoch": 22.64448,
+      "grad_norm": 1.1348381042480469,
+      "learning_rate": 1.4633853541416567e-05,
+      "loss": 0.487,
+      "step": 17691
+    },
+    {
+      "epoch": 22.64576,
+      "grad_norm": 1.1356598138809204,
+      "learning_rate": 1.4631852741096439e-05,
+      "loss": 0.5139,
+      "step": 17692
+    },
+    {
+      "epoch": 22.64704,
+      "grad_norm": 1.204661250114441,
+      "learning_rate": 1.4629851940776313e-05,
+      "loss": 0.5212,
+      "step": 17693
+    },
+    {
+      "epoch": 22.64832,
+      "grad_norm": 1.1288728713989258,
+      "learning_rate": 1.4627851140456184e-05,
+      "loss": 0.5026,
+      "step": 17694
+    },
+    {
+      "epoch": 22.6496,
+      "grad_norm": 1.1189838647842407,
+      "learning_rate": 1.4625850340136055e-05,
+      "loss": 0.4806,
+      "step": 17695
+    },
+    {
+      "epoch": 22.65088,
+      "grad_norm": 1.099488615989685,
+      "learning_rate": 1.4623849539815926e-05,
+      "loss": 0.4657,
+      "step": 17696
+    },
+    {
+      "epoch": 22.65216,
+      "grad_norm": 1.1469444036483765,
+      "learning_rate": 1.46218487394958e-05,
+      "loss": 0.5038,
+      "step": 17697
+    },
+    {
+      "epoch": 22.65344,
+      "grad_norm": 1.1008341312408447,
+      "learning_rate": 1.4619847939175672e-05,
+      "loss": 0.4696,
+      "step": 17698
+    },
+    {
+      "epoch": 22.65472,
+      "grad_norm": 1.1267304420471191,
+      "learning_rate": 1.4617847138855542e-05,
+      "loss": 0.4798,
+      "step": 17699
+    },
+    {
+      "epoch": 22.656,
+      "grad_norm": 1.0712333917617798,
+      "learning_rate": 1.4615846338535416e-05,
+      "loss": 0.4875,
+      "step": 17700
+    },
+    {
+      "epoch": 22.65728,
+      "grad_norm": 1.090133547782898,
+      "learning_rate": 1.4613845538215287e-05,
+      "loss": 0.5221,
+      "step": 17701
+    },
+    {
+      "epoch": 22.65856,
+      "grad_norm": 1.0911118984222412,
+      "learning_rate": 1.461184473789516e-05,
+      "loss": 0.4545,
+      "step": 17702
+    },
+    {
+      "epoch": 22.65984,
+      "grad_norm": 1.1019871234893799,
+      "learning_rate": 1.460984393757503e-05,
+      "loss": 0.5021,
+      "step": 17703
+    },
+    {
+      "epoch": 22.66112,
+      "grad_norm": 1.1478427648544312,
+      "learning_rate": 1.4607843137254903e-05,
+      "loss": 0.4791,
+      "step": 17704
+    },
+    {
+      "epoch": 22.6624,
+      "grad_norm": 1.038714051246643,
+      "learning_rate": 1.4605842336934775e-05,
+      "loss": 0.4635,
+      "step": 17705
+    },
+    {
+      "epoch": 22.66368,
+      "grad_norm": 1.1068960428237915,
+      "learning_rate": 1.4603841536614647e-05,
+      "loss": 0.4915,
+      "step": 17706
+    },
+    {
+      "epoch": 22.66496,
+      "grad_norm": 1.1241267919540405,
+      "learning_rate": 1.460184073629452e-05,
+      "loss": 0.4938,
+      "step": 17707
+    },
+    {
+      "epoch": 22.66624,
+      "grad_norm": 1.1513806581497192,
+      "learning_rate": 1.459983993597439e-05,
+      "loss": 0.4996,
+      "step": 17708
+    },
+    {
+      "epoch": 22.66752,
+      "grad_norm": 1.0984781980514526,
+      "learning_rate": 1.4597839135654262e-05,
+      "loss": 0.4556,
+      "step": 17709
+    },
+    {
+      "epoch": 22.6688,
+      "grad_norm": 1.1513564586639404,
+      "learning_rate": 1.4595838335334134e-05,
+      "loss": 0.4871,
+      "step": 17710
+    },
+    {
+      "epoch": 22.67008,
+      "grad_norm": 1.0893858671188354,
+      "learning_rate": 1.4593837535014008e-05,
+      "loss": 0.418,
+      "step": 17711
+    },
+    {
+      "epoch": 22.67136,
+      "grad_norm": 1.1374493837356567,
+      "learning_rate": 1.4591836734693878e-05,
+      "loss": 0.4628,
+      "step": 17712
+    },
+    {
+      "epoch": 22.67264,
+      "grad_norm": 1.066206932067871,
+      "learning_rate": 1.458983593437375e-05,
+      "loss": 0.4407,
+      "step": 17713
+    },
+    {
+      "epoch": 22.67392,
+      "grad_norm": 1.120012640953064,
+      "learning_rate": 1.4587835134053623e-05,
+      "loss": 0.4844,
+      "step": 17714
+    },
+    {
+      "epoch": 22.6752,
+      "grad_norm": 1.1568934917449951,
+      "learning_rate": 1.4585834333733495e-05,
+      "loss": 0.4558,
+      "step": 17715
+    },
+    {
+      "epoch": 22.67648,
+      "grad_norm": 1.1683546304702759,
+      "learning_rate": 1.4583833533413365e-05,
+      "loss": 0.5141,
+      "step": 17716
+    },
+    {
+      "epoch": 22.67776,
+      "grad_norm": 1.1002466678619385,
+      "learning_rate": 1.4581832733093237e-05,
+      "loss": 0.483,
+      "step": 17717
+    },
+    {
+      "epoch": 22.67904,
+      "grad_norm": 1.1254520416259766,
+      "learning_rate": 1.457983193277311e-05,
+      "loss": 0.4755,
+      "step": 17718
+    },
+    {
+      "epoch": 22.680320000000002,
+      "grad_norm": 1.0948314666748047,
+      "learning_rate": 1.4577831132452982e-05,
+      "loss": 0.4998,
+      "step": 17719
+    },
+    {
+      "epoch": 22.6816,
+      "grad_norm": 1.126795768737793,
+      "learning_rate": 1.4575830332132853e-05,
+      "loss": 0.4927,
+      "step": 17720
+    },
+    {
+      "epoch": 22.68288,
+      "grad_norm": 1.096273422241211,
+      "learning_rate": 1.4573829531812728e-05,
+      "loss": 0.4373,
+      "step": 17721
+    },
+    {
+      "epoch": 22.68416,
+      "grad_norm": 1.1614435911178589,
+      "learning_rate": 1.4571828731492598e-05,
+      "loss": 0.5023,
+      "step": 17722
+    },
+    {
+      "epoch": 22.68544,
+      "grad_norm": 1.181004524230957,
+      "learning_rate": 1.456982793117247e-05,
+      "loss": 0.5358,
+      "step": 17723
+    },
+    {
+      "epoch": 22.68672,
+      "grad_norm": 1.0640144348144531,
+      "learning_rate": 1.456782713085234e-05,
+      "loss": 0.4442,
+      "step": 17724
+    },
+    {
+      "epoch": 22.688,
+      "grad_norm": 1.1106035709381104,
+      "learning_rate": 1.4565826330532215e-05,
+      "loss": 0.4776,
+      "step": 17725
+    },
+    {
+      "epoch": 22.68928,
+      "grad_norm": 1.104239583015442,
+      "learning_rate": 1.4563825530212085e-05,
+      "loss": 0.4939,
+      "step": 17726
+    },
+    {
+      "epoch": 22.69056,
+      "grad_norm": 1.165892481803894,
+      "learning_rate": 1.4561824729891957e-05,
+      "loss": 0.5086,
+      "step": 17727
+    },
+    {
+      "epoch": 22.69184,
+      "grad_norm": 1.1454297304153442,
+      "learning_rate": 1.4559823929571828e-05,
+      "loss": 0.4816,
+      "step": 17728
+    },
+    {
+      "epoch": 22.69312,
+      "grad_norm": 1.103267788887024,
+      "learning_rate": 1.4557823129251703e-05,
+      "loss": 0.4574,
+      "step": 17729
+    },
+    {
+      "epoch": 22.6944,
+      "grad_norm": 1.1371041536331177,
+      "learning_rate": 1.4555822328931573e-05,
+      "loss": 0.4844,
+      "step": 17730
+    },
+    {
+      "epoch": 22.69568,
+      "grad_norm": 1.0860483646392822,
+      "learning_rate": 1.4553821528611445e-05,
+      "loss": 0.4337,
+      "step": 17731
+    },
+    {
+      "epoch": 22.69696,
+      "grad_norm": 1.1786482334136963,
+      "learning_rate": 1.4551820728291318e-05,
+      "loss": 0.5068,
+      "step": 17732
+    },
+    {
+      "epoch": 22.69824,
+      "grad_norm": 1.1752735376358032,
+      "learning_rate": 1.454981992797119e-05,
+      "loss": 0.5256,
+      "step": 17733
+    },
+    {
+      "epoch": 22.69952,
+      "grad_norm": 1.1000226736068726,
+      "learning_rate": 1.454781912765106e-05,
+      "loss": 0.4766,
+      "step": 17734
+    },
+    {
+      "epoch": 22.7008,
+      "grad_norm": 1.137929081916809,
+      "learning_rate": 1.4545818327330932e-05,
+      "loss": 0.4936,
+      "step": 17735
+    },
+    {
+      "epoch": 22.70208,
+      "grad_norm": 1.1989151239395142,
+      "learning_rate": 1.4543817527010806e-05,
+      "loss": 0.5027,
+      "step": 17736
+    },
+    {
+      "epoch": 22.70336,
+      "grad_norm": 1.180206537246704,
+      "learning_rate": 1.4541816726690678e-05,
+      "loss": 0.5085,
+      "step": 17737
+    },
+    {
+      "epoch": 22.70464,
+      "grad_norm": 1.0755740404129028,
+      "learning_rate": 1.4539815926370548e-05,
+      "loss": 0.4829,
+      "step": 17738
+    },
+    {
+      "epoch": 22.70592,
+      "grad_norm": 1.0281373262405396,
+      "learning_rate": 1.4537815126050421e-05,
+      "loss": 0.4234,
+      "step": 17739
+    },
+    {
+      "epoch": 22.7072,
+      "grad_norm": 1.0628975629806519,
+      "learning_rate": 1.4535814325730293e-05,
+      "loss": 0.4498,
+      "step": 17740
+    },
+    {
+      "epoch": 22.70848,
+      "grad_norm": 1.0640840530395508,
+      "learning_rate": 1.4533813525410165e-05,
+      "loss": 0.4457,
+      "step": 17741
+    },
+    {
+      "epoch": 22.70976,
+      "grad_norm": 1.1275110244750977,
+      "learning_rate": 1.4531812725090035e-05,
+      "loss": 0.5072,
+      "step": 17742
+    },
+    {
+      "epoch": 22.71104,
+      "grad_norm": 1.1140217781066895,
+      "learning_rate": 1.4529811924769909e-05,
+      "loss": 0.4612,
+      "step": 17743
+    },
+    {
+      "epoch": 22.71232,
+      "grad_norm": 1.083038568496704,
+      "learning_rate": 1.452781112444978e-05,
+      "loss": 0.4893,
+      "step": 17744
+    },
+    {
+      "epoch": 22.7136,
+      "grad_norm": 1.105229377746582,
+      "learning_rate": 1.4525810324129652e-05,
+      "loss": 0.461,
+      "step": 17745
+    },
+    {
+      "epoch": 22.71488,
+      "grad_norm": 1.2414989471435547,
+      "learning_rate": 1.4523809523809526e-05,
+      "loss": 0.5307,
+      "step": 17746
+    },
+    {
+      "epoch": 22.71616,
+      "grad_norm": 1.2143381834030151,
+      "learning_rate": 1.4521808723489396e-05,
+      "loss": 0.5498,
+      "step": 17747
+    },
+    {
+      "epoch": 22.71744,
+      "grad_norm": 1.171492099761963,
+      "learning_rate": 1.4519807923169268e-05,
+      "loss": 0.4874,
+      "step": 17748
+    },
+    {
+      "epoch": 22.71872,
+      "grad_norm": 1.1219180822372437,
+      "learning_rate": 1.451780712284914e-05,
+      "loss": 0.4637,
+      "step": 17749
+    },
+    {
+      "epoch": 22.72,
+      "grad_norm": 1.1720901727676392,
+      "learning_rate": 1.4515806322529013e-05,
+      "loss": 0.5322,
+      "step": 17750
+    },
+    {
+      "epoch": 22.72128,
+      "grad_norm": 1.137251615524292,
+      "learning_rate": 1.4513805522208884e-05,
+      "loss": 0.5004,
+      "step": 17751
+    },
+    {
+      "epoch": 22.72256,
+      "grad_norm": 1.147051215171814,
+      "learning_rate": 1.4511804721888755e-05,
+      "loss": 0.4769,
+      "step": 17752
+    },
+    {
+      "epoch": 22.72384,
+      "grad_norm": 1.1876957416534424,
+      "learning_rate": 1.4509803921568629e-05,
+      "loss": 0.4828,
+      "step": 17753
+    },
+    {
+      "epoch": 22.72512,
+      "grad_norm": 1.1537209749221802,
+      "learning_rate": 1.4507803121248501e-05,
+      "loss": 0.5377,
+      "step": 17754
+    },
+    {
+      "epoch": 22.7264,
+      "grad_norm": 1.0867969989776611,
+      "learning_rate": 1.4505802320928371e-05,
+      "loss": 0.4288,
+      "step": 17755
+    },
+    {
+      "epoch": 22.72768,
+      "grad_norm": 1.1193979978561401,
+      "learning_rate": 1.4503801520608243e-05,
+      "loss": 0.4479,
+      "step": 17756
+    },
+    {
+      "epoch": 22.72896,
+      "grad_norm": 1.113378882408142,
+      "learning_rate": 1.4501800720288116e-05,
+      "loss": 0.4729,
+      "step": 17757
+    },
+    {
+      "epoch": 22.73024,
+      "grad_norm": 1.0698660612106323,
+      "learning_rate": 1.4499799919967988e-05,
+      "loss": 0.454,
+      "step": 17758
+    },
+    {
+      "epoch": 22.73152,
+      "grad_norm": 1.0980463027954102,
+      "learning_rate": 1.4497799119647858e-05,
+      "loss": 0.512,
+      "step": 17759
+    },
+    {
+      "epoch": 22.7328,
+      "grad_norm": 1.0734245777130127,
+      "learning_rate": 1.4495798319327734e-05,
+      "loss": 0.4737,
+      "step": 17760
+    },
+    {
+      "epoch": 22.73408,
+      "grad_norm": 1.1146806478500366,
+      "learning_rate": 1.4493797519007604e-05,
+      "loss": 0.4867,
+      "step": 17761
+    },
+    {
+      "epoch": 22.73536,
+      "grad_norm": 1.1204739809036255,
+      "learning_rate": 1.4491796718687476e-05,
+      "loss": 0.4844,
+      "step": 17762
+    },
+    {
+      "epoch": 22.73664,
+      "grad_norm": 1.0988640785217285,
+      "learning_rate": 1.4489795918367346e-05,
+      "loss": 0.4449,
+      "step": 17763
+    },
+    {
+      "epoch": 22.73792,
+      "grad_norm": 1.1634745597839355,
+      "learning_rate": 1.4487795118047221e-05,
+      "loss": 0.5171,
+      "step": 17764
+    },
+    {
+      "epoch": 22.7392,
+      "grad_norm": 1.177647352218628,
+      "learning_rate": 1.4485794317727091e-05,
+      "loss": 0.5184,
+      "step": 17765
+    },
+    {
+      "epoch": 22.74048,
+      "grad_norm": 1.12690007686615,
+      "learning_rate": 1.4483793517406963e-05,
+      "loss": 0.4881,
+      "step": 17766
+    },
+    {
+      "epoch": 22.74176,
+      "grad_norm": 1.1667804718017578,
+      "learning_rate": 1.4481792717086837e-05,
+      "loss": 0.4587,
+      "step": 17767
+    },
+    {
+      "epoch": 22.74304,
+      "grad_norm": 1.1464014053344727,
+      "learning_rate": 1.4479791916766709e-05,
+      "loss": 0.4792,
+      "step": 17768
+    },
+    {
+      "epoch": 22.74432,
+      "grad_norm": 1.0807876586914062,
+      "learning_rate": 1.4477791116446579e-05,
+      "loss": 0.4861,
+      "step": 17769
+    },
+    {
+      "epoch": 22.7456,
+      "grad_norm": 1.1696687936782837,
+      "learning_rate": 1.447579031612645e-05,
+      "loss": 0.5222,
+      "step": 17770
+    },
+    {
+      "epoch": 22.74688,
+      "grad_norm": 1.1905460357666016,
+      "learning_rate": 1.4473789515806324e-05,
+      "loss": 0.5166,
+      "step": 17771
+    },
+    {
+      "epoch": 22.74816,
+      "grad_norm": 1.1877039670944214,
+      "learning_rate": 1.4471788715486196e-05,
+      "loss": 0.4591,
+      "step": 17772
+    },
+    {
+      "epoch": 22.74944,
+      "grad_norm": 1.1958463191986084,
+      "learning_rate": 1.4469787915166066e-05,
+      "loss": 0.4868,
+      "step": 17773
+    },
+    {
+      "epoch": 22.75072,
+      "grad_norm": 1.1075440645217896,
+      "learning_rate": 1.446778711484594e-05,
+      "loss": 0.4828,
+      "step": 17774
+    },
+    {
+      "epoch": 22.752,
+      "grad_norm": 1.0892101526260376,
+      "learning_rate": 1.4465786314525812e-05,
+      "loss": 0.4635,
+      "step": 17775
+    },
+    {
+      "epoch": 22.75328,
+      "grad_norm": 1.1559746265411377,
+      "learning_rate": 1.4463785514205683e-05,
+      "loss": 0.5233,
+      "step": 17776
+    },
+    {
+      "epoch": 22.75456,
+      "grad_norm": 1.0793570280075073,
+      "learning_rate": 1.4461784713885554e-05,
+      "loss": 0.4376,
+      "step": 17777
+    },
+    {
+      "epoch": 22.75584,
+      "grad_norm": 1.12493896484375,
+      "learning_rate": 1.4459783913565427e-05,
+      "loss": 0.5247,
+      "step": 17778
+    },
+    {
+      "epoch": 22.75712,
+      "grad_norm": 1.1282217502593994,
+      "learning_rate": 1.4457783113245299e-05,
+      "loss": 0.4513,
+      "step": 17779
+    },
+    {
+      "epoch": 22.7584,
+      "grad_norm": 1.1162893772125244,
+      "learning_rate": 1.445578231292517e-05,
+      "loss": 0.4873,
+      "step": 17780
+    },
+    {
+      "epoch": 22.75968,
+      "grad_norm": 1.1070144176483154,
+      "learning_rate": 1.4453781512605044e-05,
+      "loss": 0.4985,
+      "step": 17781
+    },
+    {
+      "epoch": 22.76096,
+      "grad_norm": 1.0941122770309448,
+      "learning_rate": 1.4451780712284915e-05,
+      "loss": 0.4902,
+      "step": 17782
+    },
+    {
+      "epoch": 22.76224,
+      "grad_norm": 1.1269056797027588,
+      "learning_rate": 1.4449779911964786e-05,
+      "loss": 0.469,
+      "step": 17783
+    },
+    {
+      "epoch": 22.76352,
+      "grad_norm": 1.0950514078140259,
+      "learning_rate": 1.4447779111644658e-05,
+      "loss": 0.4394,
+      "step": 17784
+    },
+    {
+      "epoch": 22.7648,
+      "grad_norm": 1.1223610639572144,
+      "learning_rate": 1.4445778311324532e-05,
+      "loss": 0.4732,
+      "step": 17785
+    },
+    {
+      "epoch": 22.76608,
+      "grad_norm": 1.0896393060684204,
+      "learning_rate": 1.4443777511004402e-05,
+      "loss": 0.4444,
+      "step": 17786
+    },
+    {
+      "epoch": 22.76736,
+      "grad_norm": 1.1810468435287476,
+      "learning_rate": 1.4441776710684274e-05,
+      "loss": 0.5342,
+      "step": 17787
+    },
+    {
+      "epoch": 22.76864,
+      "grad_norm": 1.1601206064224243,
+      "learning_rate": 1.4439775910364146e-05,
+      "loss": 0.5354,
+      "step": 17788
+    },
+    {
+      "epoch": 22.76992,
+      "grad_norm": 1.1780500411987305,
+      "learning_rate": 1.443777511004402e-05,
+      "loss": 0.4826,
+      "step": 17789
+    },
+    {
+      "epoch": 22.7712,
+      "grad_norm": 1.1130772829055786,
+      "learning_rate": 1.443577430972389e-05,
+      "loss": 0.4948,
+      "step": 17790
+    },
+    {
+      "epoch": 22.77248,
+      "grad_norm": 1.1274909973144531,
+      "learning_rate": 1.4433773509403761e-05,
+      "loss": 0.4979,
+      "step": 17791
+    },
+    {
+      "epoch": 22.77376,
+      "grad_norm": 1.1447163820266724,
+      "learning_rate": 1.4431772709083635e-05,
+      "loss": 0.4735,
+      "step": 17792
+    },
+    {
+      "epoch": 22.77504,
+      "grad_norm": 1.0660210847854614,
+      "learning_rate": 1.4429771908763507e-05,
+      "loss": 0.4555,
+      "step": 17793
+    },
+    {
+      "epoch": 22.77632,
+      "grad_norm": 1.0781277418136597,
+      "learning_rate": 1.4427771108443377e-05,
+      "loss": 0.4619,
+      "step": 17794
+    },
+    {
+      "epoch": 22.7776,
+      "grad_norm": 1.1187148094177246,
+      "learning_rate": 1.4425770308123249e-05,
+      "loss": 0.4552,
+      "step": 17795
+    },
+    {
+      "epoch": 22.77888,
+      "grad_norm": 1.1533479690551758,
+      "learning_rate": 1.4423769507803122e-05,
+      "loss": 0.4782,
+      "step": 17796
+    },
+    {
+      "epoch": 22.78016,
+      "grad_norm": 1.143250584602356,
+      "learning_rate": 1.4421768707482994e-05,
+      "loss": 0.5023,
+      "step": 17797
+    },
+    {
+      "epoch": 22.78144,
+      "grad_norm": 1.1441926956176758,
+      "learning_rate": 1.4419767907162864e-05,
+      "loss": 0.5012,
+      "step": 17798
+    },
+    {
+      "epoch": 22.78272,
+      "grad_norm": 1.1615264415740967,
+      "learning_rate": 1.441776710684274e-05,
+      "loss": 0.5388,
+      "step": 17799
+    },
+    {
+      "epoch": 22.784,
+      "grad_norm": 1.1589363813400269,
+      "learning_rate": 1.441576630652261e-05,
+      "loss": 0.505,
+      "step": 17800
+    },
+    {
+      "epoch": 22.78528,
+      "grad_norm": 1.1797236204147339,
+      "learning_rate": 1.4413765506202482e-05,
+      "loss": 0.502,
+      "step": 17801
+    },
+    {
+      "epoch": 22.78656,
+      "grad_norm": 1.2213630676269531,
+      "learning_rate": 1.4411764705882352e-05,
+      "loss": 0.5167,
+      "step": 17802
+    },
+    {
+      "epoch": 22.78784,
+      "grad_norm": 1.155617117881775,
+      "learning_rate": 1.4409763905562227e-05,
+      "loss": 0.4613,
+      "step": 17803
+    },
+    {
+      "epoch": 22.78912,
+      "grad_norm": 1.1112602949142456,
+      "learning_rate": 1.4407763105242097e-05,
+      "loss": 0.4594,
+      "step": 17804
+    },
+    {
+      "epoch": 22.790399999999998,
+      "grad_norm": 1.1475427150726318,
+      "learning_rate": 1.4405762304921969e-05,
+      "loss": 0.457,
+      "step": 17805
+    },
+    {
+      "epoch": 22.79168,
+      "grad_norm": 1.1044026613235474,
+      "learning_rate": 1.4403761504601842e-05,
+      "loss": 0.48,
+      "step": 17806
+    },
+    {
+      "epoch": 22.79296,
+      "grad_norm": 1.119400978088379,
+      "learning_rate": 1.4401760704281714e-05,
+      "loss": 0.481,
+      "step": 17807
+    },
+    {
+      "epoch": 22.79424,
+      "grad_norm": 1.1394966840744019,
+      "learning_rate": 1.4399759903961585e-05,
+      "loss": 0.5108,
+      "step": 17808
+    },
+    {
+      "epoch": 22.79552,
+      "grad_norm": 1.170421838760376,
+      "learning_rate": 1.4397759103641456e-05,
+      "loss": 0.4999,
+      "step": 17809
+    },
+    {
+      "epoch": 22.7968,
+      "grad_norm": 1.134114384651184,
+      "learning_rate": 1.439575830332133e-05,
+      "loss": 0.4723,
+      "step": 17810
+    },
+    {
+      "epoch": 22.79808,
+      "grad_norm": 1.180664300918579,
+      "learning_rate": 1.4393757503001202e-05,
+      "loss": 0.4983,
+      "step": 17811
+    },
+    {
+      "epoch": 22.79936,
+      "grad_norm": 1.2017977237701416,
+      "learning_rate": 1.4391756702681072e-05,
+      "loss": 0.5217,
+      "step": 17812
+    },
+    {
+      "epoch": 22.80064,
+      "grad_norm": 1.0838621854782104,
+      "learning_rate": 1.4389755902360945e-05,
+      "loss": 0.4477,
+      "step": 17813
+    },
+    {
+      "epoch": 22.80192,
+      "grad_norm": 1.1182233095169067,
+      "learning_rate": 1.4387755102040817e-05,
+      "loss": 0.4882,
+      "step": 17814
+    },
+    {
+      "epoch": 22.8032,
+      "grad_norm": 1.1456414461135864,
+      "learning_rate": 1.438575430172069e-05,
+      "loss": 0.4561,
+      "step": 17815
+    },
+    {
+      "epoch": 22.80448,
+      "grad_norm": 1.0864543914794922,
+      "learning_rate": 1.438375350140056e-05,
+      "loss": 0.4516,
+      "step": 17816
+    },
+    {
+      "epoch": 22.80576,
+      "grad_norm": 1.145043134689331,
+      "learning_rate": 1.4381752701080433e-05,
+      "loss": 0.4764,
+      "step": 17817
+    },
+    {
+      "epoch": 22.80704,
+      "grad_norm": 1.1899402141571045,
+      "learning_rate": 1.4379751900760305e-05,
+      "loss": 0.564,
+      "step": 17818
+    },
+    {
+      "epoch": 22.80832,
+      "grad_norm": 1.0811036825180054,
+      "learning_rate": 1.4377751100440177e-05,
+      "loss": 0.453,
+      "step": 17819
+    },
+    {
+      "epoch": 22.8096,
+      "grad_norm": 1.1460031270980835,
+      "learning_rate": 1.437575030012005e-05,
+      "loss": 0.4835,
+      "step": 17820
+    },
+    {
+      "epoch": 22.81088,
+      "grad_norm": 1.1830692291259766,
+      "learning_rate": 1.437374949979992e-05,
+      "loss": 0.4641,
+      "step": 17821
+    },
+    {
+      "epoch": 22.81216,
+      "grad_norm": 1.204338550567627,
+      "learning_rate": 1.4371748699479792e-05,
+      "loss": 0.4938,
+      "step": 17822
+    },
+    {
+      "epoch": 22.81344,
+      "grad_norm": 1.107739806175232,
+      "learning_rate": 1.4369747899159664e-05,
+      "loss": 0.459,
+      "step": 17823
+    },
+    {
+      "epoch": 22.81472,
+      "grad_norm": 1.113206386566162,
+      "learning_rate": 1.4367747098839538e-05,
+      "loss": 0.459,
+      "step": 17824
+    },
+    {
+      "epoch": 22.816,
+      "grad_norm": 1.164965033531189,
+      "learning_rate": 1.4365746298519408e-05,
+      "loss": 0.533,
+      "step": 17825
+    },
+    {
+      "epoch": 22.81728,
+      "grad_norm": 1.1349233388900757,
+      "learning_rate": 1.436374549819928e-05,
+      "loss": 0.4582,
+      "step": 17826
+    },
+    {
+      "epoch": 22.81856,
+      "grad_norm": 1.1366132497787476,
+      "learning_rate": 1.4361744697879153e-05,
+      "loss": 0.518,
+      "step": 17827
+    },
+    {
+      "epoch": 22.81984,
+      "grad_norm": 1.064651370048523,
+      "learning_rate": 1.4359743897559025e-05,
+      "loss": 0.4514,
+      "step": 17828
+    },
+    {
+      "epoch": 22.82112,
+      "grad_norm": 1.0854195356369019,
+      "learning_rate": 1.4357743097238895e-05,
+      "loss": 0.4654,
+      "step": 17829
+    },
+    {
+      "epoch": 22.822400000000002,
+      "grad_norm": 1.1058958768844604,
+      "learning_rate": 1.4355742296918767e-05,
+      "loss": 0.4712,
+      "step": 17830
+    },
+    {
+      "epoch": 22.82368,
+      "grad_norm": 1.1246731281280518,
+      "learning_rate": 1.435374149659864e-05,
+      "loss": 0.4796,
+      "step": 17831
+    },
+    {
+      "epoch": 22.82496,
+      "grad_norm": 1.123199224472046,
+      "learning_rate": 1.4351740696278512e-05,
+      "loss": 0.521,
+      "step": 17832
+    },
+    {
+      "epoch": 22.82624,
+      "grad_norm": 1.1568586826324463,
+      "learning_rate": 1.4349739895958383e-05,
+      "loss": 0.5103,
+      "step": 17833
+    },
+    {
+      "epoch": 22.82752,
+      "grad_norm": 1.0084097385406494,
+      "learning_rate": 1.4347739095638258e-05,
+      "loss": 0.4231,
+      "step": 17834
+    },
+    {
+      "epoch": 22.8288,
+      "grad_norm": 1.0467265844345093,
+      "learning_rate": 1.4345738295318128e-05,
+      "loss": 0.4596,
+      "step": 17835
+    },
+    {
+      "epoch": 22.83008,
+      "grad_norm": 1.1111983060836792,
+      "learning_rate": 1.4343737494998e-05,
+      "loss": 0.4774,
+      "step": 17836
+    },
+    {
+      "epoch": 22.83136,
+      "grad_norm": 1.1236525774002075,
+      "learning_rate": 1.434173669467787e-05,
+      "loss": 0.4637,
+      "step": 17837
+    },
+    {
+      "epoch": 22.83264,
+      "grad_norm": 1.144333839416504,
+      "learning_rate": 1.4339735894357745e-05,
+      "loss": 0.5001,
+      "step": 17838
+    },
+    {
+      "epoch": 22.83392,
+      "grad_norm": 1.1298198699951172,
+      "learning_rate": 1.4337735094037615e-05,
+      "loss": 0.4897,
+      "step": 17839
+    },
+    {
+      "epoch": 22.8352,
+      "grad_norm": 1.0742158889770508,
+      "learning_rate": 1.4335734293717487e-05,
+      "loss": 0.4371,
+      "step": 17840
+    },
+    {
+      "epoch": 22.83648,
+      "grad_norm": 1.1449413299560547,
+      "learning_rate": 1.4333733493397357e-05,
+      "loss": 0.4818,
+      "step": 17841
+    },
+    {
+      "epoch": 22.83776,
+      "grad_norm": 1.1488326787948608,
+      "learning_rate": 1.4331732693077233e-05,
+      "loss": 0.4926,
+      "step": 17842
+    },
+    {
+      "epoch": 22.83904,
+      "grad_norm": 1.1773475408554077,
+      "learning_rate": 1.4329731892757103e-05,
+      "loss": 0.5046,
+      "step": 17843
+    },
+    {
+      "epoch": 22.84032,
+      "grad_norm": 1.1405889987945557,
+      "learning_rate": 1.4327731092436975e-05,
+      "loss": 0.4876,
+      "step": 17844
+    },
+    {
+      "epoch": 22.8416,
+      "grad_norm": 1.0573958158493042,
+      "learning_rate": 1.4325730292116848e-05,
+      "loss": 0.4591,
+      "step": 17845
+    },
+    {
+      "epoch": 22.84288,
+      "grad_norm": 1.1423496007919312,
+      "learning_rate": 1.432372949179672e-05,
+      "loss": 0.5063,
+      "step": 17846
+    },
+    {
+      "epoch": 22.84416,
+      "grad_norm": 1.1753638982772827,
+      "learning_rate": 1.432172869147659e-05,
+      "loss": 0.5118,
+      "step": 17847
+    },
+    {
+      "epoch": 22.84544,
+      "grad_norm": 1.114268183708191,
+      "learning_rate": 1.4319727891156462e-05,
+      "loss": 0.4528,
+      "step": 17848
+    },
+    {
+      "epoch": 22.84672,
+      "grad_norm": 1.1355352401733398,
+      "learning_rate": 1.4317727090836336e-05,
+      "loss": 0.4828,
+      "step": 17849
+    },
+    {
+      "epoch": 22.848,
+      "grad_norm": 1.199616551399231,
+      "learning_rate": 1.4315726290516208e-05,
+      "loss": 0.4811,
+      "step": 17850
+    },
+    {
+      "epoch": 22.84928,
+      "grad_norm": 1.173221468925476,
+      "learning_rate": 1.4313725490196078e-05,
+      "loss": 0.5268,
+      "step": 17851
+    },
+    {
+      "epoch": 22.85056,
+      "grad_norm": 1.1589332818984985,
+      "learning_rate": 1.4311724689875953e-05,
+      "loss": 0.4963,
+      "step": 17852
+    },
+    {
+      "epoch": 22.85184,
+      "grad_norm": 1.137489676475525,
+      "learning_rate": 1.4309723889555823e-05,
+      "loss": 0.4951,
+      "step": 17853
+    },
+    {
+      "epoch": 22.85312,
+      "grad_norm": 1.1129984855651855,
+      "learning_rate": 1.4307723089235695e-05,
+      "loss": 0.4814,
+      "step": 17854
+    },
+    {
+      "epoch": 22.8544,
+      "grad_norm": 1.0853337049484253,
+      "learning_rate": 1.4305722288915565e-05,
+      "loss": 0.4432,
+      "step": 17855
+    },
+    {
+      "epoch": 22.85568,
+      "grad_norm": 1.1107585430145264,
+      "learning_rate": 1.430372148859544e-05,
+      "loss": 0.5175,
+      "step": 17856
+    },
+    {
+      "epoch": 22.85696,
+      "grad_norm": 1.143998146057129,
+      "learning_rate": 1.430172068827531e-05,
+      "loss": 0.4882,
+      "step": 17857
+    },
+    {
+      "epoch": 22.85824,
+      "grad_norm": 1.164641261100769,
+      "learning_rate": 1.4299719887955182e-05,
+      "loss": 0.53,
+      "step": 17858
+    },
+    {
+      "epoch": 22.85952,
+      "grad_norm": 1.1436692476272583,
+      "learning_rate": 1.4297719087635056e-05,
+      "loss": 0.4866,
+      "step": 17859
+    },
+    {
+      "epoch": 22.8608,
+      "grad_norm": 1.1730378866195679,
+      "learning_rate": 1.4295718287314928e-05,
+      "loss": 0.5107,
+      "step": 17860
+    },
+    {
+      "epoch": 22.86208,
+      "grad_norm": 1.122793436050415,
+      "learning_rate": 1.4293717486994798e-05,
+      "loss": 0.5111,
+      "step": 17861
+    },
+    {
+      "epoch": 22.86336,
+      "grad_norm": 1.0745701789855957,
+      "learning_rate": 1.429171668667467e-05,
+      "loss": 0.4485,
+      "step": 17862
+    },
+    {
+      "epoch": 22.86464,
+      "grad_norm": 1.1484006643295288,
+      "learning_rate": 1.4289715886354543e-05,
+      "loss": 0.5085,
+      "step": 17863
+    },
+    {
+      "epoch": 22.86592,
+      "grad_norm": 1.1889216899871826,
+      "learning_rate": 1.4287715086034415e-05,
+      "loss": 0.5249,
+      "step": 17864
+    },
+    {
+      "epoch": 22.8672,
+      "grad_norm": 1.1078394651412964,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.4893,
+      "step": 17865
+    },
+    {
+      "epoch": 22.86848,
+      "grad_norm": 1.0795047283172607,
+      "learning_rate": 1.4283713485394159e-05,
+      "loss": 0.4388,
+      "step": 17866
+    },
+    {
+      "epoch": 22.86976,
+      "grad_norm": 1.0773197412490845,
+      "learning_rate": 1.428171268507403e-05,
+      "loss": 0.4857,
+      "step": 17867
+    },
+    {
+      "epoch": 22.87104,
+      "grad_norm": 1.1797206401824951,
+      "learning_rate": 1.4279711884753903e-05,
+      "loss": 0.5151,
+      "step": 17868
+    },
+    {
+      "epoch": 22.87232,
+      "grad_norm": 1.1339863538742065,
+      "learning_rate": 1.4277711084433773e-05,
+      "loss": 0.457,
+      "step": 17869
+    },
+    {
+      "epoch": 22.8736,
+      "grad_norm": 1.1581915616989136,
+      "learning_rate": 1.4275710284113646e-05,
+      "loss": 0.5057,
+      "step": 17870
+    },
+    {
+      "epoch": 22.87488,
+      "grad_norm": 1.1262785196304321,
+      "learning_rate": 1.4273709483793518e-05,
+      "loss": 0.5133,
+      "step": 17871
+    },
+    {
+      "epoch": 22.87616,
+      "grad_norm": 1.1205724477767944,
+      "learning_rate": 1.427170868347339e-05,
+      "loss": 0.5048,
+      "step": 17872
+    },
+    {
+      "epoch": 22.87744,
+      "grad_norm": 1.0860368013381958,
+      "learning_rate": 1.4269707883153264e-05,
+      "loss": 0.4702,
+      "step": 17873
+    },
+    {
+      "epoch": 22.87872,
+      "grad_norm": 1.1230723857879639,
+      "learning_rate": 1.4267707082833134e-05,
+      "loss": 0.4754,
+      "step": 17874
+    },
+    {
+      "epoch": 22.88,
+      "grad_norm": 1.1550164222717285,
+      "learning_rate": 1.4265706282513006e-05,
+      "loss": 0.5318,
+      "step": 17875
+    },
+    {
+      "epoch": 22.88128,
+      "grad_norm": 1.0607014894485474,
+      "learning_rate": 1.4263705482192878e-05,
+      "loss": 0.4266,
+      "step": 17876
+    },
+    {
+      "epoch": 22.88256,
+      "grad_norm": 1.1427565813064575,
+      "learning_rate": 1.4261704681872751e-05,
+      "loss": 0.5016,
+      "step": 17877
+    },
+    {
+      "epoch": 22.88384,
+      "grad_norm": 1.082594394683838,
+      "learning_rate": 1.4259703881552621e-05,
+      "loss": 0.4745,
+      "step": 17878
+    },
+    {
+      "epoch": 22.88512,
+      "grad_norm": 1.1618205308914185,
+      "learning_rate": 1.4257703081232493e-05,
+      "loss": 0.5358,
+      "step": 17879
+    },
+    {
+      "epoch": 22.8864,
+      "grad_norm": 1.0929317474365234,
+      "learning_rate": 1.4255702280912367e-05,
+      "loss": 0.4662,
+      "step": 17880
+    },
+    {
+      "epoch": 22.88768,
+      "grad_norm": 1.1140695810317993,
+      "learning_rate": 1.4253701480592239e-05,
+      "loss": 0.4729,
+      "step": 17881
+    },
+    {
+      "epoch": 22.88896,
+      "grad_norm": 1.0910569429397583,
+      "learning_rate": 1.4251700680272109e-05,
+      "loss": 0.4473,
+      "step": 17882
+    },
+    {
+      "epoch": 22.89024,
+      "grad_norm": 1.193246841430664,
+      "learning_rate": 1.424969987995198e-05,
+      "loss": 0.541,
+      "step": 17883
+    },
+    {
+      "epoch": 22.89152,
+      "grad_norm": 1.0496735572814941,
+      "learning_rate": 1.4247699079631854e-05,
+      "loss": 0.494,
+      "step": 17884
+    },
+    {
+      "epoch": 22.8928,
+      "grad_norm": 1.1013808250427246,
+      "learning_rate": 1.4245698279311726e-05,
+      "loss": 0.4842,
+      "step": 17885
+    },
+    {
+      "epoch": 22.89408,
+      "grad_norm": 1.1200693845748901,
+      "learning_rate": 1.4243697478991596e-05,
+      "loss": 0.5158,
+      "step": 17886
+    },
+    {
+      "epoch": 22.89536,
+      "grad_norm": 1.1038991212844849,
+      "learning_rate": 1.4241696678671471e-05,
+      "loss": 0.4855,
+      "step": 17887
+    },
+    {
+      "epoch": 22.89664,
+      "grad_norm": 1.119197964668274,
+      "learning_rate": 1.4239695878351342e-05,
+      "loss": 0.5312,
+      "step": 17888
+    },
+    {
+      "epoch": 22.89792,
+      "grad_norm": 1.0970196723937988,
+      "learning_rate": 1.4237695078031213e-05,
+      "loss": 0.481,
+      "step": 17889
+    },
+    {
+      "epoch": 22.8992,
+      "grad_norm": 1.1380757093429565,
+      "learning_rate": 1.4235694277711084e-05,
+      "loss": 0.5077,
+      "step": 17890
+    },
+    {
+      "epoch": 22.90048,
+      "grad_norm": 1.0854060649871826,
+      "learning_rate": 1.4233693477390959e-05,
+      "loss": 0.4607,
+      "step": 17891
+    },
+    {
+      "epoch": 22.90176,
+      "grad_norm": 1.0660706758499146,
+      "learning_rate": 1.4231692677070829e-05,
+      "loss": 0.4662,
+      "step": 17892
+    },
+    {
+      "epoch": 22.90304,
+      "grad_norm": 1.1607842445373535,
+      "learning_rate": 1.42296918767507e-05,
+      "loss": 0.5338,
+      "step": 17893
+    },
+    {
+      "epoch": 22.90432,
+      "grad_norm": 1.1080985069274902,
+      "learning_rate": 1.4227691076430571e-05,
+      "loss": 0.4802,
+      "step": 17894
+    },
+    {
+      "epoch": 22.9056,
+      "grad_norm": 1.1644606590270996,
+      "learning_rate": 1.4225690276110446e-05,
+      "loss": 0.5258,
+      "step": 17895
+    },
+    {
+      "epoch": 22.90688,
+      "grad_norm": 1.1852686405181885,
+      "learning_rate": 1.4223689475790316e-05,
+      "loss": 0.4759,
+      "step": 17896
+    },
+    {
+      "epoch": 22.90816,
+      "grad_norm": 1.1916351318359375,
+      "learning_rate": 1.4221688675470188e-05,
+      "loss": 0.4743,
+      "step": 17897
+    },
+    {
+      "epoch": 22.90944,
+      "grad_norm": 1.1184988021850586,
+      "learning_rate": 1.4219687875150062e-05,
+      "loss": 0.5025,
+      "step": 17898
+    },
+    {
+      "epoch": 22.91072,
+      "grad_norm": 1.147409439086914,
+      "learning_rate": 1.4217687074829934e-05,
+      "loss": 0.526,
+      "step": 17899
+    },
+    {
+      "epoch": 22.912,
+      "grad_norm": 1.1477159261703491,
+      "learning_rate": 1.4215686274509804e-05,
+      "loss": 0.4882,
+      "step": 17900
+    },
+    {
+      "epoch": 22.91328,
+      "grad_norm": 1.0916913747787476,
+      "learning_rate": 1.4213685474189676e-05,
+      "loss": 0.508,
+      "step": 17901
+    },
+    {
+      "epoch": 22.91456,
+      "grad_norm": 1.0843031406402588,
+      "learning_rate": 1.421168467386955e-05,
+      "loss": 0.5053,
+      "step": 17902
+    },
+    {
+      "epoch": 22.91584,
+      "grad_norm": 1.1080353260040283,
+      "learning_rate": 1.4209683873549421e-05,
+      "loss": 0.5002,
+      "step": 17903
+    },
+    {
+      "epoch": 22.91712,
+      "grad_norm": 1.1214649677276611,
+      "learning_rate": 1.4207683073229291e-05,
+      "loss": 0.5296,
+      "step": 17904
+    },
+    {
+      "epoch": 22.9184,
+      "grad_norm": 1.0759986639022827,
+      "learning_rate": 1.4205682272909165e-05,
+      "loss": 0.4737,
+      "step": 17905
+    },
+    {
+      "epoch": 22.91968,
+      "grad_norm": 1.1020501852035522,
+      "learning_rate": 1.4203681472589037e-05,
+      "loss": 0.4915,
+      "step": 17906
+    },
+    {
+      "epoch": 22.92096,
+      "grad_norm": 1.1096168756484985,
+      "learning_rate": 1.4201680672268908e-05,
+      "loss": 0.4827,
+      "step": 17907
+    },
+    {
+      "epoch": 22.92224,
+      "grad_norm": 1.0973726511001587,
+      "learning_rate": 1.4199679871948779e-05,
+      "loss": 0.4771,
+      "step": 17908
+    },
+    {
+      "epoch": 22.92352,
+      "grad_norm": 1.143187403678894,
+      "learning_rate": 1.4197679071628652e-05,
+      "loss": 0.5155,
+      "step": 17909
+    },
+    {
+      "epoch": 22.9248,
+      "grad_norm": 1.1421252489089966,
+      "learning_rate": 1.4195678271308524e-05,
+      "loss": 0.4643,
+      "step": 17910
+    },
+    {
+      "epoch": 22.92608,
+      "grad_norm": 1.1452699899673462,
+      "learning_rate": 1.4193677470988396e-05,
+      "loss": 0.5185,
+      "step": 17911
+    },
+    {
+      "epoch": 22.92736,
+      "grad_norm": 1.0854843854904175,
+      "learning_rate": 1.419167667066827e-05,
+      "loss": 0.4865,
+      "step": 17912
+    },
+    {
+      "epoch": 22.92864,
+      "grad_norm": 1.117672324180603,
+      "learning_rate": 1.418967587034814e-05,
+      "loss": 0.4689,
+      "step": 17913
+    },
+    {
+      "epoch": 22.92992,
+      "grad_norm": 1.135145902633667,
+      "learning_rate": 1.4187675070028011e-05,
+      "loss": 0.4971,
+      "step": 17914
+    },
+    {
+      "epoch": 22.9312,
+      "grad_norm": 1.0750057697296143,
+      "learning_rate": 1.4185674269707883e-05,
+      "loss": 0.4357,
+      "step": 17915
+    },
+    {
+      "epoch": 22.932479999999998,
+      "grad_norm": 1.1267346143722534,
+      "learning_rate": 1.4183673469387757e-05,
+      "loss": 0.4957,
+      "step": 17916
+    },
+    {
+      "epoch": 22.93376,
+      "grad_norm": 1.1975816488265991,
+      "learning_rate": 1.4181672669067627e-05,
+      "loss": 0.5446,
+      "step": 17917
+    },
+    {
+      "epoch": 22.93504,
+      "grad_norm": 1.12879478931427,
+      "learning_rate": 1.4179671868747499e-05,
+      "loss": 0.4855,
+      "step": 17918
+    },
+    {
+      "epoch": 22.93632,
+      "grad_norm": 1.1235387325286865,
+      "learning_rate": 1.4177671068427372e-05,
+      "loss": 0.5129,
+      "step": 17919
+    },
+    {
+      "epoch": 22.9376,
+      "grad_norm": 1.1320890188217163,
+      "learning_rate": 1.4175670268107244e-05,
+      "loss": 0.4815,
+      "step": 17920
+    },
+    {
+      "epoch": 22.93888,
+      "grad_norm": 1.1158738136291504,
+      "learning_rate": 1.4173669467787114e-05,
+      "loss": 0.4736,
+      "step": 17921
+    },
+    {
+      "epoch": 22.94016,
+      "grad_norm": 1.0395013093948364,
+      "learning_rate": 1.4171668667466986e-05,
+      "loss": 0.4211,
+      "step": 17922
+    },
+    {
+      "epoch": 22.94144,
+      "grad_norm": 1.0959734916687012,
+      "learning_rate": 1.416966786714686e-05,
+      "loss": 0.4874,
+      "step": 17923
+    },
+    {
+      "epoch": 22.94272,
+      "grad_norm": 1.0824655294418335,
+      "learning_rate": 1.4167667066826732e-05,
+      "loss": 0.494,
+      "step": 17924
+    },
+    {
+      "epoch": 22.944,
+      "grad_norm": 1.1469032764434814,
+      "learning_rate": 1.4165666266506602e-05,
+      "loss": 0.4917,
+      "step": 17925
+    },
+    {
+      "epoch": 22.94528,
+      "grad_norm": 1.2326279878616333,
+      "learning_rate": 1.4163665466186477e-05,
+      "loss": 0.5198,
+      "step": 17926
+    },
+    {
+      "epoch": 22.94656,
+      "grad_norm": 1.1711357831954956,
+      "learning_rate": 1.4161664665866347e-05,
+      "loss": 0.5073,
+      "step": 17927
+    },
+    {
+      "epoch": 22.94784,
+      "grad_norm": 1.1785284280776978,
+      "learning_rate": 1.415966386554622e-05,
+      "loss": 0.4986,
+      "step": 17928
+    },
+    {
+      "epoch": 22.94912,
+      "grad_norm": 1.1128637790679932,
+      "learning_rate": 1.415766306522609e-05,
+      "loss": 0.4626,
+      "step": 17929
+    },
+    {
+      "epoch": 22.9504,
+      "grad_norm": 1.1575325727462769,
+      "learning_rate": 1.4155662264905965e-05,
+      "loss": 0.5185,
+      "step": 17930
+    },
+    {
+      "epoch": 22.95168,
+      "grad_norm": 1.11526620388031,
+      "learning_rate": 1.4153661464585835e-05,
+      "loss": 0.47,
+      "step": 17931
+    },
+    {
+      "epoch": 22.95296,
+      "grad_norm": 1.1603832244873047,
+      "learning_rate": 1.4151660664265707e-05,
+      "loss": 0.479,
+      "step": 17932
+    },
+    {
+      "epoch": 22.95424,
+      "grad_norm": 1.1286890506744385,
+      "learning_rate": 1.414965986394558e-05,
+      "loss": 0.4544,
+      "step": 17933
+    },
+    {
+      "epoch": 22.95552,
+      "grad_norm": 1.0917226076126099,
+      "learning_rate": 1.4147659063625452e-05,
+      "loss": 0.4819,
+      "step": 17934
+    },
+    {
+      "epoch": 22.9568,
+      "grad_norm": 1.1569414138793945,
+      "learning_rate": 1.4145658263305322e-05,
+      "loss": 0.5174,
+      "step": 17935
+    },
+    {
+      "epoch": 22.95808,
+      "grad_norm": 1.1322275400161743,
+      "learning_rate": 1.4143657462985194e-05,
+      "loss": 0.5149,
+      "step": 17936
+    },
+    {
+      "epoch": 22.95936,
+      "grad_norm": 1.1222246885299683,
+      "learning_rate": 1.4141656662665068e-05,
+      "loss": 0.5159,
+      "step": 17937
+    },
+    {
+      "epoch": 22.96064,
+      "grad_norm": 1.0865634679794312,
+      "learning_rate": 1.413965586234494e-05,
+      "loss": 0.4598,
+      "step": 17938
+    },
+    {
+      "epoch": 22.96192,
+      "grad_norm": 1.2040810585021973,
+      "learning_rate": 1.413765506202481e-05,
+      "loss": 0.5407,
+      "step": 17939
+    },
+    {
+      "epoch": 22.9632,
+      "grad_norm": 1.1161025762557983,
+      "learning_rate": 1.4135654261704683e-05,
+      "loss": 0.52,
+      "step": 17940
+    },
+    {
+      "epoch": 22.964480000000002,
+      "grad_norm": 1.1312873363494873,
+      "learning_rate": 1.4133653461384555e-05,
+      "loss": 0.5067,
+      "step": 17941
+    },
+    {
+      "epoch": 22.96576,
+      "grad_norm": 1.0572233200073242,
+      "learning_rate": 1.4131652661064427e-05,
+      "loss": 0.4639,
+      "step": 17942
+    },
+    {
+      "epoch": 22.96704,
+      "grad_norm": 1.0853049755096436,
+      "learning_rate": 1.4129651860744297e-05,
+      "loss": 0.4694,
+      "step": 17943
+    },
+    {
+      "epoch": 22.96832,
+      "grad_norm": 1.133817434310913,
+      "learning_rate": 1.412765106042417e-05,
+      "loss": 0.5031,
+      "step": 17944
+    },
+    {
+      "epoch": 22.9696,
+      "grad_norm": 1.1215367317199707,
+      "learning_rate": 1.4125650260104042e-05,
+      "loss": 0.4852,
+      "step": 17945
+    },
+    {
+      "epoch": 22.97088,
+      "grad_norm": 1.1265122890472412,
+      "learning_rate": 1.4123649459783914e-05,
+      "loss": 0.4838,
+      "step": 17946
+    },
+    {
+      "epoch": 22.97216,
+      "grad_norm": 1.1352272033691406,
+      "learning_rate": 1.4121648659463788e-05,
+      "loss": 0.4701,
+      "step": 17947
+    },
+    {
+      "epoch": 22.97344,
+      "grad_norm": 1.0938243865966797,
+      "learning_rate": 1.4119647859143658e-05,
+      "loss": 0.4614,
+      "step": 17948
+    },
+    {
+      "epoch": 22.97472,
+      "grad_norm": 1.147600769996643,
+      "learning_rate": 1.411764705882353e-05,
+      "loss": 0.4799,
+      "step": 17949
+    },
+    {
+      "epoch": 22.976,
+      "grad_norm": 1.1563359498977661,
+      "learning_rate": 1.4115646258503402e-05,
+      "loss": 0.4585,
+      "step": 17950
+    },
+    {
+      "epoch": 22.97728,
+      "grad_norm": 1.1627904176712036,
+      "learning_rate": 1.4113645458183275e-05,
+      "loss": 0.56,
+      "step": 17951
+    },
+    {
+      "epoch": 22.97856,
+      "grad_norm": 1.1048862934112549,
+      "learning_rate": 1.4111644657863145e-05,
+      "loss": 0.5129,
+      "step": 17952
+    },
+    {
+      "epoch": 22.97984,
+      "grad_norm": 1.1043338775634766,
+      "learning_rate": 1.4109643857543017e-05,
+      "loss": 0.4792,
+      "step": 17953
+    },
+    {
+      "epoch": 22.98112,
+      "grad_norm": 1.151166319847107,
+      "learning_rate": 1.4107643057222889e-05,
+      "loss": 0.5038,
+      "step": 17954
+    },
+    {
+      "epoch": 22.9824,
+      "grad_norm": 1.1335054636001587,
+      "learning_rate": 1.4105642256902763e-05,
+      "loss": 0.5076,
+      "step": 17955
+    },
+    {
+      "epoch": 22.98368,
+      "grad_norm": 1.0865764617919922,
+      "learning_rate": 1.4103641456582633e-05,
+      "loss": 0.4719,
+      "step": 17956
+    },
+    {
+      "epoch": 22.98496,
+      "grad_norm": 1.1240063905715942,
+      "learning_rate": 1.4101640656262505e-05,
+      "loss": 0.5044,
+      "step": 17957
+    },
+    {
+      "epoch": 22.98624,
+      "grad_norm": 1.142352819442749,
+      "learning_rate": 1.4099639855942378e-05,
+      "loss": 0.5157,
+      "step": 17958
+    },
+    {
+      "epoch": 22.98752,
+      "grad_norm": 1.0918866395950317,
+      "learning_rate": 1.409763905562225e-05,
+      "loss": 0.4706,
+      "step": 17959
+    },
+    {
+      "epoch": 22.9888,
+      "grad_norm": 1.0810490846633911,
+      "learning_rate": 1.409563825530212e-05,
+      "loss": 0.4711,
+      "step": 17960
+    },
+    {
+      "epoch": 22.99008,
+      "grad_norm": 1.1601754426956177,
+      "learning_rate": 1.4093637454981992e-05,
+      "loss": 0.5184,
+      "step": 17961
+    },
+    {
+      "epoch": 22.99136,
+      "grad_norm": 1.089206576347351,
+      "learning_rate": 1.4091636654661866e-05,
+      "loss": 0.5103,
+      "step": 17962
+    },
+    {
+      "epoch": 22.99264,
+      "grad_norm": 1.1088589429855347,
+      "learning_rate": 1.4089635854341738e-05,
+      "loss": 0.4609,
+      "step": 17963
+    },
+    {
+      "epoch": 22.99392,
+      "grad_norm": 1.1909695863723755,
+      "learning_rate": 1.4087635054021608e-05,
+      "loss": 0.5124,
+      "step": 17964
+    },
+    {
+      "epoch": 22.9952,
+      "grad_norm": 1.0958809852600098,
+      "learning_rate": 1.4085634253701483e-05,
+      "loss": 0.4415,
+      "step": 17965
+    },
+    {
+      "epoch": 22.99648,
+      "grad_norm": 1.120891809463501,
+      "learning_rate": 1.4083633453381353e-05,
+      "loss": 0.4856,
+      "step": 17966
+    },
+    {
+      "epoch": 22.99776,
+      "grad_norm": 1.111769437789917,
+      "learning_rate": 1.4081632653061225e-05,
+      "loss": 0.5141,
+      "step": 17967
+    },
+    {
+      "epoch": 22.99904,
+      "grad_norm": 1.1565723419189453,
+      "learning_rate": 1.4079631852741095e-05,
+      "loss": 0.5062,
+      "step": 17968
+    },
+    {
+      "epoch": 23.00032,
+      "grad_norm": 2.3065271377563477,
+      "learning_rate": 1.407763105242097e-05,
+      "loss": 0.7523,
+      "step": 17969
+    },
+    {
+      "epoch": 23.0016,
+      "grad_norm": 1.0669747591018677,
+      "learning_rate": 1.407563025210084e-05,
+      "loss": 0.4847,
+      "step": 17970
+    },
+    {
+      "epoch": 23.00288,
+      "grad_norm": 1.0902814865112305,
+      "learning_rate": 1.4073629451780712e-05,
+      "loss": 0.4821,
+      "step": 17971
+    },
+    {
+      "epoch": 23.00416,
+      "grad_norm": 1.1460057497024536,
+      "learning_rate": 1.4071628651460586e-05,
+      "loss": 0.4837,
+      "step": 17972
+    },
+    {
+      "epoch": 23.00544,
+      "grad_norm": 1.0994794368743896,
+      "learning_rate": 1.4069627851140458e-05,
+      "loss": 0.4203,
+      "step": 17973
+    },
+    {
+      "epoch": 23.00672,
+      "grad_norm": 1.093055009841919,
+      "learning_rate": 1.4067627050820328e-05,
+      "loss": 0.4731,
+      "step": 17974
+    },
+    {
+      "epoch": 23.008,
+      "grad_norm": 1.139220952987671,
+      "learning_rate": 1.40656262505002e-05,
+      "loss": 0.4932,
+      "step": 17975
+    },
+    {
+      "epoch": 23.00928,
+      "grad_norm": 1.0311874151229858,
+      "learning_rate": 1.4063625450180073e-05,
+      "loss": 0.4708,
+      "step": 17976
+    },
+    {
+      "epoch": 23.01056,
+      "grad_norm": 1.0706690549850464,
+      "learning_rate": 1.4061624649859945e-05,
+      "loss": 0.4173,
+      "step": 17977
+    },
+    {
+      "epoch": 23.01184,
+      "grad_norm": 1.192206621170044,
+      "learning_rate": 1.4059623849539815e-05,
+      "loss": 0.4742,
+      "step": 17978
+    },
+    {
+      "epoch": 23.01312,
+      "grad_norm": 1.13650381565094,
+      "learning_rate": 1.4057623049219689e-05,
+      "loss": 0.4875,
+      "step": 17979
+    },
+    {
+      "epoch": 23.0144,
+      "grad_norm": 1.1168575286865234,
+      "learning_rate": 1.405562224889956e-05,
+      "loss": 0.4488,
+      "step": 17980
+    },
+    {
+      "epoch": 23.01568,
+      "grad_norm": 1.1627042293548584,
+      "learning_rate": 1.4053621448579433e-05,
+      "loss": 0.4947,
+      "step": 17981
+    },
+    {
+      "epoch": 23.01696,
+      "grad_norm": 1.2101789712905884,
+      "learning_rate": 1.4051620648259303e-05,
+      "loss": 0.5105,
+      "step": 17982
+    },
+    {
+      "epoch": 23.01824,
+      "grad_norm": 1.080032467842102,
+      "learning_rate": 1.4049619847939176e-05,
+      "loss": 0.4658,
+      "step": 17983
+    },
+    {
+      "epoch": 23.01952,
+      "grad_norm": 1.169611930847168,
+      "learning_rate": 1.4047619047619048e-05,
+      "loss": 0.4832,
+      "step": 17984
+    },
+    {
+      "epoch": 23.0208,
+      "grad_norm": 1.147308588027954,
+      "learning_rate": 1.404561824729892e-05,
+      "loss": 0.4709,
+      "step": 17985
+    },
+    {
+      "epoch": 23.02208,
+      "grad_norm": 1.1938717365264893,
+      "learning_rate": 1.4043617446978794e-05,
+      "loss": 0.5101,
+      "step": 17986
+    },
+    {
+      "epoch": 23.02336,
+      "grad_norm": 1.174412488937378,
+      "learning_rate": 1.4041616646658664e-05,
+      "loss": 0.4884,
+      "step": 17987
+    },
+    {
+      "epoch": 23.02464,
+      "grad_norm": 1.1288552284240723,
+      "learning_rate": 1.4039615846338536e-05,
+      "loss": 0.4617,
+      "step": 17988
+    },
+    {
+      "epoch": 23.02592,
+      "grad_norm": 1.100049614906311,
+      "learning_rate": 1.4037615046018408e-05,
+      "loss": 0.4318,
+      "step": 17989
+    },
+    {
+      "epoch": 23.0272,
+      "grad_norm": 1.1527884006500244,
+      "learning_rate": 1.4035614245698281e-05,
+      "loss": 0.4885,
+      "step": 17990
+    },
+    {
+      "epoch": 23.02848,
+      "grad_norm": 1.0793335437774658,
+      "learning_rate": 1.4033613445378151e-05,
+      "loss": 0.4429,
+      "step": 17991
+    },
+    {
+      "epoch": 23.02976,
+      "grad_norm": 1.1420992612838745,
+      "learning_rate": 1.4031612645058023e-05,
+      "loss": 0.4655,
+      "step": 17992
+    },
+    {
+      "epoch": 23.03104,
+      "grad_norm": 1.115891695022583,
+      "learning_rate": 1.4029611844737897e-05,
+      "loss": 0.5128,
+      "step": 17993
+    },
+    {
+      "epoch": 23.03232,
+      "grad_norm": 1.1061939001083374,
+      "learning_rate": 1.4027611044417769e-05,
+      "loss": 0.4666,
+      "step": 17994
+    },
+    {
+      "epoch": 23.0336,
+      "grad_norm": 1.1033188104629517,
+      "learning_rate": 1.4025610244097639e-05,
+      "loss": 0.4384,
+      "step": 17995
+    },
+    {
+      "epoch": 23.03488,
+      "grad_norm": 1.1487460136413574,
+      "learning_rate": 1.402360944377751e-05,
+      "loss": 0.4253,
+      "step": 17996
+    },
+    {
+      "epoch": 23.03616,
+      "grad_norm": 1.210303544998169,
+      "learning_rate": 1.4021608643457384e-05,
+      "loss": 0.4954,
+      "step": 17997
+    },
+    {
+      "epoch": 23.03744,
+      "grad_norm": 1.119410753250122,
+      "learning_rate": 1.4019607843137256e-05,
+      "loss": 0.4598,
+      "step": 17998
+    },
+    {
+      "epoch": 23.03872,
+      "grad_norm": 1.1100172996520996,
+      "learning_rate": 1.4017607042817126e-05,
+      "loss": 0.4887,
+      "step": 17999
+    },
+    {
+      "epoch": 23.04,
+      "grad_norm": 1.1196397542953491,
+      "learning_rate": 1.4015606242497001e-05,
+      "loss": 0.4729,
+      "step": 18000
+    },
+    {
+      "epoch": 23.04128,
+      "grad_norm": 1.1445229053497314,
+      "learning_rate": 1.4013605442176872e-05,
+      "loss": 0.4818,
+      "step": 18001
+    },
+    {
+      "epoch": 23.04256,
+      "grad_norm": 1.0524245500564575,
+      "learning_rate": 1.4011604641856743e-05,
+      "loss": 0.45,
+      "step": 18002
+    },
+    {
+      "epoch": 23.04384,
+      "grad_norm": 1.1181271076202393,
+      "learning_rate": 1.4009603841536614e-05,
+      "loss": 0.4498,
+      "step": 18003
+    },
+    {
+      "epoch": 23.04512,
+      "grad_norm": 1.0702544450759888,
+      "learning_rate": 1.4007603041216489e-05,
+      "loss": 0.4782,
+      "step": 18004
+    },
+    {
+      "epoch": 23.0464,
+      "grad_norm": 1.1783769130706787,
+      "learning_rate": 1.4005602240896359e-05,
+      "loss": 0.5008,
+      "step": 18005
+    },
+    {
+      "epoch": 23.04768,
+      "grad_norm": 1.1337677240371704,
+      "learning_rate": 1.400360144057623e-05,
+      "loss": 0.465,
+      "step": 18006
+    },
+    {
+      "epoch": 23.04896,
+      "grad_norm": 1.0862562656402588,
+      "learning_rate": 1.4001600640256101e-05,
+      "loss": 0.4675,
+      "step": 18007
+    },
+    {
+      "epoch": 23.05024,
+      "grad_norm": 1.122215747833252,
+      "learning_rate": 1.3999599839935976e-05,
+      "loss": 0.4764,
+      "step": 18008
+    },
+    {
+      "epoch": 23.05152,
+      "grad_norm": 1.1443709135055542,
+      "learning_rate": 1.3997599039615846e-05,
+      "loss": 0.496,
+      "step": 18009
+    },
+    {
+      "epoch": 23.0528,
+      "grad_norm": 1.1884753704071045,
+      "learning_rate": 1.3995598239295718e-05,
+      "loss": 0.5125,
+      "step": 18010
+    },
+    {
+      "epoch": 23.05408,
+      "grad_norm": 1.1133694648742676,
+      "learning_rate": 1.3993597438975592e-05,
+      "loss": 0.4876,
+      "step": 18011
+    },
+    {
+      "epoch": 23.05536,
+      "grad_norm": 1.0614551305770874,
+      "learning_rate": 1.3991596638655464e-05,
+      "loss": 0.4339,
+      "step": 18012
+    },
+    {
+      "epoch": 23.05664,
+      "grad_norm": 1.1484719514846802,
+      "learning_rate": 1.3989595838335334e-05,
+      "loss": 0.4655,
+      "step": 18013
+    },
+    {
+      "epoch": 23.05792,
+      "grad_norm": 1.1289807558059692,
+      "learning_rate": 1.3987595038015206e-05,
+      "loss": 0.4593,
+      "step": 18014
+    },
+    {
+      "epoch": 23.0592,
+      "grad_norm": 1.1590949296951294,
+      "learning_rate": 1.398559423769508e-05,
+      "loss": 0.4633,
+      "step": 18015
+    },
+    {
+      "epoch": 23.06048,
+      "grad_norm": 1.1496042013168335,
+      "learning_rate": 1.3983593437374951e-05,
+      "loss": 0.4818,
+      "step": 18016
+    },
+    {
+      "epoch": 23.06176,
+      "grad_norm": 1.1318520307540894,
+      "learning_rate": 1.3981592637054821e-05,
+      "loss": 0.4942,
+      "step": 18017
+    },
+    {
+      "epoch": 23.06304,
+      "grad_norm": 1.1314492225646973,
+      "learning_rate": 1.3979591836734696e-05,
+      "loss": 0.4815,
+      "step": 18018
+    },
+    {
+      "epoch": 23.06432,
+      "grad_norm": 1.1428439617156982,
+      "learning_rate": 1.3977591036414567e-05,
+      "loss": 0.4794,
+      "step": 18019
+    },
+    {
+      "epoch": 23.0656,
+      "grad_norm": 1.1494139432907104,
+      "learning_rate": 1.3975590236094438e-05,
+      "loss": 0.4554,
+      "step": 18020
+    },
+    {
+      "epoch": 23.06688,
+      "grad_norm": 1.0700560808181763,
+      "learning_rate": 1.3973589435774309e-05,
+      "loss": 0.4345,
+      "step": 18021
+    },
+    {
+      "epoch": 23.06816,
+      "grad_norm": 1.1106326580047607,
+      "learning_rate": 1.3971588635454184e-05,
+      "loss": 0.4747,
+      "step": 18022
+    },
+    {
+      "epoch": 23.06944,
+      "grad_norm": 1.1512048244476318,
+      "learning_rate": 1.3969587835134054e-05,
+      "loss": 0.478,
+      "step": 18023
+    },
+    {
+      "epoch": 23.07072,
+      "grad_norm": 1.0818241834640503,
+      "learning_rate": 1.3967587034813926e-05,
+      "loss": 0.4671,
+      "step": 18024
+    },
+    {
+      "epoch": 23.072,
+      "grad_norm": 1.1178078651428223,
+      "learning_rate": 1.39655862344938e-05,
+      "loss": 0.4329,
+      "step": 18025
+    },
+    {
+      "epoch": 23.07328,
+      "grad_norm": 1.1609102487564087,
+      "learning_rate": 1.3963585434173671e-05,
+      "loss": 0.4793,
+      "step": 18026
+    },
+    {
+      "epoch": 23.07456,
+      "grad_norm": 1.1655328273773193,
+      "learning_rate": 1.3961584633853541e-05,
+      "loss": 0.4715,
+      "step": 18027
+    },
+    {
+      "epoch": 23.07584,
+      "grad_norm": 1.150871753692627,
+      "learning_rate": 1.3959583833533413e-05,
+      "loss": 0.4812,
+      "step": 18028
+    },
+    {
+      "epoch": 23.07712,
+      "grad_norm": 1.2036709785461426,
+      "learning_rate": 1.3957583033213287e-05,
+      "loss": 0.5125,
+      "step": 18029
+    },
+    {
+      "epoch": 23.0784,
+      "grad_norm": 1.169793725013733,
+      "learning_rate": 1.3955582232893159e-05,
+      "loss": 0.4995,
+      "step": 18030
+    },
+    {
+      "epoch": 23.07968,
+      "grad_norm": 1.1694694757461548,
+      "learning_rate": 1.3953581432573029e-05,
+      "loss": 0.5015,
+      "step": 18031
+    },
+    {
+      "epoch": 23.08096,
+      "grad_norm": 1.1842191219329834,
+      "learning_rate": 1.3951580632252902e-05,
+      "loss": 0.5023,
+      "step": 18032
+    },
+    {
+      "epoch": 23.08224,
+      "grad_norm": 1.1319433450698853,
+      "learning_rate": 1.3949579831932774e-05,
+      "loss": 0.4759,
+      "step": 18033
+    },
+    {
+      "epoch": 23.08352,
+      "grad_norm": 1.0593634843826294,
+      "learning_rate": 1.3947579031612646e-05,
+      "loss": 0.4504,
+      "step": 18034
+    },
+    {
+      "epoch": 23.0848,
+      "grad_norm": 1.1120679378509521,
+      "learning_rate": 1.3945578231292516e-05,
+      "loss": 0.5023,
+      "step": 18035
+    },
+    {
+      "epoch": 23.08608,
+      "grad_norm": 1.1178866624832153,
+      "learning_rate": 1.394357743097239e-05,
+      "loss": 0.4568,
+      "step": 18036
+    },
+    {
+      "epoch": 23.08736,
+      "grad_norm": 1.2306301593780518,
+      "learning_rate": 1.3941576630652262e-05,
+      "loss": 0.5722,
+      "step": 18037
+    },
+    {
+      "epoch": 23.08864,
+      "grad_norm": 1.0901416540145874,
+      "learning_rate": 1.3939575830332134e-05,
+      "loss": 0.4703,
+      "step": 18038
+    },
+    {
+      "epoch": 23.08992,
+      "grad_norm": 1.0788695812225342,
+      "learning_rate": 1.3937575030012007e-05,
+      "loss": 0.4225,
+      "step": 18039
+    },
+    {
+      "epoch": 23.0912,
+      "grad_norm": 1.1180025339126587,
+      "learning_rate": 1.3935574229691877e-05,
+      "loss": 0.4538,
+      "step": 18040
+    },
+    {
+      "epoch": 23.09248,
+      "grad_norm": 1.1626075506210327,
+      "learning_rate": 1.393357342937175e-05,
+      "loss": 0.4787,
+      "step": 18041
+    },
+    {
+      "epoch": 23.09376,
+      "grad_norm": 1.1562151908874512,
+      "learning_rate": 1.3931572629051621e-05,
+      "loss": 0.4508,
+      "step": 18042
+    },
+    {
+      "epoch": 23.09504,
+      "grad_norm": 1.0841056108474731,
+      "learning_rate": 1.3929571828731495e-05,
+      "loss": 0.4401,
+      "step": 18043
+    },
+    {
+      "epoch": 23.09632,
+      "grad_norm": 1.1172791719436646,
+      "learning_rate": 1.3927571028411365e-05,
+      "loss": 0.4876,
+      "step": 18044
+    },
+    {
+      "epoch": 23.0976,
+      "grad_norm": 1.1132711172103882,
+      "learning_rate": 1.3925570228091237e-05,
+      "loss": 0.483,
+      "step": 18045
+    },
+    {
+      "epoch": 23.09888,
+      "grad_norm": 1.1011528968811035,
+      "learning_rate": 1.392356942777111e-05,
+      "loss": 0.4601,
+      "step": 18046
+    },
+    {
+      "epoch": 23.10016,
+      "grad_norm": 1.2015907764434814,
+      "learning_rate": 1.3921568627450982e-05,
+      "loss": 0.4793,
+      "step": 18047
+    },
+    {
+      "epoch": 23.10144,
+      "grad_norm": 1.1870005130767822,
+      "learning_rate": 1.3919567827130852e-05,
+      "loss": 0.4781,
+      "step": 18048
+    },
+    {
+      "epoch": 23.10272,
+      "grad_norm": 1.1651169061660767,
+      "learning_rate": 1.3917567026810724e-05,
+      "loss": 0.4782,
+      "step": 18049
+    },
+    {
+      "epoch": 23.104,
+      "grad_norm": 1.1237787008285522,
+      "learning_rate": 1.3915566226490598e-05,
+      "loss": 0.4722,
+      "step": 18050
+    },
+    {
+      "epoch": 23.10528,
+      "grad_norm": 1.1419413089752197,
+      "learning_rate": 1.391356542617047e-05,
+      "loss": 0.4772,
+      "step": 18051
+    },
+    {
+      "epoch": 23.10656,
+      "grad_norm": 1.1415618658065796,
+      "learning_rate": 1.391156462585034e-05,
+      "loss": 0.4407,
+      "step": 18052
+    },
+    {
+      "epoch": 23.10784,
+      "grad_norm": 1.1670557260513306,
+      "learning_rate": 1.3909563825530215e-05,
+      "loss": 0.5378,
+      "step": 18053
+    },
+    {
+      "epoch": 23.10912,
+      "grad_norm": 1.1420172452926636,
+      "learning_rate": 1.3907563025210085e-05,
+      "loss": 0.4717,
+      "step": 18054
+    },
+    {
+      "epoch": 23.1104,
+      "grad_norm": 1.1076371669769287,
+      "learning_rate": 1.3905562224889957e-05,
+      "loss": 0.4571,
+      "step": 18055
+    },
+    {
+      "epoch": 23.11168,
+      "grad_norm": 1.1110548973083496,
+      "learning_rate": 1.3903561424569827e-05,
+      "loss": 0.4484,
+      "step": 18056
+    },
+    {
+      "epoch": 23.11296,
+      "grad_norm": 1.1428343057632446,
+      "learning_rate": 1.3901560624249702e-05,
+      "loss": 0.5226,
+      "step": 18057
+    },
+    {
+      "epoch": 23.11424,
+      "grad_norm": 1.1347376108169556,
+      "learning_rate": 1.3899559823929572e-05,
+      "loss": 0.4633,
+      "step": 18058
+    },
+    {
+      "epoch": 23.11552,
+      "grad_norm": 1.1556316614151,
+      "learning_rate": 1.3897559023609444e-05,
+      "loss": 0.4751,
+      "step": 18059
+    },
+    {
+      "epoch": 23.1168,
+      "grad_norm": 1.1056498289108276,
+      "learning_rate": 1.3895558223289318e-05,
+      "loss": 0.4446,
+      "step": 18060
+    },
+    {
+      "epoch": 23.11808,
+      "grad_norm": 1.194538950920105,
+      "learning_rate": 1.389355742296919e-05,
+      "loss": 0.5121,
+      "step": 18061
+    },
+    {
+      "epoch": 23.11936,
+      "grad_norm": 1.0786786079406738,
+      "learning_rate": 1.389155662264906e-05,
+      "loss": 0.482,
+      "step": 18062
+    },
+    {
+      "epoch": 23.12064,
+      "grad_norm": 1.1537976264953613,
+      "learning_rate": 1.3889555822328932e-05,
+      "loss": 0.5048,
+      "step": 18063
+    },
+    {
+      "epoch": 23.12192,
+      "grad_norm": 1.1101106405258179,
+      "learning_rate": 1.3887555022008805e-05,
+      "loss": 0.5427,
+      "step": 18064
+    },
+    {
+      "epoch": 23.1232,
+      "grad_norm": 1.0096379518508911,
+      "learning_rate": 1.3885554221688677e-05,
+      "loss": 0.4264,
+      "step": 18065
+    },
+    {
+      "epoch": 23.12448,
+      "grad_norm": 1.1329681873321533,
+      "learning_rate": 1.3883553421368547e-05,
+      "loss": 0.5005,
+      "step": 18066
+    },
+    {
+      "epoch": 23.12576,
+      "grad_norm": 1.1219654083251953,
+      "learning_rate": 1.3881552621048419e-05,
+      "loss": 0.4677,
+      "step": 18067
+    },
+    {
+      "epoch": 23.12704,
+      "grad_norm": 1.1316763162612915,
+      "learning_rate": 1.3879551820728293e-05,
+      "loss": 0.4857,
+      "step": 18068
+    },
+    {
+      "epoch": 23.12832,
+      "grad_norm": 1.1240242719650269,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 0.481,
+      "step": 18069
+    },
+    {
+      "epoch": 23.1296,
+      "grad_norm": 1.0853866338729858,
+      "learning_rate": 1.3875550220088035e-05,
+      "loss": 0.4481,
+      "step": 18070
+    },
+    {
+      "epoch": 23.13088,
+      "grad_norm": 1.1638892889022827,
+      "learning_rate": 1.3873549419767908e-05,
+      "loss": 0.5032,
+      "step": 18071
+    },
+    {
+      "epoch": 23.13216,
+      "grad_norm": 1.1282097101211548,
+      "learning_rate": 1.387154861944778e-05,
+      "loss": 0.4717,
+      "step": 18072
+    },
+    {
+      "epoch": 23.13344,
+      "grad_norm": 1.1175307035446167,
+      "learning_rate": 1.3869547819127652e-05,
+      "loss": 0.4375,
+      "step": 18073
+    },
+    {
+      "epoch": 23.13472,
+      "grad_norm": 1.0587260723114014,
+      "learning_rate": 1.3867547018807522e-05,
+      "loss": 0.4765,
+      "step": 18074
+    },
+    {
+      "epoch": 23.136,
+      "grad_norm": 1.179371953010559,
+      "learning_rate": 1.3865546218487396e-05,
+      "loss": 0.5337,
+      "step": 18075
+    },
+    {
+      "epoch": 23.13728,
+      "grad_norm": 1.153063178062439,
+      "learning_rate": 1.3863545418167268e-05,
+      "loss": 0.4959,
+      "step": 18076
+    },
+    {
+      "epoch": 23.13856,
+      "grad_norm": 1.1271103620529175,
+      "learning_rate": 1.386154461784714e-05,
+      "loss": 0.478,
+      "step": 18077
+    },
+    {
+      "epoch": 23.13984,
+      "grad_norm": 1.0915660858154297,
+      "learning_rate": 1.3859543817527013e-05,
+      "loss": 0.4598,
+      "step": 18078
+    },
+    {
+      "epoch": 23.14112,
+      "grad_norm": 1.1732710599899292,
+      "learning_rate": 1.3857543017206883e-05,
+      "loss": 0.4857,
+      "step": 18079
+    },
+    {
+      "epoch": 23.1424,
+      "grad_norm": 1.1222920417785645,
+      "learning_rate": 1.3855542216886755e-05,
+      "loss": 0.4344,
+      "step": 18080
+    },
+    {
+      "epoch": 23.14368,
+      "grad_norm": 1.1178438663482666,
+      "learning_rate": 1.3853541416566627e-05,
+      "loss": 0.4794,
+      "step": 18081
+    },
+    {
+      "epoch": 23.14496,
+      "grad_norm": 1.1883807182312012,
+      "learning_rate": 1.38515406162465e-05,
+      "loss": 0.4837,
+      "step": 18082
+    },
+    {
+      "epoch": 23.14624,
+      "grad_norm": 1.1413466930389404,
+      "learning_rate": 1.384953981592637e-05,
+      "loss": 0.4826,
+      "step": 18083
+    },
+    {
+      "epoch": 23.14752,
+      "grad_norm": 1.0779472589492798,
+      "learning_rate": 1.3847539015606242e-05,
+      "loss": 0.4544,
+      "step": 18084
+    },
+    {
+      "epoch": 23.1488,
+      "grad_norm": 1.0848368406295776,
+      "learning_rate": 1.3845538215286116e-05,
+      "loss": 0.4832,
+      "step": 18085
+    },
+    {
+      "epoch": 23.15008,
+      "grad_norm": 1.117047667503357,
+      "learning_rate": 1.3843537414965988e-05,
+      "loss": 0.4593,
+      "step": 18086
+    },
+    {
+      "epoch": 23.15136,
+      "grad_norm": 1.113674283027649,
+      "learning_rate": 1.3841536614645858e-05,
+      "loss": 0.4669,
+      "step": 18087
+    },
+    {
+      "epoch": 23.15264,
+      "grad_norm": 1.186633825302124,
+      "learning_rate": 1.383953581432573e-05,
+      "loss": 0.4913,
+      "step": 18088
+    },
+    {
+      "epoch": 23.15392,
+      "grad_norm": 1.1024609804153442,
+      "learning_rate": 1.3837535014005603e-05,
+      "loss": 0.4779,
+      "step": 18089
+    },
+    {
+      "epoch": 23.1552,
+      "grad_norm": 1.0634430646896362,
+      "learning_rate": 1.3835534213685475e-05,
+      "loss": 0.4301,
+      "step": 18090
+    },
+    {
+      "epoch": 23.15648,
+      "grad_norm": 1.10919988155365,
+      "learning_rate": 1.3833533413365345e-05,
+      "loss": 0.4733,
+      "step": 18091
+    },
+    {
+      "epoch": 23.15776,
+      "grad_norm": 1.1104084253311157,
+      "learning_rate": 1.383153261304522e-05,
+      "loss": 0.4612,
+      "step": 18092
+    },
+    {
+      "epoch": 23.15904,
+      "grad_norm": 1.1497199535369873,
+      "learning_rate": 1.382953181272509e-05,
+      "loss": 0.4389,
+      "step": 18093
+    },
+    {
+      "epoch": 23.16032,
+      "grad_norm": 1.177474856376648,
+      "learning_rate": 1.3827531012404963e-05,
+      "loss": 0.4916,
+      "step": 18094
+    },
+    {
+      "epoch": 23.1616,
+      "grad_norm": 1.1319663524627686,
+      "learning_rate": 1.3825530212084833e-05,
+      "loss": 0.4834,
+      "step": 18095
+    },
+    {
+      "epoch": 23.16288,
+      "grad_norm": 1.1392722129821777,
+      "learning_rate": 1.3823529411764708e-05,
+      "loss": 0.4928,
+      "step": 18096
+    },
+    {
+      "epoch": 23.16416,
+      "grad_norm": 1.1550222635269165,
+      "learning_rate": 1.3821528611444578e-05,
+      "loss": 0.4575,
+      "step": 18097
+    },
+    {
+      "epoch": 23.16544,
+      "grad_norm": 1.17167067527771,
+      "learning_rate": 1.381952781112445e-05,
+      "loss": 0.5154,
+      "step": 18098
+    },
+    {
+      "epoch": 23.16672,
+      "grad_norm": 1.167103886604309,
+      "learning_rate": 1.3817527010804324e-05,
+      "loss": 0.4935,
+      "step": 18099
+    },
+    {
+      "epoch": 23.168,
+      "grad_norm": 1.1817306280136108,
+      "learning_rate": 1.3815526210484195e-05,
+      "loss": 0.4877,
+      "step": 18100
+    },
+    {
+      "epoch": 23.16928,
+      "grad_norm": 1.1134272813796997,
+      "learning_rate": 1.3813525410164066e-05,
+      "loss": 0.4528,
+      "step": 18101
+    },
+    {
+      "epoch": 23.17056,
+      "grad_norm": 1.1140645742416382,
+      "learning_rate": 1.3811524609843938e-05,
+      "loss": 0.4576,
+      "step": 18102
+    },
+    {
+      "epoch": 23.17184,
+      "grad_norm": 1.140651822090149,
+      "learning_rate": 1.3809523809523811e-05,
+      "loss": 0.4729,
+      "step": 18103
+    },
+    {
+      "epoch": 23.17312,
+      "grad_norm": 1.1449460983276367,
+      "learning_rate": 1.3807523009203683e-05,
+      "loss": 0.4636,
+      "step": 18104
+    },
+    {
+      "epoch": 23.1744,
+      "grad_norm": 1.1074920892715454,
+      "learning_rate": 1.3805522208883553e-05,
+      "loss": 0.4774,
+      "step": 18105
+    },
+    {
+      "epoch": 23.17568,
+      "grad_norm": 1.1994163990020752,
+      "learning_rate": 1.3803521408563427e-05,
+      "loss": 0.4314,
+      "step": 18106
+    },
+    {
+      "epoch": 23.17696,
+      "grad_norm": 1.1801729202270508,
+      "learning_rate": 1.3801520608243298e-05,
+      "loss": 0.4872,
+      "step": 18107
+    },
+    {
+      "epoch": 23.17824,
+      "grad_norm": 1.1620689630508423,
+      "learning_rate": 1.379951980792317e-05,
+      "loss": 0.473,
+      "step": 18108
+    },
+    {
+      "epoch": 23.17952,
+      "grad_norm": 1.1328171491622925,
+      "learning_rate": 1.379751900760304e-05,
+      "loss": 0.4785,
+      "step": 18109
+    },
+    {
+      "epoch": 23.1808,
+      "grad_norm": 1.1725462675094604,
+      "learning_rate": 1.3795518207282914e-05,
+      "loss": 0.4587,
+      "step": 18110
+    },
+    {
+      "epoch": 23.18208,
+      "grad_norm": 1.1395995616912842,
+      "learning_rate": 1.3793517406962786e-05,
+      "loss": 0.4554,
+      "step": 18111
+    },
+    {
+      "epoch": 23.18336,
+      "grad_norm": 1.0973244905471802,
+      "learning_rate": 1.3791516606642658e-05,
+      "loss": 0.4621,
+      "step": 18112
+    },
+    {
+      "epoch": 23.18464,
+      "grad_norm": 1.1823533773422241,
+      "learning_rate": 1.3789515806322531e-05,
+      "loss": 0.5013,
+      "step": 18113
+    },
+    {
+      "epoch": 23.18592,
+      "grad_norm": 1.124189019203186,
+      "learning_rate": 1.3787515006002401e-05,
+      "loss": 0.4375,
+      "step": 18114
+    },
+    {
+      "epoch": 23.1872,
+      "grad_norm": 1.0450643301010132,
+      "learning_rate": 1.3785514205682273e-05,
+      "loss": 0.4448,
+      "step": 18115
+    },
+    {
+      "epoch": 23.18848,
+      "grad_norm": 1.1765981912612915,
+      "learning_rate": 1.3783513405362145e-05,
+      "loss": 0.5302,
+      "step": 18116
+    },
+    {
+      "epoch": 23.18976,
+      "grad_norm": 1.1082278490066528,
+      "learning_rate": 1.3781512605042019e-05,
+      "loss": 0.4354,
+      "step": 18117
+    },
+    {
+      "epoch": 23.19104,
+      "grad_norm": 1.123656153678894,
+      "learning_rate": 1.3779511804721889e-05,
+      "loss": 0.4896,
+      "step": 18118
+    },
+    {
+      "epoch": 23.19232,
+      "grad_norm": 1.1027441024780273,
+      "learning_rate": 1.377751100440176e-05,
+      "loss": 0.4553,
+      "step": 18119
+    },
+    {
+      "epoch": 23.1936,
+      "grad_norm": 1.1921277046203613,
+      "learning_rate": 1.3775510204081633e-05,
+      "loss": 0.5235,
+      "step": 18120
+    },
+    {
+      "epoch": 23.19488,
+      "grad_norm": 1.1635897159576416,
+      "learning_rate": 1.3773509403761506e-05,
+      "loss": 0.4774,
+      "step": 18121
+    },
+    {
+      "epoch": 23.19616,
+      "grad_norm": 1.1405202150344849,
+      "learning_rate": 1.3771508603441376e-05,
+      "loss": 0.4663,
+      "step": 18122
+    },
+    {
+      "epoch": 23.19744,
+      "grad_norm": 1.0812584161758423,
+      "learning_rate": 1.3769507803121248e-05,
+      "loss": 0.4246,
+      "step": 18123
+    },
+    {
+      "epoch": 23.19872,
+      "grad_norm": 1.1046404838562012,
+      "learning_rate": 1.3767507002801122e-05,
+      "loss": 0.444,
+      "step": 18124
+    },
+    {
+      "epoch": 23.2,
+      "grad_norm": 1.1270830631256104,
+      "learning_rate": 1.3765506202480994e-05,
+      "loss": 0.4681,
+      "step": 18125
+    },
+    {
+      "epoch": 23.20128,
+      "grad_norm": 1.092661738395691,
+      "learning_rate": 1.3763505402160864e-05,
+      "loss": 0.4537,
+      "step": 18126
+    },
+    {
+      "epoch": 23.20256,
+      "grad_norm": 1.1461021900177002,
+      "learning_rate": 1.3761504601840736e-05,
+      "loss": 0.4697,
+      "step": 18127
+    },
+    {
+      "epoch": 23.20384,
+      "grad_norm": 1.1606507301330566,
+      "learning_rate": 1.375950380152061e-05,
+      "loss": 0.5199,
+      "step": 18128
+    },
+    {
+      "epoch": 23.20512,
+      "grad_norm": 1.1296125650405884,
+      "learning_rate": 1.3757503001200481e-05,
+      "loss": 0.4839,
+      "step": 18129
+    },
+    {
+      "epoch": 23.2064,
+      "grad_norm": 1.2692511081695557,
+      "learning_rate": 1.3755502200880351e-05,
+      "loss": 0.4903,
+      "step": 18130
+    },
+    {
+      "epoch": 23.20768,
+      "grad_norm": 1.1228126287460327,
+      "learning_rate": 1.3753501400560226e-05,
+      "loss": 0.4585,
+      "step": 18131
+    },
+    {
+      "epoch": 23.20896,
+      "grad_norm": 1.1252394914627075,
+      "learning_rate": 1.3751500600240097e-05,
+      "loss": 0.5162,
+      "step": 18132
+    },
+    {
+      "epoch": 23.21024,
+      "grad_norm": 1.128462791442871,
+      "learning_rate": 1.3749499799919968e-05,
+      "loss": 0.4823,
+      "step": 18133
+    },
+    {
+      "epoch": 23.21152,
+      "grad_norm": 1.157885193824768,
+      "learning_rate": 1.3747498999599839e-05,
+      "loss": 0.5073,
+      "step": 18134
+    },
+    {
+      "epoch": 23.2128,
+      "grad_norm": 1.0952425003051758,
+      "learning_rate": 1.3745498199279714e-05,
+      "loss": 0.5003,
+      "step": 18135
+    },
+    {
+      "epoch": 23.21408,
+      "grad_norm": 1.1897740364074707,
+      "learning_rate": 1.3743497398959584e-05,
+      "loss": 0.5091,
+      "step": 18136
+    },
+    {
+      "epoch": 23.21536,
+      "grad_norm": 1.1394944190979004,
+      "learning_rate": 1.3741496598639456e-05,
+      "loss": 0.4115,
+      "step": 18137
+    },
+    {
+      "epoch": 23.21664,
+      "grad_norm": 1.1047616004943848,
+      "learning_rate": 1.373949579831933e-05,
+      "loss": 0.4888,
+      "step": 18138
+    },
+    {
+      "epoch": 23.21792,
+      "grad_norm": 1.1248316764831543,
+      "learning_rate": 1.3737494997999201e-05,
+      "loss": 0.4505,
+      "step": 18139
+    },
+    {
+      "epoch": 23.2192,
+      "grad_norm": 1.1958407163619995,
+      "learning_rate": 1.3735494197679071e-05,
+      "loss": 0.5051,
+      "step": 18140
+    },
+    {
+      "epoch": 23.22048,
+      "grad_norm": 1.186632513999939,
+      "learning_rate": 1.3733493397358943e-05,
+      "loss": 0.4692,
+      "step": 18141
+    },
+    {
+      "epoch": 23.22176,
+      "grad_norm": 1.1667543649673462,
+      "learning_rate": 1.3731492597038817e-05,
+      "loss": 0.5314,
+      "step": 18142
+    },
+    {
+      "epoch": 23.22304,
+      "grad_norm": 1.1657756567001343,
+      "learning_rate": 1.3729491796718689e-05,
+      "loss": 0.5432,
+      "step": 18143
+    },
+    {
+      "epoch": 23.22432,
+      "grad_norm": 1.1313673257827759,
+      "learning_rate": 1.3727490996398559e-05,
+      "loss": 0.4813,
+      "step": 18144
+    },
+    {
+      "epoch": 23.2256,
+      "grad_norm": 1.1224007606506348,
+      "learning_rate": 1.3725490196078432e-05,
+      "loss": 0.4471,
+      "step": 18145
+    },
+    {
+      "epoch": 23.22688,
+      "grad_norm": 1.171020746231079,
+      "learning_rate": 1.3723489395758304e-05,
+      "loss": 0.5398,
+      "step": 18146
+    },
+    {
+      "epoch": 23.22816,
+      "grad_norm": 1.0536117553710938,
+      "learning_rate": 1.3721488595438176e-05,
+      "loss": 0.4308,
+      "step": 18147
+    },
+    {
+      "epoch": 23.22944,
+      "grad_norm": 1.146742820739746,
+      "learning_rate": 1.3719487795118046e-05,
+      "loss": 0.4901,
+      "step": 18148
+    },
+    {
+      "epoch": 23.23072,
+      "grad_norm": 1.09756338596344,
+      "learning_rate": 1.371748699479792e-05,
+      "loss": 0.4469,
+      "step": 18149
+    },
+    {
+      "epoch": 23.232,
+      "grad_norm": 1.1287214756011963,
+      "learning_rate": 1.3715486194477792e-05,
+      "loss": 0.4648,
+      "step": 18150
+    },
+    {
+      "epoch": 23.23328,
+      "grad_norm": 1.138611078262329,
+      "learning_rate": 1.3713485394157664e-05,
+      "loss": 0.4304,
+      "step": 18151
+    },
+    {
+      "epoch": 23.23456,
+      "grad_norm": 1.092214822769165,
+      "learning_rate": 1.3711484593837537e-05,
+      "loss": 0.4395,
+      "step": 18152
+    },
+    {
+      "epoch": 23.23584,
+      "grad_norm": 1.1681923866271973,
+      "learning_rate": 1.3709483793517407e-05,
+      "loss": 0.5203,
+      "step": 18153
+    },
+    {
+      "epoch": 23.23712,
+      "grad_norm": 1.1423531770706177,
+      "learning_rate": 1.3707482993197279e-05,
+      "loss": 0.455,
+      "step": 18154
+    },
+    {
+      "epoch": 23.2384,
+      "grad_norm": 1.1326360702514648,
+      "learning_rate": 1.3705482192877151e-05,
+      "loss": 0.4795,
+      "step": 18155
+    },
+    {
+      "epoch": 23.23968,
+      "grad_norm": 1.1082360744476318,
+      "learning_rate": 1.3703481392557025e-05,
+      "loss": 0.4743,
+      "step": 18156
+    },
+    {
+      "epoch": 23.24096,
+      "grad_norm": 1.1555867195129395,
+      "learning_rate": 1.3701480592236895e-05,
+      "loss": 0.4985,
+      "step": 18157
+    },
+    {
+      "epoch": 23.24224,
+      "grad_norm": 1.1038802862167358,
+      "learning_rate": 1.3699479791916767e-05,
+      "loss": 0.4486,
+      "step": 18158
+    },
+    {
+      "epoch": 23.24352,
+      "grad_norm": 1.1623843908309937,
+      "learning_rate": 1.369747899159664e-05,
+      "loss": 0.4817,
+      "step": 18159
+    },
+    {
+      "epoch": 23.2448,
+      "grad_norm": 1.2444136142730713,
+      "learning_rate": 1.3695478191276512e-05,
+      "loss": 0.5373,
+      "step": 18160
+    },
+    {
+      "epoch": 23.24608,
+      "grad_norm": 1.107471227645874,
+      "learning_rate": 1.3693477390956382e-05,
+      "loss": 0.463,
+      "step": 18161
+    },
+    {
+      "epoch": 23.24736,
+      "grad_norm": 1.1224188804626465,
+      "learning_rate": 1.3691476590636254e-05,
+      "loss": 0.4517,
+      "step": 18162
+    },
+    {
+      "epoch": 23.24864,
+      "grad_norm": 1.119936466217041,
+      "learning_rate": 1.3689475790316128e-05,
+      "loss": 0.4567,
+      "step": 18163
+    },
+    {
+      "epoch": 23.24992,
+      "grad_norm": 1.1780643463134766,
+      "learning_rate": 1.3687474989996e-05,
+      "loss": 0.4465,
+      "step": 18164
+    },
+    {
+      "epoch": 23.2512,
+      "grad_norm": 1.0980896949768066,
+      "learning_rate": 1.368547418967587e-05,
+      "loss": 0.4619,
+      "step": 18165
+    },
+    {
+      "epoch": 23.25248,
+      "grad_norm": 1.1995065212249756,
+      "learning_rate": 1.3683473389355745e-05,
+      "loss": 0.5128,
+      "step": 18166
+    },
+    {
+      "epoch": 23.25376,
+      "grad_norm": 1.0773276090621948,
+      "learning_rate": 1.3681472589035615e-05,
+      "loss": 0.4548,
+      "step": 18167
+    },
+    {
+      "epoch": 23.25504,
+      "grad_norm": 1.1148123741149902,
+      "learning_rate": 1.3679471788715487e-05,
+      "loss": 0.4754,
+      "step": 18168
+    },
+    {
+      "epoch": 23.25632,
+      "grad_norm": 1.1616092920303345,
+      "learning_rate": 1.3677470988395357e-05,
+      "loss": 0.4963,
+      "step": 18169
+    },
+    {
+      "epoch": 23.2576,
+      "grad_norm": 1.1255162954330444,
+      "learning_rate": 1.3675470188075232e-05,
+      "loss": 0.4957,
+      "step": 18170
+    },
+    {
+      "epoch": 23.25888,
+      "grad_norm": 1.1024315357208252,
+      "learning_rate": 1.3673469387755102e-05,
+      "loss": 0.4764,
+      "step": 18171
+    },
+    {
+      "epoch": 23.26016,
+      "grad_norm": 1.1717227697372437,
+      "learning_rate": 1.3671468587434974e-05,
+      "loss": 0.4885,
+      "step": 18172
+    },
+    {
+      "epoch": 23.26144,
+      "grad_norm": 1.1702935695648193,
+      "learning_rate": 1.3669467787114844e-05,
+      "loss": 0.4821,
+      "step": 18173
+    },
+    {
+      "epoch": 23.26272,
+      "grad_norm": 1.051859974861145,
+      "learning_rate": 1.366746698679472e-05,
+      "loss": 0.4485,
+      "step": 18174
+    },
+    {
+      "epoch": 23.264,
+      "grad_norm": 1.042208194732666,
+      "learning_rate": 1.366546618647459e-05,
+      "loss": 0.4337,
+      "step": 18175
+    },
+    {
+      "epoch": 23.26528,
+      "grad_norm": 1.1617704629898071,
+      "learning_rate": 1.3663465386154462e-05,
+      "loss": 0.5218,
+      "step": 18176
+    },
+    {
+      "epoch": 23.26656,
+      "grad_norm": 1.151158332824707,
+      "learning_rate": 1.3661464585834335e-05,
+      "loss": 0.4289,
+      "step": 18177
+    },
+    {
+      "epoch": 23.26784,
+      "grad_norm": 1.1818608045578003,
+      "learning_rate": 1.3659463785514207e-05,
+      "loss": 0.4554,
+      "step": 18178
+    },
+    {
+      "epoch": 23.26912,
+      "grad_norm": 1.1779991388320923,
+      "learning_rate": 1.3657462985194077e-05,
+      "loss": 0.5049,
+      "step": 18179
+    },
+    {
+      "epoch": 23.2704,
+      "grad_norm": 1.209188461303711,
+      "learning_rate": 1.3655462184873949e-05,
+      "loss": 0.539,
+      "step": 18180
+    },
+    {
+      "epoch": 23.27168,
+      "grad_norm": 1.1615291833877563,
+      "learning_rate": 1.3653461384553823e-05,
+      "loss": 0.5058,
+      "step": 18181
+    },
+    {
+      "epoch": 23.27296,
+      "grad_norm": 1.1432435512542725,
+      "learning_rate": 1.3651460584233695e-05,
+      "loss": 0.4513,
+      "step": 18182
+    },
+    {
+      "epoch": 23.27424,
+      "grad_norm": 1.2107926607131958,
+      "learning_rate": 1.3649459783913565e-05,
+      "loss": 0.5399,
+      "step": 18183
+    },
+    {
+      "epoch": 23.27552,
+      "grad_norm": 1.2299079895019531,
+      "learning_rate": 1.364745898359344e-05,
+      "loss": 0.5395,
+      "step": 18184
+    },
+    {
+      "epoch": 23.2768,
+      "grad_norm": 1.1999679803848267,
+      "learning_rate": 1.364545818327331e-05,
+      "loss": 0.4976,
+      "step": 18185
+    },
+    {
+      "epoch": 23.27808,
+      "grad_norm": 1.1381257772445679,
+      "learning_rate": 1.3643457382953182e-05,
+      "loss": 0.507,
+      "step": 18186
+    },
+    {
+      "epoch": 23.27936,
+      "grad_norm": 1.1756349802017212,
+      "learning_rate": 1.3641456582633052e-05,
+      "loss": 0.5124,
+      "step": 18187
+    },
+    {
+      "epoch": 23.28064,
+      "grad_norm": 1.1597191095352173,
+      "learning_rate": 1.3639455782312927e-05,
+      "loss": 0.4758,
+      "step": 18188
+    },
+    {
+      "epoch": 23.28192,
+      "grad_norm": 1.1272660493850708,
+      "learning_rate": 1.3637454981992798e-05,
+      "loss": 0.4734,
+      "step": 18189
+    },
+    {
+      "epoch": 23.2832,
+      "grad_norm": 1.1996667385101318,
+      "learning_rate": 1.363545418167267e-05,
+      "loss": 0.4675,
+      "step": 18190
+    },
+    {
+      "epoch": 23.28448,
+      "grad_norm": 1.1502827405929565,
+      "learning_rate": 1.3633453381352543e-05,
+      "loss": 0.5163,
+      "step": 18191
+    },
+    {
+      "epoch": 23.28576,
+      "grad_norm": 1.1988962888717651,
+      "learning_rate": 1.3631452581032415e-05,
+      "loss": 0.4729,
+      "step": 18192
+    },
+    {
+      "epoch": 23.28704,
+      "grad_norm": 1.0956463813781738,
+      "learning_rate": 1.3629451780712285e-05,
+      "loss": 0.4646,
+      "step": 18193
+    },
+    {
+      "epoch": 23.28832,
+      "grad_norm": 1.0955307483673096,
+      "learning_rate": 1.3627450980392157e-05,
+      "loss": 0.4796,
+      "step": 18194
+    },
+    {
+      "epoch": 23.2896,
+      "grad_norm": 1.2165777683258057,
+      "learning_rate": 1.362545018007203e-05,
+      "loss": 0.5016,
+      "step": 18195
+    },
+    {
+      "epoch": 23.29088,
+      "grad_norm": 1.2035049200057983,
+      "learning_rate": 1.3623449379751902e-05,
+      "loss": 0.4793,
+      "step": 18196
+    },
+    {
+      "epoch": 23.29216,
+      "grad_norm": 1.1386168003082275,
+      "learning_rate": 1.3621448579431772e-05,
+      "loss": 0.4938,
+      "step": 18197
+    },
+    {
+      "epoch": 23.29344,
+      "grad_norm": 1.1671661138534546,
+      "learning_rate": 1.3619447779111646e-05,
+      "loss": 0.523,
+      "step": 18198
+    },
+    {
+      "epoch": 23.29472,
+      "grad_norm": 1.096612811088562,
+      "learning_rate": 1.3617446978791518e-05,
+      "loss": 0.4371,
+      "step": 18199
+    },
+    {
+      "epoch": 23.296,
+      "grad_norm": 1.1467458009719849,
+      "learning_rate": 1.361544617847139e-05,
+      "loss": 0.4933,
+      "step": 18200
+    },
+    {
+      "epoch": 23.29728,
+      "grad_norm": 1.131923794746399,
+      "learning_rate": 1.361344537815126e-05,
+      "loss": 0.4217,
+      "step": 18201
+    },
+    {
+      "epoch": 23.29856,
+      "grad_norm": 1.149901032447815,
+      "learning_rate": 1.3611444577831133e-05,
+      "loss": 0.4558,
+      "step": 18202
+    },
+    {
+      "epoch": 23.29984,
+      "grad_norm": 1.162813663482666,
+      "learning_rate": 1.3609443777511005e-05,
+      "loss": 0.4712,
+      "step": 18203
+    },
+    {
+      "epoch": 23.30112,
+      "grad_norm": 1.11738920211792,
+      "learning_rate": 1.3607442977190877e-05,
+      "loss": 0.4764,
+      "step": 18204
+    },
+    {
+      "epoch": 23.3024,
+      "grad_norm": 1.2097214460372925,
+      "learning_rate": 1.360544217687075e-05,
+      "loss": 0.4811,
+      "step": 18205
+    },
+    {
+      "epoch": 23.30368,
+      "grad_norm": 1.095917820930481,
+      "learning_rate": 1.360344137655062e-05,
+      "loss": 0.4281,
+      "step": 18206
+    },
+    {
+      "epoch": 23.30496,
+      "grad_norm": 1.1422666311264038,
+      "learning_rate": 1.3601440576230493e-05,
+      "loss": 0.4902,
+      "step": 18207
+    },
+    {
+      "epoch": 23.30624,
+      "grad_norm": 1.118310809135437,
+      "learning_rate": 1.3599439775910364e-05,
+      "loss": 0.4364,
+      "step": 18208
+    },
+    {
+      "epoch": 23.30752,
+      "grad_norm": 1.0865938663482666,
+      "learning_rate": 1.3597438975590238e-05,
+      "loss": 0.4709,
+      "step": 18209
+    },
+    {
+      "epoch": 23.3088,
+      "grad_norm": 1.0885086059570312,
+      "learning_rate": 1.3595438175270108e-05,
+      "loss": 0.489,
+      "step": 18210
+    },
+    {
+      "epoch": 23.31008,
+      "grad_norm": 1.1577430963516235,
+      "learning_rate": 1.359343737494998e-05,
+      "loss": 0.5023,
+      "step": 18211
+    },
+    {
+      "epoch": 23.31136,
+      "grad_norm": 1.162643551826477,
+      "learning_rate": 1.3591436574629854e-05,
+      "loss": 0.4846,
+      "step": 18212
+    },
+    {
+      "epoch": 23.31264,
+      "grad_norm": 1.1046801805496216,
+      "learning_rate": 1.3589435774309725e-05,
+      "loss": 0.4425,
+      "step": 18213
+    },
+    {
+      "epoch": 23.31392,
+      "grad_norm": 1.1700000762939453,
+      "learning_rate": 1.3587434973989596e-05,
+      "loss": 0.4799,
+      "step": 18214
+    },
+    {
+      "epoch": 23.3152,
+      "grad_norm": 1.0989649295806885,
+      "learning_rate": 1.3585434173669467e-05,
+      "loss": 0.4446,
+      "step": 18215
+    },
+    {
+      "epoch": 23.31648,
+      "grad_norm": 1.168190598487854,
+      "learning_rate": 1.3583433373349341e-05,
+      "loss": 0.5183,
+      "step": 18216
+    },
+    {
+      "epoch": 23.31776,
+      "grad_norm": 1.081773281097412,
+      "learning_rate": 1.3581432573029213e-05,
+      "loss": 0.4795,
+      "step": 18217
+    },
+    {
+      "epoch": 23.31904,
+      "grad_norm": 1.1346309185028076,
+      "learning_rate": 1.3579431772709083e-05,
+      "loss": 0.4431,
+      "step": 18218
+    },
+    {
+      "epoch": 23.32032,
+      "grad_norm": 1.150037169456482,
+      "learning_rate": 1.3577430972388958e-05,
+      "loss": 0.4687,
+      "step": 18219
+    },
+    {
+      "epoch": 23.3216,
+      "grad_norm": 1.1243027448654175,
+      "learning_rate": 1.3575430172068828e-05,
+      "loss": 0.4564,
+      "step": 18220
+    },
+    {
+      "epoch": 23.32288,
+      "grad_norm": 1.1913601160049438,
+      "learning_rate": 1.35734293717487e-05,
+      "loss": 0.5253,
+      "step": 18221
+    },
+    {
+      "epoch": 23.32416,
+      "grad_norm": 1.142009973526001,
+      "learning_rate": 1.357142857142857e-05,
+      "loss": 0.4782,
+      "step": 18222
+    },
+    {
+      "epoch": 23.32544,
+      "grad_norm": 1.1879467964172363,
+      "learning_rate": 1.3569427771108446e-05,
+      "loss": 0.4572,
+      "step": 18223
+    },
+    {
+      "epoch": 23.32672,
+      "grad_norm": 1.1096800565719604,
+      "learning_rate": 1.3567426970788316e-05,
+      "loss": 0.4476,
+      "step": 18224
+    },
+    {
+      "epoch": 23.328,
+      "grad_norm": 1.2172260284423828,
+      "learning_rate": 1.3565426170468188e-05,
+      "loss": 0.4931,
+      "step": 18225
+    },
+    {
+      "epoch": 23.32928,
+      "grad_norm": 1.1828625202178955,
+      "learning_rate": 1.3563425370148061e-05,
+      "loss": 0.4673,
+      "step": 18226
+    },
+    {
+      "epoch": 23.33056,
+      "grad_norm": 1.162265419960022,
+      "learning_rate": 1.3561424569827933e-05,
+      "loss": 0.4658,
+      "step": 18227
+    },
+    {
+      "epoch": 23.33184,
+      "grad_norm": 1.1894097328186035,
+      "learning_rate": 1.3559423769507803e-05,
+      "loss": 0.5361,
+      "step": 18228
+    },
+    {
+      "epoch": 23.33312,
+      "grad_norm": 1.0926059484481812,
+      "learning_rate": 1.3557422969187675e-05,
+      "loss": 0.4379,
+      "step": 18229
+    },
+    {
+      "epoch": 23.3344,
+      "grad_norm": 1.1234076023101807,
+      "learning_rate": 1.3555422168867549e-05,
+      "loss": 0.4882,
+      "step": 18230
+    },
+    {
+      "epoch": 23.33568,
+      "grad_norm": 1.1318891048431396,
+      "learning_rate": 1.355342136854742e-05,
+      "loss": 0.4784,
+      "step": 18231
+    },
+    {
+      "epoch": 23.33696,
+      "grad_norm": 1.2062749862670898,
+      "learning_rate": 1.355142056822729e-05,
+      "loss": 0.5137,
+      "step": 18232
+    },
+    {
+      "epoch": 23.33824,
+      "grad_norm": 1.1662986278533936,
+      "learning_rate": 1.3549419767907163e-05,
+      "loss": 0.4596,
+      "step": 18233
+    },
+    {
+      "epoch": 23.33952,
+      "grad_norm": 1.1790847778320312,
+      "learning_rate": 1.3547418967587036e-05,
+      "loss": 0.482,
+      "step": 18234
+    },
+    {
+      "epoch": 23.3408,
+      "grad_norm": 1.079903244972229,
+      "learning_rate": 1.3545418167266908e-05,
+      "loss": 0.4931,
+      "step": 18235
+    },
+    {
+      "epoch": 23.34208,
+      "grad_norm": 1.0966075658798218,
+      "learning_rate": 1.3543417366946778e-05,
+      "loss": 0.4332,
+      "step": 18236
+    },
+    {
+      "epoch": 23.34336,
+      "grad_norm": 1.104612946510315,
+      "learning_rate": 1.3541416566626652e-05,
+      "loss": 0.4597,
+      "step": 18237
+    },
+    {
+      "epoch": 23.34464,
+      "grad_norm": 1.1291465759277344,
+      "learning_rate": 1.3539415766306524e-05,
+      "loss": 0.4864,
+      "step": 18238
+    },
+    {
+      "epoch": 23.34592,
+      "grad_norm": 1.205235242843628,
+      "learning_rate": 1.3537414965986395e-05,
+      "loss": 0.5488,
+      "step": 18239
+    },
+    {
+      "epoch": 23.3472,
+      "grad_norm": 1.1906328201293945,
+      "learning_rate": 1.3535414165666266e-05,
+      "loss": 0.4874,
+      "step": 18240
+    },
+    {
+      "epoch": 23.34848,
+      "grad_norm": 1.1368838548660278,
+      "learning_rate": 1.3533413365346139e-05,
+      "loss": 0.4336,
+      "step": 18241
+    },
+    {
+      "epoch": 23.34976,
+      "grad_norm": 1.2331318855285645,
+      "learning_rate": 1.3531412565026011e-05,
+      "loss": 0.5252,
+      "step": 18242
+    },
+    {
+      "epoch": 23.35104,
+      "grad_norm": 1.147335410118103,
+      "learning_rate": 1.3529411764705883e-05,
+      "loss": 0.4733,
+      "step": 18243
+    },
+    {
+      "epoch": 23.35232,
+      "grad_norm": 1.1000919342041016,
+      "learning_rate": 1.3527410964385756e-05,
+      "loss": 0.4743,
+      "step": 18244
+    },
+    {
+      "epoch": 23.3536,
+      "grad_norm": 1.0781689882278442,
+      "learning_rate": 1.3525410164065627e-05,
+      "loss": 0.4291,
+      "step": 18245
+    },
+    {
+      "epoch": 23.35488,
+      "grad_norm": 1.218489170074463,
+      "learning_rate": 1.3523409363745498e-05,
+      "loss": 0.4633,
+      "step": 18246
+    },
+    {
+      "epoch": 23.35616,
+      "grad_norm": 1.2166152000427246,
+      "learning_rate": 1.352140856342537e-05,
+      "loss": 0.5128,
+      "step": 18247
+    },
+    {
+      "epoch": 23.35744,
+      "grad_norm": 1.1179983615875244,
+      "learning_rate": 1.3519407763105244e-05,
+      "loss": 0.4939,
+      "step": 18248
+    },
+    {
+      "epoch": 23.35872,
+      "grad_norm": 1.1847084760665894,
+      "learning_rate": 1.3517406962785114e-05,
+      "loss": 0.5327,
+      "step": 18249
+    },
+    {
+      "epoch": 23.36,
+      "grad_norm": 1.1466864347457886,
+      "learning_rate": 1.3515406162464986e-05,
+      "loss": 0.4945,
+      "step": 18250
+    },
+    {
+      "epoch": 23.36128,
+      "grad_norm": 1.1324851512908936,
+      "learning_rate": 1.351340536214486e-05,
+      "loss": 0.4895,
+      "step": 18251
+    },
+    {
+      "epoch": 23.36256,
+      "grad_norm": 1.1512887477874756,
+      "learning_rate": 1.3511404561824731e-05,
+      "loss": 0.4707,
+      "step": 18252
+    },
+    {
+      "epoch": 23.36384,
+      "grad_norm": 1.1378182172775269,
+      "learning_rate": 1.3509403761504601e-05,
+      "loss": 0.5104,
+      "step": 18253
+    },
+    {
+      "epoch": 23.36512,
+      "grad_norm": 1.1355574131011963,
+      "learning_rate": 1.3507402961184473e-05,
+      "loss": 0.4774,
+      "step": 18254
+    },
+    {
+      "epoch": 23.3664,
+      "grad_norm": 1.095694899559021,
+      "learning_rate": 1.3505402160864347e-05,
+      "loss": 0.4631,
+      "step": 18255
+    },
+    {
+      "epoch": 23.36768,
+      "grad_norm": 1.0908712148666382,
+      "learning_rate": 1.3503401360544219e-05,
+      "loss": 0.5036,
+      "step": 18256
+    },
+    {
+      "epoch": 23.36896,
+      "grad_norm": 1.1985548734664917,
+      "learning_rate": 1.3501400560224089e-05,
+      "loss": 0.5001,
+      "step": 18257
+    },
+    {
+      "epoch": 23.37024,
+      "grad_norm": 1.1489444971084595,
+      "learning_rate": 1.3499399759903964e-05,
+      "loss": 0.4526,
+      "step": 18258
+    },
+    {
+      "epoch": 23.37152,
+      "grad_norm": 1.183395266532898,
+      "learning_rate": 1.3497398959583834e-05,
+      "loss": 0.5232,
+      "step": 18259
+    },
+    {
+      "epoch": 23.3728,
+      "grad_norm": 1.1368277072906494,
+      "learning_rate": 1.3495398159263706e-05,
+      "loss": 0.4497,
+      "step": 18260
+    },
+    {
+      "epoch": 23.37408,
+      "grad_norm": 1.1291542053222656,
+      "learning_rate": 1.3493397358943576e-05,
+      "loss": 0.4875,
+      "step": 18261
+    },
+    {
+      "epoch": 23.37536,
+      "grad_norm": 1.132755160331726,
+      "learning_rate": 1.3491396558623452e-05,
+      "loss": 0.4763,
+      "step": 18262
+    },
+    {
+      "epoch": 23.37664,
+      "grad_norm": 1.107372522354126,
+      "learning_rate": 1.3489395758303322e-05,
+      "loss": 0.4821,
+      "step": 18263
+    },
+    {
+      "epoch": 23.37792,
+      "grad_norm": 1.1369836330413818,
+      "learning_rate": 1.3487394957983194e-05,
+      "loss": 0.4666,
+      "step": 18264
+    },
+    {
+      "epoch": 23.3792,
+      "grad_norm": 1.1723092794418335,
+      "learning_rate": 1.3485394157663067e-05,
+      "loss": 0.5537,
+      "step": 18265
+    },
+    {
+      "epoch": 23.38048,
+      "grad_norm": 1.1863315105438232,
+      "learning_rate": 1.3483393357342939e-05,
+      "loss": 0.5479,
+      "step": 18266
+    },
+    {
+      "epoch": 23.38176,
+      "grad_norm": 1.1658673286437988,
+      "learning_rate": 1.3481392557022809e-05,
+      "loss": 0.5388,
+      "step": 18267
+    },
+    {
+      "epoch": 23.38304,
+      "grad_norm": 1.1437945365905762,
+      "learning_rate": 1.3479391756702681e-05,
+      "loss": 0.4794,
+      "step": 18268
+    },
+    {
+      "epoch": 23.38432,
+      "grad_norm": 1.149343729019165,
+      "learning_rate": 1.3477390956382555e-05,
+      "loss": 0.4792,
+      "step": 18269
+    },
+    {
+      "epoch": 23.3856,
+      "grad_norm": 1.0931353569030762,
+      "learning_rate": 1.3475390156062426e-05,
+      "loss": 0.4807,
+      "step": 18270
+    },
+    {
+      "epoch": 23.38688,
+      "grad_norm": 1.122345209121704,
+      "learning_rate": 1.3473389355742297e-05,
+      "loss": 0.4718,
+      "step": 18271
+    },
+    {
+      "epoch": 23.38816,
+      "grad_norm": 1.0672874450683594,
+      "learning_rate": 1.347138855542217e-05,
+      "loss": 0.4547,
+      "step": 18272
+    },
+    {
+      "epoch": 23.38944,
+      "grad_norm": 1.1067068576812744,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 0.4553,
+      "step": 18273
+    },
+    {
+      "epoch": 23.39072,
+      "grad_norm": 1.193006992340088,
+      "learning_rate": 1.3467386954781914e-05,
+      "loss": 0.5064,
+      "step": 18274
+    },
+    {
+      "epoch": 23.392,
+      "grad_norm": 1.1643834114074707,
+      "learning_rate": 1.3465386154461784e-05,
+      "loss": 0.4489,
+      "step": 18275
+    },
+    {
+      "epoch": 23.39328,
+      "grad_norm": 1.0767834186553955,
+      "learning_rate": 1.3463385354141658e-05,
+      "loss": 0.4177,
+      "step": 18276
+    },
+    {
+      "epoch": 23.39456,
+      "grad_norm": 1.1161274909973145,
+      "learning_rate": 1.346138455382153e-05,
+      "loss": 0.4918,
+      "step": 18277
+    },
+    {
+      "epoch": 23.39584,
+      "grad_norm": 1.2007757425308228,
+      "learning_rate": 1.3459383753501401e-05,
+      "loss": 0.5226,
+      "step": 18278
+    },
+    {
+      "epoch": 23.39712,
+      "grad_norm": 1.1643115282058716,
+      "learning_rate": 1.3457382953181275e-05,
+      "loss": 0.4366,
+      "step": 18279
+    },
+    {
+      "epoch": 23.3984,
+      "grad_norm": 1.183032512664795,
+      "learning_rate": 1.3455382152861145e-05,
+      "loss": 0.4606,
+      "step": 18280
+    },
+    {
+      "epoch": 23.39968,
+      "grad_norm": 1.0798239707946777,
+      "learning_rate": 1.3453381352541017e-05,
+      "loss": 0.4546,
+      "step": 18281
+    },
+    {
+      "epoch": 23.40096,
+      "grad_norm": 1.117011547088623,
+      "learning_rate": 1.3451380552220889e-05,
+      "loss": 0.4951,
+      "step": 18282
+    },
+    {
+      "epoch": 23.40224,
+      "grad_norm": 1.2203407287597656,
+      "learning_rate": 1.3449379751900762e-05,
+      "loss": 0.4415,
+      "step": 18283
+    },
+    {
+      "epoch": 23.40352,
+      "grad_norm": 1.1237221956253052,
+      "learning_rate": 1.3447378951580632e-05,
+      "loss": 0.4756,
+      "step": 18284
+    },
+    {
+      "epoch": 23.4048,
+      "grad_norm": 1.107589840888977,
+      "learning_rate": 1.3445378151260504e-05,
+      "loss": 0.4927,
+      "step": 18285
+    },
+    {
+      "epoch": 23.40608,
+      "grad_norm": 1.1575672626495361,
+      "learning_rate": 1.3443377350940376e-05,
+      "loss": 0.4813,
+      "step": 18286
+    },
+    {
+      "epoch": 23.40736,
+      "grad_norm": 1.074341893196106,
+      "learning_rate": 1.344137655062025e-05,
+      "loss": 0.4872,
+      "step": 18287
+    },
+    {
+      "epoch": 23.40864,
+      "grad_norm": 1.0716972351074219,
+      "learning_rate": 1.343937575030012e-05,
+      "loss": 0.4199,
+      "step": 18288
+    },
+    {
+      "epoch": 23.40992,
+      "grad_norm": 1.1021279096603394,
+      "learning_rate": 1.3437374949979992e-05,
+      "loss": 0.5047,
+      "step": 18289
+    },
+    {
+      "epoch": 23.4112,
+      "grad_norm": 1.1311216354370117,
+      "learning_rate": 1.3435374149659865e-05,
+      "loss": 0.5052,
+      "step": 18290
+    },
+    {
+      "epoch": 23.41248,
+      "grad_norm": 1.1251327991485596,
+      "learning_rate": 1.3433373349339737e-05,
+      "loss": 0.4459,
+      "step": 18291
+    },
+    {
+      "epoch": 23.41376,
+      "grad_norm": 1.1681146621704102,
+      "learning_rate": 1.3431372549019607e-05,
+      "loss": 0.4603,
+      "step": 18292
+    },
+    {
+      "epoch": 23.41504,
+      "grad_norm": 1.1599884033203125,
+      "learning_rate": 1.3429371748699479e-05,
+      "loss": 0.4869,
+      "step": 18293
+    },
+    {
+      "epoch": 23.41632,
+      "grad_norm": 1.170707106590271,
+      "learning_rate": 1.3427370948379353e-05,
+      "loss": 0.4833,
+      "step": 18294
+    },
+    {
+      "epoch": 23.4176,
+      "grad_norm": 1.1624923944473267,
+      "learning_rate": 1.3425370148059224e-05,
+      "loss": 0.472,
+      "step": 18295
+    },
+    {
+      "epoch": 23.41888,
+      "grad_norm": 1.1000187397003174,
+      "learning_rate": 1.3423369347739095e-05,
+      "loss": 0.4675,
+      "step": 18296
+    },
+    {
+      "epoch": 23.42016,
+      "grad_norm": 1.2124167680740356,
+      "learning_rate": 1.342136854741897e-05,
+      "loss": 0.5053,
+      "step": 18297
+    },
+    {
+      "epoch": 23.42144,
+      "grad_norm": 1.0960698127746582,
+      "learning_rate": 1.341936774709884e-05,
+      "loss": 0.4545,
+      "step": 18298
+    },
+    {
+      "epoch": 23.422719999999998,
+      "grad_norm": 1.1535124778747559,
+      "learning_rate": 1.3417366946778712e-05,
+      "loss": 0.4919,
+      "step": 18299
+    },
+    {
+      "epoch": 23.424,
+      "grad_norm": 1.0561548471450806,
+      "learning_rate": 1.3415366146458582e-05,
+      "loss": 0.4315,
+      "step": 18300
+    },
+    {
+      "epoch": 23.42528,
+      "grad_norm": 1.136558175086975,
+      "learning_rate": 1.3413365346138457e-05,
+      "loss": 0.4742,
+      "step": 18301
+    },
+    {
+      "epoch": 23.42656,
+      "grad_norm": 1.1573458909988403,
+      "learning_rate": 1.3411364545818327e-05,
+      "loss": 0.4753,
+      "step": 18302
+    },
+    {
+      "epoch": 23.42784,
+      "grad_norm": 1.1722750663757324,
+      "learning_rate": 1.34093637454982e-05,
+      "loss": 0.4666,
+      "step": 18303
+    },
+    {
+      "epoch": 23.42912,
+      "grad_norm": 1.1860383749008179,
+      "learning_rate": 1.3407362945178073e-05,
+      "loss": 0.4841,
+      "step": 18304
+    },
+    {
+      "epoch": 23.4304,
+      "grad_norm": 1.1146502494812012,
+      "learning_rate": 1.3405362144857945e-05,
+      "loss": 0.4435,
+      "step": 18305
+    },
+    {
+      "epoch": 23.43168,
+      "grad_norm": 1.1159223318099976,
+      "learning_rate": 1.3403361344537815e-05,
+      "loss": 0.4846,
+      "step": 18306
+    },
+    {
+      "epoch": 23.43296,
+      "grad_norm": 1.122043490409851,
+      "learning_rate": 1.3401360544217687e-05,
+      "loss": 0.4639,
+      "step": 18307
+    },
+    {
+      "epoch": 23.43424,
+      "grad_norm": 1.105041742324829,
+      "learning_rate": 1.339935974389756e-05,
+      "loss": 0.4886,
+      "step": 18308
+    },
+    {
+      "epoch": 23.43552,
+      "grad_norm": 1.1351252794265747,
+      "learning_rate": 1.3397358943577432e-05,
+      "loss": 0.4366,
+      "step": 18309
+    },
+    {
+      "epoch": 23.4368,
+      "grad_norm": 1.1929394006729126,
+      "learning_rate": 1.3395358143257302e-05,
+      "loss": 0.4917,
+      "step": 18310
+    },
+    {
+      "epoch": 23.43808,
+      "grad_norm": 1.1288291215896606,
+      "learning_rate": 1.3393357342937176e-05,
+      "loss": 0.4498,
+      "step": 18311
+    },
+    {
+      "epoch": 23.43936,
+      "grad_norm": 1.1538920402526855,
+      "learning_rate": 1.3391356542617048e-05,
+      "loss": 0.4655,
+      "step": 18312
+    },
+    {
+      "epoch": 23.44064,
+      "grad_norm": 1.0849840641021729,
+      "learning_rate": 1.338935574229692e-05,
+      "loss": 0.472,
+      "step": 18313
+    },
+    {
+      "epoch": 23.44192,
+      "grad_norm": 1.0807019472122192,
+      "learning_rate": 1.338735494197679e-05,
+      "loss": 0.4488,
+      "step": 18314
+    },
+    {
+      "epoch": 23.4432,
+      "grad_norm": 1.1522600650787354,
+      "learning_rate": 1.3385354141656663e-05,
+      "loss": 0.5164,
+      "step": 18315
+    },
+    {
+      "epoch": 23.44448,
+      "grad_norm": 1.2187221050262451,
+      "learning_rate": 1.3383353341336535e-05,
+      "loss": 0.5471,
+      "step": 18316
+    },
+    {
+      "epoch": 23.44576,
+      "grad_norm": 1.1273307800292969,
+      "learning_rate": 1.3381352541016407e-05,
+      "loss": 0.4396,
+      "step": 18317
+    },
+    {
+      "epoch": 23.44704,
+      "grad_norm": 1.0828683376312256,
+      "learning_rate": 1.337935174069628e-05,
+      "loss": 0.4686,
+      "step": 18318
+    },
+    {
+      "epoch": 23.44832,
+      "grad_norm": 1.0796676874160767,
+      "learning_rate": 1.337735094037615e-05,
+      "loss": 0.4329,
+      "step": 18319
+    },
+    {
+      "epoch": 23.4496,
+      "grad_norm": 1.2199808359146118,
+      "learning_rate": 1.3375350140056023e-05,
+      "loss": 0.5222,
+      "step": 18320
+    },
+    {
+      "epoch": 23.45088,
+      "grad_norm": 1.096962571144104,
+      "learning_rate": 1.3373349339735894e-05,
+      "loss": 0.4696,
+      "step": 18321
+    },
+    {
+      "epoch": 23.45216,
+      "grad_norm": 1.1377143859863281,
+      "learning_rate": 1.3371348539415768e-05,
+      "loss": 0.5229,
+      "step": 18322
+    },
+    {
+      "epoch": 23.45344,
+      "grad_norm": 1.1810131072998047,
+      "learning_rate": 1.3369347739095638e-05,
+      "loss": 0.476,
+      "step": 18323
+    },
+    {
+      "epoch": 23.454720000000002,
+      "grad_norm": 1.1354953050613403,
+      "learning_rate": 1.336734693877551e-05,
+      "loss": 0.4791,
+      "step": 18324
+    },
+    {
+      "epoch": 23.456,
+      "grad_norm": 1.0756906270980835,
+      "learning_rate": 1.3365346138455384e-05,
+      "loss": 0.44,
+      "step": 18325
+    },
+    {
+      "epoch": 23.45728,
+      "grad_norm": 1.1659843921661377,
+      "learning_rate": 1.3363345338135255e-05,
+      "loss": 0.4829,
+      "step": 18326
+    },
+    {
+      "epoch": 23.45856,
+      "grad_norm": 1.0880708694458008,
+      "learning_rate": 1.3361344537815126e-05,
+      "loss": 0.4278,
+      "step": 18327
+    },
+    {
+      "epoch": 23.45984,
+      "grad_norm": 1.161241888999939,
+      "learning_rate": 1.3359343737494997e-05,
+      "loss": 0.4762,
+      "step": 18328
+    },
+    {
+      "epoch": 23.46112,
+      "grad_norm": 1.1863901615142822,
+      "learning_rate": 1.3357342937174871e-05,
+      "loss": 0.4841,
+      "step": 18329
+    },
+    {
+      "epoch": 23.4624,
+      "grad_norm": 1.1466909646987915,
+      "learning_rate": 1.3355342136854743e-05,
+      "loss": 0.5037,
+      "step": 18330
+    },
+    {
+      "epoch": 23.46368,
+      "grad_norm": 1.2050119638442993,
+      "learning_rate": 1.3353341336534613e-05,
+      "loss": 0.4718,
+      "step": 18331
+    },
+    {
+      "epoch": 23.46496,
+      "grad_norm": 1.173220157623291,
+      "learning_rate": 1.3351340536214488e-05,
+      "loss": 0.4654,
+      "step": 18332
+    },
+    {
+      "epoch": 23.46624,
+      "grad_norm": 1.1541483402252197,
+      "learning_rate": 1.3349339735894358e-05,
+      "loss": 0.455,
+      "step": 18333
+    },
+    {
+      "epoch": 23.46752,
+      "grad_norm": 1.1780107021331787,
+      "learning_rate": 1.334733893557423e-05,
+      "loss": 0.5024,
+      "step": 18334
+    },
+    {
+      "epoch": 23.4688,
+      "grad_norm": 1.1591427326202393,
+      "learning_rate": 1.33453381352541e-05,
+      "loss": 0.4447,
+      "step": 18335
+    },
+    {
+      "epoch": 23.47008,
+      "grad_norm": 1.058050274848938,
+      "learning_rate": 1.3343337334933976e-05,
+      "loss": 0.4141,
+      "step": 18336
+    },
+    {
+      "epoch": 23.47136,
+      "grad_norm": 1.1635574102401733,
+      "learning_rate": 1.3341336534613846e-05,
+      "loss": 0.4718,
+      "step": 18337
+    },
+    {
+      "epoch": 23.47264,
+      "grad_norm": 1.1959178447723389,
+      "learning_rate": 1.3339335734293718e-05,
+      "loss": 0.5097,
+      "step": 18338
+    },
+    {
+      "epoch": 23.47392,
+      "grad_norm": 1.1117192506790161,
+      "learning_rate": 1.3337334933973591e-05,
+      "loss": 0.4703,
+      "step": 18339
+    },
+    {
+      "epoch": 23.4752,
+      "grad_norm": 1.1207165718078613,
+      "learning_rate": 1.3335334133653463e-05,
+      "loss": 0.4286,
+      "step": 18340
+    },
+    {
+      "epoch": 23.47648,
+      "grad_norm": 1.1271799802780151,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4686,
+      "step": 18341
+    },
+    {
+      "epoch": 23.47776,
+      "grad_norm": 1.1951183080673218,
+      "learning_rate": 1.3331332533013205e-05,
+      "loss": 0.5247,
+      "step": 18342
+    },
+    {
+      "epoch": 23.47904,
+      "grad_norm": 1.1584644317626953,
+      "learning_rate": 1.3329331732693079e-05,
+      "loss": 0.4629,
+      "step": 18343
+    },
+    {
+      "epoch": 23.48032,
+      "grad_norm": 1.1701948642730713,
+      "learning_rate": 1.332733093237295e-05,
+      "loss": 0.4776,
+      "step": 18344
+    },
+    {
+      "epoch": 23.4816,
+      "grad_norm": 1.1780883073806763,
+      "learning_rate": 1.332533013205282e-05,
+      "loss": 0.4967,
+      "step": 18345
+    },
+    {
+      "epoch": 23.48288,
+      "grad_norm": 1.11570143699646,
+      "learning_rate": 1.3323329331732693e-05,
+      "loss": 0.486,
+      "step": 18346
+    },
+    {
+      "epoch": 23.48416,
+      "grad_norm": 1.060181975364685,
+      "learning_rate": 1.3321328531412566e-05,
+      "loss": 0.4462,
+      "step": 18347
+    },
+    {
+      "epoch": 23.48544,
+      "grad_norm": 1.0786004066467285,
+      "learning_rate": 1.3319327731092438e-05,
+      "loss": 0.4475,
+      "step": 18348
+    },
+    {
+      "epoch": 23.48672,
+      "grad_norm": 1.0986952781677246,
+      "learning_rate": 1.3317326930772308e-05,
+      "loss": 0.4567,
+      "step": 18349
+    },
+    {
+      "epoch": 23.488,
+      "grad_norm": 1.1361345052719116,
+      "learning_rate": 1.3315326130452183e-05,
+      "loss": 0.4504,
+      "step": 18350
+    },
+    {
+      "epoch": 23.48928,
+      "grad_norm": 1.145846962928772,
+      "learning_rate": 1.3313325330132054e-05,
+      "loss": 0.4997,
+      "step": 18351
+    },
+    {
+      "epoch": 23.49056,
+      "grad_norm": 1.1790693998336792,
+      "learning_rate": 1.3311324529811925e-05,
+      "loss": 0.504,
+      "step": 18352
+    },
+    {
+      "epoch": 23.49184,
+      "grad_norm": 1.1805227994918823,
+      "learning_rate": 1.3309323729491796e-05,
+      "loss": 0.4856,
+      "step": 18353
+    },
+    {
+      "epoch": 23.49312,
+      "grad_norm": 1.1480375528335571,
+      "learning_rate": 1.330732292917167e-05,
+      "loss": 0.4626,
+      "step": 18354
+    },
+    {
+      "epoch": 23.4944,
+      "grad_norm": 1.1622354984283447,
+      "learning_rate": 1.3305322128851541e-05,
+      "loss": 0.5071,
+      "step": 18355
+    },
+    {
+      "epoch": 23.49568,
+      "grad_norm": 1.1230262517929077,
+      "learning_rate": 1.3303321328531413e-05,
+      "loss": 0.4853,
+      "step": 18356
+    },
+    {
+      "epoch": 23.49696,
+      "grad_norm": 1.181242823600769,
+      "learning_rate": 1.3301320528211286e-05,
+      "loss": 0.4681,
+      "step": 18357
+    },
+    {
+      "epoch": 23.49824,
+      "grad_norm": 1.1128627061843872,
+      "learning_rate": 1.3299319727891158e-05,
+      "loss": 0.5194,
+      "step": 18358
+    },
+    {
+      "epoch": 23.49952,
+      "grad_norm": 1.192732572555542,
+      "learning_rate": 1.3297318927571028e-05,
+      "loss": 0.5138,
+      "step": 18359
+    },
+    {
+      "epoch": 23.5008,
+      "grad_norm": 1.1375622749328613,
+      "learning_rate": 1.32953181272509e-05,
+      "loss": 0.5004,
+      "step": 18360
+    },
+    {
+      "epoch": 23.50208,
+      "grad_norm": 1.13736093044281,
+      "learning_rate": 1.3293317326930774e-05,
+      "loss": 0.5064,
+      "step": 18361
+    },
+    {
+      "epoch": 23.50336,
+      "grad_norm": 1.1759809255599976,
+      "learning_rate": 1.3291316526610646e-05,
+      "loss": 0.4812,
+      "step": 18362
+    },
+    {
+      "epoch": 23.50464,
+      "grad_norm": 1.115228295326233,
+      "learning_rate": 1.3289315726290516e-05,
+      "loss": 0.4502,
+      "step": 18363
+    },
+    {
+      "epoch": 23.50592,
+      "grad_norm": 1.1482048034667969,
+      "learning_rate": 1.328731492597039e-05,
+      "loss": 0.4791,
+      "step": 18364
+    },
+    {
+      "epoch": 23.5072,
+      "grad_norm": 1.1504039764404297,
+      "learning_rate": 1.3285314125650261e-05,
+      "loss": 0.4734,
+      "step": 18365
+    },
+    {
+      "epoch": 23.50848,
+      "grad_norm": 1.0897183418273926,
+      "learning_rate": 1.3283313325330133e-05,
+      "loss": 0.4615,
+      "step": 18366
+    },
+    {
+      "epoch": 23.50976,
+      "grad_norm": 1.1377613544464111,
+      "learning_rate": 1.3281312525010003e-05,
+      "loss": 0.482,
+      "step": 18367
+    },
+    {
+      "epoch": 23.51104,
+      "grad_norm": 1.1299495697021484,
+      "learning_rate": 1.3279311724689877e-05,
+      "loss": 0.4337,
+      "step": 18368
+    },
+    {
+      "epoch": 23.51232,
+      "grad_norm": 1.1434617042541504,
+      "learning_rate": 1.3277310924369749e-05,
+      "loss": 0.5078,
+      "step": 18369
+    },
+    {
+      "epoch": 23.5136,
+      "grad_norm": 1.1693207025527954,
+      "learning_rate": 1.327531012404962e-05,
+      "loss": 0.4764,
+      "step": 18370
+    },
+    {
+      "epoch": 23.51488,
+      "grad_norm": 1.1742783784866333,
+      "learning_rate": 1.3273309323729494e-05,
+      "loss": 0.4831,
+      "step": 18371
+    },
+    {
+      "epoch": 23.51616,
+      "grad_norm": 1.1278483867645264,
+      "learning_rate": 1.3271308523409364e-05,
+      "loss": 0.4247,
+      "step": 18372
+    },
+    {
+      "epoch": 23.51744,
+      "grad_norm": 1.1402909755706787,
+      "learning_rate": 1.3269307723089236e-05,
+      "loss": 0.4449,
+      "step": 18373
+    },
+    {
+      "epoch": 23.51872,
+      "grad_norm": 1.1170767545700073,
+      "learning_rate": 1.3267306922769108e-05,
+      "loss": 0.4955,
+      "step": 18374
+    },
+    {
+      "epoch": 23.52,
+      "grad_norm": 1.194703221321106,
+      "learning_rate": 1.3265306122448982e-05,
+      "loss": 0.5139,
+      "step": 18375
+    },
+    {
+      "epoch": 23.52128,
+      "grad_norm": 1.1340879201889038,
+      "learning_rate": 1.3263305322128852e-05,
+      "loss": 0.4611,
+      "step": 18376
+    },
+    {
+      "epoch": 23.52256,
+      "grad_norm": 1.1601988077163696,
+      "learning_rate": 1.3261304521808724e-05,
+      "loss": 0.4484,
+      "step": 18377
+    },
+    {
+      "epoch": 23.52384,
+      "grad_norm": 1.1433758735656738,
+      "learning_rate": 1.3259303721488597e-05,
+      "loss": 0.4465,
+      "step": 18378
+    },
+    {
+      "epoch": 23.52512,
+      "grad_norm": 1.2454715967178345,
+      "learning_rate": 1.3257302921168469e-05,
+      "loss": 0.528,
+      "step": 18379
+    },
+    {
+      "epoch": 23.5264,
+      "grad_norm": 1.1691670417785645,
+      "learning_rate": 1.3255302120848339e-05,
+      "loss": 0.4814,
+      "step": 18380
+    },
+    {
+      "epoch": 23.52768,
+      "grad_norm": 1.140670657157898,
+      "learning_rate": 1.3253301320528211e-05,
+      "loss": 0.4821,
+      "step": 18381
+    },
+    {
+      "epoch": 23.52896,
+      "grad_norm": 1.1309688091278076,
+      "learning_rate": 1.3251300520208085e-05,
+      "loss": 0.4691,
+      "step": 18382
+    },
+    {
+      "epoch": 23.53024,
+      "grad_norm": 1.1939188241958618,
+      "learning_rate": 1.3249299719887956e-05,
+      "loss": 0.5132,
+      "step": 18383
+    },
+    {
+      "epoch": 23.53152,
+      "grad_norm": 1.214198112487793,
+      "learning_rate": 1.3247298919567827e-05,
+      "loss": 0.5292,
+      "step": 18384
+    },
+    {
+      "epoch": 23.5328,
+      "grad_norm": 1.1498759984970093,
+      "learning_rate": 1.3245298119247702e-05,
+      "loss": 0.4803,
+      "step": 18385
+    },
+    {
+      "epoch": 23.53408,
+      "grad_norm": 1.1071444749832153,
+      "learning_rate": 1.3243297318927572e-05,
+      "loss": 0.466,
+      "step": 18386
+    },
+    {
+      "epoch": 23.53536,
+      "grad_norm": 1.1929941177368164,
+      "learning_rate": 1.3241296518607444e-05,
+      "loss": 0.4975,
+      "step": 18387
+    },
+    {
+      "epoch": 23.53664,
+      "grad_norm": 1.1311265230178833,
+      "learning_rate": 1.3239295718287314e-05,
+      "loss": 0.4715,
+      "step": 18388
+    },
+    {
+      "epoch": 23.53792,
+      "grad_norm": 1.0508058071136475,
+      "learning_rate": 1.323729491796719e-05,
+      "loss": 0.4312,
+      "step": 18389
+    },
+    {
+      "epoch": 23.5392,
+      "grad_norm": 1.1475504636764526,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 0.5074,
+      "step": 18390
+    },
+    {
+      "epoch": 23.54048,
+      "grad_norm": 1.1296319961547852,
+      "learning_rate": 1.3233293317326931e-05,
+      "loss": 0.5169,
+      "step": 18391
+    },
+    {
+      "epoch": 23.54176,
+      "grad_norm": 1.1883400678634644,
+      "learning_rate": 1.3231292517006805e-05,
+      "loss": 0.4944,
+      "step": 18392
+    },
+    {
+      "epoch": 23.54304,
+      "grad_norm": 1.1157552003860474,
+      "learning_rate": 1.3229291716686677e-05,
+      "loss": 0.4619,
+      "step": 18393
+    },
+    {
+      "epoch": 23.54432,
+      "grad_norm": 1.1468461751937866,
+      "learning_rate": 1.3227290916366547e-05,
+      "loss": 0.5149,
+      "step": 18394
+    },
+    {
+      "epoch": 23.5456,
+      "grad_norm": 1.0889732837677002,
+      "learning_rate": 1.3225290116046419e-05,
+      "loss": 0.4268,
+      "step": 18395
+    },
+    {
+      "epoch": 23.54688,
+      "grad_norm": 1.142313838005066,
+      "learning_rate": 1.3223289315726292e-05,
+      "loss": 0.5009,
+      "step": 18396
+    },
+    {
+      "epoch": 23.54816,
+      "grad_norm": 1.1910713911056519,
+      "learning_rate": 1.3221288515406164e-05,
+      "loss": 0.5089,
+      "step": 18397
+    },
+    {
+      "epoch": 23.54944,
+      "grad_norm": 1.1199548244476318,
+      "learning_rate": 1.3219287715086034e-05,
+      "loss": 0.4684,
+      "step": 18398
+    },
+    {
+      "epoch": 23.55072,
+      "grad_norm": 1.0918095111846924,
+      "learning_rate": 1.3217286914765906e-05,
+      "loss": 0.462,
+      "step": 18399
+    },
+    {
+      "epoch": 23.552,
+      "grad_norm": 1.1307764053344727,
+      "learning_rate": 1.321528611444578e-05,
+      "loss": 0.4839,
+      "step": 18400
+    },
+    {
+      "epoch": 23.55328,
+      "grad_norm": 1.1751610040664673,
+      "learning_rate": 1.3213285314125651e-05,
+      "loss": 0.4891,
+      "step": 18401
+    },
+    {
+      "epoch": 23.55456,
+      "grad_norm": 1.1532877683639526,
+      "learning_rate": 1.3211284513805522e-05,
+      "loss": 0.4607,
+      "step": 18402
+    },
+    {
+      "epoch": 23.55584,
+      "grad_norm": 1.1767176389694214,
+      "learning_rate": 1.3209283713485395e-05,
+      "loss": 0.5076,
+      "step": 18403
+    },
+    {
+      "epoch": 23.55712,
+      "grad_norm": 1.2673394680023193,
+      "learning_rate": 1.3207282913165267e-05,
+      "loss": 0.5038,
+      "step": 18404
+    },
+    {
+      "epoch": 23.5584,
+      "grad_norm": 1.1632843017578125,
+      "learning_rate": 1.3205282112845139e-05,
+      "loss": 0.5004,
+      "step": 18405
+    },
+    {
+      "epoch": 23.55968,
+      "grad_norm": 1.1043983697891235,
+      "learning_rate": 1.3203281312525009e-05,
+      "loss": 0.4669,
+      "step": 18406
+    },
+    {
+      "epoch": 23.56096,
+      "grad_norm": 1.1622287034988403,
+      "learning_rate": 1.3201280512204883e-05,
+      "loss": 0.4538,
+      "step": 18407
+    },
+    {
+      "epoch": 23.56224,
+      "grad_norm": 1.1268742084503174,
+      "learning_rate": 1.3199279711884754e-05,
+      "loss": 0.4563,
+      "step": 18408
+    },
+    {
+      "epoch": 23.56352,
+      "grad_norm": 1.1703455448150635,
+      "learning_rate": 1.3197278911564626e-05,
+      "loss": 0.5168,
+      "step": 18409
+    },
+    {
+      "epoch": 23.564799999999998,
+      "grad_norm": 1.1342973709106445,
+      "learning_rate": 1.31952781112445e-05,
+      "loss": 0.505,
+      "step": 18410
+    },
+    {
+      "epoch": 23.56608,
+      "grad_norm": 1.1047439575195312,
+      "learning_rate": 1.319327731092437e-05,
+      "loss": 0.4839,
+      "step": 18411
+    },
+    {
+      "epoch": 23.56736,
+      "grad_norm": 1.1439480781555176,
+      "learning_rate": 1.3191276510604242e-05,
+      "loss": 0.462,
+      "step": 18412
+    },
+    {
+      "epoch": 23.56864,
+      "grad_norm": 1.1610522270202637,
+      "learning_rate": 1.3189275710284114e-05,
+      "loss": 0.498,
+      "step": 18413
+    },
+    {
+      "epoch": 23.56992,
+      "grad_norm": 1.1288456916809082,
+      "learning_rate": 1.3187274909963987e-05,
+      "loss": 0.4751,
+      "step": 18414
+    },
+    {
+      "epoch": 23.5712,
+      "grad_norm": 1.204026460647583,
+      "learning_rate": 1.3185274109643857e-05,
+      "loss": 0.517,
+      "step": 18415
+    },
+    {
+      "epoch": 23.57248,
+      "grad_norm": 1.1533409357070923,
+      "learning_rate": 1.318327330932373e-05,
+      "loss": 0.4879,
+      "step": 18416
+    },
+    {
+      "epoch": 23.57376,
+      "grad_norm": 1.1280791759490967,
+      "learning_rate": 1.3181272509003603e-05,
+      "loss": 0.4578,
+      "step": 18417
+    },
+    {
+      "epoch": 23.57504,
+      "grad_norm": 1.1419196128845215,
+      "learning_rate": 1.3179271708683475e-05,
+      "loss": 0.4827,
+      "step": 18418
+    },
+    {
+      "epoch": 23.57632,
+      "grad_norm": 1.1568984985351562,
+      "learning_rate": 1.3177270908363345e-05,
+      "loss": 0.4784,
+      "step": 18419
+    },
+    {
+      "epoch": 23.5776,
+      "grad_norm": 1.1242620944976807,
+      "learning_rate": 1.3175270108043217e-05,
+      "loss": 0.4755,
+      "step": 18420
+    },
+    {
+      "epoch": 23.57888,
+      "grad_norm": 1.1650310754776,
+      "learning_rate": 1.317326930772309e-05,
+      "loss": 0.5058,
+      "step": 18421
+    },
+    {
+      "epoch": 23.58016,
+      "grad_norm": 1.2178586721420288,
+      "learning_rate": 1.3171268507402962e-05,
+      "loss": 0.457,
+      "step": 18422
+    },
+    {
+      "epoch": 23.58144,
+      "grad_norm": 1.105704665184021,
+      "learning_rate": 1.3169267707082832e-05,
+      "loss": 0.4333,
+      "step": 18423
+    },
+    {
+      "epoch": 23.58272,
+      "grad_norm": 1.1392953395843506,
+      "learning_rate": 1.3167266906762708e-05,
+      "loss": 0.517,
+      "step": 18424
+    },
+    {
+      "epoch": 23.584,
+      "grad_norm": 1.1252771615982056,
+      "learning_rate": 1.3165266106442578e-05,
+      "loss": 0.4966,
+      "step": 18425
+    },
+    {
+      "epoch": 23.58528,
+      "grad_norm": 1.2022629976272583,
+      "learning_rate": 1.316326530612245e-05,
+      "loss": 0.4968,
+      "step": 18426
+    },
+    {
+      "epoch": 23.58656,
+      "grad_norm": 1.1347490549087524,
+      "learning_rate": 1.316126450580232e-05,
+      "loss": 0.4637,
+      "step": 18427
+    },
+    {
+      "epoch": 23.58784,
+      "grad_norm": 1.175969123840332,
+      "learning_rate": 1.3159263705482195e-05,
+      "loss": 0.4954,
+      "step": 18428
+    },
+    {
+      "epoch": 23.58912,
+      "grad_norm": 1.1564298868179321,
+      "learning_rate": 1.3157262905162065e-05,
+      "loss": 0.5057,
+      "step": 18429
+    },
+    {
+      "epoch": 23.5904,
+      "grad_norm": 1.1712132692337036,
+      "learning_rate": 1.3155262104841937e-05,
+      "loss": 0.49,
+      "step": 18430
+    },
+    {
+      "epoch": 23.59168,
+      "grad_norm": 1.1701792478561401,
+      "learning_rate": 1.315326130452181e-05,
+      "loss": 0.485,
+      "step": 18431
+    },
+    {
+      "epoch": 23.59296,
+      "grad_norm": 1.1181670427322388,
+      "learning_rate": 1.3151260504201682e-05,
+      "loss": 0.447,
+      "step": 18432
+    },
+    {
+      "epoch": 23.59424,
+      "grad_norm": 1.1028802394866943,
+      "learning_rate": 1.3149259703881553e-05,
+      "loss": 0.4823,
+      "step": 18433
+    },
+    {
+      "epoch": 23.59552,
+      "grad_norm": 1.191518783569336,
+      "learning_rate": 1.3147258903561424e-05,
+      "loss": 0.4561,
+      "step": 18434
+    },
+    {
+      "epoch": 23.5968,
+      "grad_norm": 1.1290074586868286,
+      "learning_rate": 1.3145258103241298e-05,
+      "loss": 0.4843,
+      "step": 18435
+    },
+    {
+      "epoch": 23.59808,
+      "grad_norm": 1.1779063940048218,
+      "learning_rate": 1.314325730292117e-05,
+      "loss": 0.4708,
+      "step": 18436
+    },
+    {
+      "epoch": 23.59936,
+      "grad_norm": 1.1097599267959595,
+      "learning_rate": 1.314125650260104e-05,
+      "loss": 0.4734,
+      "step": 18437
+    },
+    {
+      "epoch": 23.60064,
+      "grad_norm": 1.1385622024536133,
+      "learning_rate": 1.3139255702280914e-05,
+      "loss": 0.4822,
+      "step": 18438
+    },
+    {
+      "epoch": 23.60192,
+      "grad_norm": 1.0946968793869019,
+      "learning_rate": 1.3137254901960785e-05,
+      "loss": 0.4713,
+      "step": 18439
+    },
+    {
+      "epoch": 23.6032,
+      "grad_norm": 1.1536378860473633,
+      "learning_rate": 1.3135254101640657e-05,
+      "loss": 0.5073,
+      "step": 18440
+    },
+    {
+      "epoch": 23.60448,
+      "grad_norm": 1.1673475503921509,
+      "learning_rate": 1.3133253301320527e-05,
+      "loss": 0.4681,
+      "step": 18441
+    },
+    {
+      "epoch": 23.60576,
+      "grad_norm": 1.163110375404358,
+      "learning_rate": 1.3131252501000401e-05,
+      "loss": 0.5031,
+      "step": 18442
+    },
+    {
+      "epoch": 23.60704,
+      "grad_norm": 1.131412386894226,
+      "learning_rate": 1.3129251700680273e-05,
+      "loss": 0.4597,
+      "step": 18443
+    },
+    {
+      "epoch": 23.60832,
+      "grad_norm": 1.1555569171905518,
+      "learning_rate": 1.3127250900360145e-05,
+      "loss": 0.5379,
+      "step": 18444
+    },
+    {
+      "epoch": 23.6096,
+      "grad_norm": 1.1479874849319458,
+      "learning_rate": 1.3125250100040018e-05,
+      "loss": 0.4431,
+      "step": 18445
+    },
+    {
+      "epoch": 23.61088,
+      "grad_norm": 1.140120029449463,
+      "learning_rate": 1.3123249299719888e-05,
+      "loss": 0.4916,
+      "step": 18446
+    },
+    {
+      "epoch": 23.61216,
+      "grad_norm": 1.1265473365783691,
+      "learning_rate": 1.312124849939976e-05,
+      "loss": 0.5152,
+      "step": 18447
+    },
+    {
+      "epoch": 23.61344,
+      "grad_norm": 1.043741226196289,
+      "learning_rate": 1.3119247699079632e-05,
+      "loss": 0.4195,
+      "step": 18448
+    },
+    {
+      "epoch": 23.61472,
+      "grad_norm": 1.1078736782073975,
+      "learning_rate": 1.3117246898759506e-05,
+      "loss": 0.4793,
+      "step": 18449
+    },
+    {
+      "epoch": 23.616,
+      "grad_norm": 1.1717230081558228,
+      "learning_rate": 1.3115246098439376e-05,
+      "loss": 0.4681,
+      "step": 18450
+    },
+    {
+      "epoch": 23.61728,
+      "grad_norm": 1.174607753753662,
+      "learning_rate": 1.3113245298119248e-05,
+      "loss": 0.4738,
+      "step": 18451
+    },
+    {
+      "epoch": 23.61856,
+      "grad_norm": 1.1562086343765259,
+      "learning_rate": 1.311124449779912e-05,
+      "loss": 0.4716,
+      "step": 18452
+    },
+    {
+      "epoch": 23.61984,
+      "grad_norm": 1.1882222890853882,
+      "learning_rate": 1.3109243697478993e-05,
+      "loss": 0.5002,
+      "step": 18453
+    },
+    {
+      "epoch": 23.62112,
+      "grad_norm": 1.1770943403244019,
+      "learning_rate": 1.3107242897158863e-05,
+      "loss": 0.4787,
+      "step": 18454
+    },
+    {
+      "epoch": 23.6224,
+      "grad_norm": 1.16677987575531,
+      "learning_rate": 1.3105242096838735e-05,
+      "loss": 0.494,
+      "step": 18455
+    },
+    {
+      "epoch": 23.62368,
+      "grad_norm": 1.190652847290039,
+      "learning_rate": 1.3103241296518609e-05,
+      "loss": 0.4809,
+      "step": 18456
+    },
+    {
+      "epoch": 23.62496,
+      "grad_norm": 1.1706172227859497,
+      "learning_rate": 1.310124049619848e-05,
+      "loss": 0.4633,
+      "step": 18457
+    },
+    {
+      "epoch": 23.62624,
+      "grad_norm": 1.0963094234466553,
+      "learning_rate": 1.309923969587835e-05,
+      "loss": 0.4892,
+      "step": 18458
+    },
+    {
+      "epoch": 23.62752,
+      "grad_norm": 1.1064554452896118,
+      "learning_rate": 1.3097238895558223e-05,
+      "loss": 0.4748,
+      "step": 18459
+    },
+    {
+      "epoch": 23.6288,
+      "grad_norm": 1.1423083543777466,
+      "learning_rate": 1.3095238095238096e-05,
+      "loss": 0.4798,
+      "step": 18460
+    },
+    {
+      "epoch": 23.63008,
+      "grad_norm": 1.1238229274749756,
+      "learning_rate": 1.3093237294917968e-05,
+      "loss": 0.4674,
+      "step": 18461
+    },
+    {
+      "epoch": 23.63136,
+      "grad_norm": 1.1107577085494995,
+      "learning_rate": 1.3091236494597838e-05,
+      "loss": 0.438,
+      "step": 18462
+    },
+    {
+      "epoch": 23.63264,
+      "grad_norm": 1.1175166368484497,
+      "learning_rate": 1.3089235694277713e-05,
+      "loss": 0.4601,
+      "step": 18463
+    },
+    {
+      "epoch": 23.63392,
+      "grad_norm": 1.1670434474945068,
+      "learning_rate": 1.3087234893957584e-05,
+      "loss": 0.4494,
+      "step": 18464
+    },
+    {
+      "epoch": 23.6352,
+      "grad_norm": 1.2020723819732666,
+      "learning_rate": 1.3085234093637455e-05,
+      "loss": 0.4967,
+      "step": 18465
+    },
+    {
+      "epoch": 23.63648,
+      "grad_norm": 1.1246027946472168,
+      "learning_rate": 1.3083233293317326e-05,
+      "loss": 0.4723,
+      "step": 18466
+    },
+    {
+      "epoch": 23.63776,
+      "grad_norm": 1.1464457511901855,
+      "learning_rate": 1.30812324929972e-05,
+      "loss": 0.5051,
+      "step": 18467
+    },
+    {
+      "epoch": 23.63904,
+      "grad_norm": 1.152032732963562,
+      "learning_rate": 1.3079231692677071e-05,
+      "loss": 0.4749,
+      "step": 18468
+    },
+    {
+      "epoch": 23.64032,
+      "grad_norm": 1.150923490524292,
+      "learning_rate": 1.3077230892356943e-05,
+      "loss": 0.4738,
+      "step": 18469
+    },
+    {
+      "epoch": 23.6416,
+      "grad_norm": 1.1413908004760742,
+      "learning_rate": 1.3075230092036816e-05,
+      "loss": 0.4809,
+      "step": 18470
+    },
+    {
+      "epoch": 23.64288,
+      "grad_norm": 1.130707025527954,
+      "learning_rate": 1.3073229291716688e-05,
+      "loss": 0.4826,
+      "step": 18471
+    },
+    {
+      "epoch": 23.64416,
+      "grad_norm": 1.115024209022522,
+      "learning_rate": 1.3071228491396558e-05,
+      "loss": 0.4714,
+      "step": 18472
+    },
+    {
+      "epoch": 23.64544,
+      "grad_norm": 1.1373069286346436,
+      "learning_rate": 1.306922769107643e-05,
+      "loss": 0.4832,
+      "step": 18473
+    },
+    {
+      "epoch": 23.64672,
+      "grad_norm": 1.1076675653457642,
+      "learning_rate": 1.3067226890756304e-05,
+      "loss": 0.466,
+      "step": 18474
+    },
+    {
+      "epoch": 23.648,
+      "grad_norm": 1.190407156944275,
+      "learning_rate": 1.3065226090436176e-05,
+      "loss": 0.5365,
+      "step": 18475
+    },
+    {
+      "epoch": 23.64928,
+      "grad_norm": 1.0926904678344727,
+      "learning_rate": 1.3063225290116046e-05,
+      "loss": 0.5145,
+      "step": 18476
+    },
+    {
+      "epoch": 23.65056,
+      "grad_norm": 1.066358208656311,
+      "learning_rate": 1.306122448979592e-05,
+      "loss": 0.4459,
+      "step": 18477
+    },
+    {
+      "epoch": 23.65184,
+      "grad_norm": 1.129427194595337,
+      "learning_rate": 1.3059223689475791e-05,
+      "loss": 0.4508,
+      "step": 18478
+    },
+    {
+      "epoch": 23.65312,
+      "grad_norm": 1.1428444385528564,
+      "learning_rate": 1.3057222889155663e-05,
+      "loss": 0.4887,
+      "step": 18479
+    },
+    {
+      "epoch": 23.6544,
+      "grad_norm": 1.1522812843322754,
+      "learning_rate": 1.3055222088835533e-05,
+      "loss": 0.4857,
+      "step": 18480
+    },
+    {
+      "epoch": 23.65568,
+      "grad_norm": 1.1594774723052979,
+      "learning_rate": 1.3053221288515407e-05,
+      "loss": 0.483,
+      "step": 18481
+    },
+    {
+      "epoch": 23.65696,
+      "grad_norm": 1.1202539205551147,
+      "learning_rate": 1.3051220488195279e-05,
+      "loss": 0.4884,
+      "step": 18482
+    },
+    {
+      "epoch": 23.65824,
+      "grad_norm": 1.184862732887268,
+      "learning_rate": 1.304921968787515e-05,
+      "loss": 0.5249,
+      "step": 18483
+    },
+    {
+      "epoch": 23.65952,
+      "grad_norm": 1.2333064079284668,
+      "learning_rate": 1.3047218887555024e-05,
+      "loss": 0.5485,
+      "step": 18484
+    },
+    {
+      "epoch": 23.660800000000002,
+      "grad_norm": 1.1727824211120605,
+      "learning_rate": 1.3045218087234894e-05,
+      "loss": 0.4687,
+      "step": 18485
+    },
+    {
+      "epoch": 23.66208,
+      "grad_norm": 1.1470680236816406,
+      "learning_rate": 1.3043217286914766e-05,
+      "loss": 0.5179,
+      "step": 18486
+    },
+    {
+      "epoch": 23.66336,
+      "grad_norm": 1.1859335899353027,
+      "learning_rate": 1.3041216486594638e-05,
+      "loss": 0.4971,
+      "step": 18487
+    },
+    {
+      "epoch": 23.66464,
+      "grad_norm": 1.2277741432189941,
+      "learning_rate": 1.3039215686274511e-05,
+      "loss": 0.4887,
+      "step": 18488
+    },
+    {
+      "epoch": 23.66592,
+      "grad_norm": 1.1842122077941895,
+      "learning_rate": 1.3037214885954382e-05,
+      "loss": 0.4908,
+      "step": 18489
+    },
+    {
+      "epoch": 23.6672,
+      "grad_norm": 1.0600404739379883,
+      "learning_rate": 1.3035214085634254e-05,
+      "loss": 0.4354,
+      "step": 18490
+    },
+    {
+      "epoch": 23.66848,
+      "grad_norm": 1.13839590549469,
+      "learning_rate": 1.3033213285314127e-05,
+      "loss": 0.4953,
+      "step": 18491
+    },
+    {
+      "epoch": 23.66976,
+      "grad_norm": 1.1359751224517822,
+      "learning_rate": 1.3031212484993999e-05,
+      "loss": 0.4667,
+      "step": 18492
+    },
+    {
+      "epoch": 23.67104,
+      "grad_norm": 1.168168067932129,
+      "learning_rate": 1.3029211684673869e-05,
+      "loss": 0.4623,
+      "step": 18493
+    },
+    {
+      "epoch": 23.67232,
+      "grad_norm": 1.1284360885620117,
+      "learning_rate": 1.3027210884353741e-05,
+      "loss": 0.4694,
+      "step": 18494
+    },
+    {
+      "epoch": 23.6736,
+      "grad_norm": 1.1159340143203735,
+      "learning_rate": 1.3025210084033614e-05,
+      "loss": 0.4493,
+      "step": 18495
+    },
+    {
+      "epoch": 23.67488,
+      "grad_norm": 1.236001968383789,
+      "learning_rate": 1.3023209283713486e-05,
+      "loss": 0.508,
+      "step": 18496
+    },
+    {
+      "epoch": 23.67616,
+      "grad_norm": 1.2399630546569824,
+      "learning_rate": 1.3021208483393357e-05,
+      "loss": 0.494,
+      "step": 18497
+    },
+    {
+      "epoch": 23.67744,
+      "grad_norm": 1.181467890739441,
+      "learning_rate": 1.3019207683073232e-05,
+      "loss": 0.4886,
+      "step": 18498
+    },
+    {
+      "epoch": 23.67872,
+      "grad_norm": 1.2008730173110962,
+      "learning_rate": 1.3017206882753102e-05,
+      "loss": 0.5143,
+      "step": 18499
+    },
+    {
+      "epoch": 23.68,
+      "grad_norm": 1.1191836595535278,
+      "learning_rate": 1.3015206082432974e-05,
+      "loss": 0.4378,
+      "step": 18500
+    },
+    {
+      "epoch": 23.68128,
+      "grad_norm": 1.0821129083633423,
+      "learning_rate": 1.3013205282112844e-05,
+      "loss": 0.4534,
+      "step": 18501
+    },
+    {
+      "epoch": 23.68256,
+      "grad_norm": 1.1685720682144165,
+      "learning_rate": 1.301120448179272e-05,
+      "loss": 0.5053,
+      "step": 18502
+    },
+    {
+      "epoch": 23.68384,
+      "grad_norm": 1.1220587491989136,
+      "learning_rate": 1.300920368147259e-05,
+      "loss": 0.5203,
+      "step": 18503
+    },
+    {
+      "epoch": 23.68512,
+      "grad_norm": 1.1965205669403076,
+      "learning_rate": 1.3007202881152461e-05,
+      "loss": 0.535,
+      "step": 18504
+    },
+    {
+      "epoch": 23.6864,
+      "grad_norm": 1.161190152168274,
+      "learning_rate": 1.3005202080832335e-05,
+      "loss": 0.4718,
+      "step": 18505
+    },
+    {
+      "epoch": 23.68768,
+      "grad_norm": 1.1077537536621094,
+      "learning_rate": 1.3003201280512207e-05,
+      "loss": 0.4594,
+      "step": 18506
+    },
+    {
+      "epoch": 23.68896,
+      "grad_norm": 1.136878490447998,
+      "learning_rate": 1.3001200480192077e-05,
+      "loss": 0.5092,
+      "step": 18507
+    },
+    {
+      "epoch": 23.69024,
+      "grad_norm": 1.1334853172302246,
+      "learning_rate": 1.2999199679871949e-05,
+      "loss": 0.4772,
+      "step": 18508
+    },
+    {
+      "epoch": 23.69152,
+      "grad_norm": 1.1531234979629517,
+      "learning_rate": 1.2997198879551822e-05,
+      "loss": 0.4892,
+      "step": 18509
+    },
+    {
+      "epoch": 23.6928,
+      "grad_norm": 1.186574101448059,
+      "learning_rate": 1.2995198079231694e-05,
+      "loss": 0.5056,
+      "step": 18510
+    },
+    {
+      "epoch": 23.69408,
+      "grad_norm": 1.0875732898712158,
+      "learning_rate": 1.2993197278911564e-05,
+      "loss": 0.458,
+      "step": 18511
+    },
+    {
+      "epoch": 23.69536,
+      "grad_norm": 1.11167311668396,
+      "learning_rate": 1.2991196478591436e-05,
+      "loss": 0.4613,
+      "step": 18512
+    },
+    {
+      "epoch": 23.69664,
+      "grad_norm": 1.0831806659698486,
+      "learning_rate": 1.298919567827131e-05,
+      "loss": 0.4726,
+      "step": 18513
+    },
+    {
+      "epoch": 23.69792,
+      "grad_norm": 1.1323453187942505,
+      "learning_rate": 1.2987194877951181e-05,
+      "loss": 0.449,
+      "step": 18514
+    },
+    {
+      "epoch": 23.6992,
+      "grad_norm": 1.2333152294158936,
+      "learning_rate": 1.2985194077631052e-05,
+      "loss": 0.5372,
+      "step": 18515
+    },
+    {
+      "epoch": 23.70048,
+      "grad_norm": 1.1051863431930542,
+      "learning_rate": 1.2983193277310927e-05,
+      "loss": 0.4527,
+      "step": 18516
+    },
+    {
+      "epoch": 23.70176,
+      "grad_norm": 1.1067372560501099,
+      "learning_rate": 1.2981192476990797e-05,
+      "loss": 0.4401,
+      "step": 18517
+    },
+    {
+      "epoch": 23.70304,
+      "grad_norm": 1.1563383340835571,
+      "learning_rate": 1.2979191676670669e-05,
+      "loss": 0.4822,
+      "step": 18518
+    },
+    {
+      "epoch": 23.70432,
+      "grad_norm": 1.1573394536972046,
+      "learning_rate": 1.2977190876350539e-05,
+      "loss": 0.4711,
+      "step": 18519
+    },
+    {
+      "epoch": 23.7056,
+      "grad_norm": 1.1465578079223633,
+      "learning_rate": 1.2975190076030414e-05,
+      "loss": 0.4794,
+      "step": 18520
+    },
+    {
+      "epoch": 23.706879999999998,
+      "grad_norm": 1.093738317489624,
+      "learning_rate": 1.2973189275710284e-05,
+      "loss": 0.4742,
+      "step": 18521
+    },
+    {
+      "epoch": 23.70816,
+      "grad_norm": 1.1131172180175781,
+      "learning_rate": 1.2971188475390156e-05,
+      "loss": 0.5315,
+      "step": 18522
+    },
+    {
+      "epoch": 23.70944,
+      "grad_norm": 1.1380599737167358,
+      "learning_rate": 1.296918767507003e-05,
+      "loss": 0.4672,
+      "step": 18523
+    },
+    {
+      "epoch": 23.71072,
+      "grad_norm": 1.1121565103530884,
+      "learning_rate": 1.2967186874749902e-05,
+      "loss": 0.4578,
+      "step": 18524
+    },
+    {
+      "epoch": 23.712,
+      "grad_norm": 1.1249797344207764,
+      "learning_rate": 1.2965186074429772e-05,
+      "loss": 0.4734,
+      "step": 18525
+    },
+    {
+      "epoch": 23.71328,
+      "grad_norm": 1.141965389251709,
+      "learning_rate": 1.2963185274109644e-05,
+      "loss": 0.4883,
+      "step": 18526
+    },
+    {
+      "epoch": 23.71456,
+      "grad_norm": 1.1544233560562134,
+      "learning_rate": 1.2961184473789517e-05,
+      "loss": 0.4956,
+      "step": 18527
+    },
+    {
+      "epoch": 23.71584,
+      "grad_norm": 1.1456880569458008,
+      "learning_rate": 1.2959183673469389e-05,
+      "loss": 0.4913,
+      "step": 18528
+    },
+    {
+      "epoch": 23.71712,
+      "grad_norm": 1.1444575786590576,
+      "learning_rate": 1.295718287314926e-05,
+      "loss": 0.4707,
+      "step": 18529
+    },
+    {
+      "epoch": 23.7184,
+      "grad_norm": 1.050844669342041,
+      "learning_rate": 1.2955182072829133e-05,
+      "loss": 0.4341,
+      "step": 18530
+    },
+    {
+      "epoch": 23.71968,
+      "grad_norm": 1.1426994800567627,
+      "learning_rate": 1.2953181272509005e-05,
+      "loss": 0.4679,
+      "step": 18531
+    },
+    {
+      "epoch": 23.72096,
+      "grad_norm": 1.1515729427337646,
+      "learning_rate": 1.2951180472188877e-05,
+      "loss": 0.4746,
+      "step": 18532
+    },
+    {
+      "epoch": 23.72224,
+      "grad_norm": 1.2546969652175903,
+      "learning_rate": 1.2949179671868747e-05,
+      "loss": 0.4886,
+      "step": 18533
+    },
+    {
+      "epoch": 23.72352,
+      "grad_norm": 1.2079882621765137,
+      "learning_rate": 1.294717887154862e-05,
+      "loss": 0.4708,
+      "step": 18534
+    },
+    {
+      "epoch": 23.7248,
+      "grad_norm": 1.2106804847717285,
+      "learning_rate": 1.2945178071228492e-05,
+      "loss": 0.5217,
+      "step": 18535
+    },
+    {
+      "epoch": 23.72608,
+      "grad_norm": 1.2002469301223755,
+      "learning_rate": 1.2943177270908364e-05,
+      "loss": 0.4444,
+      "step": 18536
+    },
+    {
+      "epoch": 23.72736,
+      "grad_norm": 1.2250186204910278,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 0.5409,
+      "step": 18537
+    },
+    {
+      "epoch": 23.72864,
+      "grad_norm": 1.1869670152664185,
+      "learning_rate": 1.2939175670268108e-05,
+      "loss": 0.4845,
+      "step": 18538
+    },
+    {
+      "epoch": 23.72992,
+      "grad_norm": 1.2370437383651733,
+      "learning_rate": 1.293717486994798e-05,
+      "loss": 0.4883,
+      "step": 18539
+    },
+    {
+      "epoch": 23.7312,
+      "grad_norm": 1.15911066532135,
+      "learning_rate": 1.2935174069627851e-05,
+      "loss": 0.4697,
+      "step": 18540
+    },
+    {
+      "epoch": 23.73248,
+      "grad_norm": 1.1930956840515137,
+      "learning_rate": 1.2933173269307725e-05,
+      "loss": 0.4824,
+      "step": 18541
+    },
+    {
+      "epoch": 23.73376,
+      "grad_norm": 1.1702487468719482,
+      "learning_rate": 1.2931172468987595e-05,
+      "loss": 0.4525,
+      "step": 18542
+    },
+    {
+      "epoch": 23.73504,
+      "grad_norm": 1.1612924337387085,
+      "learning_rate": 1.2929171668667467e-05,
+      "loss": 0.5488,
+      "step": 18543
+    },
+    {
+      "epoch": 23.73632,
+      "grad_norm": 1.1951484680175781,
+      "learning_rate": 1.292717086834734e-05,
+      "loss": 0.4683,
+      "step": 18544
+    },
+    {
+      "epoch": 23.7376,
+      "grad_norm": 1.1368860006332397,
+      "learning_rate": 1.2925170068027212e-05,
+      "loss": 0.4735,
+      "step": 18545
+    },
+    {
+      "epoch": 23.73888,
+      "grad_norm": 1.100229263305664,
+      "learning_rate": 1.2923169267707083e-05,
+      "loss": 0.4805,
+      "step": 18546
+    },
+    {
+      "epoch": 23.74016,
+      "grad_norm": 1.1962560415267944,
+      "learning_rate": 1.2921168467386954e-05,
+      "loss": 0.488,
+      "step": 18547
+    },
+    {
+      "epoch": 23.74144,
+      "grad_norm": 1.1619665622711182,
+      "learning_rate": 1.2919167667066828e-05,
+      "loss": 0.4893,
+      "step": 18548
+    },
+    {
+      "epoch": 23.74272,
+      "grad_norm": 1.1612014770507812,
+      "learning_rate": 1.29171668667467e-05,
+      "loss": 0.4735,
+      "step": 18549
+    },
+    {
+      "epoch": 23.744,
+      "grad_norm": 1.1319211721420288,
+      "learning_rate": 1.291516606642657e-05,
+      "loss": 0.4683,
+      "step": 18550
+    },
+    {
+      "epoch": 23.74528,
+      "grad_norm": 1.1409507989883423,
+      "learning_rate": 1.2913165266106445e-05,
+      "loss": 0.4702,
+      "step": 18551
+    },
+    {
+      "epoch": 23.74656,
+      "grad_norm": 1.1675480604171753,
+      "learning_rate": 1.2911164465786315e-05,
+      "loss": 0.4782,
+      "step": 18552
+    },
+    {
+      "epoch": 23.74784,
+      "grad_norm": 1.1573346853256226,
+      "learning_rate": 1.2909163665466187e-05,
+      "loss": 0.5024,
+      "step": 18553
+    },
+    {
+      "epoch": 23.74912,
+      "grad_norm": 1.1228611469268799,
+      "learning_rate": 1.2907162865146057e-05,
+      "loss": 0.4761,
+      "step": 18554
+    },
+    {
+      "epoch": 23.7504,
+      "grad_norm": 1.0913078784942627,
+      "learning_rate": 1.2905162064825933e-05,
+      "loss": 0.4681,
+      "step": 18555
+    },
+    {
+      "epoch": 23.75168,
+      "grad_norm": 1.1557594537734985,
+      "learning_rate": 1.2903161264505803e-05,
+      "loss": 0.4563,
+      "step": 18556
+    },
+    {
+      "epoch": 23.75296,
+      "grad_norm": 1.1187933683395386,
+      "learning_rate": 1.2901160464185675e-05,
+      "loss": 0.4953,
+      "step": 18557
+    },
+    {
+      "epoch": 23.75424,
+      "grad_norm": 1.18485426902771,
+      "learning_rate": 1.2899159663865548e-05,
+      "loss": 0.5024,
+      "step": 18558
+    },
+    {
+      "epoch": 23.75552,
+      "grad_norm": 1.158681035041809,
+      "learning_rate": 1.289715886354542e-05,
+      "loss": 0.471,
+      "step": 18559
+    },
+    {
+      "epoch": 23.7568,
+      "grad_norm": 1.1719639301300049,
+      "learning_rate": 1.289515806322529e-05,
+      "loss": 0.4878,
+      "step": 18560
+    },
+    {
+      "epoch": 23.75808,
+      "grad_norm": 1.1346675157546997,
+      "learning_rate": 1.2893157262905162e-05,
+      "loss": 0.4737,
+      "step": 18561
+    },
+    {
+      "epoch": 23.75936,
+      "grad_norm": 1.1762566566467285,
+      "learning_rate": 1.2891156462585036e-05,
+      "loss": 0.4824,
+      "step": 18562
+    },
+    {
+      "epoch": 23.76064,
+      "grad_norm": 1.190079689025879,
+      "learning_rate": 1.2889155662264908e-05,
+      "loss": 0.4976,
+      "step": 18563
+    },
+    {
+      "epoch": 23.76192,
+      "grad_norm": 1.1472417116165161,
+      "learning_rate": 1.2887154861944778e-05,
+      "loss": 0.4598,
+      "step": 18564
+    },
+    {
+      "epoch": 23.7632,
+      "grad_norm": 1.1223433017730713,
+      "learning_rate": 1.288515406162465e-05,
+      "loss": 0.4866,
+      "step": 18565
+    },
+    {
+      "epoch": 23.76448,
+      "grad_norm": 1.154817819595337,
+      "learning_rate": 1.2883153261304523e-05,
+      "loss": 0.4892,
+      "step": 18566
+    },
+    {
+      "epoch": 23.76576,
+      "grad_norm": 1.1164178848266602,
+      "learning_rate": 1.2881152460984395e-05,
+      "loss": 0.5027,
+      "step": 18567
+    },
+    {
+      "epoch": 23.76704,
+      "grad_norm": 1.171064019203186,
+      "learning_rate": 1.2879151660664265e-05,
+      "loss": 0.4635,
+      "step": 18568
+    },
+    {
+      "epoch": 23.76832,
+      "grad_norm": 1.1405155658721924,
+      "learning_rate": 1.2877150860344139e-05,
+      "loss": 0.4721,
+      "step": 18569
+    },
+    {
+      "epoch": 23.7696,
+      "grad_norm": 1.0519278049468994,
+      "learning_rate": 1.287515006002401e-05,
+      "loss": 0.4368,
+      "step": 18570
+    },
+    {
+      "epoch": 23.77088,
+      "grad_norm": 1.1556754112243652,
+      "learning_rate": 1.2873149259703882e-05,
+      "loss": 0.5245,
+      "step": 18571
+    },
+    {
+      "epoch": 23.77216,
+      "grad_norm": 1.1321344375610352,
+      "learning_rate": 1.2871148459383753e-05,
+      "loss": 0.4937,
+      "step": 18572
+    },
+    {
+      "epoch": 23.77344,
+      "grad_norm": 1.14230477809906,
+      "learning_rate": 1.2869147659063626e-05,
+      "loss": 0.4789,
+      "step": 18573
+    },
+    {
+      "epoch": 23.77472,
+      "grad_norm": 1.082296371459961,
+      "learning_rate": 1.2867146858743498e-05,
+      "loss": 0.4855,
+      "step": 18574
+    },
+    {
+      "epoch": 23.776,
+      "grad_norm": 1.1363306045532227,
+      "learning_rate": 1.286514605842337e-05,
+      "loss": 0.4706,
+      "step": 18575
+    },
+    {
+      "epoch": 23.77728,
+      "grad_norm": 1.180460810661316,
+      "learning_rate": 1.2863145258103243e-05,
+      "loss": 0.503,
+      "step": 18576
+    },
+    {
+      "epoch": 23.77856,
+      "grad_norm": 1.1838196516036987,
+      "learning_rate": 1.2861144457783114e-05,
+      "loss": 0.5241,
+      "step": 18577
+    },
+    {
+      "epoch": 23.77984,
+      "grad_norm": 1.1658860445022583,
+      "learning_rate": 1.2859143657462985e-05,
+      "loss": 0.4915,
+      "step": 18578
+    },
+    {
+      "epoch": 23.78112,
+      "grad_norm": 1.1486824750900269,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.4803,
+      "step": 18579
+    },
+    {
+      "epoch": 23.7824,
+      "grad_norm": 1.0898503065109253,
+      "learning_rate": 1.285514205682273e-05,
+      "loss": 0.4444,
+      "step": 18580
+    },
+    {
+      "epoch": 23.78368,
+      "grad_norm": 1.1599661111831665,
+      "learning_rate": 1.2853141256502601e-05,
+      "loss": 0.539,
+      "step": 18581
+    },
+    {
+      "epoch": 23.78496,
+      "grad_norm": 1.151951551437378,
+      "learning_rate": 1.2851140456182473e-05,
+      "loss": 0.4741,
+      "step": 18582
+    },
+    {
+      "epoch": 23.78624,
+      "grad_norm": 1.2248902320861816,
+      "learning_rate": 1.2849139655862346e-05,
+      "loss": 0.4911,
+      "step": 18583
+    },
+    {
+      "epoch": 23.78752,
+      "grad_norm": 1.1011284589767456,
+      "learning_rate": 1.2847138855542218e-05,
+      "loss": 0.491,
+      "step": 18584
+    },
+    {
+      "epoch": 23.7888,
+      "grad_norm": 1.1119115352630615,
+      "learning_rate": 1.2845138055222088e-05,
+      "loss": 0.47,
+      "step": 18585
+    },
+    {
+      "epoch": 23.79008,
+      "grad_norm": 1.1536645889282227,
+      "learning_rate": 1.284313725490196e-05,
+      "loss": 0.4986,
+      "step": 18586
+    },
+    {
+      "epoch": 23.79136,
+      "grad_norm": 1.1889305114746094,
+      "learning_rate": 1.2841136454581834e-05,
+      "loss": 0.4693,
+      "step": 18587
+    },
+    {
+      "epoch": 23.79264,
+      "grad_norm": 1.0987606048583984,
+      "learning_rate": 1.2839135654261706e-05,
+      "loss": 0.4626,
+      "step": 18588
+    },
+    {
+      "epoch": 23.79392,
+      "grad_norm": 1.1800764799118042,
+      "learning_rate": 1.2837134853941576e-05,
+      "loss": 0.5212,
+      "step": 18589
+    },
+    {
+      "epoch": 23.7952,
+      "grad_norm": 1.1225427389144897,
+      "learning_rate": 1.2835134053621451e-05,
+      "loss": 0.4598,
+      "step": 18590
+    },
+    {
+      "epoch": 23.79648,
+      "grad_norm": 1.1654258966445923,
+      "learning_rate": 1.2833133253301321e-05,
+      "loss": 0.491,
+      "step": 18591
+    },
+    {
+      "epoch": 23.79776,
+      "grad_norm": 1.114014983177185,
+      "learning_rate": 1.2831132452981193e-05,
+      "loss": 0.4837,
+      "step": 18592
+    },
+    {
+      "epoch": 23.79904,
+      "grad_norm": 1.117835521697998,
+      "learning_rate": 1.2829131652661063e-05,
+      "loss": 0.4643,
+      "step": 18593
+    },
+    {
+      "epoch": 23.80032,
+      "grad_norm": 1.181115746498108,
+      "learning_rate": 1.2827130852340938e-05,
+      "loss": 0.5055,
+      "step": 18594
+    },
+    {
+      "epoch": 23.8016,
+      "grad_norm": 1.112195611000061,
+      "learning_rate": 1.2825130052020809e-05,
+      "loss": 0.4422,
+      "step": 18595
+    },
+    {
+      "epoch": 23.802880000000002,
+      "grad_norm": 1.1325654983520508,
+      "learning_rate": 1.282312925170068e-05,
+      "loss": 0.4486,
+      "step": 18596
+    },
+    {
+      "epoch": 23.80416,
+      "grad_norm": 1.103108286857605,
+      "learning_rate": 1.2821128451380554e-05,
+      "loss": 0.4953,
+      "step": 18597
+    },
+    {
+      "epoch": 23.80544,
+      "grad_norm": 1.0995981693267822,
+      "learning_rate": 1.2819127651060426e-05,
+      "loss": 0.4517,
+      "step": 18598
+    },
+    {
+      "epoch": 23.80672,
+      "grad_norm": 1.0977530479431152,
+      "learning_rate": 1.2817126850740296e-05,
+      "loss": 0.4939,
+      "step": 18599
+    },
+    {
+      "epoch": 23.808,
+      "grad_norm": 1.1470292806625366,
+      "learning_rate": 1.2815126050420168e-05,
+      "loss": 0.4998,
+      "step": 18600
+    },
+    {
+      "epoch": 23.80928,
+      "grad_norm": 1.1226685047149658,
+      "learning_rate": 1.2813125250100041e-05,
+      "loss": 0.4792,
+      "step": 18601
+    },
+    {
+      "epoch": 23.81056,
+      "grad_norm": 1.1221965551376343,
+      "learning_rate": 1.2811124449779913e-05,
+      "loss": 0.4889,
+      "step": 18602
+    },
+    {
+      "epoch": 23.81184,
+      "grad_norm": 1.1775237321853638,
+      "learning_rate": 1.2809123649459783e-05,
+      "loss": 0.4841,
+      "step": 18603
+    },
+    {
+      "epoch": 23.81312,
+      "grad_norm": 1.0999358892440796,
+      "learning_rate": 1.2807122849139657e-05,
+      "loss": 0.4482,
+      "step": 18604
+    },
+    {
+      "epoch": 23.8144,
+      "grad_norm": 1.074433445930481,
+      "learning_rate": 1.2805122048819529e-05,
+      "loss": 0.4656,
+      "step": 18605
+    },
+    {
+      "epoch": 23.81568,
+      "grad_norm": 1.0768699645996094,
+      "learning_rate": 1.28031212484994e-05,
+      "loss": 0.4976,
+      "step": 18606
+    },
+    {
+      "epoch": 23.81696,
+      "grad_norm": 1.0591034889221191,
+      "learning_rate": 1.2801120448179271e-05,
+      "loss": 0.4466,
+      "step": 18607
+    },
+    {
+      "epoch": 23.81824,
+      "grad_norm": 1.2163914442062378,
+      "learning_rate": 1.2799119647859144e-05,
+      "loss": 0.5735,
+      "step": 18608
+    },
+    {
+      "epoch": 23.81952,
+      "grad_norm": 1.1105905771255493,
+      "learning_rate": 1.2797118847539016e-05,
+      "loss": 0.4582,
+      "step": 18609
+    },
+    {
+      "epoch": 23.8208,
+      "grad_norm": 1.1055454015731812,
+      "learning_rate": 1.2795118047218888e-05,
+      "loss": 0.4706,
+      "step": 18610
+    },
+    {
+      "epoch": 23.82208,
+      "grad_norm": 1.1523383855819702,
+      "learning_rate": 1.2793117246898762e-05,
+      "loss": 0.4749,
+      "step": 18611
+    },
+    {
+      "epoch": 23.82336,
+      "grad_norm": 1.1723835468292236,
+      "learning_rate": 1.2791116446578632e-05,
+      "loss": 0.4854,
+      "step": 18612
+    },
+    {
+      "epoch": 23.82464,
+      "grad_norm": 1.1401787996292114,
+      "learning_rate": 1.2789115646258504e-05,
+      "loss": 0.5144,
+      "step": 18613
+    },
+    {
+      "epoch": 23.82592,
+      "grad_norm": 1.1540292501449585,
+      "learning_rate": 1.2787114845938376e-05,
+      "loss": 0.4866,
+      "step": 18614
+    },
+    {
+      "epoch": 23.8272,
+      "grad_norm": 1.1869522333145142,
+      "learning_rate": 1.2785114045618249e-05,
+      "loss": 0.4653,
+      "step": 18615
+    },
+    {
+      "epoch": 23.82848,
+      "grad_norm": 1.1412826776504517,
+      "learning_rate": 1.278311324529812e-05,
+      "loss": 0.4968,
+      "step": 18616
+    },
+    {
+      "epoch": 23.82976,
+      "grad_norm": 1.15664803981781,
+      "learning_rate": 1.2781112444977991e-05,
+      "loss": 0.4903,
+      "step": 18617
+    },
+    {
+      "epoch": 23.83104,
+      "grad_norm": 1.1142537593841553,
+      "learning_rate": 1.2779111644657865e-05,
+      "loss": 0.4709,
+      "step": 18618
+    },
+    {
+      "epoch": 23.83232,
+      "grad_norm": 1.0486506223678589,
+      "learning_rate": 1.2777110844337737e-05,
+      "loss": 0.4785,
+      "step": 18619
+    },
+    {
+      "epoch": 23.8336,
+      "grad_norm": 1.0911049842834473,
+      "learning_rate": 1.2775110044017607e-05,
+      "loss": 0.4762,
+      "step": 18620
+    },
+    {
+      "epoch": 23.83488,
+      "grad_norm": 1.146394968032837,
+      "learning_rate": 1.2773109243697479e-05,
+      "loss": 0.521,
+      "step": 18621
+    },
+    {
+      "epoch": 23.83616,
+      "grad_norm": 1.053382396697998,
+      "learning_rate": 1.2771108443377352e-05,
+      "loss": 0.4642,
+      "step": 18622
+    },
+    {
+      "epoch": 23.83744,
+      "grad_norm": 1.1303876638412476,
+      "learning_rate": 1.2769107643057224e-05,
+      "loss": 0.4674,
+      "step": 18623
+    },
+    {
+      "epoch": 23.83872,
+      "grad_norm": 1.1761562824249268,
+      "learning_rate": 1.2767106842737094e-05,
+      "loss": 0.5115,
+      "step": 18624
+    },
+    {
+      "epoch": 23.84,
+      "grad_norm": 1.120778203010559,
+      "learning_rate": 1.2765106042416966e-05,
+      "loss": 0.4848,
+      "step": 18625
+    },
+    {
+      "epoch": 23.84128,
+      "grad_norm": 1.110851764678955,
+      "learning_rate": 1.276310524209684e-05,
+      "loss": 0.4846,
+      "step": 18626
+    },
+    {
+      "epoch": 23.84256,
+      "grad_norm": 1.1596050262451172,
+      "learning_rate": 1.2761104441776711e-05,
+      "loss": 0.4922,
+      "step": 18627
+    },
+    {
+      "epoch": 23.84384,
+      "grad_norm": 1.1609506607055664,
+      "learning_rate": 1.2759103641456582e-05,
+      "loss": 0.5059,
+      "step": 18628
+    },
+    {
+      "epoch": 23.84512,
+      "grad_norm": 1.1542946100234985,
+      "learning_rate": 1.2757102841136457e-05,
+      "loss": 0.5106,
+      "step": 18629
+    },
+    {
+      "epoch": 23.8464,
+      "grad_norm": 1.1738747358322144,
+      "learning_rate": 1.2755102040816327e-05,
+      "loss": 0.4977,
+      "step": 18630
+    },
+    {
+      "epoch": 23.84768,
+      "grad_norm": 1.1270424127578735,
+      "learning_rate": 1.2753101240496199e-05,
+      "loss": 0.4735,
+      "step": 18631
+    },
+    {
+      "epoch": 23.84896,
+      "grad_norm": 1.1536674499511719,
+      "learning_rate": 1.2751100440176069e-05,
+      "loss": 0.506,
+      "step": 18632
+    },
+    {
+      "epoch": 23.85024,
+      "grad_norm": 1.1659085750579834,
+      "learning_rate": 1.2749099639855944e-05,
+      "loss": 0.5192,
+      "step": 18633
+    },
+    {
+      "epoch": 23.85152,
+      "grad_norm": 1.1134508848190308,
+      "learning_rate": 1.2747098839535814e-05,
+      "loss": 0.4403,
+      "step": 18634
+    },
+    {
+      "epoch": 23.8528,
+      "grad_norm": 1.1781237125396729,
+      "learning_rate": 1.2745098039215686e-05,
+      "loss": 0.4842,
+      "step": 18635
+    },
+    {
+      "epoch": 23.85408,
+      "grad_norm": 1.1390173435211182,
+      "learning_rate": 1.274309723889556e-05,
+      "loss": 0.4649,
+      "step": 18636
+    },
+    {
+      "epoch": 23.85536,
+      "grad_norm": 1.1824352741241455,
+      "learning_rate": 1.2741096438575432e-05,
+      "loss": 0.4635,
+      "step": 18637
+    },
+    {
+      "epoch": 23.85664,
+      "grad_norm": 1.1931419372558594,
+      "learning_rate": 1.2739095638255302e-05,
+      "loss": 0.5115,
+      "step": 18638
+    },
+    {
+      "epoch": 23.85792,
+      "grad_norm": 1.082920789718628,
+      "learning_rate": 1.2737094837935174e-05,
+      "loss": 0.4455,
+      "step": 18639
+    },
+    {
+      "epoch": 23.8592,
+      "grad_norm": 1.1064939498901367,
+      "learning_rate": 1.2735094037615047e-05,
+      "loss": 0.4162,
+      "step": 18640
+    },
+    {
+      "epoch": 23.86048,
+      "grad_norm": 1.1889894008636475,
+      "learning_rate": 1.2733093237294919e-05,
+      "loss": 0.4835,
+      "step": 18641
+    },
+    {
+      "epoch": 23.86176,
+      "grad_norm": 1.2531040906906128,
+      "learning_rate": 1.273109243697479e-05,
+      "loss": 0.5775,
+      "step": 18642
+    },
+    {
+      "epoch": 23.86304,
+      "grad_norm": 1.176379919052124,
+      "learning_rate": 1.2729091636654663e-05,
+      "loss": 0.5473,
+      "step": 18643
+    },
+    {
+      "epoch": 23.86432,
+      "grad_norm": 1.1142444610595703,
+      "learning_rate": 1.2727090836334535e-05,
+      "loss": 0.4867,
+      "step": 18644
+    },
+    {
+      "epoch": 23.8656,
+      "grad_norm": 1.1500768661499023,
+      "learning_rate": 1.2725090036014407e-05,
+      "loss": 0.4787,
+      "step": 18645
+    },
+    {
+      "epoch": 23.86688,
+      "grad_norm": 1.1695806980133057,
+      "learning_rate": 1.2723089235694277e-05,
+      "loss": 0.4926,
+      "step": 18646
+    },
+    {
+      "epoch": 23.86816,
+      "grad_norm": 1.1269073486328125,
+      "learning_rate": 1.272108843537415e-05,
+      "loss": 0.4831,
+      "step": 18647
+    },
+    {
+      "epoch": 23.86944,
+      "grad_norm": 1.213214635848999,
+      "learning_rate": 1.2719087635054022e-05,
+      "loss": 0.5235,
+      "step": 18648
+    },
+    {
+      "epoch": 23.87072,
+      "grad_norm": 1.134529709815979,
+      "learning_rate": 1.2717086834733894e-05,
+      "loss": 0.5038,
+      "step": 18649
+    },
+    {
+      "epoch": 23.872,
+      "grad_norm": 1.142578363418579,
+      "learning_rate": 1.2715086034413768e-05,
+      "loss": 0.4529,
+      "step": 18650
+    },
+    {
+      "epoch": 23.87328,
+      "grad_norm": 1.2363348007202148,
+      "learning_rate": 1.2713085234093638e-05,
+      "loss": 0.5431,
+      "step": 18651
+    },
+    {
+      "epoch": 23.87456,
+      "grad_norm": 1.0975717306137085,
+      "learning_rate": 1.271108443377351e-05,
+      "loss": 0.4614,
+      "step": 18652
+    },
+    {
+      "epoch": 23.87584,
+      "grad_norm": 1.1565914154052734,
+      "learning_rate": 1.2709083633453381e-05,
+      "loss": 0.5314,
+      "step": 18653
+    },
+    {
+      "epoch": 23.87712,
+      "grad_norm": 1.1390055418014526,
+      "learning_rate": 1.2707082833133255e-05,
+      "loss": 0.4763,
+      "step": 18654
+    },
+    {
+      "epoch": 23.8784,
+      "grad_norm": 1.120939016342163,
+      "learning_rate": 1.2705082032813125e-05,
+      "loss": 0.4826,
+      "step": 18655
+    },
+    {
+      "epoch": 23.87968,
+      "grad_norm": 1.1649311780929565,
+      "learning_rate": 1.2703081232492997e-05,
+      "loss": 0.5242,
+      "step": 18656
+    },
+    {
+      "epoch": 23.88096,
+      "grad_norm": 1.1487033367156982,
+      "learning_rate": 1.270108043217287e-05,
+      "loss": 0.4823,
+      "step": 18657
+    },
+    {
+      "epoch": 23.88224,
+      "grad_norm": 1.161218523979187,
+      "learning_rate": 1.2699079631852742e-05,
+      "loss": 0.4973,
+      "step": 18658
+    },
+    {
+      "epoch": 23.88352,
+      "grad_norm": 1.1504417657852173,
+      "learning_rate": 1.2697078831532613e-05,
+      "loss": 0.4817,
+      "step": 18659
+    },
+    {
+      "epoch": 23.8848,
+      "grad_norm": 1.204056978225708,
+      "learning_rate": 1.2695078031212484e-05,
+      "loss": 0.531,
+      "step": 18660
+    },
+    {
+      "epoch": 23.88608,
+      "grad_norm": 1.180108666419983,
+      "learning_rate": 1.2693077230892358e-05,
+      "loss": 0.4767,
+      "step": 18661
+    },
+    {
+      "epoch": 23.88736,
+      "grad_norm": 1.0854483842849731,
+      "learning_rate": 1.269107643057223e-05,
+      "loss": 0.4293,
+      "step": 18662
+    },
+    {
+      "epoch": 23.88864,
+      "grad_norm": 1.11992609500885,
+      "learning_rate": 1.26890756302521e-05,
+      "loss": 0.4731,
+      "step": 18663
+    },
+    {
+      "epoch": 23.88992,
+      "grad_norm": 1.1562358140945435,
+      "learning_rate": 1.2687074829931975e-05,
+      "loss": 0.5066,
+      "step": 18664
+    },
+    {
+      "epoch": 23.8912,
+      "grad_norm": 1.1416101455688477,
+      "learning_rate": 1.2685074029611845e-05,
+      "loss": 0.517,
+      "step": 18665
+    },
+    {
+      "epoch": 23.89248,
+      "grad_norm": 1.144882082939148,
+      "learning_rate": 1.2683073229291717e-05,
+      "loss": 0.4406,
+      "step": 18666
+    },
+    {
+      "epoch": 23.89376,
+      "grad_norm": 1.1081516742706299,
+      "learning_rate": 1.2681072428971587e-05,
+      "loss": 0.4549,
+      "step": 18667
+    },
+    {
+      "epoch": 23.89504,
+      "grad_norm": 1.1630297899246216,
+      "learning_rate": 1.2679071628651463e-05,
+      "loss": 0.4835,
+      "step": 18668
+    },
+    {
+      "epoch": 23.89632,
+      "grad_norm": 1.1396640539169312,
+      "learning_rate": 1.2677070828331333e-05,
+      "loss": 0.4788,
+      "step": 18669
+    },
+    {
+      "epoch": 23.8976,
+      "grad_norm": 1.2122644186019897,
+      "learning_rate": 1.2675070028011205e-05,
+      "loss": 0.5103,
+      "step": 18670
+    },
+    {
+      "epoch": 23.89888,
+      "grad_norm": 1.224592924118042,
+      "learning_rate": 1.2673069227691078e-05,
+      "loss": 0.5471,
+      "step": 18671
+    },
+    {
+      "epoch": 23.90016,
+      "grad_norm": 1.184139370918274,
+      "learning_rate": 1.267106842737095e-05,
+      "loss": 0.5388,
+      "step": 18672
+    },
+    {
+      "epoch": 23.90144,
+      "grad_norm": 1.2130299806594849,
+      "learning_rate": 1.266906762705082e-05,
+      "loss": 0.5117,
+      "step": 18673
+    },
+    {
+      "epoch": 23.90272,
+      "grad_norm": 1.1781811714172363,
+      "learning_rate": 1.2667066826730692e-05,
+      "loss": 0.4856,
+      "step": 18674
+    },
+    {
+      "epoch": 23.904,
+      "grad_norm": 1.1690351963043213,
+      "learning_rate": 1.2665066026410566e-05,
+      "loss": 0.4863,
+      "step": 18675
+    },
+    {
+      "epoch": 23.90528,
+      "grad_norm": 1.0583910942077637,
+      "learning_rate": 1.2663065226090437e-05,
+      "loss": 0.4572,
+      "step": 18676
+    },
+    {
+      "epoch": 23.90656,
+      "grad_norm": 1.0969510078430176,
+      "learning_rate": 1.2661064425770308e-05,
+      "loss": 0.4255,
+      "step": 18677
+    },
+    {
+      "epoch": 23.90784,
+      "grad_norm": 1.1258211135864258,
+      "learning_rate": 1.265906362545018e-05,
+      "loss": 0.4855,
+      "step": 18678
+    },
+    {
+      "epoch": 23.90912,
+      "grad_norm": 1.175527572631836,
+      "learning_rate": 1.2657062825130053e-05,
+      "loss": 0.5177,
+      "step": 18679
+    },
+    {
+      "epoch": 23.9104,
+      "grad_norm": 1.1303937435150146,
+      "learning_rate": 1.2655062024809925e-05,
+      "loss": 0.4979,
+      "step": 18680
+    },
+    {
+      "epoch": 23.91168,
+      "grad_norm": 1.0970633029937744,
+      "learning_rate": 1.2653061224489795e-05,
+      "loss": 0.4648,
+      "step": 18681
+    },
+    {
+      "epoch": 23.912959999999998,
+      "grad_norm": 1.12120521068573,
+      "learning_rate": 1.265106042416967e-05,
+      "loss": 0.5126,
+      "step": 18682
+    },
+    {
+      "epoch": 23.91424,
+      "grad_norm": 1.1222971677780151,
+      "learning_rate": 1.264905962384954e-05,
+      "loss": 0.4831,
+      "step": 18683
+    },
+    {
+      "epoch": 23.91552,
+      "grad_norm": 1.146567702293396,
+      "learning_rate": 1.2647058823529412e-05,
+      "loss": 0.4479,
+      "step": 18684
+    },
+    {
+      "epoch": 23.9168,
+      "grad_norm": 1.0587197542190552,
+      "learning_rate": 1.2645058023209283e-05,
+      "loss": 0.4675,
+      "step": 18685
+    },
+    {
+      "epoch": 23.91808,
+      "grad_norm": 1.0969828367233276,
+      "learning_rate": 1.2643057222889158e-05,
+      "loss": 0.5139,
+      "step": 18686
+    },
+    {
+      "epoch": 23.91936,
+      "grad_norm": 1.1403480768203735,
+      "learning_rate": 1.2641056422569028e-05,
+      "loss": 0.5023,
+      "step": 18687
+    },
+    {
+      "epoch": 23.92064,
+      "grad_norm": 1.137335181236267,
+      "learning_rate": 1.26390556222489e-05,
+      "loss": 0.4405,
+      "step": 18688
+    },
+    {
+      "epoch": 23.92192,
+      "grad_norm": 1.1394611597061157,
+      "learning_rate": 1.2637054821928773e-05,
+      "loss": 0.5089,
+      "step": 18689
+    },
+    {
+      "epoch": 23.9232,
+      "grad_norm": 1.1698044538497925,
+      "learning_rate": 1.2635054021608645e-05,
+      "loss": 0.436,
+      "step": 18690
+    },
+    {
+      "epoch": 23.92448,
+      "grad_norm": 1.1128085851669312,
+      "learning_rate": 1.2633053221288515e-05,
+      "loss": 0.4677,
+      "step": 18691
+    },
+    {
+      "epoch": 23.92576,
+      "grad_norm": 1.1220570802688599,
+      "learning_rate": 1.2631052420968387e-05,
+      "loss": 0.4621,
+      "step": 18692
+    },
+    {
+      "epoch": 23.92704,
+      "grad_norm": 1.0818321704864502,
+      "learning_rate": 1.262905162064826e-05,
+      "loss": 0.4531,
+      "step": 18693
+    },
+    {
+      "epoch": 23.92832,
+      "grad_norm": 1.1508997678756714,
+      "learning_rate": 1.2627050820328133e-05,
+      "loss": 0.4856,
+      "step": 18694
+    },
+    {
+      "epoch": 23.9296,
+      "grad_norm": 1.146302580833435,
+      "learning_rate": 1.2625050020008003e-05,
+      "loss": 0.4544,
+      "step": 18695
+    },
+    {
+      "epoch": 23.93088,
+      "grad_norm": 1.1530104875564575,
+      "learning_rate": 1.2623049219687876e-05,
+      "loss": 0.4572,
+      "step": 18696
+    },
+    {
+      "epoch": 23.93216,
+      "grad_norm": 1.2042316198349,
+      "learning_rate": 1.2621048419367748e-05,
+      "loss": 0.521,
+      "step": 18697
+    },
+    {
+      "epoch": 23.93344,
+      "grad_norm": 1.1437196731567383,
+      "learning_rate": 1.261904761904762e-05,
+      "loss": 0.4874,
+      "step": 18698
+    },
+    {
+      "epoch": 23.93472,
+      "grad_norm": 1.0899137258529663,
+      "learning_rate": 1.261704681872749e-05,
+      "loss": 0.4506,
+      "step": 18699
+    },
+    {
+      "epoch": 23.936,
+      "grad_norm": 1.153517723083496,
+      "learning_rate": 1.2615046018407364e-05,
+      "loss": 0.4816,
+      "step": 18700
+    },
+    {
+      "epoch": 23.93728,
+      "grad_norm": 1.1165791749954224,
+      "learning_rate": 1.2613045218087236e-05,
+      "loss": 0.4971,
+      "step": 18701
+    },
+    {
+      "epoch": 23.93856,
+      "grad_norm": 1.0416091680526733,
+      "learning_rate": 1.2611044417767107e-05,
+      "loss": 0.4636,
+      "step": 18702
+    },
+    {
+      "epoch": 23.93984,
+      "grad_norm": 1.1278401613235474,
+      "learning_rate": 1.2609043617446981e-05,
+      "loss": 0.4784,
+      "step": 18703
+    },
+    {
+      "epoch": 23.94112,
+      "grad_norm": 1.0824394226074219,
+      "learning_rate": 1.2607042817126851e-05,
+      "loss": 0.4231,
+      "step": 18704
+    },
+    {
+      "epoch": 23.9424,
+      "grad_norm": 1.1062698364257812,
+      "learning_rate": 1.2605042016806723e-05,
+      "loss": 0.4636,
+      "step": 18705
+    },
+    {
+      "epoch": 23.94368,
+      "grad_norm": 1.1416337490081787,
+      "learning_rate": 1.2603041216486595e-05,
+      "loss": 0.4995,
+      "step": 18706
+    },
+    {
+      "epoch": 23.944960000000002,
+      "grad_norm": 1.0663871765136719,
+      "learning_rate": 1.2601040416166468e-05,
+      "loss": 0.4747,
+      "step": 18707
+    },
+    {
+      "epoch": 23.94624,
+      "grad_norm": 1.2235000133514404,
+      "learning_rate": 1.2599039615846339e-05,
+      "loss": 0.4992,
+      "step": 18708
+    },
+    {
+      "epoch": 23.94752,
+      "grad_norm": 1.0971728563308716,
+      "learning_rate": 1.259703881552621e-05,
+      "loss": 0.4431,
+      "step": 18709
+    },
+    {
+      "epoch": 23.9488,
+      "grad_norm": 1.1560825109481812,
+      "learning_rate": 1.2595038015206084e-05,
+      "loss": 0.4653,
+      "step": 18710
+    },
+    {
+      "epoch": 23.95008,
+      "grad_norm": 1.0700130462646484,
+      "learning_rate": 1.2593037214885956e-05,
+      "loss": 0.4259,
+      "step": 18711
+    },
+    {
+      "epoch": 23.95136,
+      "grad_norm": 1.120007038116455,
+      "learning_rate": 1.2591036414565826e-05,
+      "loss": 0.5081,
+      "step": 18712
+    },
+    {
+      "epoch": 23.95264,
+      "grad_norm": 1.1066792011260986,
+      "learning_rate": 1.2589035614245698e-05,
+      "loss": 0.4736,
+      "step": 18713
+    },
+    {
+      "epoch": 23.95392,
+      "grad_norm": 1.1956183910369873,
+      "learning_rate": 1.2587034813925571e-05,
+      "loss": 0.5158,
+      "step": 18714
+    },
+    {
+      "epoch": 23.9552,
+      "grad_norm": 1.1309118270874023,
+      "learning_rate": 1.2585034013605443e-05,
+      "loss": 0.459,
+      "step": 18715
+    },
+    {
+      "epoch": 23.95648,
+      "grad_norm": 1.1367197036743164,
+      "learning_rate": 1.2583033213285313e-05,
+      "loss": 0.5102,
+      "step": 18716
+    },
+    {
+      "epoch": 23.95776,
+      "grad_norm": 1.1677851676940918,
+      "learning_rate": 1.2581032412965189e-05,
+      "loss": 0.4909,
+      "step": 18717
+    },
+    {
+      "epoch": 23.95904,
+      "grad_norm": 1.1207268238067627,
+      "learning_rate": 1.2579031612645059e-05,
+      "loss": 0.5059,
+      "step": 18718
+    },
+    {
+      "epoch": 23.96032,
+      "grad_norm": 1.18141770362854,
+      "learning_rate": 1.257703081232493e-05,
+      "loss": 0.4715,
+      "step": 18719
+    },
+    {
+      "epoch": 23.9616,
+      "grad_norm": 1.203596591949463,
+      "learning_rate": 1.2575030012004801e-05,
+      "loss": 0.5205,
+      "step": 18720
+    },
+    {
+      "epoch": 23.96288,
+      "grad_norm": 1.124710202217102,
+      "learning_rate": 1.2573029211684676e-05,
+      "loss": 0.478,
+      "step": 18721
+    },
+    {
+      "epoch": 23.96416,
+      "grad_norm": 1.1432015895843506,
+      "learning_rate": 1.2571028411364546e-05,
+      "loss": 0.4859,
+      "step": 18722
+    },
+    {
+      "epoch": 23.96544,
+      "grad_norm": 1.1251646280288696,
+      "learning_rate": 1.2569027611044418e-05,
+      "loss": 0.4917,
+      "step": 18723
+    },
+    {
+      "epoch": 23.96672,
+      "grad_norm": 1.1594064235687256,
+      "learning_rate": 1.2567026810724292e-05,
+      "loss": 0.4919,
+      "step": 18724
+    },
+    {
+      "epoch": 23.968,
+      "grad_norm": 1.253121256828308,
+      "learning_rate": 1.2565026010404164e-05,
+      "loss": 0.467,
+      "step": 18725
+    },
+    {
+      "epoch": 23.96928,
+      "grad_norm": 1.188087821006775,
+      "learning_rate": 1.2563025210084034e-05,
+      "loss": 0.5015,
+      "step": 18726
+    },
+    {
+      "epoch": 23.97056,
+      "grad_norm": 1.1531556844711304,
+      "learning_rate": 1.2561024409763906e-05,
+      "loss": 0.4871,
+      "step": 18727
+    },
+    {
+      "epoch": 23.97184,
+      "grad_norm": 1.1422969102859497,
+      "learning_rate": 1.2559023609443779e-05,
+      "loss": 0.4797,
+      "step": 18728
+    },
+    {
+      "epoch": 23.97312,
+      "grad_norm": 1.1510505676269531,
+      "learning_rate": 1.2557022809123651e-05,
+      "loss": 0.4865,
+      "step": 18729
+    },
+    {
+      "epoch": 23.9744,
+      "grad_norm": 1.1790677309036255,
+      "learning_rate": 1.2555022008803521e-05,
+      "loss": 0.5234,
+      "step": 18730
+    },
+    {
+      "epoch": 23.97568,
+      "grad_norm": 1.141440987586975,
+      "learning_rate": 1.2553021208483393e-05,
+      "loss": 0.4965,
+      "step": 18731
+    },
+    {
+      "epoch": 23.97696,
+      "grad_norm": 1.1835546493530273,
+      "learning_rate": 1.2551020408163267e-05,
+      "loss": 0.4611,
+      "step": 18732
+    },
+    {
+      "epoch": 23.97824,
+      "grad_norm": 1.1586164236068726,
+      "learning_rate": 1.2549019607843138e-05,
+      "loss": 0.5029,
+      "step": 18733
+    },
+    {
+      "epoch": 23.97952,
+      "grad_norm": 1.1706522703170776,
+      "learning_rate": 1.2547018807523009e-05,
+      "loss": 0.4507,
+      "step": 18734
+    },
+    {
+      "epoch": 23.9808,
+      "grad_norm": 1.1484572887420654,
+      "learning_rate": 1.2545018007202882e-05,
+      "loss": 0.5009,
+      "step": 18735
+    },
+    {
+      "epoch": 23.98208,
+      "grad_norm": 1.1552718877792358,
+      "learning_rate": 1.2543017206882754e-05,
+      "loss": 0.4575,
+      "step": 18736
+    },
+    {
+      "epoch": 23.98336,
+      "grad_norm": 1.15525221824646,
+      "learning_rate": 1.2541016406562626e-05,
+      "loss": 0.5026,
+      "step": 18737
+    },
+    {
+      "epoch": 23.98464,
+      "grad_norm": 1.107552170753479,
+      "learning_rate": 1.2539015606242496e-05,
+      "loss": 0.4756,
+      "step": 18738
+    },
+    {
+      "epoch": 23.98592,
+      "grad_norm": 1.2044583559036255,
+      "learning_rate": 1.253701480592237e-05,
+      "loss": 0.4688,
+      "step": 18739
+    },
+    {
+      "epoch": 23.9872,
+      "grad_norm": 1.1863468885421753,
+      "learning_rate": 1.2535014005602241e-05,
+      "loss": 0.4841,
+      "step": 18740
+    },
+    {
+      "epoch": 23.98848,
+      "grad_norm": 1.1284416913986206,
+      "learning_rate": 1.2533013205282113e-05,
+      "loss": 0.4416,
+      "step": 18741
+    },
+    {
+      "epoch": 23.98976,
+      "grad_norm": 1.1152154207229614,
+      "learning_rate": 1.2531012404961987e-05,
+      "loss": 0.4756,
+      "step": 18742
+    },
+    {
+      "epoch": 23.99104,
+      "grad_norm": 1.1176108121871948,
+      "learning_rate": 1.2529011604641857e-05,
+      "loss": 0.449,
+      "step": 18743
+    },
+    {
+      "epoch": 23.99232,
+      "grad_norm": 1.2353702783584595,
+      "learning_rate": 1.2527010804321729e-05,
+      "loss": 0.513,
+      "step": 18744
+    },
+    {
+      "epoch": 23.9936,
+      "grad_norm": 1.1160780191421509,
+      "learning_rate": 1.25250100040016e-05,
+      "loss": 0.4815,
+      "step": 18745
+    },
+    {
+      "epoch": 23.99488,
+      "grad_norm": 1.2114685773849487,
+      "learning_rate": 1.2523009203681474e-05,
+      "loss": 0.5117,
+      "step": 18746
+    },
+    {
+      "epoch": 23.99616,
+      "grad_norm": 1.226145625114441,
+      "learning_rate": 1.2521008403361344e-05,
+      "loss": 0.5135,
+      "step": 18747
+    },
+    {
+      "epoch": 23.99744,
+      "grad_norm": 1.1571670770645142,
+      "learning_rate": 1.2519007603041216e-05,
+      "loss": 0.4824,
+      "step": 18748
+    },
+    {
+      "epoch": 23.99872,
+      "grad_norm": 1.0984230041503906,
+      "learning_rate": 1.251700680272109e-05,
+      "loss": 0.4727,
+      "step": 18749
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 2.419872283935547,
+      "learning_rate": 1.2515006002400962e-05,
+      "loss": 0.8476,
+      "step": 18750
+    },
+    {
+      "epoch": 24.00128,
+      "grad_norm": 1.1103284358978271,
+      "learning_rate": 1.2513005202080832e-05,
+      "loss": 0.5034,
+      "step": 18751
+    },
+    {
+      "epoch": 24.00256,
+      "grad_norm": 1.077684760093689,
+      "learning_rate": 1.2511004401760704e-05,
+      "loss": 0.4662,
+      "step": 18752
+    },
+    {
+      "epoch": 24.00384,
+      "grad_norm": 1.0543673038482666,
+      "learning_rate": 1.2509003601440577e-05,
+      "loss": 0.4556,
+      "step": 18753
+    },
+    {
+      "epoch": 24.00512,
+      "grad_norm": 1.18900728225708,
+      "learning_rate": 1.2507002801120449e-05,
+      "loss": 0.5031,
+      "step": 18754
+    },
+    {
+      "epoch": 24.0064,
+      "grad_norm": 1.0953742265701294,
+      "learning_rate": 1.250500200080032e-05,
+      "loss": 0.4672,
+      "step": 18755
+    },
+    {
+      "epoch": 24.00768,
+      "grad_norm": 1.1897867918014526,
+      "learning_rate": 1.2503001200480195e-05,
+      "loss": 0.4988,
+      "step": 18756
+    },
+    {
+      "epoch": 24.00896,
+      "grad_norm": 1.134912371635437,
+      "learning_rate": 1.2501000400160065e-05,
+      "loss": 0.5086,
+      "step": 18757
+    },
+    {
+      "epoch": 24.01024,
+      "grad_norm": 1.1343998908996582,
+      "learning_rate": 1.2498999599839937e-05,
+      "loss": 0.4737,
+      "step": 18758
+    },
+    {
+      "epoch": 24.01152,
+      "grad_norm": 1.0810625553131104,
+      "learning_rate": 1.2496998799519808e-05,
+      "loss": 0.4678,
+      "step": 18759
+    },
+    {
+      "epoch": 24.0128,
+      "grad_norm": 1.1055142879486084,
+      "learning_rate": 1.249499799919968e-05,
+      "loss": 0.459,
+      "step": 18760
+    },
+    {
+      "epoch": 24.01408,
+      "grad_norm": 1.1549721956253052,
+      "learning_rate": 1.2492997198879552e-05,
+      "loss": 0.4943,
+      "step": 18761
+    },
+    {
+      "epoch": 24.01536,
+      "grad_norm": 1.121549129486084,
+      "learning_rate": 1.2490996398559426e-05,
+      "loss": 0.4437,
+      "step": 18762
+    },
+    {
+      "epoch": 24.01664,
+      "grad_norm": 1.0666358470916748,
+      "learning_rate": 1.2488995598239296e-05,
+      "loss": 0.4201,
+      "step": 18763
+    },
+    {
+      "epoch": 24.01792,
+      "grad_norm": 1.134900689125061,
+      "learning_rate": 1.248699479791917e-05,
+      "loss": 0.4853,
+      "step": 18764
+    },
+    {
+      "epoch": 24.0192,
+      "grad_norm": 1.1466809511184692,
+      "learning_rate": 1.248499399759904e-05,
+      "loss": 0.4803,
+      "step": 18765
+    },
+    {
+      "epoch": 24.02048,
+      "grad_norm": 1.1177642345428467,
+      "learning_rate": 1.2482993197278913e-05,
+      "loss": 0.4439,
+      "step": 18766
+    },
+    {
+      "epoch": 24.02176,
+      "grad_norm": 1.1064014434814453,
+      "learning_rate": 1.2480992396958783e-05,
+      "loss": 0.4562,
+      "step": 18767
+    },
+    {
+      "epoch": 24.02304,
+      "grad_norm": 1.103027105331421,
+      "learning_rate": 1.2478991596638657e-05,
+      "loss": 0.4495,
+      "step": 18768
+    },
+    {
+      "epoch": 24.02432,
+      "grad_norm": 1.165516972541809,
+      "learning_rate": 1.2476990796318529e-05,
+      "loss": 0.4858,
+      "step": 18769
+    },
+    {
+      "epoch": 24.0256,
+      "grad_norm": 1.037825107574463,
+      "learning_rate": 1.24749899959984e-05,
+      "loss": 0.4153,
+      "step": 18770
+    },
+    {
+      "epoch": 24.02688,
+      "grad_norm": 1.0688804388046265,
+      "learning_rate": 1.2472989195678272e-05,
+      "loss": 0.4368,
+      "step": 18771
+    },
+    {
+      "epoch": 24.02816,
+      "grad_norm": 1.1697540283203125,
+      "learning_rate": 1.2470988395358144e-05,
+      "loss": 0.4791,
+      "step": 18772
+    },
+    {
+      "epoch": 24.02944,
+      "grad_norm": 1.1564234495162964,
+      "learning_rate": 1.2468987595038016e-05,
+      "loss": 0.4714,
+      "step": 18773
+    },
+    {
+      "epoch": 24.03072,
+      "grad_norm": 1.1580915451049805,
+      "learning_rate": 1.2466986794717888e-05,
+      "loss": 0.4786,
+      "step": 18774
+    },
+    {
+      "epoch": 24.032,
+      "grad_norm": 1.172911286354065,
+      "learning_rate": 1.246498599439776e-05,
+      "loss": 0.4536,
+      "step": 18775
+    },
+    {
+      "epoch": 24.03328,
+      "grad_norm": 1.1058646440505981,
+      "learning_rate": 1.2462985194077632e-05,
+      "loss": 0.436,
+      "step": 18776
+    },
+    {
+      "epoch": 24.03456,
+      "grad_norm": 1.1083683967590332,
+      "learning_rate": 1.2460984393757503e-05,
+      "loss": 0.4947,
+      "step": 18777
+    },
+    {
+      "epoch": 24.03584,
+      "grad_norm": 1.1440143585205078,
+      "learning_rate": 1.2458983593437375e-05,
+      "loss": 0.4808,
+      "step": 18778
+    },
+    {
+      "epoch": 24.03712,
+      "grad_norm": 1.11277174949646,
+      "learning_rate": 1.2456982793117247e-05,
+      "loss": 0.4342,
+      "step": 18779
+    },
+    {
+      "epoch": 24.0384,
+      "grad_norm": 1.054457426071167,
+      "learning_rate": 1.2454981992797119e-05,
+      "loss": 0.4136,
+      "step": 18780
+    },
+    {
+      "epoch": 24.03968,
+      "grad_norm": 1.1549383401870728,
+      "learning_rate": 1.2452981192476991e-05,
+      "loss": 0.4787,
+      "step": 18781
+    },
+    {
+      "epoch": 24.04096,
+      "grad_norm": 1.158250093460083,
+      "learning_rate": 1.2450980392156863e-05,
+      "loss": 0.4401,
+      "step": 18782
+    },
+    {
+      "epoch": 24.04224,
+      "grad_norm": 1.170989990234375,
+      "learning_rate": 1.2448979591836735e-05,
+      "loss": 0.4973,
+      "step": 18783
+    },
+    {
+      "epoch": 24.04352,
+      "grad_norm": 1.1259979009628296,
+      "learning_rate": 1.2446978791516606e-05,
+      "loss": 0.4472,
+      "step": 18784
+    },
+    {
+      "epoch": 24.0448,
+      "grad_norm": 1.147057056427002,
+      "learning_rate": 1.244497799119648e-05,
+      "loss": 0.5124,
+      "step": 18785
+    },
+    {
+      "epoch": 24.04608,
+      "grad_norm": 1.2855243682861328,
+      "learning_rate": 1.244297719087635e-05,
+      "loss": 0.543,
+      "step": 18786
+    },
+    {
+      "epoch": 24.04736,
+      "grad_norm": 1.1440112590789795,
+      "learning_rate": 1.2440976390556224e-05,
+      "loss": 0.4572,
+      "step": 18787
+    },
+    {
+      "epoch": 24.04864,
+      "grad_norm": 1.1109403371810913,
+      "learning_rate": 1.2438975590236094e-05,
+      "loss": 0.4864,
+      "step": 18788
+    },
+    {
+      "epoch": 24.04992,
+      "grad_norm": 1.129292368888855,
+      "learning_rate": 1.2436974789915967e-05,
+      "loss": 0.4461,
+      "step": 18789
+    },
+    {
+      "epoch": 24.0512,
+      "grad_norm": 1.091012716293335,
+      "learning_rate": 1.2434973989595838e-05,
+      "loss": 0.4387,
+      "step": 18790
+    },
+    {
+      "epoch": 24.05248,
+      "grad_norm": 1.1717545986175537,
+      "learning_rate": 1.2432973189275711e-05,
+      "loss": 0.461,
+      "step": 18791
+    },
+    {
+      "epoch": 24.05376,
+      "grad_norm": 1.1226803064346313,
+      "learning_rate": 1.2430972388955583e-05,
+      "loss": 0.4271,
+      "step": 18792
+    },
+    {
+      "epoch": 24.05504,
+      "grad_norm": 1.1514590978622437,
+      "learning_rate": 1.2428971588635455e-05,
+      "loss": 0.4739,
+      "step": 18793
+    },
+    {
+      "epoch": 24.05632,
+      "grad_norm": 1.157591700553894,
+      "learning_rate": 1.2426970788315327e-05,
+      "loss": 0.4825,
+      "step": 18794
+    },
+    {
+      "epoch": 24.0576,
+      "grad_norm": 1.17417311668396,
+      "learning_rate": 1.2424969987995199e-05,
+      "loss": 0.5034,
+      "step": 18795
+    },
+    {
+      "epoch": 24.05888,
+      "grad_norm": 1.0604926347732544,
+      "learning_rate": 1.242296918767507e-05,
+      "loss": 0.451,
+      "step": 18796
+    },
+    {
+      "epoch": 24.06016,
+      "grad_norm": 1.0973352193832397,
+      "learning_rate": 1.2420968387354942e-05,
+      "loss": 0.4632,
+      "step": 18797
+    },
+    {
+      "epoch": 24.06144,
+      "grad_norm": 1.1098352670669556,
+      "learning_rate": 1.2418967587034814e-05,
+      "loss": 0.4198,
+      "step": 18798
+    },
+    {
+      "epoch": 24.06272,
+      "grad_norm": 1.117893934249878,
+      "learning_rate": 1.2416966786714688e-05,
+      "loss": 0.4585,
+      "step": 18799
+    },
+    {
+      "epoch": 24.064,
+      "grad_norm": 1.1006295680999756,
+      "learning_rate": 1.2414965986394558e-05,
+      "loss": 0.474,
+      "step": 18800
+    },
+    {
+      "epoch": 24.06528,
+      "grad_norm": 1.1561734676361084,
+      "learning_rate": 1.2412965186074431e-05,
+      "loss": 0.4768,
+      "step": 18801
+    },
+    {
+      "epoch": 24.06656,
+      "grad_norm": 1.1429786682128906,
+      "learning_rate": 1.2410964385754302e-05,
+      "loss": 0.4457,
+      "step": 18802
+    },
+    {
+      "epoch": 24.06784,
+      "grad_norm": 1.1339454650878906,
+      "learning_rate": 1.2408963585434175e-05,
+      "loss": 0.4879,
+      "step": 18803
+    },
+    {
+      "epoch": 24.06912,
+      "grad_norm": 1.198805332183838,
+      "learning_rate": 1.2406962785114045e-05,
+      "loss": 0.5238,
+      "step": 18804
+    },
+    {
+      "epoch": 24.0704,
+      "grad_norm": 1.101431131362915,
+      "learning_rate": 1.2404961984793919e-05,
+      "loss": 0.4727,
+      "step": 18805
+    },
+    {
+      "epoch": 24.07168,
+      "grad_norm": 1.0786443948745728,
+      "learning_rate": 1.2402961184473789e-05,
+      "loss": 0.4213,
+      "step": 18806
+    },
+    {
+      "epoch": 24.07296,
+      "grad_norm": 1.1256358623504639,
+      "learning_rate": 1.2400960384153663e-05,
+      "loss": 0.4796,
+      "step": 18807
+    },
+    {
+      "epoch": 24.07424,
+      "grad_norm": 1.1662980318069458,
+      "learning_rate": 1.2398959583833534e-05,
+      "loss": 0.4701,
+      "step": 18808
+    },
+    {
+      "epoch": 24.07552,
+      "grad_norm": 1.2283040285110474,
+      "learning_rate": 1.2396958783513406e-05,
+      "loss": 0.5541,
+      "step": 18809
+    },
+    {
+      "epoch": 24.0768,
+      "grad_norm": 1.189424753189087,
+      "learning_rate": 1.2394957983193278e-05,
+      "loss": 0.4847,
+      "step": 18810
+    },
+    {
+      "epoch": 24.07808,
+      "grad_norm": 1.106939673423767,
+      "learning_rate": 1.239295718287315e-05,
+      "loss": 0.4491,
+      "step": 18811
+    },
+    {
+      "epoch": 24.07936,
+      "grad_norm": 1.059758186340332,
+      "learning_rate": 1.2390956382553022e-05,
+      "loss": 0.4485,
+      "step": 18812
+    },
+    {
+      "epoch": 24.08064,
+      "grad_norm": 1.2013357877731323,
+      "learning_rate": 1.2388955582232894e-05,
+      "loss": 0.502,
+      "step": 18813
+    },
+    {
+      "epoch": 24.08192,
+      "grad_norm": 1.1451191902160645,
+      "learning_rate": 1.2386954781912766e-05,
+      "loss": 0.4661,
+      "step": 18814
+    },
+    {
+      "epoch": 24.0832,
+      "grad_norm": 1.110666275024414,
+      "learning_rate": 1.2384953981592637e-05,
+      "loss": 0.453,
+      "step": 18815
+    },
+    {
+      "epoch": 24.08448,
+      "grad_norm": 1.1258022785186768,
+      "learning_rate": 1.238295318127251e-05,
+      "loss": 0.4725,
+      "step": 18816
+    },
+    {
+      "epoch": 24.08576,
+      "grad_norm": 1.1745010614395142,
+      "learning_rate": 1.2380952380952381e-05,
+      "loss": 0.4724,
+      "step": 18817
+    },
+    {
+      "epoch": 24.087040000000002,
+      "grad_norm": 1.1408393383026123,
+      "learning_rate": 1.2378951580632253e-05,
+      "loss": 0.485,
+      "step": 18818
+    },
+    {
+      "epoch": 24.08832,
+      "grad_norm": 1.2080360651016235,
+      "learning_rate": 1.2376950780312125e-05,
+      "loss": 0.5133,
+      "step": 18819
+    },
+    {
+      "epoch": 24.0896,
+      "grad_norm": 1.2680752277374268,
+      "learning_rate": 1.2374949979991997e-05,
+      "loss": 0.4442,
+      "step": 18820
+    },
+    {
+      "epoch": 24.09088,
+      "grad_norm": 1.0254223346710205,
+      "learning_rate": 1.2372949179671869e-05,
+      "loss": 0.4307,
+      "step": 18821
+    },
+    {
+      "epoch": 24.09216,
+      "grad_norm": 1.1265034675598145,
+      "learning_rate": 1.2370948379351742e-05,
+      "loss": 0.468,
+      "step": 18822
+    },
+    {
+      "epoch": 24.09344,
+      "grad_norm": 1.1478646993637085,
+      "learning_rate": 1.2368947579031612e-05,
+      "loss": 0.456,
+      "step": 18823
+    },
+    {
+      "epoch": 24.09472,
+      "grad_norm": 1.2165275812149048,
+      "learning_rate": 1.2366946778711486e-05,
+      "loss": 0.4712,
+      "step": 18824
+    },
+    {
+      "epoch": 24.096,
+      "grad_norm": 1.1693812608718872,
+      "learning_rate": 1.2364945978391356e-05,
+      "loss": 0.4956,
+      "step": 18825
+    },
+    {
+      "epoch": 24.09728,
+      "grad_norm": 1.1482276916503906,
+      "learning_rate": 1.236294517807123e-05,
+      "loss": 0.4687,
+      "step": 18826
+    },
+    {
+      "epoch": 24.09856,
+      "grad_norm": 1.1738667488098145,
+      "learning_rate": 1.23609443777511e-05,
+      "loss": 0.4096,
+      "step": 18827
+    },
+    {
+      "epoch": 24.09984,
+      "grad_norm": 1.1686372756958008,
+      "learning_rate": 1.2358943577430973e-05,
+      "loss": 0.4708,
+      "step": 18828
+    },
+    {
+      "epoch": 24.10112,
+      "grad_norm": 1.1600852012634277,
+      "learning_rate": 1.2356942777110843e-05,
+      "loss": 0.5038,
+      "step": 18829
+    },
+    {
+      "epoch": 24.1024,
+      "grad_norm": 1.1481022834777832,
+      "learning_rate": 1.2354941976790717e-05,
+      "loss": 0.4626,
+      "step": 18830
+    },
+    {
+      "epoch": 24.10368,
+      "grad_norm": 1.158562421798706,
+      "learning_rate": 1.2352941176470589e-05,
+      "loss": 0.4819,
+      "step": 18831
+    },
+    {
+      "epoch": 24.10496,
+      "grad_norm": 1.2289602756500244,
+      "learning_rate": 1.235094037615046e-05,
+      "loss": 0.4865,
+      "step": 18832
+    },
+    {
+      "epoch": 24.10624,
+      "grad_norm": 1.2404357194900513,
+      "learning_rate": 1.2348939575830333e-05,
+      "loss": 0.4778,
+      "step": 18833
+    },
+    {
+      "epoch": 24.10752,
+      "grad_norm": 1.0343823432922363,
+      "learning_rate": 1.2346938775510204e-05,
+      "loss": 0.4364,
+      "step": 18834
+    },
+    {
+      "epoch": 24.1088,
+      "grad_norm": 1.1348611116409302,
+      "learning_rate": 1.2344937975190076e-05,
+      "loss": 0.4825,
+      "step": 18835
+    },
+    {
+      "epoch": 24.11008,
+      "grad_norm": 1.1395764350891113,
+      "learning_rate": 1.2342937174869948e-05,
+      "loss": 0.4656,
+      "step": 18836
+    },
+    {
+      "epoch": 24.11136,
+      "grad_norm": 1.1095606088638306,
+      "learning_rate": 1.234093637454982e-05,
+      "loss": 0.4568,
+      "step": 18837
+    },
+    {
+      "epoch": 24.11264,
+      "grad_norm": 1.184478521347046,
+      "learning_rate": 1.2338935574229694e-05,
+      "loss": 0.4887,
+      "step": 18838
+    },
+    {
+      "epoch": 24.11392,
+      "grad_norm": 1.1172800064086914,
+      "learning_rate": 1.2336934773909564e-05,
+      "loss": 0.4558,
+      "step": 18839
+    },
+    {
+      "epoch": 24.1152,
+      "grad_norm": 1.2466896772384644,
+      "learning_rate": 1.2334933973589437e-05,
+      "loss": 0.5224,
+      "step": 18840
+    },
+    {
+      "epoch": 24.11648,
+      "grad_norm": 1.1181126832962036,
+      "learning_rate": 1.2332933173269307e-05,
+      "loss": 0.4622,
+      "step": 18841
+    },
+    {
+      "epoch": 24.11776,
+      "grad_norm": 1.1256959438323975,
+      "learning_rate": 1.2330932372949181e-05,
+      "loss": 0.4646,
+      "step": 18842
+    },
+    {
+      "epoch": 24.11904,
+      "grad_norm": 1.1790109872817993,
+      "learning_rate": 1.2328931572629051e-05,
+      "loss": 0.4708,
+      "step": 18843
+    },
+    {
+      "epoch": 24.12032,
+      "grad_norm": 1.1937501430511475,
+      "learning_rate": 1.2326930772308925e-05,
+      "loss": 0.4791,
+      "step": 18844
+    },
+    {
+      "epoch": 24.1216,
+      "grad_norm": 1.2300196886062622,
+      "learning_rate": 1.2324929971988797e-05,
+      "loss": 0.4995,
+      "step": 18845
+    },
+    {
+      "epoch": 24.12288,
+      "grad_norm": 1.1216380596160889,
+      "learning_rate": 1.2322929171668668e-05,
+      "loss": 0.4832,
+      "step": 18846
+    },
+    {
+      "epoch": 24.12416,
+      "grad_norm": 1.1680253744125366,
+      "learning_rate": 1.232092837134854e-05,
+      "loss": 0.4863,
+      "step": 18847
+    },
+    {
+      "epoch": 24.12544,
+      "grad_norm": 1.191275715827942,
+      "learning_rate": 1.2318927571028412e-05,
+      "loss": 0.4561,
+      "step": 18848
+    },
+    {
+      "epoch": 24.12672,
+      "grad_norm": 1.1473134756088257,
+      "learning_rate": 1.2316926770708284e-05,
+      "loss": 0.466,
+      "step": 18849
+    },
+    {
+      "epoch": 24.128,
+      "grad_norm": 1.181347370147705,
+      "learning_rate": 1.2314925970388156e-05,
+      "loss": 0.478,
+      "step": 18850
+    },
+    {
+      "epoch": 24.12928,
+      "grad_norm": 1.1565207242965698,
+      "learning_rate": 1.2312925170068028e-05,
+      "loss": 0.5205,
+      "step": 18851
+    },
+    {
+      "epoch": 24.13056,
+      "grad_norm": 1.136231541633606,
+      "learning_rate": 1.2310924369747901e-05,
+      "loss": 0.4984,
+      "step": 18852
+    },
+    {
+      "epoch": 24.13184,
+      "grad_norm": 1.1483358144760132,
+      "learning_rate": 1.2308923569427771e-05,
+      "loss": 0.4857,
+      "step": 18853
+    },
+    {
+      "epoch": 24.13312,
+      "grad_norm": 1.1635894775390625,
+      "learning_rate": 1.2306922769107645e-05,
+      "loss": 0.4728,
+      "step": 18854
+    },
+    {
+      "epoch": 24.1344,
+      "grad_norm": 1.1853755712509155,
+      "learning_rate": 1.2304921968787515e-05,
+      "loss": 0.4908,
+      "step": 18855
+    },
+    {
+      "epoch": 24.13568,
+      "grad_norm": 1.1830283403396606,
+      "learning_rate": 1.2302921168467389e-05,
+      "loss": 0.4935,
+      "step": 18856
+    },
+    {
+      "epoch": 24.13696,
+      "grad_norm": 1.1618894338607788,
+      "learning_rate": 1.2300920368147259e-05,
+      "loss": 0.4388,
+      "step": 18857
+    },
+    {
+      "epoch": 24.13824,
+      "grad_norm": 1.1691558361053467,
+      "learning_rate": 1.2298919567827132e-05,
+      "loss": 0.459,
+      "step": 18858
+    },
+    {
+      "epoch": 24.13952,
+      "grad_norm": 1.1906380653381348,
+      "learning_rate": 1.2296918767507003e-05,
+      "loss": 0.4695,
+      "step": 18859
+    },
+    {
+      "epoch": 24.1408,
+      "grad_norm": 1.0874199867248535,
+      "learning_rate": 1.2294917967186876e-05,
+      "loss": 0.4549,
+      "step": 18860
+    },
+    {
+      "epoch": 24.14208,
+      "grad_norm": 1.1504427194595337,
+      "learning_rate": 1.2292917166866748e-05,
+      "loss": 0.4616,
+      "step": 18861
+    },
+    {
+      "epoch": 24.14336,
+      "grad_norm": 1.1930842399597168,
+      "learning_rate": 1.229091636654662e-05,
+      "loss": 0.5141,
+      "step": 18862
+    },
+    {
+      "epoch": 24.14464,
+      "grad_norm": 1.1488827466964722,
+      "learning_rate": 1.2288915566226492e-05,
+      "loss": 0.4666,
+      "step": 18863
+    },
+    {
+      "epoch": 24.14592,
+      "grad_norm": 1.1404608488082886,
+      "learning_rate": 1.2286914765906364e-05,
+      "loss": 0.4479,
+      "step": 18864
+    },
+    {
+      "epoch": 24.1472,
+      "grad_norm": 1.1836938858032227,
+      "learning_rate": 1.2284913965586235e-05,
+      "loss": 0.5085,
+      "step": 18865
+    },
+    {
+      "epoch": 24.14848,
+      "grad_norm": 1.1096590757369995,
+      "learning_rate": 1.2282913165266107e-05,
+      "loss": 0.4589,
+      "step": 18866
+    },
+    {
+      "epoch": 24.14976,
+      "grad_norm": 1.1787697076797485,
+      "learning_rate": 1.2280912364945979e-05,
+      "loss": 0.4746,
+      "step": 18867
+    },
+    {
+      "epoch": 24.15104,
+      "grad_norm": 1.0832207202911377,
+      "learning_rate": 1.2278911564625851e-05,
+      "loss": 0.4693,
+      "step": 18868
+    },
+    {
+      "epoch": 24.15232,
+      "grad_norm": 1.1375176906585693,
+      "learning_rate": 1.2276910764305723e-05,
+      "loss": 0.4311,
+      "step": 18869
+    },
+    {
+      "epoch": 24.1536,
+      "grad_norm": 1.1903811693191528,
+      "learning_rate": 1.2274909963985595e-05,
+      "loss": 0.4851,
+      "step": 18870
+    },
+    {
+      "epoch": 24.15488,
+      "grad_norm": 1.169784665107727,
+      "learning_rate": 1.2272909163665467e-05,
+      "loss": 0.495,
+      "step": 18871
+    },
+    {
+      "epoch": 24.15616,
+      "grad_norm": 1.154515027999878,
+      "learning_rate": 1.2270908363345338e-05,
+      "loss": 0.4999,
+      "step": 18872
+    },
+    {
+      "epoch": 24.15744,
+      "grad_norm": 1.141190528869629,
+      "learning_rate": 1.226890756302521e-05,
+      "loss": 0.4566,
+      "step": 18873
+    },
+    {
+      "epoch": 24.15872,
+      "grad_norm": 1.1099512577056885,
+      "learning_rate": 1.2266906762705082e-05,
+      "loss": 0.3997,
+      "step": 18874
+    },
+    {
+      "epoch": 24.16,
+      "grad_norm": 1.129675030708313,
+      "learning_rate": 1.2264905962384956e-05,
+      "loss": 0.4474,
+      "step": 18875
+    },
+    {
+      "epoch": 24.16128,
+      "grad_norm": 1.2040431499481201,
+      "learning_rate": 1.2262905162064826e-05,
+      "loss": 0.4952,
+      "step": 18876
+    },
+    {
+      "epoch": 24.16256,
+      "grad_norm": 1.1369742155075073,
+      "learning_rate": 1.22609043617447e-05,
+      "loss": 0.4785,
+      "step": 18877
+    },
+    {
+      "epoch": 24.16384,
+      "grad_norm": 1.1882468461990356,
+      "learning_rate": 1.225890356142457e-05,
+      "loss": 0.4668,
+      "step": 18878
+    },
+    {
+      "epoch": 24.16512,
+      "grad_norm": 1.1586542129516602,
+      "learning_rate": 1.2256902761104443e-05,
+      "loss": 0.4616,
+      "step": 18879
+    },
+    {
+      "epoch": 24.1664,
+      "grad_norm": 1.1391338109970093,
+      "learning_rate": 1.2254901960784313e-05,
+      "loss": 0.4558,
+      "step": 18880
+    },
+    {
+      "epoch": 24.16768,
+      "grad_norm": 1.2286574840545654,
+      "learning_rate": 1.2252901160464187e-05,
+      "loss": 0.5128,
+      "step": 18881
+    },
+    {
+      "epoch": 24.16896,
+      "grad_norm": 1.1499093770980835,
+      "learning_rate": 1.2250900360144059e-05,
+      "loss": 0.4686,
+      "step": 18882
+    },
+    {
+      "epoch": 24.17024,
+      "grad_norm": 1.1170378923416138,
+      "learning_rate": 1.224889955982393e-05,
+      "loss": 0.5008,
+      "step": 18883
+    },
+    {
+      "epoch": 24.17152,
+      "grad_norm": 1.1456178426742554,
+      "learning_rate": 1.2246898759503802e-05,
+      "loss": 0.4888,
+      "step": 18884
+    },
+    {
+      "epoch": 24.1728,
+      "grad_norm": 1.1081149578094482,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 0.4932,
+      "step": 18885
+    },
+    {
+      "epoch": 24.17408,
+      "grad_norm": 1.250626564025879,
+      "learning_rate": 1.2242897158863546e-05,
+      "loss": 0.4693,
+      "step": 18886
+    },
+    {
+      "epoch": 24.17536,
+      "grad_norm": 1.2053821086883545,
+      "learning_rate": 1.2240896358543418e-05,
+      "loss": 0.5102,
+      "step": 18887
+    },
+    {
+      "epoch": 24.17664,
+      "grad_norm": 1.1828999519348145,
+      "learning_rate": 1.223889555822329e-05,
+      "loss": 0.4694,
+      "step": 18888
+    },
+    {
+      "epoch": 24.17792,
+      "grad_norm": 1.1308258771896362,
+      "learning_rate": 1.2236894757903162e-05,
+      "loss": 0.4473,
+      "step": 18889
+    },
+    {
+      "epoch": 24.1792,
+      "grad_norm": 1.1865490674972534,
+      "learning_rate": 1.2234893957583033e-05,
+      "loss": 0.5123,
+      "step": 18890
+    },
+    {
+      "epoch": 24.18048,
+      "grad_norm": 1.1763098239898682,
+      "learning_rate": 1.2232893157262907e-05,
+      "loss": 0.5011,
+      "step": 18891
+    },
+    {
+      "epoch": 24.18176,
+      "grad_norm": 1.1184165477752686,
+      "learning_rate": 1.2230892356942777e-05,
+      "loss": 0.4935,
+      "step": 18892
+    },
+    {
+      "epoch": 24.18304,
+      "grad_norm": 1.2041736841201782,
+      "learning_rate": 1.222889155662265e-05,
+      "loss": 0.5107,
+      "step": 18893
+    },
+    {
+      "epoch": 24.18432,
+      "grad_norm": 1.1142528057098389,
+      "learning_rate": 1.2226890756302521e-05,
+      "loss": 0.4647,
+      "step": 18894
+    },
+    {
+      "epoch": 24.1856,
+      "grad_norm": 1.1789337396621704,
+      "learning_rate": 1.2224889955982394e-05,
+      "loss": 0.4476,
+      "step": 18895
+    },
+    {
+      "epoch": 24.18688,
+      "grad_norm": 1.0955803394317627,
+      "learning_rate": 1.2222889155662265e-05,
+      "loss": 0.4594,
+      "step": 18896
+    },
+    {
+      "epoch": 24.18816,
+      "grad_norm": 1.1010346412658691,
+      "learning_rate": 1.2220888355342138e-05,
+      "loss": 0.4668,
+      "step": 18897
+    },
+    {
+      "epoch": 24.18944,
+      "grad_norm": 1.0665899515151978,
+      "learning_rate": 1.221888755502201e-05,
+      "loss": 0.4563,
+      "step": 18898
+    },
+    {
+      "epoch": 24.19072,
+      "grad_norm": 1.1239908933639526,
+      "learning_rate": 1.2216886754701882e-05,
+      "loss": 0.4785,
+      "step": 18899
+    },
+    {
+      "epoch": 24.192,
+      "grad_norm": 1.1581279039382935,
+      "learning_rate": 1.2214885954381754e-05,
+      "loss": 0.501,
+      "step": 18900
+    },
+    {
+      "epoch": 24.19328,
+      "grad_norm": 1.1129977703094482,
+      "learning_rate": 1.2212885154061626e-05,
+      "loss": 0.4571,
+      "step": 18901
+    },
+    {
+      "epoch": 24.19456,
+      "grad_norm": 1.1460938453674316,
+      "learning_rate": 1.2210884353741497e-05,
+      "loss": 0.4706,
+      "step": 18902
+    },
+    {
+      "epoch": 24.19584,
+      "grad_norm": 1.2374320030212402,
+      "learning_rate": 1.220888355342137e-05,
+      "loss": 0.5082,
+      "step": 18903
+    },
+    {
+      "epoch": 24.19712,
+      "grad_norm": 1.1086463928222656,
+      "learning_rate": 1.2206882753101241e-05,
+      "loss": 0.4465,
+      "step": 18904
+    },
+    {
+      "epoch": 24.1984,
+      "grad_norm": 1.147619605064392,
+      "learning_rate": 1.2204881952781113e-05,
+      "loss": 0.4718,
+      "step": 18905
+    },
+    {
+      "epoch": 24.19968,
+      "grad_norm": 1.117663025856018,
+      "learning_rate": 1.2202881152460985e-05,
+      "loss": 0.5058,
+      "step": 18906
+    },
+    {
+      "epoch": 24.20096,
+      "grad_norm": 1.1890056133270264,
+      "learning_rate": 1.2200880352140857e-05,
+      "loss": 0.4962,
+      "step": 18907
+    },
+    {
+      "epoch": 24.20224,
+      "grad_norm": 1.0809297561645508,
+      "learning_rate": 1.2198879551820729e-05,
+      "loss": 0.4143,
+      "step": 18908
+    },
+    {
+      "epoch": 24.20352,
+      "grad_norm": 1.146008014678955,
+      "learning_rate": 1.21968787515006e-05,
+      "loss": 0.4672,
+      "step": 18909
+    },
+    {
+      "epoch": 24.2048,
+      "grad_norm": 1.1988301277160645,
+      "learning_rate": 1.2194877951180472e-05,
+      "loss": 0.4645,
+      "step": 18910
+    },
+    {
+      "epoch": 24.20608,
+      "grad_norm": 1.2274937629699707,
+      "learning_rate": 1.2192877150860344e-05,
+      "loss": 0.5402,
+      "step": 18911
+    },
+    {
+      "epoch": 24.20736,
+      "grad_norm": 1.1610702276229858,
+      "learning_rate": 1.2190876350540216e-05,
+      "loss": 0.5225,
+      "step": 18912
+    },
+    {
+      "epoch": 24.20864,
+      "grad_norm": 1.2057294845581055,
+      "learning_rate": 1.2188875550220088e-05,
+      "loss": 0.4622,
+      "step": 18913
+    },
+    {
+      "epoch": 24.20992,
+      "grad_norm": 1.1343713998794556,
+      "learning_rate": 1.2186874749899961e-05,
+      "loss": 0.4997,
+      "step": 18914
+    },
+    {
+      "epoch": 24.2112,
+      "grad_norm": 1.0980840921401978,
+      "learning_rate": 1.2184873949579832e-05,
+      "loss": 0.4442,
+      "step": 18915
+    },
+    {
+      "epoch": 24.21248,
+      "grad_norm": 1.099935531616211,
+      "learning_rate": 1.2182873149259705e-05,
+      "loss": 0.4833,
+      "step": 18916
+    },
+    {
+      "epoch": 24.21376,
+      "grad_norm": 1.2026798725128174,
+      "learning_rate": 1.2180872348939575e-05,
+      "loss": 0.4885,
+      "step": 18917
+    },
+    {
+      "epoch": 24.21504,
+      "grad_norm": 1.138269066810608,
+      "learning_rate": 1.2178871548619449e-05,
+      "loss": 0.4539,
+      "step": 18918
+    },
+    {
+      "epoch": 24.21632,
+      "grad_norm": 1.233717441558838,
+      "learning_rate": 1.2176870748299319e-05,
+      "loss": 0.5163,
+      "step": 18919
+    },
+    {
+      "epoch": 24.2176,
+      "grad_norm": 1.2078992128372192,
+      "learning_rate": 1.2174869947979193e-05,
+      "loss": 0.4496,
+      "step": 18920
+    },
+    {
+      "epoch": 24.21888,
+      "grad_norm": 1.226100206375122,
+      "learning_rate": 1.2172869147659064e-05,
+      "loss": 0.511,
+      "step": 18921
+    },
+    {
+      "epoch": 24.22016,
+      "grad_norm": 1.183384895324707,
+      "learning_rate": 1.2170868347338936e-05,
+      "loss": 0.4633,
+      "step": 18922
+    },
+    {
+      "epoch": 24.22144,
+      "grad_norm": 1.2039517164230347,
+      "learning_rate": 1.2168867547018808e-05,
+      "loss": 0.5063,
+      "step": 18923
+    },
+    {
+      "epoch": 24.22272,
+      "grad_norm": 1.2347691059112549,
+      "learning_rate": 1.216686674669868e-05,
+      "loss": 0.4843,
+      "step": 18924
+    },
+    {
+      "epoch": 24.224,
+      "grad_norm": 1.2100071907043457,
+      "learning_rate": 1.2164865946378552e-05,
+      "loss": 0.4907,
+      "step": 18925
+    },
+    {
+      "epoch": 24.22528,
+      "grad_norm": 1.1934117078781128,
+      "learning_rate": 1.2162865146058424e-05,
+      "loss": 0.4824,
+      "step": 18926
+    },
+    {
+      "epoch": 24.22656,
+      "grad_norm": 1.1818937063217163,
+      "learning_rate": 1.2160864345738296e-05,
+      "loss": 0.4746,
+      "step": 18927
+    },
+    {
+      "epoch": 24.22784,
+      "grad_norm": 1.0990982055664062,
+      "learning_rate": 1.2158863545418169e-05,
+      "loss": 0.4461,
+      "step": 18928
+    },
+    {
+      "epoch": 24.22912,
+      "grad_norm": 1.1781893968582153,
+      "learning_rate": 1.215686274509804e-05,
+      "loss": 0.5018,
+      "step": 18929
+    },
+    {
+      "epoch": 24.2304,
+      "grad_norm": 1.147173523902893,
+      "learning_rate": 1.2154861944777913e-05,
+      "loss": 0.4664,
+      "step": 18930
+    },
+    {
+      "epoch": 24.23168,
+      "grad_norm": 1.0861092805862427,
+      "learning_rate": 1.2152861144457783e-05,
+      "loss": 0.4398,
+      "step": 18931
+    },
+    {
+      "epoch": 24.23296,
+      "grad_norm": 1.103131651878357,
+      "learning_rate": 1.2150860344137657e-05,
+      "loss": 0.4301,
+      "step": 18932
+    },
+    {
+      "epoch": 24.23424,
+      "grad_norm": 1.0709394216537476,
+      "learning_rate": 1.2148859543817527e-05,
+      "loss": 0.4424,
+      "step": 18933
+    },
+    {
+      "epoch": 24.23552,
+      "grad_norm": 1.1708987951278687,
+      "learning_rate": 1.21468587434974e-05,
+      "loss": 0.4815,
+      "step": 18934
+    },
+    {
+      "epoch": 24.2368,
+      "grad_norm": 1.1951818466186523,
+      "learning_rate": 1.2144857943177272e-05,
+      "loss": 0.4715,
+      "step": 18935
+    },
+    {
+      "epoch": 24.23808,
+      "grad_norm": 1.169890284538269,
+      "learning_rate": 1.2142857142857144e-05,
+      "loss": 0.4762,
+      "step": 18936
+    },
+    {
+      "epoch": 24.23936,
+      "grad_norm": 1.164069652557373,
+      "learning_rate": 1.2140856342537016e-05,
+      "loss": 0.4739,
+      "step": 18937
+    },
+    {
+      "epoch": 24.24064,
+      "grad_norm": 1.2312345504760742,
+      "learning_rate": 1.2138855542216888e-05,
+      "loss": 0.5088,
+      "step": 18938
+    },
+    {
+      "epoch": 24.24192,
+      "grad_norm": 1.0831817388534546,
+      "learning_rate": 1.213685474189676e-05,
+      "loss": 0.4128,
+      "step": 18939
+    },
+    {
+      "epoch": 24.2432,
+      "grad_norm": 1.1156847476959229,
+      "learning_rate": 1.2134853941576631e-05,
+      "loss": 0.4483,
+      "step": 18940
+    },
+    {
+      "epoch": 24.24448,
+      "grad_norm": 1.2278788089752197,
+      "learning_rate": 1.2132853141256503e-05,
+      "loss": 0.5139,
+      "step": 18941
+    },
+    {
+      "epoch": 24.24576,
+      "grad_norm": 1.1557207107543945,
+      "learning_rate": 1.2130852340936375e-05,
+      "loss": 0.5051,
+      "step": 18942
+    },
+    {
+      "epoch": 24.24704,
+      "grad_norm": 1.1289258003234863,
+      "learning_rate": 1.2128851540616247e-05,
+      "loss": 0.473,
+      "step": 18943
+    },
+    {
+      "epoch": 24.24832,
+      "grad_norm": 1.245375633239746,
+      "learning_rate": 1.2126850740296119e-05,
+      "loss": 0.4962,
+      "step": 18944
+    },
+    {
+      "epoch": 24.2496,
+      "grad_norm": 1.1361275911331177,
+      "learning_rate": 1.212484993997599e-05,
+      "loss": 0.4961,
+      "step": 18945
+    },
+    {
+      "epoch": 24.25088,
+      "grad_norm": 1.0928107500076294,
+      "learning_rate": 1.2122849139655863e-05,
+      "loss": 0.4831,
+      "step": 18946
+    },
+    {
+      "epoch": 24.25216,
+      "grad_norm": 1.0881211757659912,
+      "learning_rate": 1.2120848339335734e-05,
+      "loss": 0.4388,
+      "step": 18947
+    },
+    {
+      "epoch": 24.25344,
+      "grad_norm": 1.1172457933425903,
+      "learning_rate": 1.2118847539015606e-05,
+      "loss": 0.4584,
+      "step": 18948
+    },
+    {
+      "epoch": 24.25472,
+      "grad_norm": 1.1269158124923706,
+      "learning_rate": 1.2116846738695478e-05,
+      "loss": 0.4362,
+      "step": 18949
+    },
+    {
+      "epoch": 24.256,
+      "grad_norm": 1.0757176876068115,
+      "learning_rate": 1.211484593837535e-05,
+      "loss": 0.4515,
+      "step": 18950
+    },
+    {
+      "epoch": 24.25728,
+      "grad_norm": 1.0964628458023071,
+      "learning_rate": 1.2112845138055224e-05,
+      "loss": 0.4624,
+      "step": 18951
+    },
+    {
+      "epoch": 24.25856,
+      "grad_norm": 1.1697242259979248,
+      "learning_rate": 1.2110844337735094e-05,
+      "loss": 0.4753,
+      "step": 18952
+    },
+    {
+      "epoch": 24.25984,
+      "grad_norm": 1.098563313484192,
+      "learning_rate": 1.2108843537414967e-05,
+      "loss": 0.4689,
+      "step": 18953
+    },
+    {
+      "epoch": 24.26112,
+      "grad_norm": 1.081687092781067,
+      "learning_rate": 1.2106842737094837e-05,
+      "loss": 0.467,
+      "step": 18954
+    },
+    {
+      "epoch": 24.2624,
+      "grad_norm": 1.130462884902954,
+      "learning_rate": 1.2104841936774711e-05,
+      "loss": 0.4772,
+      "step": 18955
+    },
+    {
+      "epoch": 24.26368,
+      "grad_norm": 1.1619678735733032,
+      "learning_rate": 1.2102841136454581e-05,
+      "loss": 0.4805,
+      "step": 18956
+    },
+    {
+      "epoch": 24.26496,
+      "grad_norm": 1.2183215618133545,
+      "learning_rate": 1.2100840336134455e-05,
+      "loss": 0.4988,
+      "step": 18957
+    },
+    {
+      "epoch": 24.26624,
+      "grad_norm": 1.1362768411636353,
+      "learning_rate": 1.2098839535814327e-05,
+      "loss": 0.4678,
+      "step": 18958
+    },
+    {
+      "epoch": 24.26752,
+      "grad_norm": 1.1488263607025146,
+      "learning_rate": 1.2096838735494198e-05,
+      "loss": 0.4232,
+      "step": 18959
+    },
+    {
+      "epoch": 24.2688,
+      "grad_norm": 1.1357872486114502,
+      "learning_rate": 1.209483793517407e-05,
+      "loss": 0.4901,
+      "step": 18960
+    },
+    {
+      "epoch": 24.27008,
+      "grad_norm": 1.1027874946594238,
+      "learning_rate": 1.2092837134853942e-05,
+      "loss": 0.4982,
+      "step": 18961
+    },
+    {
+      "epoch": 24.27136,
+      "grad_norm": 1.157282829284668,
+      "learning_rate": 1.2090836334533814e-05,
+      "loss": 0.4774,
+      "step": 18962
+    },
+    {
+      "epoch": 24.27264,
+      "grad_norm": 1.1591331958770752,
+      "learning_rate": 1.2088835534213686e-05,
+      "loss": 0.4684,
+      "step": 18963
+    },
+    {
+      "epoch": 24.27392,
+      "grad_norm": 1.0676262378692627,
+      "learning_rate": 1.2086834733893558e-05,
+      "loss": 0.4187,
+      "step": 18964
+    },
+    {
+      "epoch": 24.2752,
+      "grad_norm": 1.143548846244812,
+      "learning_rate": 1.2084833933573431e-05,
+      "loss": 0.4558,
+      "step": 18965
+    },
+    {
+      "epoch": 24.27648,
+      "grad_norm": 1.1962380409240723,
+      "learning_rate": 1.2082833133253301e-05,
+      "loss": 0.5233,
+      "step": 18966
+    },
+    {
+      "epoch": 24.27776,
+      "grad_norm": 1.1192182302474976,
+      "learning_rate": 1.2080832332933175e-05,
+      "loss": 0.4541,
+      "step": 18967
+    },
+    {
+      "epoch": 24.27904,
+      "grad_norm": 1.1944539546966553,
+      "learning_rate": 1.2078831532613045e-05,
+      "loss": 0.4669,
+      "step": 18968
+    },
+    {
+      "epoch": 24.28032,
+      "grad_norm": 1.0787324905395508,
+      "learning_rate": 1.2076830732292919e-05,
+      "loss": 0.4497,
+      "step": 18969
+    },
+    {
+      "epoch": 24.2816,
+      "grad_norm": 1.12637197971344,
+      "learning_rate": 1.2074829931972789e-05,
+      "loss": 0.4599,
+      "step": 18970
+    },
+    {
+      "epoch": 24.28288,
+      "grad_norm": 1.1529687643051147,
+      "learning_rate": 1.2072829131652662e-05,
+      "loss": 0.5066,
+      "step": 18971
+    },
+    {
+      "epoch": 24.28416,
+      "grad_norm": 1.202721357345581,
+      "learning_rate": 1.2070828331332533e-05,
+      "loss": 0.494,
+      "step": 18972
+    },
+    {
+      "epoch": 24.28544,
+      "grad_norm": 1.1301331520080566,
+      "learning_rate": 1.2068827531012406e-05,
+      "loss": 0.4652,
+      "step": 18973
+    },
+    {
+      "epoch": 24.28672,
+      "grad_norm": 1.128435492515564,
+      "learning_rate": 1.2066826730692278e-05,
+      "loss": 0.4296,
+      "step": 18974
+    },
+    {
+      "epoch": 24.288,
+      "grad_norm": 1.1200505495071411,
+      "learning_rate": 1.206482593037215e-05,
+      "loss": 0.474,
+      "step": 18975
+    },
+    {
+      "epoch": 24.28928,
+      "grad_norm": 1.2111964225769043,
+      "learning_rate": 1.2062825130052022e-05,
+      "loss": 0.4765,
+      "step": 18976
+    },
+    {
+      "epoch": 24.29056,
+      "grad_norm": 1.1505967378616333,
+      "learning_rate": 1.2060824329731893e-05,
+      "loss": 0.4658,
+      "step": 18977
+    },
+    {
+      "epoch": 24.29184,
+      "grad_norm": 1.0870449542999268,
+      "learning_rate": 1.2058823529411765e-05,
+      "loss": 0.4378,
+      "step": 18978
+    },
+    {
+      "epoch": 24.29312,
+      "grad_norm": 1.1326916217803955,
+      "learning_rate": 1.2056822729091637e-05,
+      "loss": 0.4744,
+      "step": 18979
+    },
+    {
+      "epoch": 24.2944,
+      "grad_norm": 1.1708118915557861,
+      "learning_rate": 1.2054821928771509e-05,
+      "loss": 0.4676,
+      "step": 18980
+    },
+    {
+      "epoch": 24.29568,
+      "grad_norm": 1.1142549514770508,
+      "learning_rate": 1.2052821128451381e-05,
+      "loss": 0.4682,
+      "step": 18981
+    },
+    {
+      "epoch": 24.29696,
+      "grad_norm": 1.2015085220336914,
+      "learning_rate": 1.2050820328131253e-05,
+      "loss": 0.5163,
+      "step": 18982
+    },
+    {
+      "epoch": 24.29824,
+      "grad_norm": 1.0846655368804932,
+      "learning_rate": 1.2048819527811125e-05,
+      "loss": 0.411,
+      "step": 18983
+    },
+    {
+      "epoch": 24.29952,
+      "grad_norm": 1.1354156732559204,
+      "learning_rate": 1.2046818727490996e-05,
+      "loss": 0.4529,
+      "step": 18984
+    },
+    {
+      "epoch": 24.3008,
+      "grad_norm": 1.1821706295013428,
+      "learning_rate": 1.2044817927170868e-05,
+      "loss": 0.4717,
+      "step": 18985
+    },
+    {
+      "epoch": 24.30208,
+      "grad_norm": 1.1429351568222046,
+      "learning_rate": 1.204281712685074e-05,
+      "loss": 0.4679,
+      "step": 18986
+    },
+    {
+      "epoch": 24.30336,
+      "grad_norm": 1.1987007856369019,
+      "learning_rate": 1.2040816326530612e-05,
+      "loss": 0.479,
+      "step": 18987
+    },
+    {
+      "epoch": 24.30464,
+      "grad_norm": 1.1858272552490234,
+      "learning_rate": 1.2038815526210486e-05,
+      "loss": 0.4528,
+      "step": 18988
+    },
+    {
+      "epoch": 24.30592,
+      "grad_norm": 1.1735626459121704,
+      "learning_rate": 1.2036814725890356e-05,
+      "loss": 0.4459,
+      "step": 18989
+    },
+    {
+      "epoch": 24.3072,
+      "grad_norm": 1.2804443836212158,
+      "learning_rate": 1.203481392557023e-05,
+      "loss": 0.5244,
+      "step": 18990
+    },
+    {
+      "epoch": 24.30848,
+      "grad_norm": 1.2020955085754395,
+      "learning_rate": 1.20328131252501e-05,
+      "loss": 0.4598,
+      "step": 18991
+    },
+    {
+      "epoch": 24.30976,
+      "grad_norm": 1.2442394495010376,
+      "learning_rate": 1.2030812324929973e-05,
+      "loss": 0.4843,
+      "step": 18992
+    },
+    {
+      "epoch": 24.31104,
+      "grad_norm": 1.137221097946167,
+      "learning_rate": 1.2028811524609843e-05,
+      "loss": 0.4527,
+      "step": 18993
+    },
+    {
+      "epoch": 24.31232,
+      "grad_norm": 1.1395703554153442,
+      "learning_rate": 1.2026810724289717e-05,
+      "loss": 0.5018,
+      "step": 18994
+    },
+    {
+      "epoch": 24.3136,
+      "grad_norm": 1.1961019039154053,
+      "learning_rate": 1.2024809923969587e-05,
+      "loss": 0.5364,
+      "step": 18995
+    },
+    {
+      "epoch": 24.31488,
+      "grad_norm": 1.1549445390701294,
+      "learning_rate": 1.202280912364946e-05,
+      "loss": 0.4855,
+      "step": 18996
+    },
+    {
+      "epoch": 24.31616,
+      "grad_norm": 1.1084400415420532,
+      "learning_rate": 1.2020808323329332e-05,
+      "loss": 0.4317,
+      "step": 18997
+    },
+    {
+      "epoch": 24.31744,
+      "grad_norm": 1.1791542768478394,
+      "learning_rate": 1.2018807523009204e-05,
+      "loss": 0.5045,
+      "step": 18998
+    },
+    {
+      "epoch": 24.31872,
+      "grad_norm": 1.1057560443878174,
+      "learning_rate": 1.2016806722689076e-05,
+      "loss": 0.4459,
+      "step": 18999
+    },
+    {
+      "epoch": 24.32,
+      "grad_norm": 1.1355136632919312,
+      "learning_rate": 1.2014805922368948e-05,
+      "loss": 0.4693,
+      "step": 19000
+    },
+    {
+      "epoch": 24.32128,
+      "grad_norm": 1.204648494720459,
+      "learning_rate": 1.201280512204882e-05,
+      "loss": 0.5096,
+      "step": 19001
+    },
+    {
+      "epoch": 24.32256,
+      "grad_norm": 1.1683499813079834,
+      "learning_rate": 1.2010804321728692e-05,
+      "loss": 0.4805,
+      "step": 19002
+    },
+    {
+      "epoch": 24.32384,
+      "grad_norm": 1.0788731575012207,
+      "learning_rate": 1.2008803521408563e-05,
+      "loss": 0.4542,
+      "step": 19003
+    },
+    {
+      "epoch": 24.32512,
+      "grad_norm": 1.1736633777618408,
+      "learning_rate": 1.2006802721088437e-05,
+      "loss": 0.4771,
+      "step": 19004
+    },
+    {
+      "epoch": 24.3264,
+      "grad_norm": 1.1012755632400513,
+      "learning_rate": 1.2004801920768307e-05,
+      "loss": 0.4554,
+      "step": 19005
+    },
+    {
+      "epoch": 24.32768,
+      "grad_norm": 1.0851147174835205,
+      "learning_rate": 1.200280112044818e-05,
+      "loss": 0.4286,
+      "step": 19006
+    },
+    {
+      "epoch": 24.32896,
+      "grad_norm": 1.1383137702941895,
+      "learning_rate": 1.2000800320128051e-05,
+      "loss": 0.4664,
+      "step": 19007
+    },
+    {
+      "epoch": 24.33024,
+      "grad_norm": 1.1844596862792969,
+      "learning_rate": 1.1998799519807924e-05,
+      "loss": 0.5101,
+      "step": 19008
+    },
+    {
+      "epoch": 24.33152,
+      "grad_norm": 1.212876319885254,
+      "learning_rate": 1.1996798719487795e-05,
+      "loss": 0.524,
+      "step": 19009
+    },
+    {
+      "epoch": 24.3328,
+      "grad_norm": 1.1917188167572021,
+      "learning_rate": 1.1994797919167668e-05,
+      "loss": 0.5007,
+      "step": 19010
+    },
+    {
+      "epoch": 24.33408,
+      "grad_norm": 1.0979135036468506,
+      "learning_rate": 1.199279711884754e-05,
+      "loss": 0.4251,
+      "step": 19011
+    },
+    {
+      "epoch": 24.33536,
+      "grad_norm": 1.197851300239563,
+      "learning_rate": 1.1990796318527412e-05,
+      "loss": 0.4987,
+      "step": 19012
+    },
+    {
+      "epoch": 24.33664,
+      "grad_norm": 1.1661128997802734,
+      "learning_rate": 1.1988795518207284e-05,
+      "loss": 0.4864,
+      "step": 19013
+    },
+    {
+      "epoch": 24.33792,
+      "grad_norm": 1.1655899286270142,
+      "learning_rate": 1.1986794717887156e-05,
+      "loss": 0.4645,
+      "step": 19014
+    },
+    {
+      "epoch": 24.3392,
+      "grad_norm": 1.1260086297988892,
+      "learning_rate": 1.1984793917567027e-05,
+      "loss": 0.4571,
+      "step": 19015
+    },
+    {
+      "epoch": 24.34048,
+      "grad_norm": 1.1551868915557861,
+      "learning_rate": 1.19827931172469e-05,
+      "loss": 0.454,
+      "step": 19016
+    },
+    {
+      "epoch": 24.34176,
+      "grad_norm": 1.09392249584198,
+      "learning_rate": 1.1980792316926771e-05,
+      "loss": 0.4698,
+      "step": 19017
+    },
+    {
+      "epoch": 24.34304,
+      "grad_norm": 1.2250076532363892,
+      "learning_rate": 1.1978791516606645e-05,
+      "loss": 0.4851,
+      "step": 19018
+    },
+    {
+      "epoch": 24.34432,
+      "grad_norm": 1.1403043270111084,
+      "learning_rate": 1.1976790716286515e-05,
+      "loss": 0.4895,
+      "step": 19019
+    },
+    {
+      "epoch": 24.3456,
+      "grad_norm": 1.113724946975708,
+      "learning_rate": 1.1974789915966388e-05,
+      "loss": 0.4718,
+      "step": 19020
+    },
+    {
+      "epoch": 24.34688,
+      "grad_norm": 1.1426780223846436,
+      "learning_rate": 1.1972789115646259e-05,
+      "loss": 0.4889,
+      "step": 19021
+    },
+    {
+      "epoch": 24.34816,
+      "grad_norm": 1.1902352571487427,
+      "learning_rate": 1.1970788315326132e-05,
+      "loss": 0.4702,
+      "step": 19022
+    },
+    {
+      "epoch": 24.34944,
+      "grad_norm": 1.1023898124694824,
+      "learning_rate": 1.1968787515006002e-05,
+      "loss": 0.4402,
+      "step": 19023
+    },
+    {
+      "epoch": 24.35072,
+      "grad_norm": 1.1217379570007324,
+      "learning_rate": 1.1966786714685876e-05,
+      "loss": 0.453,
+      "step": 19024
+    },
+    {
+      "epoch": 24.352,
+      "grad_norm": 1.115411639213562,
+      "learning_rate": 1.1964785914365746e-05,
+      "loss": 0.4485,
+      "step": 19025
+    },
+    {
+      "epoch": 24.35328,
+      "grad_norm": 1.209621787071228,
+      "learning_rate": 1.196278511404562e-05,
+      "loss": 0.5006,
+      "step": 19026
+    },
+    {
+      "epoch": 24.35456,
+      "grad_norm": 1.1467214822769165,
+      "learning_rate": 1.1960784313725491e-05,
+      "loss": 0.4591,
+      "step": 19027
+    },
+    {
+      "epoch": 24.35584,
+      "grad_norm": 1.1208250522613525,
+      "learning_rate": 1.1958783513405363e-05,
+      "loss": 0.4738,
+      "step": 19028
+    },
+    {
+      "epoch": 24.35712,
+      "grad_norm": 1.1195433139801025,
+      "learning_rate": 1.1956782713085235e-05,
+      "loss": 0.4478,
+      "step": 19029
+    },
+    {
+      "epoch": 24.3584,
+      "grad_norm": 1.1476173400878906,
+      "learning_rate": 1.1954781912765107e-05,
+      "loss": 0.4901,
+      "step": 19030
+    },
+    {
+      "epoch": 24.35968,
+      "grad_norm": 1.2079092264175415,
+      "learning_rate": 1.1952781112444979e-05,
+      "loss": 0.4439,
+      "step": 19031
+    },
+    {
+      "epoch": 24.36096,
+      "grad_norm": 1.1161270141601562,
+      "learning_rate": 1.195078031212485e-05,
+      "loss": 0.4548,
+      "step": 19032
+    },
+    {
+      "epoch": 24.36224,
+      "grad_norm": 1.1574740409851074,
+      "learning_rate": 1.1948779511804723e-05,
+      "loss": 0.4633,
+      "step": 19033
+    },
+    {
+      "epoch": 24.36352,
+      "grad_norm": 1.1372309923171997,
+      "learning_rate": 1.1946778711484594e-05,
+      "loss": 0.4352,
+      "step": 19034
+    },
+    {
+      "epoch": 24.3648,
+      "grad_norm": 1.1243098974227905,
+      "learning_rate": 1.1944777911164466e-05,
+      "loss": 0.4584,
+      "step": 19035
+    },
+    {
+      "epoch": 24.36608,
+      "grad_norm": 1.1075938940048218,
+      "learning_rate": 1.1942777110844338e-05,
+      "loss": 0.4726,
+      "step": 19036
+    },
+    {
+      "epoch": 24.36736,
+      "grad_norm": 1.1360459327697754,
+      "learning_rate": 1.194077631052421e-05,
+      "loss": 0.423,
+      "step": 19037
+    },
+    {
+      "epoch": 24.36864,
+      "grad_norm": 1.1323789358139038,
+      "learning_rate": 1.1938775510204082e-05,
+      "loss": 0.4819,
+      "step": 19038
+    },
+    {
+      "epoch": 24.36992,
+      "grad_norm": 1.1539483070373535,
+      "learning_rate": 1.1936774709883954e-05,
+      "loss": 0.4689,
+      "step": 19039
+    },
+    {
+      "epoch": 24.3712,
+      "grad_norm": 1.1773465871810913,
+      "learning_rate": 1.1934773909563826e-05,
+      "loss": 0.5437,
+      "step": 19040
+    },
+    {
+      "epoch": 24.37248,
+      "grad_norm": 1.0829496383666992,
+      "learning_rate": 1.1932773109243699e-05,
+      "loss": 0.4514,
+      "step": 19041
+    },
+    {
+      "epoch": 24.37376,
+      "grad_norm": 1.1402915716171265,
+      "learning_rate": 1.193077230892357e-05,
+      "loss": 0.4494,
+      "step": 19042
+    },
+    {
+      "epoch": 24.37504,
+      "grad_norm": 1.1688295602798462,
+      "learning_rate": 1.1928771508603443e-05,
+      "loss": 0.4596,
+      "step": 19043
+    },
+    {
+      "epoch": 24.37632,
+      "grad_norm": 1.1912732124328613,
+      "learning_rate": 1.1926770708283313e-05,
+      "loss": 0.4838,
+      "step": 19044
+    },
+    {
+      "epoch": 24.3776,
+      "grad_norm": 1.1036256551742554,
+      "learning_rate": 1.1924769907963187e-05,
+      "loss": 0.4478,
+      "step": 19045
+    },
+    {
+      "epoch": 24.37888,
+      "grad_norm": 1.1427876949310303,
+      "learning_rate": 1.1922769107643057e-05,
+      "loss": 0.4632,
+      "step": 19046
+    },
+    {
+      "epoch": 24.38016,
+      "grad_norm": 1.2275477647781372,
+      "learning_rate": 1.192076830732293e-05,
+      "loss": 0.5451,
+      "step": 19047
+    },
+    {
+      "epoch": 24.38144,
+      "grad_norm": 1.0969351530075073,
+      "learning_rate": 1.1918767507002802e-05,
+      "loss": 0.4578,
+      "step": 19048
+    },
+    {
+      "epoch": 24.38272,
+      "grad_norm": 1.117645502090454,
+      "learning_rate": 1.1916766706682674e-05,
+      "loss": 0.4327,
+      "step": 19049
+    },
+    {
+      "epoch": 24.384,
+      "grad_norm": 1.157266616821289,
+      "learning_rate": 1.1914765906362546e-05,
+      "loss": 0.478,
+      "step": 19050
+    },
+    {
+      "epoch": 24.38528,
+      "grad_norm": 1.1453224420547485,
+      "learning_rate": 1.1912765106042418e-05,
+      "loss": 0.4923,
+      "step": 19051
+    },
+    {
+      "epoch": 24.38656,
+      "grad_norm": 1.1940408945083618,
+      "learning_rate": 1.191076430572229e-05,
+      "loss": 0.5081,
+      "step": 19052
+    },
+    {
+      "epoch": 24.38784,
+      "grad_norm": 1.0976282358169556,
+      "learning_rate": 1.1908763505402161e-05,
+      "loss": 0.4342,
+      "step": 19053
+    },
+    {
+      "epoch": 24.38912,
+      "grad_norm": 1.211582899093628,
+      "learning_rate": 1.1906762705082033e-05,
+      "loss": 0.4941,
+      "step": 19054
+    },
+    {
+      "epoch": 24.3904,
+      "grad_norm": 1.1101171970367432,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.4526,
+      "step": 19055
+    },
+    {
+      "epoch": 24.39168,
+      "grad_norm": 1.213545560836792,
+      "learning_rate": 1.1902761104441777e-05,
+      "loss": 0.4828,
+      "step": 19056
+    },
+    {
+      "epoch": 24.39296,
+      "grad_norm": 1.1441152095794678,
+      "learning_rate": 1.190076030412165e-05,
+      "loss": 0.4691,
+      "step": 19057
+    },
+    {
+      "epoch": 24.39424,
+      "grad_norm": 1.1789923906326294,
+      "learning_rate": 1.189875950380152e-05,
+      "loss": 0.4947,
+      "step": 19058
+    },
+    {
+      "epoch": 24.39552,
+      "grad_norm": 1.031014323234558,
+      "learning_rate": 1.1896758703481394e-05,
+      "loss": 0.3981,
+      "step": 19059
+    },
+    {
+      "epoch": 24.3968,
+      "grad_norm": 1.2153488397598267,
+      "learning_rate": 1.1894757903161264e-05,
+      "loss": 0.5261,
+      "step": 19060
+    },
+    {
+      "epoch": 24.39808,
+      "grad_norm": 1.2073795795440674,
+      "learning_rate": 1.1892757102841138e-05,
+      "loss": 0.49,
+      "step": 19061
+    },
+    {
+      "epoch": 24.39936,
+      "grad_norm": 1.1584200859069824,
+      "learning_rate": 1.1890756302521008e-05,
+      "loss": 0.4965,
+      "step": 19062
+    },
+    {
+      "epoch": 24.40064,
+      "grad_norm": 1.1565824747085571,
+      "learning_rate": 1.1888755502200882e-05,
+      "loss": 0.4811,
+      "step": 19063
+    },
+    {
+      "epoch": 24.40192,
+      "grad_norm": 1.1859803199768066,
+      "learning_rate": 1.1886754701880753e-05,
+      "loss": 0.4927,
+      "step": 19064
+    },
+    {
+      "epoch": 24.4032,
+      "grad_norm": 1.1612244844436646,
+      "learning_rate": 1.1884753901560625e-05,
+      "loss": 0.5152,
+      "step": 19065
+    },
+    {
+      "epoch": 24.40448,
+      "grad_norm": 1.1404790878295898,
+      "learning_rate": 1.1882753101240497e-05,
+      "loss": 0.4933,
+      "step": 19066
+    },
+    {
+      "epoch": 24.40576,
+      "grad_norm": 1.1987690925598145,
+      "learning_rate": 1.1880752300920369e-05,
+      "loss": 0.4743,
+      "step": 19067
+    },
+    {
+      "epoch": 24.40704,
+      "grad_norm": 1.145716905593872,
+      "learning_rate": 1.1878751500600241e-05,
+      "loss": 0.4787,
+      "step": 19068
+    },
+    {
+      "epoch": 24.40832,
+      "grad_norm": 1.1686224937438965,
+      "learning_rate": 1.1876750700280113e-05,
+      "loss": 0.485,
+      "step": 19069
+    },
+    {
+      "epoch": 24.4096,
+      "grad_norm": 1.1903284788131714,
+      "learning_rate": 1.1874749899959985e-05,
+      "loss": 0.4924,
+      "step": 19070
+    },
+    {
+      "epoch": 24.41088,
+      "grad_norm": 1.1217108964920044,
+      "learning_rate": 1.1872749099639856e-05,
+      "loss": 0.4669,
+      "step": 19071
+    },
+    {
+      "epoch": 24.41216,
+      "grad_norm": 1.1774252653121948,
+      "learning_rate": 1.1870748299319728e-05,
+      "loss": 0.4977,
+      "step": 19072
+    },
+    {
+      "epoch": 24.41344,
+      "grad_norm": 1.2087725400924683,
+      "learning_rate": 1.18687474989996e-05,
+      "loss": 0.5188,
+      "step": 19073
+    },
+    {
+      "epoch": 24.41472,
+      "grad_norm": 1.1201173067092896,
+      "learning_rate": 1.1866746698679472e-05,
+      "loss": 0.4849,
+      "step": 19074
+    },
+    {
+      "epoch": 24.416,
+      "grad_norm": 1.0546908378601074,
+      "learning_rate": 1.1864745898359344e-05,
+      "loss": 0.4321,
+      "step": 19075
+    },
+    {
+      "epoch": 24.41728,
+      "grad_norm": 1.2529699802398682,
+      "learning_rate": 1.1862745098039216e-05,
+      "loss": 0.5136,
+      "step": 19076
+    },
+    {
+      "epoch": 24.41856,
+      "grad_norm": 1.12168550491333,
+      "learning_rate": 1.1860744297719088e-05,
+      "loss": 0.484,
+      "step": 19077
+    },
+    {
+      "epoch": 24.41984,
+      "grad_norm": 1.2384254932403564,
+      "learning_rate": 1.1858743497398961e-05,
+      "loss": 0.4956,
+      "step": 19078
+    },
+    {
+      "epoch": 24.42112,
+      "grad_norm": 1.1256715059280396,
+      "learning_rate": 1.1856742697078831e-05,
+      "loss": 0.512,
+      "step": 19079
+    },
+    {
+      "epoch": 24.4224,
+      "grad_norm": 1.2130337953567505,
+      "learning_rate": 1.1854741896758705e-05,
+      "loss": 0.4678,
+      "step": 19080
+    },
+    {
+      "epoch": 24.42368,
+      "grad_norm": 1.1287142038345337,
+      "learning_rate": 1.1852741096438575e-05,
+      "loss": 0.4757,
+      "step": 19081
+    },
+    {
+      "epoch": 24.42496,
+      "grad_norm": 1.1295257806777954,
+      "learning_rate": 1.1850740296118449e-05,
+      "loss": 0.4708,
+      "step": 19082
+    },
+    {
+      "epoch": 24.42624,
+      "grad_norm": 1.1850224733352661,
+      "learning_rate": 1.1848739495798319e-05,
+      "loss": 0.493,
+      "step": 19083
+    },
+    {
+      "epoch": 24.42752,
+      "grad_norm": 1.1218152046203613,
+      "learning_rate": 1.1846738695478192e-05,
+      "loss": 0.4252,
+      "step": 19084
+    },
+    {
+      "epoch": 24.4288,
+      "grad_norm": 1.257108449935913,
+      "learning_rate": 1.1844737895158062e-05,
+      "loss": 0.5102,
+      "step": 19085
+    },
+    {
+      "epoch": 24.43008,
+      "grad_norm": 1.2210159301757812,
+      "learning_rate": 1.1842737094837936e-05,
+      "loss": 0.4865,
+      "step": 19086
+    },
+    {
+      "epoch": 24.43136,
+      "grad_norm": 1.188494086265564,
+      "learning_rate": 1.1840736294517808e-05,
+      "loss": 0.5042,
+      "step": 19087
+    },
+    {
+      "epoch": 24.43264,
+      "grad_norm": 1.1470962762832642,
+      "learning_rate": 1.183873549419768e-05,
+      "loss": 0.4539,
+      "step": 19088
+    },
+    {
+      "epoch": 24.43392,
+      "grad_norm": 1.181655764579773,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 0.4374,
+      "step": 19089
+    },
+    {
+      "epoch": 24.4352,
+      "grad_norm": 1.1574394702911377,
+      "learning_rate": 1.1834733893557423e-05,
+      "loss": 0.4477,
+      "step": 19090
+    },
+    {
+      "epoch": 24.43648,
+      "grad_norm": 1.2077171802520752,
+      "learning_rate": 1.1832733093237295e-05,
+      "loss": 0.4863,
+      "step": 19091
+    },
+    {
+      "epoch": 24.43776,
+      "grad_norm": 1.1384170055389404,
+      "learning_rate": 1.1830732292917167e-05,
+      "loss": 0.4579,
+      "step": 19092
+    },
+    {
+      "epoch": 24.43904,
+      "grad_norm": 1.1912693977355957,
+      "learning_rate": 1.1828731492597039e-05,
+      "loss": 0.5061,
+      "step": 19093
+    },
+    {
+      "epoch": 24.44032,
+      "grad_norm": 1.2225050926208496,
+      "learning_rate": 1.1826730692276913e-05,
+      "loss": 0.5158,
+      "step": 19094
+    },
+    {
+      "epoch": 24.4416,
+      "grad_norm": 1.105109691619873,
+      "learning_rate": 1.1824729891956783e-05,
+      "loss": 0.4698,
+      "step": 19095
+    },
+    {
+      "epoch": 24.44288,
+      "grad_norm": 1.2286189794540405,
+      "learning_rate": 1.1822729091636656e-05,
+      "loss": 0.5637,
+      "step": 19096
+    },
+    {
+      "epoch": 24.44416,
+      "grad_norm": 1.2093946933746338,
+      "learning_rate": 1.1820728291316526e-05,
+      "loss": 0.5066,
+      "step": 19097
+    },
+    {
+      "epoch": 24.44544,
+      "grad_norm": 1.1525238752365112,
+      "learning_rate": 1.18187274909964e-05,
+      "loss": 0.4576,
+      "step": 19098
+    },
+    {
+      "epoch": 24.44672,
+      "grad_norm": 1.128960371017456,
+      "learning_rate": 1.181672669067627e-05,
+      "loss": 0.4811,
+      "step": 19099
+    },
+    {
+      "epoch": 24.448,
+      "grad_norm": 1.11859130859375,
+      "learning_rate": 1.1814725890356144e-05,
+      "loss": 0.4477,
+      "step": 19100
+    },
+    {
+      "epoch": 24.44928,
+      "grad_norm": 1.2186490297317505,
+      "learning_rate": 1.1812725090036016e-05,
+      "loss": 0.492,
+      "step": 19101
+    },
+    {
+      "epoch": 24.45056,
+      "grad_norm": 1.1721785068511963,
+      "learning_rate": 1.1810724289715887e-05,
+      "loss": 0.4775,
+      "step": 19102
+    },
+    {
+      "epoch": 24.45184,
+      "grad_norm": 1.1152302026748657,
+      "learning_rate": 1.180872348939576e-05,
+      "loss": 0.445,
+      "step": 19103
+    },
+    {
+      "epoch": 24.45312,
+      "grad_norm": 1.2016420364379883,
+      "learning_rate": 1.1806722689075631e-05,
+      "loss": 0.4517,
+      "step": 19104
+    },
+    {
+      "epoch": 24.4544,
+      "grad_norm": 1.1429190635681152,
+      "learning_rate": 1.1804721888755503e-05,
+      "loss": 0.4761,
+      "step": 19105
+    },
+    {
+      "epoch": 24.45568,
+      "grad_norm": 1.1792242527008057,
+      "learning_rate": 1.1802721088435375e-05,
+      "loss": 0.4446,
+      "step": 19106
+    },
+    {
+      "epoch": 24.45696,
+      "grad_norm": 1.086751103401184,
+      "learning_rate": 1.1800720288115247e-05,
+      "loss": 0.4721,
+      "step": 19107
+    },
+    {
+      "epoch": 24.45824,
+      "grad_norm": 1.1734704971313477,
+      "learning_rate": 1.1798719487795119e-05,
+      "loss": 0.4855,
+      "step": 19108
+    },
+    {
+      "epoch": 24.45952,
+      "grad_norm": 1.1557416915893555,
+      "learning_rate": 1.179671868747499e-05,
+      "loss": 0.4758,
+      "step": 19109
+    },
+    {
+      "epoch": 24.4608,
+      "grad_norm": 1.1353665590286255,
+      "learning_rate": 1.1794717887154862e-05,
+      "loss": 0.5025,
+      "step": 19110
+    },
+    {
+      "epoch": 24.46208,
+      "grad_norm": 1.1943861246109009,
+      "learning_rate": 1.1792717086834734e-05,
+      "loss": 0.4488,
+      "step": 19111
+    },
+    {
+      "epoch": 24.46336,
+      "grad_norm": 1.1474003791809082,
+      "learning_rate": 1.1790716286514606e-05,
+      "loss": 0.4272,
+      "step": 19112
+    },
+    {
+      "epoch": 24.46464,
+      "grad_norm": 1.1481927633285522,
+      "learning_rate": 1.1788715486194478e-05,
+      "loss": 0.4674,
+      "step": 19113
+    },
+    {
+      "epoch": 24.46592,
+      "grad_norm": 1.2008520364761353,
+      "learning_rate": 1.178671468587435e-05,
+      "loss": 0.47,
+      "step": 19114
+    },
+    {
+      "epoch": 24.4672,
+      "grad_norm": 1.1071596145629883,
+      "learning_rate": 1.1784713885554222e-05,
+      "loss": 0.4489,
+      "step": 19115
+    },
+    {
+      "epoch": 24.46848,
+      "grad_norm": 1.1181803941726685,
+      "learning_rate": 1.1782713085234093e-05,
+      "loss": 0.4641,
+      "step": 19116
+    },
+    {
+      "epoch": 24.46976,
+      "grad_norm": 1.163570523262024,
+      "learning_rate": 1.1780712284913967e-05,
+      "loss": 0.492,
+      "step": 19117
+    },
+    {
+      "epoch": 24.47104,
+      "grad_norm": 1.1192409992218018,
+      "learning_rate": 1.1778711484593837e-05,
+      "loss": 0.4299,
+      "step": 19118
+    },
+    {
+      "epoch": 24.47232,
+      "grad_norm": 1.1589858531951904,
+      "learning_rate": 1.177671068427371e-05,
+      "loss": 0.4381,
+      "step": 19119
+    },
+    {
+      "epoch": 24.4736,
+      "grad_norm": 1.183703064918518,
+      "learning_rate": 1.1774709883953581e-05,
+      "loss": 0.4732,
+      "step": 19120
+    },
+    {
+      "epoch": 24.47488,
+      "grad_norm": 1.0970691442489624,
+      "learning_rate": 1.1772709083633454e-05,
+      "loss": 0.4542,
+      "step": 19121
+    },
+    {
+      "epoch": 24.47616,
+      "grad_norm": 1.1885040998458862,
+      "learning_rate": 1.1770708283313325e-05,
+      "loss": 0.5307,
+      "step": 19122
+    },
+    {
+      "epoch": 24.47744,
+      "grad_norm": 1.1278455257415771,
+      "learning_rate": 1.1768707482993198e-05,
+      "loss": 0.4681,
+      "step": 19123
+    },
+    {
+      "epoch": 24.47872,
+      "grad_norm": 1.0988664627075195,
+      "learning_rate": 1.176670668267307e-05,
+      "loss": 0.4867,
+      "step": 19124
+    },
+    {
+      "epoch": 24.48,
+      "grad_norm": 1.0938818454742432,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.4929,
+      "step": 19125
+    },
+    {
+      "epoch": 24.48128,
+      "grad_norm": 1.1701966524124146,
+      "learning_rate": 1.1762705082032814e-05,
+      "loss": 0.4634,
+      "step": 19126
+    },
+    {
+      "epoch": 24.48256,
+      "grad_norm": 1.2021418809890747,
+      "learning_rate": 1.1760704281712686e-05,
+      "loss": 0.4588,
+      "step": 19127
+    },
+    {
+      "epoch": 24.48384,
+      "grad_norm": 1.1727592945098877,
+      "learning_rate": 1.1758703481392557e-05,
+      "loss": 0.4352,
+      "step": 19128
+    },
+    {
+      "epoch": 24.48512,
+      "grad_norm": 1.174090027809143,
+      "learning_rate": 1.175670268107243e-05,
+      "loss": 0.4726,
+      "step": 19129
+    },
+    {
+      "epoch": 24.4864,
+      "grad_norm": 1.1474714279174805,
+      "learning_rate": 1.1754701880752301e-05,
+      "loss": 0.4448,
+      "step": 19130
+    },
+    {
+      "epoch": 24.48768,
+      "grad_norm": 1.2793716192245483,
+      "learning_rate": 1.1752701080432175e-05,
+      "loss": 0.5175,
+      "step": 19131
+    },
+    {
+      "epoch": 24.48896,
+      "grad_norm": 1.201866626739502,
+      "learning_rate": 1.1750700280112045e-05,
+      "loss": 0.4494,
+      "step": 19132
+    },
+    {
+      "epoch": 24.49024,
+      "grad_norm": 1.1244724988937378,
+      "learning_rate": 1.1748699479791918e-05,
+      "loss": 0.4488,
+      "step": 19133
+    },
+    {
+      "epoch": 24.49152,
+      "grad_norm": 1.1583161354064941,
+      "learning_rate": 1.1746698679471789e-05,
+      "loss": 0.4516,
+      "step": 19134
+    },
+    {
+      "epoch": 24.4928,
+      "grad_norm": 1.1337168216705322,
+      "learning_rate": 1.1744697879151662e-05,
+      "loss": 0.4695,
+      "step": 19135
+    },
+    {
+      "epoch": 24.49408,
+      "grad_norm": 1.064746618270874,
+      "learning_rate": 1.1742697078831532e-05,
+      "loss": 0.4032,
+      "step": 19136
+    },
+    {
+      "epoch": 24.49536,
+      "grad_norm": 1.1812145709991455,
+      "learning_rate": 1.1740696278511406e-05,
+      "loss": 0.5035,
+      "step": 19137
+    },
+    {
+      "epoch": 24.49664,
+      "grad_norm": 1.156087875366211,
+      "learning_rate": 1.1738695478191276e-05,
+      "loss": 0.4852,
+      "step": 19138
+    },
+    {
+      "epoch": 24.49792,
+      "grad_norm": 1.1552976369857788,
+      "learning_rate": 1.173669467787115e-05,
+      "loss": 0.4727,
+      "step": 19139
+    },
+    {
+      "epoch": 24.4992,
+      "grad_norm": 1.160605788230896,
+      "learning_rate": 1.1734693877551021e-05,
+      "loss": 0.4935,
+      "step": 19140
+    },
+    {
+      "epoch": 24.50048,
+      "grad_norm": 1.1020570993423462,
+      "learning_rate": 1.1732693077230893e-05,
+      "loss": 0.453,
+      "step": 19141
+    },
+    {
+      "epoch": 24.50176,
+      "grad_norm": 1.234426498413086,
+      "learning_rate": 1.1730692276910765e-05,
+      "loss": 0.5184,
+      "step": 19142
+    },
+    {
+      "epoch": 24.50304,
+      "grad_norm": 1.2169140577316284,
+      "learning_rate": 1.1728691476590637e-05,
+      "loss": 0.509,
+      "step": 19143
+    },
+    {
+      "epoch": 24.50432,
+      "grad_norm": 1.199015736579895,
+      "learning_rate": 1.1726690676270509e-05,
+      "loss": 0.4727,
+      "step": 19144
+    },
+    {
+      "epoch": 24.5056,
+      "grad_norm": 1.1079201698303223,
+      "learning_rate": 1.172468987595038e-05,
+      "loss": 0.4845,
+      "step": 19145
+    },
+    {
+      "epoch": 24.50688,
+      "grad_norm": 1.146178126335144,
+      "learning_rate": 1.1722689075630253e-05,
+      "loss": 0.4743,
+      "step": 19146
+    },
+    {
+      "epoch": 24.50816,
+      "grad_norm": 1.0882790088653564,
+      "learning_rate": 1.1720688275310124e-05,
+      "loss": 0.4622,
+      "step": 19147
+    },
+    {
+      "epoch": 24.50944,
+      "grad_norm": 1.2238930463790894,
+      "learning_rate": 1.1718687474989996e-05,
+      "loss": 0.4771,
+      "step": 19148
+    },
+    {
+      "epoch": 24.51072,
+      "grad_norm": 1.1269352436065674,
+      "learning_rate": 1.1716686674669868e-05,
+      "loss": 0.5029,
+      "step": 19149
+    },
+    {
+      "epoch": 24.512,
+      "grad_norm": 1.1483230590820312,
+      "learning_rate": 1.171468587434974e-05,
+      "loss": 0.4726,
+      "step": 19150
+    },
+    {
+      "epoch": 24.51328,
+      "grad_norm": 1.1776012182235718,
+      "learning_rate": 1.1712685074029612e-05,
+      "loss": 0.497,
+      "step": 19151
+    },
+    {
+      "epoch": 24.51456,
+      "grad_norm": 1.1069426536560059,
+      "learning_rate": 1.1710684273709484e-05,
+      "loss": 0.4583,
+      "step": 19152
+    },
+    {
+      "epoch": 24.51584,
+      "grad_norm": 1.1324975490570068,
+      "learning_rate": 1.1708683473389356e-05,
+      "loss": 0.4128,
+      "step": 19153
+    },
+    {
+      "epoch": 24.51712,
+      "grad_norm": 1.190555214881897,
+      "learning_rate": 1.1706682673069229e-05,
+      "loss": 0.4871,
+      "step": 19154
+    },
+    {
+      "epoch": 24.5184,
+      "grad_norm": 1.1613634824752808,
+      "learning_rate": 1.17046818727491e-05,
+      "loss": 0.4866,
+      "step": 19155
+    },
+    {
+      "epoch": 24.51968,
+      "grad_norm": 1.1718133687973022,
+      "learning_rate": 1.1702681072428973e-05,
+      "loss": 0.4818,
+      "step": 19156
+    },
+    {
+      "epoch": 24.52096,
+      "grad_norm": 1.146981954574585,
+      "learning_rate": 1.1700680272108843e-05,
+      "loss": 0.5019,
+      "step": 19157
+    },
+    {
+      "epoch": 24.52224,
+      "grad_norm": 1.192657709121704,
+      "learning_rate": 1.1698679471788717e-05,
+      "loss": 0.4981,
+      "step": 19158
+    },
+    {
+      "epoch": 24.52352,
+      "grad_norm": 1.0763636827468872,
+      "learning_rate": 1.1696678671468587e-05,
+      "loss": 0.4314,
+      "step": 19159
+    },
+    {
+      "epoch": 24.5248,
+      "grad_norm": 1.1718692779541016,
+      "learning_rate": 1.169467787114846e-05,
+      "loss": 0.4831,
+      "step": 19160
+    },
+    {
+      "epoch": 24.52608,
+      "grad_norm": 1.0933730602264404,
+      "learning_rate": 1.1692677070828332e-05,
+      "loss": 0.4976,
+      "step": 19161
+    },
+    {
+      "epoch": 24.52736,
+      "grad_norm": 1.0974316596984863,
+      "learning_rate": 1.1690676270508204e-05,
+      "loss": 0.4816,
+      "step": 19162
+    },
+    {
+      "epoch": 24.52864,
+      "grad_norm": 1.2042176723480225,
+      "learning_rate": 1.1688675470188076e-05,
+      "loss": 0.4928,
+      "step": 19163
+    },
+    {
+      "epoch": 24.52992,
+      "grad_norm": 1.1524473428726196,
+      "learning_rate": 1.1686674669867948e-05,
+      "loss": 0.4771,
+      "step": 19164
+    },
+    {
+      "epoch": 24.5312,
+      "grad_norm": 1.2503443956375122,
+      "learning_rate": 1.168467386954782e-05,
+      "loss": 0.4702,
+      "step": 19165
+    },
+    {
+      "epoch": 24.53248,
+      "grad_norm": 1.1078400611877441,
+      "learning_rate": 1.1682673069227691e-05,
+      "loss": 0.4361,
+      "step": 19166
+    },
+    {
+      "epoch": 24.53376,
+      "grad_norm": 1.1431939601898193,
+      "learning_rate": 1.1680672268907563e-05,
+      "loss": 0.4693,
+      "step": 19167
+    },
+    {
+      "epoch": 24.53504,
+      "grad_norm": 1.201981782913208,
+      "learning_rate": 1.1678671468587435e-05,
+      "loss": 0.4996,
+      "step": 19168
+    },
+    {
+      "epoch": 24.53632,
+      "grad_norm": 1.2216880321502686,
+      "learning_rate": 1.1676670668267307e-05,
+      "loss": 0.4964,
+      "step": 19169
+    },
+    {
+      "epoch": 24.5376,
+      "grad_norm": 1.199825644493103,
+      "learning_rate": 1.167466986794718e-05,
+      "loss": 0.484,
+      "step": 19170
+    },
+    {
+      "epoch": 24.53888,
+      "grad_norm": 1.2285641431808472,
+      "learning_rate": 1.167266906762705e-05,
+      "loss": 0.509,
+      "step": 19171
+    },
+    {
+      "epoch": 24.54016,
+      "grad_norm": 1.1986514329910278,
+      "learning_rate": 1.1670668267306924e-05,
+      "loss": 0.49,
+      "step": 19172
+    },
+    {
+      "epoch": 24.54144,
+      "grad_norm": 1.1682507991790771,
+      "learning_rate": 1.1668667466986794e-05,
+      "loss": 0.4695,
+      "step": 19173
+    },
+    {
+      "epoch": 24.54272,
+      "grad_norm": 1.216668963432312,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.508,
+      "step": 19174
+    },
+    {
+      "epoch": 24.544,
+      "grad_norm": 1.08146071434021,
+      "learning_rate": 1.1664665866346538e-05,
+      "loss": 0.4698,
+      "step": 19175
+    },
+    {
+      "epoch": 24.545279999999998,
+      "grad_norm": 1.1765329837799072,
+      "learning_rate": 1.1662665066026412e-05,
+      "loss": 0.4394,
+      "step": 19176
+    },
+    {
+      "epoch": 24.54656,
+      "grad_norm": 1.1449719667434692,
+      "learning_rate": 1.1660664265706283e-05,
+      "loss": 0.4574,
+      "step": 19177
+    },
+    {
+      "epoch": 24.54784,
+      "grad_norm": 1.0752500295639038,
+      "learning_rate": 1.1658663465386155e-05,
+      "loss": 0.4515,
+      "step": 19178
+    },
+    {
+      "epoch": 24.54912,
+      "grad_norm": 1.214896559715271,
+      "learning_rate": 1.1656662665066027e-05,
+      "loss": 0.5097,
+      "step": 19179
+    },
+    {
+      "epoch": 24.5504,
+      "grad_norm": 1.113995909690857,
+      "learning_rate": 1.1654661864745899e-05,
+      "loss": 0.4797,
+      "step": 19180
+    },
+    {
+      "epoch": 24.55168,
+      "grad_norm": 1.1679980754852295,
+      "learning_rate": 1.1652661064425771e-05,
+      "loss": 0.4592,
+      "step": 19181
+    },
+    {
+      "epoch": 24.55296,
+      "grad_norm": 1.154405117034912,
+      "learning_rate": 1.1650660264105643e-05,
+      "loss": 0.4735,
+      "step": 19182
+    },
+    {
+      "epoch": 24.55424,
+      "grad_norm": 1.2269680500030518,
+      "learning_rate": 1.1648659463785515e-05,
+      "loss": 0.499,
+      "step": 19183
+    },
+    {
+      "epoch": 24.55552,
+      "grad_norm": 1.146951675415039,
+      "learning_rate": 1.1646658663465388e-05,
+      "loss": 0.4796,
+      "step": 19184
+    },
+    {
+      "epoch": 24.5568,
+      "grad_norm": 1.2074452638626099,
+      "learning_rate": 1.1644657863145258e-05,
+      "loss": 0.4796,
+      "step": 19185
+    },
+    {
+      "epoch": 24.55808,
+      "grad_norm": 1.1281073093414307,
+      "learning_rate": 1.1642657062825132e-05,
+      "loss": 0.4737,
+      "step": 19186
+    },
+    {
+      "epoch": 24.55936,
+      "grad_norm": 1.1057974100112915,
+      "learning_rate": 1.1640656262505002e-05,
+      "loss": 0.4699,
+      "step": 19187
+    },
+    {
+      "epoch": 24.56064,
+      "grad_norm": 1.1083946228027344,
+      "learning_rate": 1.1638655462184876e-05,
+      "loss": 0.4705,
+      "step": 19188
+    },
+    {
+      "epoch": 24.56192,
+      "grad_norm": 1.1685590744018555,
+      "learning_rate": 1.1636654661864746e-05,
+      "loss": 0.4922,
+      "step": 19189
+    },
+    {
+      "epoch": 24.5632,
+      "grad_norm": 1.1086533069610596,
+      "learning_rate": 1.163465386154462e-05,
+      "loss": 0.4473,
+      "step": 19190
+    },
+    {
+      "epoch": 24.56448,
+      "grad_norm": 1.156948447227478,
+      "learning_rate": 1.163265306122449e-05,
+      "loss": 0.4685,
+      "step": 19191
+    },
+    {
+      "epoch": 24.56576,
+      "grad_norm": 1.1937439441680908,
+      "learning_rate": 1.1630652260904363e-05,
+      "loss": 0.4974,
+      "step": 19192
+    },
+    {
+      "epoch": 24.56704,
+      "grad_norm": 1.1433395147323608,
+      "learning_rate": 1.1628651460584235e-05,
+      "loss": 0.4823,
+      "step": 19193
+    },
+    {
+      "epoch": 24.56832,
+      "grad_norm": 1.1796252727508545,
+      "learning_rate": 1.1626650660264107e-05,
+      "loss": 0.4591,
+      "step": 19194
+    },
+    {
+      "epoch": 24.5696,
+      "grad_norm": 1.1520229578018188,
+      "learning_rate": 1.1624649859943979e-05,
+      "loss": 0.4805,
+      "step": 19195
+    },
+    {
+      "epoch": 24.57088,
+      "grad_norm": 1.0709677934646606,
+      "learning_rate": 1.162264905962385e-05,
+      "loss": 0.4397,
+      "step": 19196
+    },
+    {
+      "epoch": 24.57216,
+      "grad_norm": 1.1727806329727173,
+      "learning_rate": 1.1620648259303722e-05,
+      "loss": 0.5408,
+      "step": 19197
+    },
+    {
+      "epoch": 24.57344,
+      "grad_norm": 1.1703429222106934,
+      "learning_rate": 1.1618647458983594e-05,
+      "loss": 0.5012,
+      "step": 19198
+    },
+    {
+      "epoch": 24.57472,
+      "grad_norm": 1.1146478652954102,
+      "learning_rate": 1.1616646658663466e-05,
+      "loss": 0.4456,
+      "step": 19199
+    },
+    {
+      "epoch": 24.576,
+      "grad_norm": 1.1429121494293213,
+      "learning_rate": 1.1614645858343338e-05,
+      "loss": 0.4617,
+      "step": 19200
+    },
+    {
+      "epoch": 24.577280000000002,
+      "grad_norm": 1.0925348997116089,
+      "learning_rate": 1.161264505802321e-05,
+      "loss": 0.487,
+      "step": 19201
+    },
+    {
+      "epoch": 24.57856,
+      "grad_norm": 1.1159874200820923,
+      "learning_rate": 1.1610644257703082e-05,
+      "loss": 0.4982,
+      "step": 19202
+    },
+    {
+      "epoch": 24.57984,
+      "grad_norm": 1.1333866119384766,
+      "learning_rate": 1.1608643457382953e-05,
+      "loss": 0.4537,
+      "step": 19203
+    },
+    {
+      "epoch": 24.58112,
+      "grad_norm": 1.134494662284851,
+      "learning_rate": 1.1606642657062825e-05,
+      "loss": 0.4516,
+      "step": 19204
+    },
+    {
+      "epoch": 24.5824,
+      "grad_norm": 1.1905516386032104,
+      "learning_rate": 1.1604641856742697e-05,
+      "loss": 0.4714,
+      "step": 19205
+    },
+    {
+      "epoch": 24.58368,
+      "grad_norm": 1.208512544631958,
+      "learning_rate": 1.1602641056422569e-05,
+      "loss": 0.4834,
+      "step": 19206
+    },
+    {
+      "epoch": 24.58496,
+      "grad_norm": 1.2601969242095947,
+      "learning_rate": 1.1600640256102443e-05,
+      "loss": 0.501,
+      "step": 19207
+    },
+    {
+      "epoch": 24.58624,
+      "grad_norm": 1.1832962036132812,
+      "learning_rate": 1.1598639455782313e-05,
+      "loss": 0.459,
+      "step": 19208
+    },
+    {
+      "epoch": 24.58752,
+      "grad_norm": 1.2461732625961304,
+      "learning_rate": 1.1596638655462186e-05,
+      "loss": 0.5219,
+      "step": 19209
+    },
+    {
+      "epoch": 24.5888,
+      "grad_norm": 1.1089426279067993,
+      "learning_rate": 1.1594637855142056e-05,
+      "loss": 0.5035,
+      "step": 19210
+    },
+    {
+      "epoch": 24.59008,
+      "grad_norm": 1.2491692304611206,
+      "learning_rate": 1.159263705482193e-05,
+      "loss": 0.5015,
+      "step": 19211
+    },
+    {
+      "epoch": 24.59136,
+      "grad_norm": 1.1578104496002197,
+      "learning_rate": 1.15906362545018e-05,
+      "loss": 0.4683,
+      "step": 19212
+    },
+    {
+      "epoch": 24.59264,
+      "grad_norm": 1.1508525609970093,
+      "learning_rate": 1.1588635454181674e-05,
+      "loss": 0.4839,
+      "step": 19213
+    },
+    {
+      "epoch": 24.59392,
+      "grad_norm": 1.249266266822815,
+      "learning_rate": 1.1586634653861546e-05,
+      "loss": 0.4757,
+      "step": 19214
+    },
+    {
+      "epoch": 24.5952,
+      "grad_norm": 1.13655686378479,
+      "learning_rate": 1.1584633853541417e-05,
+      "loss": 0.4996,
+      "step": 19215
+    },
+    {
+      "epoch": 24.59648,
+      "grad_norm": 1.1786251068115234,
+      "learning_rate": 1.158263305322129e-05,
+      "loss": 0.4798,
+      "step": 19216
+    },
+    {
+      "epoch": 24.59776,
+      "grad_norm": 1.1763968467712402,
+      "learning_rate": 1.1580632252901161e-05,
+      "loss": 0.4782,
+      "step": 19217
+    },
+    {
+      "epoch": 24.59904,
+      "grad_norm": 1.097705602645874,
+      "learning_rate": 1.1578631452581033e-05,
+      "loss": 0.4706,
+      "step": 19218
+    },
+    {
+      "epoch": 24.60032,
+      "grad_norm": 1.1707555055618286,
+      "learning_rate": 1.1576630652260905e-05,
+      "loss": 0.4738,
+      "step": 19219
+    },
+    {
+      "epoch": 24.6016,
+      "grad_norm": 1.1376193761825562,
+      "learning_rate": 1.1574629851940777e-05,
+      "loss": 0.505,
+      "step": 19220
+    },
+    {
+      "epoch": 24.60288,
+      "grad_norm": 1.26047945022583,
+      "learning_rate": 1.1572629051620649e-05,
+      "loss": 0.5244,
+      "step": 19221
+    },
+    {
+      "epoch": 24.60416,
+      "grad_norm": 1.1913353204727173,
+      "learning_rate": 1.157062825130052e-05,
+      "loss": 0.4879,
+      "step": 19222
+    },
+    {
+      "epoch": 24.60544,
+      "grad_norm": 1.1769354343414307,
+      "learning_rate": 1.1568627450980394e-05,
+      "loss": 0.5038,
+      "step": 19223
+    },
+    {
+      "epoch": 24.60672,
+      "grad_norm": 1.16858971118927,
+      "learning_rate": 1.1566626650660264e-05,
+      "loss": 0.4796,
+      "step": 19224
+    },
+    {
+      "epoch": 24.608,
+      "grad_norm": 1.1839061975479126,
+      "learning_rate": 1.1564625850340138e-05,
+      "loss": 0.5132,
+      "step": 19225
+    },
+    {
+      "epoch": 24.60928,
+      "grad_norm": 1.2122026681900024,
+      "learning_rate": 1.1562625050020008e-05,
+      "loss": 0.5023,
+      "step": 19226
+    },
+    {
+      "epoch": 24.61056,
+      "grad_norm": 1.1488457918167114,
+      "learning_rate": 1.1560624249699881e-05,
+      "loss": 0.4586,
+      "step": 19227
+    },
+    {
+      "epoch": 24.61184,
+      "grad_norm": 1.1649909019470215,
+      "learning_rate": 1.1558623449379752e-05,
+      "loss": 0.4974,
+      "step": 19228
+    },
+    {
+      "epoch": 24.61312,
+      "grad_norm": 1.180234670639038,
+      "learning_rate": 1.1556622649059625e-05,
+      "loss": 0.4725,
+      "step": 19229
+    },
+    {
+      "epoch": 24.6144,
+      "grad_norm": 1.1081424951553345,
+      "learning_rate": 1.1554621848739497e-05,
+      "loss": 0.47,
+      "step": 19230
+    },
+    {
+      "epoch": 24.61568,
+      "grad_norm": 1.1423239707946777,
+      "learning_rate": 1.1552621048419369e-05,
+      "loss": 0.4692,
+      "step": 19231
+    },
+    {
+      "epoch": 24.61696,
+      "grad_norm": 1.1524152755737305,
+      "learning_rate": 1.155062024809924e-05,
+      "loss": 0.4718,
+      "step": 19232
+    },
+    {
+      "epoch": 24.61824,
+      "grad_norm": 1.2075049877166748,
+      "learning_rate": 1.1548619447779113e-05,
+      "loss": 0.4985,
+      "step": 19233
+    },
+    {
+      "epoch": 24.61952,
+      "grad_norm": 1.1373603343963623,
+      "learning_rate": 1.1546618647458984e-05,
+      "loss": 0.5022,
+      "step": 19234
+    },
+    {
+      "epoch": 24.6208,
+      "grad_norm": 1.1606000661849976,
+      "learning_rate": 1.1544617847138856e-05,
+      "loss": 0.4829,
+      "step": 19235
+    },
+    {
+      "epoch": 24.62208,
+      "grad_norm": 1.1605327129364014,
+      "learning_rate": 1.1542617046818728e-05,
+      "loss": 0.4924,
+      "step": 19236
+    },
+    {
+      "epoch": 24.62336,
+      "grad_norm": 1.228710651397705,
+      "learning_rate": 1.15406162464986e-05,
+      "loss": 0.5122,
+      "step": 19237
+    },
+    {
+      "epoch": 24.62464,
+      "grad_norm": 1.0705116987228394,
+      "learning_rate": 1.1538615446178472e-05,
+      "loss": 0.4644,
+      "step": 19238
+    },
+    {
+      "epoch": 24.62592,
+      "grad_norm": 1.2077350616455078,
+      "learning_rate": 1.1536614645858344e-05,
+      "loss": 0.4765,
+      "step": 19239
+    },
+    {
+      "epoch": 24.6272,
+      "grad_norm": 1.1825919151306152,
+      "learning_rate": 1.1534613845538216e-05,
+      "loss": 0.47,
+      "step": 19240
+    },
+    {
+      "epoch": 24.62848,
+      "grad_norm": 1.1296800374984741,
+      "learning_rate": 1.1532613045218087e-05,
+      "loss": 0.4958,
+      "step": 19241
+    },
+    {
+      "epoch": 24.62976,
+      "grad_norm": 1.184819221496582,
+      "learning_rate": 1.153061224489796e-05,
+      "loss": 0.4932,
+      "step": 19242
+    },
+    {
+      "epoch": 24.63104,
+      "grad_norm": 1.2108218669891357,
+      "learning_rate": 1.1528611444577831e-05,
+      "loss": 0.4847,
+      "step": 19243
+    },
+    {
+      "epoch": 24.63232,
+      "grad_norm": 1.1210412979125977,
+      "learning_rate": 1.1526610644257705e-05,
+      "loss": 0.4636,
+      "step": 19244
+    },
+    {
+      "epoch": 24.6336,
+      "grad_norm": 1.1325141191482544,
+      "learning_rate": 1.1524609843937575e-05,
+      "loss": 0.4885,
+      "step": 19245
+    },
+    {
+      "epoch": 24.63488,
+      "grad_norm": 1.198911428451538,
+      "learning_rate": 1.1522609043617448e-05,
+      "loss": 0.4902,
+      "step": 19246
+    },
+    {
+      "epoch": 24.63616,
+      "grad_norm": 1.1128379106521606,
+      "learning_rate": 1.1520608243297319e-05,
+      "loss": 0.4737,
+      "step": 19247
+    },
+    {
+      "epoch": 24.63744,
+      "grad_norm": 1.147925853729248,
+      "learning_rate": 1.1518607442977192e-05,
+      "loss": 0.4618,
+      "step": 19248
+    },
+    {
+      "epoch": 24.63872,
+      "grad_norm": 1.10917329788208,
+      "learning_rate": 1.1516606642657062e-05,
+      "loss": 0.437,
+      "step": 19249
+    },
+    {
+      "epoch": 24.64,
+      "grad_norm": 1.1810179948806763,
+      "learning_rate": 1.1514605842336936e-05,
+      "loss": 0.4697,
+      "step": 19250
+    },
+    {
+      "epoch": 24.64128,
+      "grad_norm": 1.15571928024292,
+      "learning_rate": 1.1512605042016806e-05,
+      "loss": 0.4482,
+      "step": 19251
+    },
+    {
+      "epoch": 24.64256,
+      "grad_norm": 1.2140380144119263,
+      "learning_rate": 1.151060424169668e-05,
+      "loss": 0.4722,
+      "step": 19252
+    },
+    {
+      "epoch": 24.64384,
+      "grad_norm": 1.1459300518035889,
+      "learning_rate": 1.1508603441376551e-05,
+      "loss": 0.4415,
+      "step": 19253
+    },
+    {
+      "epoch": 24.64512,
+      "grad_norm": 1.2092753648757935,
+      "learning_rate": 1.1506602641056423e-05,
+      "loss": 0.4927,
+      "step": 19254
+    },
+    {
+      "epoch": 24.6464,
+      "grad_norm": 1.0979299545288086,
+      "learning_rate": 1.1504601840736295e-05,
+      "loss": 0.474,
+      "step": 19255
+    },
+    {
+      "epoch": 24.64768,
+      "grad_norm": 1.2090137004852295,
+      "learning_rate": 1.1502601040416167e-05,
+      "loss": 0.4667,
+      "step": 19256
+    },
+    {
+      "epoch": 24.64896,
+      "grad_norm": 1.0773359537124634,
+      "learning_rate": 1.1500600240096039e-05,
+      "loss": 0.4631,
+      "step": 19257
+    },
+    {
+      "epoch": 24.65024,
+      "grad_norm": 1.1589806079864502,
+      "learning_rate": 1.149859943977591e-05,
+      "loss": 0.4615,
+      "step": 19258
+    },
+    {
+      "epoch": 24.65152,
+      "grad_norm": 1.1646665334701538,
+      "learning_rate": 1.1496598639455783e-05,
+      "loss": 0.5144,
+      "step": 19259
+    },
+    {
+      "epoch": 24.6528,
+      "grad_norm": 1.1850485801696777,
+      "learning_rate": 1.1494597839135656e-05,
+      "loss": 0.4873,
+      "step": 19260
+    },
+    {
+      "epoch": 24.65408,
+      "grad_norm": 1.111702561378479,
+      "learning_rate": 1.1492597038815526e-05,
+      "loss": 0.462,
+      "step": 19261
+    },
+    {
+      "epoch": 24.65536,
+      "grad_norm": 1.0776835680007935,
+      "learning_rate": 1.14905962384954e-05,
+      "loss": 0.4624,
+      "step": 19262
+    },
+    {
+      "epoch": 24.65664,
+      "grad_norm": 1.1940804719924927,
+      "learning_rate": 1.148859543817527e-05,
+      "loss": 0.4731,
+      "step": 19263
+    },
+    {
+      "epoch": 24.65792,
+      "grad_norm": 1.0899372100830078,
+      "learning_rate": 1.1486594637855143e-05,
+      "loss": 0.4729,
+      "step": 19264
+    },
+    {
+      "epoch": 24.6592,
+      "grad_norm": 1.2058807611465454,
+      "learning_rate": 1.1484593837535014e-05,
+      "loss": 0.4518,
+      "step": 19265
+    },
+    {
+      "epoch": 24.66048,
+      "grad_norm": 1.14639413356781,
+      "learning_rate": 1.1482593037214887e-05,
+      "loss": 0.4715,
+      "step": 19266
+    },
+    {
+      "epoch": 24.66176,
+      "grad_norm": 1.0863559246063232,
+      "learning_rate": 1.1480592236894759e-05,
+      "loss": 0.4479,
+      "step": 19267
+    },
+    {
+      "epoch": 24.66304,
+      "grad_norm": 1.1600267887115479,
+      "learning_rate": 1.1478591436574631e-05,
+      "loss": 0.5148,
+      "step": 19268
+    },
+    {
+      "epoch": 24.66432,
+      "grad_norm": 1.1542725563049316,
+      "learning_rate": 1.1476590636254503e-05,
+      "loss": 0.4798,
+      "step": 19269
+    },
+    {
+      "epoch": 24.6656,
+      "grad_norm": 1.1546802520751953,
+      "learning_rate": 1.1474589835934375e-05,
+      "loss": 0.4433,
+      "step": 19270
+    },
+    {
+      "epoch": 24.66688,
+      "grad_norm": 1.123247742652893,
+      "learning_rate": 1.1472589035614246e-05,
+      "loss": 0.4422,
+      "step": 19271
+    },
+    {
+      "epoch": 24.66816,
+      "grad_norm": 1.2257593870162964,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 0.4726,
+      "step": 19272
+    },
+    {
+      "epoch": 24.66944,
+      "grad_norm": 1.196321725845337,
+      "learning_rate": 1.146858743497399e-05,
+      "loss": 0.5249,
+      "step": 19273
+    },
+    {
+      "epoch": 24.67072,
+      "grad_norm": 1.1580753326416016,
+      "learning_rate": 1.1466586634653862e-05,
+      "loss": 0.4793,
+      "step": 19274
+    },
+    {
+      "epoch": 24.672,
+      "grad_norm": 1.10203218460083,
+      "learning_rate": 1.1464585834333734e-05,
+      "loss": 0.416,
+      "step": 19275
+    },
+    {
+      "epoch": 24.67328,
+      "grad_norm": 1.0155160427093506,
+      "learning_rate": 1.1462585034013606e-05,
+      "loss": 0.4385,
+      "step": 19276
+    },
+    {
+      "epoch": 24.67456,
+      "grad_norm": 1.1510674953460693,
+      "learning_rate": 1.1460584233693478e-05,
+      "loss": 0.4783,
+      "step": 19277
+    },
+    {
+      "epoch": 24.67584,
+      "grad_norm": 1.1276487112045288,
+      "learning_rate": 1.145858343337335e-05,
+      "loss": 0.4613,
+      "step": 19278
+    },
+    {
+      "epoch": 24.67712,
+      "grad_norm": 1.1388295888900757,
+      "learning_rate": 1.1456582633053221e-05,
+      "loss": 0.4549,
+      "step": 19279
+    },
+    {
+      "epoch": 24.6784,
+      "grad_norm": 1.1739778518676758,
+      "learning_rate": 1.1454581832733093e-05,
+      "loss": 0.4569,
+      "step": 19280
+    },
+    {
+      "epoch": 24.67968,
+      "grad_norm": 1.0705592632293701,
+      "learning_rate": 1.1452581032412965e-05,
+      "loss": 0.4347,
+      "step": 19281
+    },
+    {
+      "epoch": 24.68096,
+      "grad_norm": 1.2039315700531006,
+      "learning_rate": 1.1450580232092837e-05,
+      "loss": 0.4751,
+      "step": 19282
+    },
+    {
+      "epoch": 24.68224,
+      "grad_norm": 1.1473381519317627,
+      "learning_rate": 1.144857943177271e-05,
+      "loss": 0.4583,
+      "step": 19283
+    },
+    {
+      "epoch": 24.68352,
+      "grad_norm": 1.2741856575012207,
+      "learning_rate": 1.144657863145258e-05,
+      "loss": 0.4695,
+      "step": 19284
+    },
+    {
+      "epoch": 24.6848,
+      "grad_norm": 1.1236690282821655,
+      "learning_rate": 1.1444577831132454e-05,
+      "loss": 0.4626,
+      "step": 19285
+    },
+    {
+      "epoch": 24.68608,
+      "grad_norm": 1.2310682535171509,
+      "learning_rate": 1.1442577030812324e-05,
+      "loss": 0.4727,
+      "step": 19286
+    },
+    {
+      "epoch": 24.687359999999998,
+      "grad_norm": 1.1304185390472412,
+      "learning_rate": 1.1440576230492198e-05,
+      "loss": 0.4577,
+      "step": 19287
+    },
+    {
+      "epoch": 24.68864,
+      "grad_norm": 1.2843713760375977,
+      "learning_rate": 1.1438575430172068e-05,
+      "loss": 0.4966,
+      "step": 19288
+    },
+    {
+      "epoch": 24.68992,
+      "grad_norm": 1.1501768827438354,
+      "learning_rate": 1.1436574629851942e-05,
+      "loss": 0.4763,
+      "step": 19289
+    },
+    {
+      "epoch": 24.6912,
+      "grad_norm": 1.2315309047698975,
+      "learning_rate": 1.1434573829531813e-05,
+      "loss": 0.4693,
+      "step": 19290
+    },
+    {
+      "epoch": 24.69248,
+      "grad_norm": 1.1766899824142456,
+      "learning_rate": 1.1432573029211685e-05,
+      "loss": 0.4635,
+      "step": 19291
+    },
+    {
+      "epoch": 24.69376,
+      "grad_norm": 1.2574455738067627,
+      "learning_rate": 1.1430572228891557e-05,
+      "loss": 0.5019,
+      "step": 19292
+    },
+    {
+      "epoch": 24.69504,
+      "grad_norm": 1.1796600818634033,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.4645,
+      "step": 19293
+    },
+    {
+      "epoch": 24.69632,
+      "grad_norm": 1.1451575756072998,
+      "learning_rate": 1.1426570628251301e-05,
+      "loss": 0.4712,
+      "step": 19294
+    },
+    {
+      "epoch": 24.6976,
+      "grad_norm": 1.1572251319885254,
+      "learning_rate": 1.1424569827931173e-05,
+      "loss": 0.5001,
+      "step": 19295
+    },
+    {
+      "epoch": 24.69888,
+      "grad_norm": 1.1308900117874146,
+      "learning_rate": 1.1422569027611045e-05,
+      "loss": 0.4832,
+      "step": 19296
+    },
+    {
+      "epoch": 24.70016,
+      "grad_norm": 1.1736658811569214,
+      "learning_rate": 1.1420568227290918e-05,
+      "loss": 0.4795,
+      "step": 19297
+    },
+    {
+      "epoch": 24.70144,
+      "grad_norm": 1.182158350944519,
+      "learning_rate": 1.1418567426970788e-05,
+      "loss": 0.4893,
+      "step": 19298
+    },
+    {
+      "epoch": 24.70272,
+      "grad_norm": 1.1963363885879517,
+      "learning_rate": 1.1416566626650662e-05,
+      "loss": 0.4335,
+      "step": 19299
+    },
+    {
+      "epoch": 24.704,
+      "grad_norm": 1.1854673624038696,
+      "learning_rate": 1.1414565826330532e-05,
+      "loss": 0.4765,
+      "step": 19300
+    },
+    {
+      "epoch": 24.70528,
+      "grad_norm": 1.1632603406906128,
+      "learning_rate": 1.1412565026010406e-05,
+      "loss": 0.453,
+      "step": 19301
+    },
+    {
+      "epoch": 24.70656,
+      "grad_norm": 1.1224347352981567,
+      "learning_rate": 1.1410564225690276e-05,
+      "loss": 0.4329,
+      "step": 19302
+    },
+    {
+      "epoch": 24.70784,
+      "grad_norm": 1.083587408065796,
+      "learning_rate": 1.140856342537015e-05,
+      "loss": 0.4908,
+      "step": 19303
+    },
+    {
+      "epoch": 24.70912,
+      "grad_norm": 1.1031328439712524,
+      "learning_rate": 1.140656262505002e-05,
+      "loss": 0.4961,
+      "step": 19304
+    },
+    {
+      "epoch": 24.7104,
+      "grad_norm": 1.1741230487823486,
+      "learning_rate": 1.1404561824729893e-05,
+      "loss": 0.4992,
+      "step": 19305
+    },
+    {
+      "epoch": 24.71168,
+      "grad_norm": 1.1455093622207642,
+      "learning_rate": 1.1402561024409765e-05,
+      "loss": 0.4462,
+      "step": 19306
+    },
+    {
+      "epoch": 24.71296,
+      "grad_norm": 1.1792043447494507,
+      "learning_rate": 1.1400560224089637e-05,
+      "loss": 0.4428,
+      "step": 19307
+    },
+    {
+      "epoch": 24.71424,
+      "grad_norm": 1.1586357355117798,
+      "learning_rate": 1.1398559423769509e-05,
+      "loss": 0.4899,
+      "step": 19308
+    },
+    {
+      "epoch": 24.71552,
+      "grad_norm": 1.1558829545974731,
+      "learning_rate": 1.139655862344938e-05,
+      "loss": 0.4433,
+      "step": 19309
+    },
+    {
+      "epoch": 24.7168,
+      "grad_norm": 1.1648732423782349,
+      "learning_rate": 1.1394557823129252e-05,
+      "loss": 0.4847,
+      "step": 19310
+    },
+    {
+      "epoch": 24.71808,
+      "grad_norm": 1.1810020208358765,
+      "learning_rate": 1.1392557022809124e-05,
+      "loss": 0.524,
+      "step": 19311
+    },
+    {
+      "epoch": 24.71936,
+      "grad_norm": 1.210523247718811,
+      "learning_rate": 1.1390556222488996e-05,
+      "loss": 0.4705,
+      "step": 19312
+    },
+    {
+      "epoch": 24.72064,
+      "grad_norm": 1.1746503114700317,
+      "learning_rate": 1.1388555422168868e-05,
+      "loss": 0.4782,
+      "step": 19313
+    },
+    {
+      "epoch": 24.72192,
+      "grad_norm": 1.1971492767333984,
+      "learning_rate": 1.138655462184874e-05,
+      "loss": 0.5247,
+      "step": 19314
+    },
+    {
+      "epoch": 24.7232,
+      "grad_norm": 1.219415307044983,
+      "learning_rate": 1.1384553821528612e-05,
+      "loss": 0.5261,
+      "step": 19315
+    },
+    {
+      "epoch": 24.72448,
+      "grad_norm": 1.157728910446167,
+      "learning_rate": 1.1382553021208483e-05,
+      "loss": 0.5033,
+      "step": 19316
+    },
+    {
+      "epoch": 24.72576,
+      "grad_norm": 1.21958327293396,
+      "learning_rate": 1.1380552220888355e-05,
+      "loss": 0.5063,
+      "step": 19317
+    },
+    {
+      "epoch": 24.72704,
+      "grad_norm": 1.183083176612854,
+      "learning_rate": 1.1378551420568227e-05,
+      "loss": 0.5095,
+      "step": 19318
+    },
+    {
+      "epoch": 24.72832,
+      "grad_norm": 1.1700811386108398,
+      "learning_rate": 1.1376550620248099e-05,
+      "loss": 0.4605,
+      "step": 19319
+    },
+    {
+      "epoch": 24.7296,
+      "grad_norm": 1.1539628505706787,
+      "learning_rate": 1.1374549819927973e-05,
+      "loss": 0.49,
+      "step": 19320
+    },
+    {
+      "epoch": 24.73088,
+      "grad_norm": 1.0613603591918945,
+      "learning_rate": 1.1372549019607843e-05,
+      "loss": 0.4647,
+      "step": 19321
+    },
+    {
+      "epoch": 24.73216,
+      "grad_norm": 1.124666690826416,
+      "learning_rate": 1.1370548219287716e-05,
+      "loss": 0.4497,
+      "step": 19322
+    },
+    {
+      "epoch": 24.73344,
+      "grad_norm": 1.1756181716918945,
+      "learning_rate": 1.1368547418967586e-05,
+      "loss": 0.4863,
+      "step": 19323
+    },
+    {
+      "epoch": 24.73472,
+      "grad_norm": 1.0413018465042114,
+      "learning_rate": 1.136654661864746e-05,
+      "loss": 0.4702,
+      "step": 19324
+    },
+    {
+      "epoch": 24.736,
+      "grad_norm": 1.0692871809005737,
+      "learning_rate": 1.136454581832733e-05,
+      "loss": 0.4444,
+      "step": 19325
+    },
+    {
+      "epoch": 24.73728,
+      "grad_norm": 1.1003504991531372,
+      "learning_rate": 1.1362545018007204e-05,
+      "loss": 0.4332,
+      "step": 19326
+    },
+    {
+      "epoch": 24.73856,
+      "grad_norm": 1.126057744026184,
+      "learning_rate": 1.1360544217687076e-05,
+      "loss": 0.4527,
+      "step": 19327
+    },
+    {
+      "epoch": 24.73984,
+      "grad_norm": 1.1299209594726562,
+      "learning_rate": 1.1358543417366947e-05,
+      "loss": 0.4338,
+      "step": 19328
+    },
+    {
+      "epoch": 24.74112,
+      "grad_norm": 1.0575133562088013,
+      "learning_rate": 1.135654261704682e-05,
+      "loss": 0.4591,
+      "step": 19329
+    },
+    {
+      "epoch": 24.7424,
+      "grad_norm": 1.1829102039337158,
+      "learning_rate": 1.1354541816726691e-05,
+      "loss": 0.4912,
+      "step": 19330
+    },
+    {
+      "epoch": 24.74368,
+      "grad_norm": 1.154536247253418,
+      "learning_rate": 1.1352541016406563e-05,
+      "loss": 0.5155,
+      "step": 19331
+    },
+    {
+      "epoch": 24.74496,
+      "grad_norm": 1.1477676630020142,
+      "learning_rate": 1.1350540216086435e-05,
+      "loss": 0.4689,
+      "step": 19332
+    },
+    {
+      "epoch": 24.74624,
+      "grad_norm": 1.1423364877700806,
+      "learning_rate": 1.1348539415766307e-05,
+      "loss": 0.4852,
+      "step": 19333
+    },
+    {
+      "epoch": 24.74752,
+      "grad_norm": 1.1712870597839355,
+      "learning_rate": 1.1346538615446179e-05,
+      "loss": 0.4899,
+      "step": 19334
+    },
+    {
+      "epoch": 24.7488,
+      "grad_norm": 1.229701042175293,
+      "learning_rate": 1.134453781512605e-05,
+      "loss": 0.4972,
+      "step": 19335
+    },
+    {
+      "epoch": 24.75008,
+      "grad_norm": 1.098663330078125,
+      "learning_rate": 1.1342537014805924e-05,
+      "loss": 0.4796,
+      "step": 19336
+    },
+    {
+      "epoch": 24.75136,
+      "grad_norm": 1.0754587650299072,
+      "learning_rate": 1.1340536214485794e-05,
+      "loss": 0.4549,
+      "step": 19337
+    },
+    {
+      "epoch": 24.75264,
+      "grad_norm": 1.1610110998153687,
+      "learning_rate": 1.1338535414165668e-05,
+      "loss": 0.4837,
+      "step": 19338
+    },
+    {
+      "epoch": 24.75392,
+      "grad_norm": 1.2524155378341675,
+      "learning_rate": 1.1336534613845538e-05,
+      "loss": 0.5235,
+      "step": 19339
+    },
+    {
+      "epoch": 24.7552,
+      "grad_norm": 1.139482021331787,
+      "learning_rate": 1.1334533813525411e-05,
+      "loss": 0.5018,
+      "step": 19340
+    },
+    {
+      "epoch": 24.75648,
+      "grad_norm": 1.2066160440444946,
+      "learning_rate": 1.1332533013205282e-05,
+      "loss": 0.5218,
+      "step": 19341
+    },
+    {
+      "epoch": 24.75776,
+      "grad_norm": 1.0574979782104492,
+      "learning_rate": 1.1330532212885155e-05,
+      "loss": 0.4569,
+      "step": 19342
+    },
+    {
+      "epoch": 24.75904,
+      "grad_norm": 1.1374906301498413,
+      "learning_rate": 1.1328531412565027e-05,
+      "loss": 0.4353,
+      "step": 19343
+    },
+    {
+      "epoch": 24.76032,
+      "grad_norm": 1.135045051574707,
+      "learning_rate": 1.1326530612244899e-05,
+      "loss": 0.4811,
+      "step": 19344
+    },
+    {
+      "epoch": 24.7616,
+      "grad_norm": 1.1101502180099487,
+      "learning_rate": 1.132452981192477e-05,
+      "loss": 0.4816,
+      "step": 19345
+    },
+    {
+      "epoch": 24.76288,
+      "grad_norm": 1.1571624279022217,
+      "learning_rate": 1.1322529011604643e-05,
+      "loss": 0.4853,
+      "step": 19346
+    },
+    {
+      "epoch": 24.76416,
+      "grad_norm": 1.216991662979126,
+      "learning_rate": 1.1320528211284514e-05,
+      "loss": 0.4746,
+      "step": 19347
+    },
+    {
+      "epoch": 24.76544,
+      "grad_norm": 1.1886634826660156,
+      "learning_rate": 1.1318527410964386e-05,
+      "loss": 0.4852,
+      "step": 19348
+    },
+    {
+      "epoch": 24.76672,
+      "grad_norm": 1.2373932600021362,
+      "learning_rate": 1.1316526610644258e-05,
+      "loss": 0.463,
+      "step": 19349
+    },
+    {
+      "epoch": 24.768,
+      "grad_norm": 1.2170404195785522,
+      "learning_rate": 1.1314525810324132e-05,
+      "loss": 0.5278,
+      "step": 19350
+    },
+    {
+      "epoch": 24.76928,
+      "grad_norm": 1.1271648406982422,
+      "learning_rate": 1.1312525010004002e-05,
+      "loss": 0.4666,
+      "step": 19351
+    },
+    {
+      "epoch": 24.77056,
+      "grad_norm": 1.2704542875289917,
+      "learning_rate": 1.1310524209683875e-05,
+      "loss": 0.541,
+      "step": 19352
+    },
+    {
+      "epoch": 24.77184,
+      "grad_norm": 1.1266765594482422,
+      "learning_rate": 1.1308523409363746e-05,
+      "loss": 0.4432,
+      "step": 19353
+    },
+    {
+      "epoch": 24.77312,
+      "grad_norm": 1.1655040979385376,
+      "learning_rate": 1.1306522609043619e-05,
+      "loss": 0.4932,
+      "step": 19354
+    },
+    {
+      "epoch": 24.7744,
+      "grad_norm": 1.0827305316925049,
+      "learning_rate": 1.130452180872349e-05,
+      "loss": 0.4499,
+      "step": 19355
+    },
+    {
+      "epoch": 24.77568,
+      "grad_norm": 1.1636383533477783,
+      "learning_rate": 1.1302521008403363e-05,
+      "loss": 0.5046,
+      "step": 19356
+    },
+    {
+      "epoch": 24.77696,
+      "grad_norm": 1.1613502502441406,
+      "learning_rate": 1.1300520208083235e-05,
+      "loss": 0.454,
+      "step": 19357
+    },
+    {
+      "epoch": 24.77824,
+      "grad_norm": 1.1521443128585815,
+      "learning_rate": 1.1298519407763106e-05,
+      "loss": 0.4479,
+      "step": 19358
+    },
+    {
+      "epoch": 24.77952,
+      "grad_norm": 1.129206657409668,
+      "learning_rate": 1.1296518607442978e-05,
+      "loss": 0.4722,
+      "step": 19359
+    },
+    {
+      "epoch": 24.7808,
+      "grad_norm": 1.1790378093719482,
+      "learning_rate": 1.129451780712285e-05,
+      "loss": 0.4682,
+      "step": 19360
+    },
+    {
+      "epoch": 24.78208,
+      "grad_norm": 1.1063263416290283,
+      "learning_rate": 1.1292517006802722e-05,
+      "loss": 0.4484,
+      "step": 19361
+    },
+    {
+      "epoch": 24.78336,
+      "grad_norm": 1.1789052486419678,
+      "learning_rate": 1.1290516206482594e-05,
+      "loss": 0.4894,
+      "step": 19362
+    },
+    {
+      "epoch": 24.78464,
+      "grad_norm": 1.210119605064392,
+      "learning_rate": 1.1288515406162466e-05,
+      "loss": 0.533,
+      "step": 19363
+    },
+    {
+      "epoch": 24.78592,
+      "grad_norm": 1.0903270244598389,
+      "learning_rate": 1.1286514605842338e-05,
+      "loss": 0.4602,
+      "step": 19364
+    },
+    {
+      "epoch": 24.7872,
+      "grad_norm": 1.1570169925689697,
+      "learning_rate": 1.128451380552221e-05,
+      "loss": 0.4611,
+      "step": 19365
+    },
+    {
+      "epoch": 24.78848,
+      "grad_norm": 1.0663702487945557,
+      "learning_rate": 1.1282513005202081e-05,
+      "loss": 0.4379,
+      "step": 19366
+    },
+    {
+      "epoch": 24.78976,
+      "grad_norm": 1.067586898803711,
+      "learning_rate": 1.1280512204881953e-05,
+      "loss": 0.4306,
+      "step": 19367
+    },
+    {
+      "epoch": 24.79104,
+      "grad_norm": 1.1584056615829468,
+      "learning_rate": 1.1278511404561825e-05,
+      "loss": 0.4455,
+      "step": 19368
+    },
+    {
+      "epoch": 24.79232,
+      "grad_norm": 1.188567042350769,
+      "learning_rate": 1.1276510604241697e-05,
+      "loss": 0.5147,
+      "step": 19369
+    },
+    {
+      "epoch": 24.7936,
+      "grad_norm": 1.2147189378738403,
+      "learning_rate": 1.1274509803921569e-05,
+      "loss": 0.4957,
+      "step": 19370
+    },
+    {
+      "epoch": 24.79488,
+      "grad_norm": 1.1530333757400513,
+      "learning_rate": 1.127250900360144e-05,
+      "loss": 0.4771,
+      "step": 19371
+    },
+    {
+      "epoch": 24.79616,
+      "grad_norm": 1.1799629926681519,
+      "learning_rate": 1.1270508203281312e-05,
+      "loss": 0.5172,
+      "step": 19372
+    },
+    {
+      "epoch": 24.79744,
+      "grad_norm": 1.1808135509490967,
+      "learning_rate": 1.1268507402961186e-05,
+      "loss": 0.4899,
+      "step": 19373
+    },
+    {
+      "epoch": 24.79872,
+      "grad_norm": 1.169081687927246,
+      "learning_rate": 1.1266506602641056e-05,
+      "loss": 0.4794,
+      "step": 19374
+    },
+    {
+      "epoch": 24.8,
+      "grad_norm": 1.2258483171463013,
+      "learning_rate": 1.126450580232093e-05,
+      "loss": 0.4697,
+      "step": 19375
+    },
+    {
+      "epoch": 24.80128,
+      "grad_norm": 1.1741260290145874,
+      "learning_rate": 1.12625050020008e-05,
+      "loss": 0.5146,
+      "step": 19376
+    },
+    {
+      "epoch": 24.80256,
+      "grad_norm": 1.1346204280853271,
+      "learning_rate": 1.1260504201680673e-05,
+      "loss": 0.4403,
+      "step": 19377
+    },
+    {
+      "epoch": 24.80384,
+      "grad_norm": 1.1797999143600464,
+      "learning_rate": 1.1258503401360544e-05,
+      "loss": 0.449,
+      "step": 19378
+    },
+    {
+      "epoch": 24.80512,
+      "grad_norm": 1.2282897233963013,
+      "learning_rate": 1.1256502601040417e-05,
+      "loss": 0.476,
+      "step": 19379
+    },
+    {
+      "epoch": 24.8064,
+      "grad_norm": 1.1009891033172607,
+      "learning_rate": 1.1254501800720289e-05,
+      "loss": 0.498,
+      "step": 19380
+    },
+    {
+      "epoch": 24.80768,
+      "grad_norm": 1.0904237031936646,
+      "learning_rate": 1.1252501000400161e-05,
+      "loss": 0.4462,
+      "step": 19381
+    },
+    {
+      "epoch": 24.80896,
+      "grad_norm": 1.162895679473877,
+      "learning_rate": 1.1250500200080033e-05,
+      "loss": 0.4788,
+      "step": 19382
+    },
+    {
+      "epoch": 24.81024,
+      "grad_norm": 1.1506489515304565,
+      "learning_rate": 1.1248499399759905e-05,
+      "loss": 0.4834,
+      "step": 19383
+    },
+    {
+      "epoch": 24.81152,
+      "grad_norm": 1.1401734352111816,
+      "learning_rate": 1.1246498599439776e-05,
+      "loss": 0.4822,
+      "step": 19384
+    },
+    {
+      "epoch": 24.8128,
+      "grad_norm": 1.0999462604522705,
+      "learning_rate": 1.1244497799119648e-05,
+      "loss": 0.4681,
+      "step": 19385
+    },
+    {
+      "epoch": 24.81408,
+      "grad_norm": 1.060306429862976,
+      "learning_rate": 1.124249699879952e-05,
+      "loss": 0.4561,
+      "step": 19386
+    },
+    {
+      "epoch": 24.81536,
+      "grad_norm": 1.2425814867019653,
+      "learning_rate": 1.1240496198479392e-05,
+      "loss": 0.5074,
+      "step": 19387
+    },
+    {
+      "epoch": 24.81664,
+      "grad_norm": 1.209997534751892,
+      "learning_rate": 1.1238495398159264e-05,
+      "loss": 0.5151,
+      "step": 19388
+    },
+    {
+      "epoch": 24.81792,
+      "grad_norm": 1.1693835258483887,
+      "learning_rate": 1.1236494597839137e-05,
+      "loss": 0.5006,
+      "step": 19389
+    },
+    {
+      "epoch": 24.8192,
+      "grad_norm": 1.1733357906341553,
+      "learning_rate": 1.1234493797519008e-05,
+      "loss": 0.5059,
+      "step": 19390
+    },
+    {
+      "epoch": 24.82048,
+      "grad_norm": 1.16240656375885,
+      "learning_rate": 1.1232492997198881e-05,
+      "loss": 0.458,
+      "step": 19391
+    },
+    {
+      "epoch": 24.82176,
+      "grad_norm": 1.0904651880264282,
+      "learning_rate": 1.1230492196878751e-05,
+      "loss": 0.4414,
+      "step": 19392
+    },
+    {
+      "epoch": 24.82304,
+      "grad_norm": 1.1327263116836548,
+      "learning_rate": 1.1228491396558625e-05,
+      "loss": 0.4849,
+      "step": 19393
+    },
+    {
+      "epoch": 24.82432,
+      "grad_norm": 1.2412240505218506,
+      "learning_rate": 1.1226490596238495e-05,
+      "loss": 0.5405,
+      "step": 19394
+    },
+    {
+      "epoch": 24.8256,
+      "grad_norm": 1.1290397644042969,
+      "learning_rate": 1.1224489795918369e-05,
+      "loss": 0.4745,
+      "step": 19395
+    },
+    {
+      "epoch": 24.82688,
+      "grad_norm": 1.1640607118606567,
+      "learning_rate": 1.122248899559824e-05,
+      "loss": 0.4839,
+      "step": 19396
+    },
+    {
+      "epoch": 24.82816,
+      "grad_norm": 1.1949489116668701,
+      "learning_rate": 1.1220488195278112e-05,
+      "loss": 0.4592,
+      "step": 19397
+    },
+    {
+      "epoch": 24.829439999999998,
+      "grad_norm": 1.164475440979004,
+      "learning_rate": 1.1218487394957984e-05,
+      "loss": 0.4737,
+      "step": 19398
+    },
+    {
+      "epoch": 24.83072,
+      "grad_norm": 1.185512661933899,
+      "learning_rate": 1.1216486594637856e-05,
+      "loss": 0.451,
+      "step": 19399
+    },
+    {
+      "epoch": 24.832,
+      "grad_norm": 1.096878170967102,
+      "learning_rate": 1.1214485794317728e-05,
+      "loss": 0.5043,
+      "step": 19400
+    },
+    {
+      "epoch": 24.83328,
+      "grad_norm": 1.2109886407852173,
+      "learning_rate": 1.12124849939976e-05,
+      "loss": 0.569,
+      "step": 19401
+    },
+    {
+      "epoch": 24.83456,
+      "grad_norm": 1.1665548086166382,
+      "learning_rate": 1.1210484193677472e-05,
+      "loss": 0.5026,
+      "step": 19402
+    },
+    {
+      "epoch": 24.83584,
+      "grad_norm": 1.1184903383255005,
+      "learning_rate": 1.1208483393357343e-05,
+      "loss": 0.4835,
+      "step": 19403
+    },
+    {
+      "epoch": 24.83712,
+      "grad_norm": 1.2306338548660278,
+      "learning_rate": 1.1206482593037215e-05,
+      "loss": 0.4726,
+      "step": 19404
+    },
+    {
+      "epoch": 24.8384,
+      "grad_norm": 1.2000937461853027,
+      "learning_rate": 1.1204481792717087e-05,
+      "loss": 0.4827,
+      "step": 19405
+    },
+    {
+      "epoch": 24.83968,
+      "grad_norm": 1.1614446640014648,
+      "learning_rate": 1.1202480992396959e-05,
+      "loss": 0.4524,
+      "step": 19406
+    },
+    {
+      "epoch": 24.84096,
+      "grad_norm": 1.1615883111953735,
+      "learning_rate": 1.1200480192076831e-05,
+      "loss": 0.5126,
+      "step": 19407
+    },
+    {
+      "epoch": 24.84224,
+      "grad_norm": 1.1704049110412598,
+      "learning_rate": 1.1198479391756703e-05,
+      "loss": 0.4715,
+      "step": 19408
+    },
+    {
+      "epoch": 24.84352,
+      "grad_norm": 1.2115988731384277,
+      "learning_rate": 1.1196478591436575e-05,
+      "loss": 0.4983,
+      "step": 19409
+    },
+    {
+      "epoch": 24.8448,
+      "grad_norm": 1.1601941585540771,
+      "learning_rate": 1.1194477791116448e-05,
+      "loss": 0.5095,
+      "step": 19410
+    },
+    {
+      "epoch": 24.84608,
+      "grad_norm": 1.1305478811264038,
+      "learning_rate": 1.1192476990796318e-05,
+      "loss": 0.4968,
+      "step": 19411
+    },
+    {
+      "epoch": 24.84736,
+      "grad_norm": 1.243231177330017,
+      "learning_rate": 1.1190476190476192e-05,
+      "loss": 0.4912,
+      "step": 19412
+    },
+    {
+      "epoch": 24.84864,
+      "grad_norm": 1.0542876720428467,
+      "learning_rate": 1.1188475390156062e-05,
+      "loss": 0.4489,
+      "step": 19413
+    },
+    {
+      "epoch": 24.84992,
+      "grad_norm": 1.184417486190796,
+      "learning_rate": 1.1186474589835936e-05,
+      "loss": 0.4938,
+      "step": 19414
+    },
+    {
+      "epoch": 24.8512,
+      "grad_norm": 1.2762627601623535,
+      "learning_rate": 1.1184473789515806e-05,
+      "loss": 0.5049,
+      "step": 19415
+    },
+    {
+      "epoch": 24.85248,
+      "grad_norm": 1.0796477794647217,
+      "learning_rate": 1.118247298919568e-05,
+      "loss": 0.4241,
+      "step": 19416
+    },
+    {
+      "epoch": 24.85376,
+      "grad_norm": 1.2161225080490112,
+      "learning_rate": 1.118047218887555e-05,
+      "loss": 0.5025,
+      "step": 19417
+    },
+    {
+      "epoch": 24.85504,
+      "grad_norm": 1.180956482887268,
+      "learning_rate": 1.1178471388555423e-05,
+      "loss": 0.4798,
+      "step": 19418
+    },
+    {
+      "epoch": 24.85632,
+      "grad_norm": 1.1922006607055664,
+      "learning_rate": 1.1176470588235295e-05,
+      "loss": 0.4928,
+      "step": 19419
+    },
+    {
+      "epoch": 24.8576,
+      "grad_norm": 1.1288496255874634,
+      "learning_rate": 1.1174469787915167e-05,
+      "loss": 0.4416,
+      "step": 19420
+    },
+    {
+      "epoch": 24.85888,
+      "grad_norm": 1.1524951457977295,
+      "learning_rate": 1.1172468987595039e-05,
+      "loss": 0.4827,
+      "step": 19421
+    },
+    {
+      "epoch": 24.86016,
+      "grad_norm": 1.239288330078125,
+      "learning_rate": 1.117046818727491e-05,
+      "loss": 0.4972,
+      "step": 19422
+    },
+    {
+      "epoch": 24.86144,
+      "grad_norm": 1.1284489631652832,
+      "learning_rate": 1.1168467386954782e-05,
+      "loss": 0.4293,
+      "step": 19423
+    },
+    {
+      "epoch": 24.86272,
+      "grad_norm": 1.178391456604004,
+      "learning_rate": 1.1166466586634654e-05,
+      "loss": 0.4698,
+      "step": 19424
+    },
+    {
+      "epoch": 24.864,
+      "grad_norm": 1.1781879663467407,
+      "learning_rate": 1.1164465786314526e-05,
+      "loss": 0.4427,
+      "step": 19425
+    },
+    {
+      "epoch": 24.86528,
+      "grad_norm": 1.1736154556274414,
+      "learning_rate": 1.11624649859944e-05,
+      "loss": 0.4899,
+      "step": 19426
+    },
+    {
+      "epoch": 24.86656,
+      "grad_norm": 1.210067868232727,
+      "learning_rate": 1.116046418567427e-05,
+      "loss": 0.4746,
+      "step": 19427
+    },
+    {
+      "epoch": 24.86784,
+      "grad_norm": 1.1524900197982788,
+      "learning_rate": 1.1158463385354143e-05,
+      "loss": 0.4755,
+      "step": 19428
+    },
+    {
+      "epoch": 24.86912,
+      "grad_norm": 1.077675461769104,
+      "learning_rate": 1.1156462585034013e-05,
+      "loss": 0.4248,
+      "step": 19429
+    },
+    {
+      "epoch": 24.8704,
+      "grad_norm": 1.1400989294052124,
+      "learning_rate": 1.1154461784713887e-05,
+      "loss": 0.4719,
+      "step": 19430
+    },
+    {
+      "epoch": 24.87168,
+      "grad_norm": 1.0998234748840332,
+      "learning_rate": 1.1152460984393757e-05,
+      "loss": 0.4563,
+      "step": 19431
+    },
+    {
+      "epoch": 24.87296,
+      "grad_norm": 1.2079157829284668,
+      "learning_rate": 1.115046018407363e-05,
+      "loss": 0.508,
+      "step": 19432
+    },
+    {
+      "epoch": 24.87424,
+      "grad_norm": 1.118981122970581,
+      "learning_rate": 1.1148459383753503e-05,
+      "loss": 0.4577,
+      "step": 19433
+    },
+    {
+      "epoch": 24.87552,
+      "grad_norm": 1.1195893287658691,
+      "learning_rate": 1.1146458583433374e-05,
+      "loss": 0.4688,
+      "step": 19434
+    },
+    {
+      "epoch": 24.8768,
+      "grad_norm": 1.1069542169570923,
+      "learning_rate": 1.1144457783113246e-05,
+      "loss": 0.4522,
+      "step": 19435
+    },
+    {
+      "epoch": 24.87808,
+      "grad_norm": 1.2050132751464844,
+      "learning_rate": 1.1142456982793118e-05,
+      "loss": 0.4965,
+      "step": 19436
+    },
+    {
+      "epoch": 24.87936,
+      "grad_norm": 1.0464030504226685,
+      "learning_rate": 1.114045618247299e-05,
+      "loss": 0.4366,
+      "step": 19437
+    },
+    {
+      "epoch": 24.88064,
+      "grad_norm": 1.1153632402420044,
+      "learning_rate": 1.1138455382152862e-05,
+      "loss": 0.454,
+      "step": 19438
+    },
+    {
+      "epoch": 24.88192,
+      "grad_norm": 1.178065299987793,
+      "learning_rate": 1.1136454581832734e-05,
+      "loss": 0.4914,
+      "step": 19439
+    },
+    {
+      "epoch": 24.8832,
+      "grad_norm": 1.2401835918426514,
+      "learning_rate": 1.1134453781512606e-05,
+      "loss": 0.4877,
+      "step": 19440
+    },
+    {
+      "epoch": 24.88448,
+      "grad_norm": 1.2349568605422974,
+      "learning_rate": 1.1132452981192477e-05,
+      "loss": 0.5027,
+      "step": 19441
+    },
+    {
+      "epoch": 24.88576,
+      "grad_norm": 1.188691258430481,
+      "learning_rate": 1.113045218087235e-05,
+      "loss": 0.4862,
+      "step": 19442
+    },
+    {
+      "epoch": 24.88704,
+      "grad_norm": 1.2274821996688843,
+      "learning_rate": 1.1128451380552221e-05,
+      "loss": 0.489,
+      "step": 19443
+    },
+    {
+      "epoch": 24.88832,
+      "grad_norm": 1.2157435417175293,
+      "learning_rate": 1.1126450580232093e-05,
+      "loss": 0.518,
+      "step": 19444
+    },
+    {
+      "epoch": 24.8896,
+      "grad_norm": 1.1647244691848755,
+      "learning_rate": 1.1124449779911965e-05,
+      "loss": 0.5051,
+      "step": 19445
+    },
+    {
+      "epoch": 24.89088,
+      "grad_norm": 1.1653064489364624,
+      "learning_rate": 1.1122448979591837e-05,
+      "loss": 0.4646,
+      "step": 19446
+    },
+    {
+      "epoch": 24.89216,
+      "grad_norm": 1.1139925718307495,
+      "learning_rate": 1.1120448179271709e-05,
+      "loss": 0.4621,
+      "step": 19447
+    },
+    {
+      "epoch": 24.89344,
+      "grad_norm": 1.1649415493011475,
+      "learning_rate": 1.111844737895158e-05,
+      "loss": 0.4833,
+      "step": 19448
+    },
+    {
+      "epoch": 24.89472,
+      "grad_norm": 1.2173422574996948,
+      "learning_rate": 1.1116446578631454e-05,
+      "loss": 0.5032,
+      "step": 19449
+    },
+    {
+      "epoch": 24.896,
+      "grad_norm": 1.1215828657150269,
+      "learning_rate": 1.1114445778311324e-05,
+      "loss": 0.4903,
+      "step": 19450
+    },
+    {
+      "epoch": 24.89728,
+      "grad_norm": 1.179179310798645,
+      "learning_rate": 1.1112444977991198e-05,
+      "loss": 0.4854,
+      "step": 19451
+    },
+    {
+      "epoch": 24.89856,
+      "grad_norm": 1.1343804597854614,
+      "learning_rate": 1.1110444177671068e-05,
+      "loss": 0.4477,
+      "step": 19452
+    },
+    {
+      "epoch": 24.89984,
+      "grad_norm": 1.1283519268035889,
+      "learning_rate": 1.1108443377350941e-05,
+      "loss": 0.4814,
+      "step": 19453
+    },
+    {
+      "epoch": 24.90112,
+      "grad_norm": 1.117431879043579,
+      "learning_rate": 1.1106442577030812e-05,
+      "loss": 0.4512,
+      "step": 19454
+    },
+    {
+      "epoch": 24.9024,
+      "grad_norm": 1.1315174102783203,
+      "learning_rate": 1.1104441776710685e-05,
+      "loss": 0.4423,
+      "step": 19455
+    },
+    {
+      "epoch": 24.90368,
+      "grad_norm": 1.2112675905227661,
+      "learning_rate": 1.1102440976390557e-05,
+      "loss": 0.5169,
+      "step": 19456
+    },
+    {
+      "epoch": 24.90496,
+      "grad_norm": 1.1385278701782227,
+      "learning_rate": 1.1100440176070429e-05,
+      "loss": 0.4717,
+      "step": 19457
+    },
+    {
+      "epoch": 24.90624,
+      "grad_norm": 1.1794296503067017,
+      "learning_rate": 1.10984393757503e-05,
+      "loss": 0.503,
+      "step": 19458
+    },
+    {
+      "epoch": 24.90752,
+      "grad_norm": 1.1522889137268066,
+      "learning_rate": 1.1096438575430172e-05,
+      "loss": 0.4582,
+      "step": 19459
+    },
+    {
+      "epoch": 24.9088,
+      "grad_norm": 1.233756422996521,
+      "learning_rate": 1.1094437775110044e-05,
+      "loss": 0.521,
+      "step": 19460
+    },
+    {
+      "epoch": 24.91008,
+      "grad_norm": 1.1709715127944946,
+      "learning_rate": 1.1092436974789916e-05,
+      "loss": 0.492,
+      "step": 19461
+    },
+    {
+      "epoch": 24.91136,
+      "grad_norm": 1.193808913230896,
+      "learning_rate": 1.1090436174469788e-05,
+      "loss": 0.4568,
+      "step": 19462
+    },
+    {
+      "epoch": 24.91264,
+      "grad_norm": 1.14645254611969,
+      "learning_rate": 1.1088435374149662e-05,
+      "loss": 0.4904,
+      "step": 19463
+    },
+    {
+      "epoch": 24.91392,
+      "grad_norm": 1.100979208946228,
+      "learning_rate": 1.1086434573829532e-05,
+      "loss": 0.4828,
+      "step": 19464
+    },
+    {
+      "epoch": 24.9152,
+      "grad_norm": 1.1360437870025635,
+      "learning_rate": 1.1084433773509405e-05,
+      "loss": 0.4519,
+      "step": 19465
+    },
+    {
+      "epoch": 24.91648,
+      "grad_norm": 1.0801247358322144,
+      "learning_rate": 1.1082432973189275e-05,
+      "loss": 0.4426,
+      "step": 19466
+    },
+    {
+      "epoch": 24.91776,
+      "grad_norm": 1.2544835805892944,
+      "learning_rate": 1.1080432172869149e-05,
+      "loss": 0.5123,
+      "step": 19467
+    },
+    {
+      "epoch": 24.91904,
+      "grad_norm": 1.1919646263122559,
+      "learning_rate": 1.107843137254902e-05,
+      "loss": 0.4854,
+      "step": 19468
+    },
+    {
+      "epoch": 24.92032,
+      "grad_norm": 1.1021586656570435,
+      "learning_rate": 1.1076430572228893e-05,
+      "loss": 0.5155,
+      "step": 19469
+    },
+    {
+      "epoch": 24.9216,
+      "grad_norm": 1.230746865272522,
+      "learning_rate": 1.1074429771908763e-05,
+      "loss": 0.5456,
+      "step": 19470
+    },
+    {
+      "epoch": 24.92288,
+      "grad_norm": 1.0849162340164185,
+      "learning_rate": 1.1072428971588636e-05,
+      "loss": 0.4524,
+      "step": 19471
+    },
+    {
+      "epoch": 24.92416,
+      "grad_norm": 1.1267224550247192,
+      "learning_rate": 1.1070428171268508e-05,
+      "loss": 0.4821,
+      "step": 19472
+    },
+    {
+      "epoch": 24.925440000000002,
+      "grad_norm": 1.134163737297058,
+      "learning_rate": 1.106842737094838e-05,
+      "loss": 0.4881,
+      "step": 19473
+    },
+    {
+      "epoch": 24.92672,
+      "grad_norm": 1.142830729484558,
+      "learning_rate": 1.1066426570628252e-05,
+      "loss": 0.5139,
+      "step": 19474
+    },
+    {
+      "epoch": 24.928,
+      "grad_norm": 1.1527884006500244,
+      "learning_rate": 1.1064425770308124e-05,
+      "loss": 0.4652,
+      "step": 19475
+    },
+    {
+      "epoch": 24.92928,
+      "grad_norm": 1.1849658489227295,
+      "learning_rate": 1.1062424969987996e-05,
+      "loss": 0.5047,
+      "step": 19476
+    },
+    {
+      "epoch": 24.93056,
+      "grad_norm": 1.1960779428482056,
+      "learning_rate": 1.1060424169667868e-05,
+      "loss": 0.4927,
+      "step": 19477
+    },
+    {
+      "epoch": 24.93184,
+      "grad_norm": 1.118734359741211,
+      "learning_rate": 1.105842336934774e-05,
+      "loss": 0.463,
+      "step": 19478
+    },
+    {
+      "epoch": 24.93312,
+      "grad_norm": 1.1411648988723755,
+      "learning_rate": 1.1056422569027611e-05,
+      "loss": 0.4535,
+      "step": 19479
+    },
+    {
+      "epoch": 24.9344,
+      "grad_norm": 1.1188914775848389,
+      "learning_rate": 1.1054421768707483e-05,
+      "loss": 0.4437,
+      "step": 19480
+    },
+    {
+      "epoch": 24.93568,
+      "grad_norm": 1.1471185684204102,
+      "learning_rate": 1.1052420968387355e-05,
+      "loss": 0.4782,
+      "step": 19481
+    },
+    {
+      "epoch": 24.93696,
+      "grad_norm": 1.2301186323165894,
+      "learning_rate": 1.1050420168067227e-05,
+      "loss": 0.4759,
+      "step": 19482
+    },
+    {
+      "epoch": 24.93824,
+      "grad_norm": 1.175561547279358,
+      "learning_rate": 1.1048419367747099e-05,
+      "loss": 0.4798,
+      "step": 19483
+    },
+    {
+      "epoch": 24.93952,
+      "grad_norm": 1.172258734703064,
+      "learning_rate": 1.104641856742697e-05,
+      "loss": 0.465,
+      "step": 19484
+    },
+    {
+      "epoch": 24.9408,
+      "grad_norm": 1.1110254526138306,
+      "learning_rate": 1.1044417767106842e-05,
+      "loss": 0.4597,
+      "step": 19485
+    },
+    {
+      "epoch": 24.94208,
+      "grad_norm": 1.0964692831039429,
+      "learning_rate": 1.1042416966786716e-05,
+      "loss": 0.4654,
+      "step": 19486
+    },
+    {
+      "epoch": 24.94336,
+      "grad_norm": 1.187997579574585,
+      "learning_rate": 1.1040416166466586e-05,
+      "loss": 0.4786,
+      "step": 19487
+    },
+    {
+      "epoch": 24.94464,
+      "grad_norm": 1.1386297941207886,
+      "learning_rate": 1.103841536614646e-05,
+      "loss": 0.4923,
+      "step": 19488
+    },
+    {
+      "epoch": 24.94592,
+      "grad_norm": 1.0917891263961792,
+      "learning_rate": 1.103641456582633e-05,
+      "loss": 0.4856,
+      "step": 19489
+    },
+    {
+      "epoch": 24.9472,
+      "grad_norm": 1.1903581619262695,
+      "learning_rate": 1.1034413765506203e-05,
+      "loss": 0.5102,
+      "step": 19490
+    },
+    {
+      "epoch": 24.94848,
+      "grad_norm": 1.118826985359192,
+      "learning_rate": 1.1032412965186074e-05,
+      "loss": 0.4864,
+      "step": 19491
+    },
+    {
+      "epoch": 24.94976,
+      "grad_norm": 1.0741007328033447,
+      "learning_rate": 1.1030412164865947e-05,
+      "loss": 0.4401,
+      "step": 19492
+    },
+    {
+      "epoch": 24.95104,
+      "grad_norm": 1.154763102531433,
+      "learning_rate": 1.1028411364545819e-05,
+      "loss": 0.493,
+      "step": 19493
+    },
+    {
+      "epoch": 24.95232,
+      "grad_norm": 1.167617678642273,
+      "learning_rate": 1.1026410564225691e-05,
+      "loss": 0.4935,
+      "step": 19494
+    },
+    {
+      "epoch": 24.9536,
+      "grad_norm": 1.1910767555236816,
+      "learning_rate": 1.1024409763905563e-05,
+      "loss": 0.5077,
+      "step": 19495
+    },
+    {
+      "epoch": 24.95488,
+      "grad_norm": 1.1297580003738403,
+      "learning_rate": 1.1022408963585435e-05,
+      "loss": 0.5074,
+      "step": 19496
+    },
+    {
+      "epoch": 24.95616,
+      "grad_norm": 1.1995117664337158,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 0.4817,
+      "step": 19497
+    },
+    {
+      "epoch": 24.95744,
+      "grad_norm": 1.1976964473724365,
+      "learning_rate": 1.1018407362945178e-05,
+      "loss": 0.5004,
+      "step": 19498
+    },
+    {
+      "epoch": 24.95872,
+      "grad_norm": 1.1336443424224854,
+      "learning_rate": 1.101640656262505e-05,
+      "loss": 0.46,
+      "step": 19499
+    },
+    {
+      "epoch": 24.96,
+      "grad_norm": 1.1351187229156494,
+      "learning_rate": 1.1014405762304922e-05,
+      "loss": 0.4802,
+      "step": 19500
+    },
+    {
+      "epoch": 24.96128,
+      "grad_norm": 1.1822195053100586,
+      "learning_rate": 1.1012404961984794e-05,
+      "loss": 0.5001,
+      "step": 19501
+    },
+    {
+      "epoch": 24.96256,
+      "grad_norm": 1.1518561840057373,
+      "learning_rate": 1.1010404161664667e-05,
+      "loss": 0.4872,
+      "step": 19502
+    },
+    {
+      "epoch": 24.96384,
+      "grad_norm": 1.2118054628372192,
+      "learning_rate": 1.1008403361344538e-05,
+      "loss": 0.4995,
+      "step": 19503
+    },
+    {
+      "epoch": 24.96512,
+      "grad_norm": 1.1399140357971191,
+      "learning_rate": 1.1006402561024411e-05,
+      "loss": 0.4491,
+      "step": 19504
+    },
+    {
+      "epoch": 24.9664,
+      "grad_norm": 1.1353464126586914,
+      "learning_rate": 1.1004401760704281e-05,
+      "loss": 0.4953,
+      "step": 19505
+    },
+    {
+      "epoch": 24.96768,
+      "grad_norm": 1.1443747282028198,
+      "learning_rate": 1.1002400960384155e-05,
+      "loss": 0.4445,
+      "step": 19506
+    },
+    {
+      "epoch": 24.96896,
+      "grad_norm": 1.1765371561050415,
+      "learning_rate": 1.1000400160064025e-05,
+      "loss": 0.5095,
+      "step": 19507
+    },
+    {
+      "epoch": 24.97024,
+      "grad_norm": 1.1671823263168335,
+      "learning_rate": 1.0998399359743899e-05,
+      "loss": 0.5449,
+      "step": 19508
+    },
+    {
+      "epoch": 24.97152,
+      "grad_norm": 1.2038657665252686,
+      "learning_rate": 1.099639855942377e-05,
+      "loss": 0.4931,
+      "step": 19509
+    },
+    {
+      "epoch": 24.9728,
+      "grad_norm": 1.1690441370010376,
+      "learning_rate": 1.0994397759103642e-05,
+      "loss": 0.4586,
+      "step": 19510
+    },
+    {
+      "epoch": 24.97408,
+      "grad_norm": 1.184372901916504,
+      "learning_rate": 1.0992396958783514e-05,
+      "loss": 0.4714,
+      "step": 19511
+    },
+    {
+      "epoch": 24.97536,
+      "grad_norm": 1.1089961528778076,
+      "learning_rate": 1.0990396158463386e-05,
+      "loss": 0.4356,
+      "step": 19512
+    },
+    {
+      "epoch": 24.97664,
+      "grad_norm": 1.1306204795837402,
+      "learning_rate": 1.0988395358143258e-05,
+      "loss": 0.4918,
+      "step": 19513
+    },
+    {
+      "epoch": 24.97792,
+      "grad_norm": 1.166355013847351,
+      "learning_rate": 1.098639455782313e-05,
+      "loss": 0.5125,
+      "step": 19514
+    },
+    {
+      "epoch": 24.9792,
+      "grad_norm": 1.1938666105270386,
+      "learning_rate": 1.0984393757503002e-05,
+      "loss": 0.5012,
+      "step": 19515
+    },
+    {
+      "epoch": 24.98048,
+      "grad_norm": 1.2338180541992188,
+      "learning_rate": 1.0982392957182875e-05,
+      "loss": 0.5217,
+      "step": 19516
+    },
+    {
+      "epoch": 24.98176,
+      "grad_norm": 1.154515027999878,
+      "learning_rate": 1.0980392156862745e-05,
+      "loss": 0.497,
+      "step": 19517
+    },
+    {
+      "epoch": 24.98304,
+      "grad_norm": 1.1719876527786255,
+      "learning_rate": 1.0978391356542619e-05,
+      "loss": 0.4643,
+      "step": 19518
+    },
+    {
+      "epoch": 24.98432,
+      "grad_norm": 1.1815974712371826,
+      "learning_rate": 1.0976390556222489e-05,
+      "loss": 0.5102,
+      "step": 19519
+    },
+    {
+      "epoch": 24.9856,
+      "grad_norm": 1.095108985900879,
+      "learning_rate": 1.0974389755902363e-05,
+      "loss": 0.4735,
+      "step": 19520
+    },
+    {
+      "epoch": 24.98688,
+      "grad_norm": 1.1639820337295532,
+      "learning_rate": 1.0972388955582233e-05,
+      "loss": 0.4904,
+      "step": 19521
+    },
+    {
+      "epoch": 24.98816,
+      "grad_norm": 1.1341108083724976,
+      "learning_rate": 1.0970388155262106e-05,
+      "loss": 0.4568,
+      "step": 19522
+    },
+    {
+      "epoch": 24.98944,
+      "grad_norm": 1.2175263166427612,
+      "learning_rate": 1.0968387354941978e-05,
+      "loss": 0.4845,
+      "step": 19523
+    },
+    {
+      "epoch": 24.99072,
+      "grad_norm": 1.1433970928192139,
+      "learning_rate": 1.096638655462185e-05,
+      "loss": 0.4664,
+      "step": 19524
+    },
+    {
+      "epoch": 24.992,
+      "grad_norm": 1.1078377962112427,
+      "learning_rate": 1.0964385754301722e-05,
+      "loss": 0.4772,
+      "step": 19525
+    },
+    {
+      "epoch": 24.99328,
+      "grad_norm": 1.181159257888794,
+      "learning_rate": 1.0962384953981594e-05,
+      "loss": 0.4969,
+      "step": 19526
+    },
+    {
+      "epoch": 24.99456,
+      "grad_norm": 1.0896167755126953,
+      "learning_rate": 1.0960384153661466e-05,
+      "loss": 0.4691,
+      "step": 19527
+    },
+    {
+      "epoch": 24.99584,
+      "grad_norm": 1.1512247323989868,
+      "learning_rate": 1.0958383353341337e-05,
+      "loss": 0.4367,
+      "step": 19528
+    },
+    {
+      "epoch": 24.99712,
+      "grad_norm": 1.1622196435928345,
+      "learning_rate": 1.095638255302121e-05,
+      "loss": 0.4856,
+      "step": 19529
+    },
+    {
+      "epoch": 24.9984,
+      "grad_norm": 1.1247361898422241,
+      "learning_rate": 1.0954381752701081e-05,
+      "loss": 0.4831,
+      "step": 19530
+    },
+    {
+      "epoch": 24.99968,
+      "grad_norm": 1.0839389562606812,
+      "learning_rate": 1.0952380952380953e-05,
+      "loss": 0.4514,
+      "step": 19531
+    },
+    {
+      "epoch": 25.00096,
+      "grad_norm": Infinity,
+      "learning_rate": 1.0952380952380953e-05,
+      "loss": 0.8962,
+      "step": 19532
+    },
+    {
+      "epoch": 25.00224,
+      "grad_norm": 1.1649938821792603,
+      "learning_rate": 1.0950380152060825e-05,
+      "loss": 0.4679,
+      "step": 19533
+    },
+    {
+      "epoch": 25.00352,
+      "grad_norm": 1.159871220588684,
+      "learning_rate": 1.0948379351740697e-05,
+      "loss": 0.4861,
+      "step": 19534
+    },
+    {
+      "epoch": 25.0048,
+      "grad_norm": 1.1205157041549683,
+      "learning_rate": 1.0946378551420569e-05,
+      "loss": 0.4473,
+      "step": 19535
+    },
+    {
+      "epoch": 25.00608,
+      "grad_norm": 1.156924843788147,
+      "learning_rate": 1.094437775110044e-05,
+      "loss": 0.4897,
+      "step": 19536
+    },
+    {
+      "epoch": 25.00736,
+      "grad_norm": 1.0987516641616821,
+      "learning_rate": 1.0942376950780312e-05,
+      "loss": 0.4423,
+      "step": 19537
+    },
+    {
+      "epoch": 25.00864,
+      "grad_norm": 1.0584564208984375,
+      "learning_rate": 1.0940376150460184e-05,
+      "loss": 0.4282,
+      "step": 19538
+    },
+    {
+      "epoch": 25.00992,
+      "grad_norm": 1.0948853492736816,
+      "learning_rate": 1.0938375350140056e-05,
+      "loss": 0.453,
+      "step": 19539
+    },
+    {
+      "epoch": 25.0112,
+      "grad_norm": 1.1634764671325684,
+      "learning_rate": 1.093637454981993e-05,
+      "loss": 0.4743,
+      "step": 19540
+    },
+    {
+      "epoch": 25.01248,
+      "grad_norm": 1.087397575378418,
+      "learning_rate": 1.09343737494998e-05,
+      "loss": 0.4398,
+      "step": 19541
+    },
+    {
+      "epoch": 25.01376,
+      "grad_norm": 1.1633117198944092,
+      "learning_rate": 1.0932372949179673e-05,
+      "loss": 0.4746,
+      "step": 19542
+    },
+    {
+      "epoch": 25.01504,
+      "grad_norm": 1.2032607793807983,
+      "learning_rate": 1.0930372148859543e-05,
+      "loss": 0.4951,
+      "step": 19543
+    },
+    {
+      "epoch": 25.01632,
+      "grad_norm": 1.1877236366271973,
+      "learning_rate": 1.0928371348539417e-05,
+      "loss": 0.5303,
+      "step": 19544
+    },
+    {
+      "epoch": 25.0176,
+      "grad_norm": 1.0547975301742554,
+      "learning_rate": 1.0926370548219287e-05,
+      "loss": 0.4098,
+      "step": 19545
+    },
+    {
+      "epoch": 25.01888,
+      "grad_norm": 1.1014307737350464,
+      "learning_rate": 1.092436974789916e-05,
+      "loss": 0.4382,
+      "step": 19546
+    },
+    {
+      "epoch": 25.02016,
+      "grad_norm": 1.15842866897583,
+      "learning_rate": 1.0922368947579032e-05,
+      "loss": 0.4791,
+      "step": 19547
+    },
+    {
+      "epoch": 25.02144,
+      "grad_norm": 1.2171627283096313,
+      "learning_rate": 1.0920368147258904e-05,
+      "loss": 0.4589,
+      "step": 19548
+    },
+    {
+      "epoch": 25.02272,
+      "grad_norm": 1.2597172260284424,
+      "learning_rate": 1.0918367346938776e-05,
+      "loss": 0.4976,
+      "step": 19549
+    },
+    {
+      "epoch": 25.024,
+      "grad_norm": 1.1223315000534058,
+      "learning_rate": 1.0916366546618648e-05,
+      "loss": 0.4765,
+      "step": 19550
+    },
+    {
+      "epoch": 25.02528,
+      "grad_norm": 1.0641728639602661,
+      "learning_rate": 1.091436574629852e-05,
+      "loss": 0.4378,
+      "step": 19551
+    },
+    {
+      "epoch": 25.02656,
+      "grad_norm": 1.0940122604370117,
+      "learning_rate": 1.0912364945978392e-05,
+      "loss": 0.4407,
+      "step": 19552
+    },
+    {
+      "epoch": 25.02784,
+      "grad_norm": 1.1914441585540771,
+      "learning_rate": 1.0910364145658264e-05,
+      "loss": 0.4541,
+      "step": 19553
+    },
+    {
+      "epoch": 25.02912,
+      "grad_norm": 1.1707878112792969,
+      "learning_rate": 1.0908363345338135e-05,
+      "loss": 0.4925,
+      "step": 19554
+    },
+    {
+      "epoch": 25.0304,
+      "grad_norm": 1.0737802982330322,
+      "learning_rate": 1.0906362545018007e-05,
+      "loss": 0.4441,
+      "step": 19555
+    },
+    {
+      "epoch": 25.03168,
+      "grad_norm": 1.1500251293182373,
+      "learning_rate": 1.0904361744697881e-05,
+      "loss": 0.4752,
+      "step": 19556
+    },
+    {
+      "epoch": 25.03296,
+      "grad_norm": 1.1668425798416138,
+      "learning_rate": 1.0902360944377751e-05,
+      "loss": 0.4919,
+      "step": 19557
+    },
+    {
+      "epoch": 25.03424,
+      "grad_norm": 1.0954525470733643,
+      "learning_rate": 1.0900360144057625e-05,
+      "loss": 0.4363,
+      "step": 19558
+    },
+    {
+      "epoch": 25.03552,
+      "grad_norm": 1.1638271808624268,
+      "learning_rate": 1.0898359343737495e-05,
+      "loss": 0.488,
+      "step": 19559
+    },
+    {
+      "epoch": 25.0368,
+      "grad_norm": 1.1640788316726685,
+      "learning_rate": 1.0896358543417368e-05,
+      "loss": 0.4723,
+      "step": 19560
+    },
+    {
+      "epoch": 25.03808,
+      "grad_norm": 1.1712232828140259,
+      "learning_rate": 1.0894357743097238e-05,
+      "loss": 0.4883,
+      "step": 19561
+    },
+    {
+      "epoch": 25.03936,
+      "grad_norm": 1.1707978248596191,
+      "learning_rate": 1.0892356942777112e-05,
+      "loss": 0.5356,
+      "step": 19562
+    },
+    {
+      "epoch": 25.04064,
+      "grad_norm": 1.095038890838623,
+      "learning_rate": 1.0890356142456984e-05,
+      "loss": 0.4394,
+      "step": 19563
+    },
+    {
+      "epoch": 25.04192,
+      "grad_norm": 1.108280062675476,
+      "learning_rate": 1.0888355342136856e-05,
+      "loss": 0.4608,
+      "step": 19564
+    },
+    {
+      "epoch": 25.0432,
+      "grad_norm": 1.1268192529678345,
+      "learning_rate": 1.0886354541816728e-05,
+      "loss": 0.4504,
+      "step": 19565
+    },
+    {
+      "epoch": 25.04448,
+      "grad_norm": 1.1444103717803955,
+      "learning_rate": 1.08843537414966e-05,
+      "loss": 0.4217,
+      "step": 19566
+    },
+    {
+      "epoch": 25.04576,
+      "grad_norm": 1.1157702207565308,
+      "learning_rate": 1.0882352941176471e-05,
+      "loss": 0.4607,
+      "step": 19567
+    },
+    {
+      "epoch": 25.04704,
+      "grad_norm": 1.1346677541732788,
+      "learning_rate": 1.0880352140856343e-05,
+      "loss": 0.4446,
+      "step": 19568
+    },
+    {
+      "epoch": 25.04832,
+      "grad_norm": 1.1954710483551025,
+      "learning_rate": 1.0878351340536215e-05,
+      "loss": 0.4784,
+      "step": 19569
+    },
+    {
+      "epoch": 25.0496,
+      "grad_norm": 1.1586507558822632,
+      "learning_rate": 1.0876350540216087e-05,
+      "loss": 0.457,
+      "step": 19570
+    },
+    {
+      "epoch": 25.05088,
+      "grad_norm": 1.112011432647705,
+      "learning_rate": 1.0874349739895959e-05,
+      "loss": 0.4386,
+      "step": 19571
+    },
+    {
+      "epoch": 25.05216,
+      "grad_norm": 1.1521170139312744,
+      "learning_rate": 1.087234893957583e-05,
+      "loss": 0.4442,
+      "step": 19572
+    },
+    {
+      "epoch": 25.05344,
+      "grad_norm": 1.1750129461288452,
+      "learning_rate": 1.0870348139255702e-05,
+      "loss": 0.4451,
+      "step": 19573
+    },
+    {
+      "epoch": 25.05472,
+      "grad_norm": 1.1142715215682983,
+      "learning_rate": 1.0868347338935574e-05,
+      "loss": 0.4397,
+      "step": 19574
+    },
+    {
+      "epoch": 25.056,
+      "grad_norm": 1.1580311059951782,
+      "learning_rate": 1.0866346538615446e-05,
+      "loss": 0.5067,
+      "step": 19575
+    },
+    {
+      "epoch": 25.05728,
+      "grad_norm": 1.1484664678573608,
+      "learning_rate": 1.0864345738295318e-05,
+      "loss": 0.461,
+      "step": 19576
+    },
+    {
+      "epoch": 25.05856,
+      "grad_norm": 1.169391393661499,
+      "learning_rate": 1.0862344937975192e-05,
+      "loss": 0.4694,
+      "step": 19577
+    },
+    {
+      "epoch": 25.05984,
+      "grad_norm": 1.1013165712356567,
+      "learning_rate": 1.0860344137655062e-05,
+      "loss": 0.4279,
+      "step": 19578
+    },
+    {
+      "epoch": 25.06112,
+      "grad_norm": 1.248039960861206,
+      "learning_rate": 1.0858343337334935e-05,
+      "loss": 0.4867,
+      "step": 19579
+    },
+    {
+      "epoch": 25.0624,
+      "grad_norm": 1.119541883468628,
+      "learning_rate": 1.0856342537014805e-05,
+      "loss": 0.455,
+      "step": 19580
+    },
+    {
+      "epoch": 25.06368,
+      "grad_norm": 1.1457856893539429,
+      "learning_rate": 1.0854341736694679e-05,
+      "loss": 0.4475,
+      "step": 19581
+    },
+    {
+      "epoch": 25.06496,
+      "grad_norm": 1.1471192836761475,
+      "learning_rate": 1.085234093637455e-05,
+      "loss": 0.4766,
+      "step": 19582
+    },
+    {
+      "epoch": 25.06624,
+      "grad_norm": 1.1262675523757935,
+      "learning_rate": 1.0850340136054423e-05,
+      "loss": 0.4753,
+      "step": 19583
+    },
+    {
+      "epoch": 25.06752,
+      "grad_norm": 1.182495355606079,
+      "learning_rate": 1.0848339335734293e-05,
+      "loss": 0.4704,
+      "step": 19584
+    },
+    {
+      "epoch": 25.0688,
+      "grad_norm": 1.0830098390579224,
+      "learning_rate": 1.0846338535414166e-05,
+      "loss": 0.4645,
+      "step": 19585
+    },
+    {
+      "epoch": 25.07008,
+      "grad_norm": 1.1021302938461304,
+      "learning_rate": 1.0844337735094038e-05,
+      "loss": 0.4325,
+      "step": 19586
+    },
+    {
+      "epoch": 25.07136,
+      "grad_norm": 1.192583441734314,
+      "learning_rate": 1.084233693477391e-05,
+      "loss": 0.5081,
+      "step": 19587
+    },
+    {
+      "epoch": 25.07264,
+      "grad_norm": 1.1611496210098267,
+      "learning_rate": 1.0840336134453782e-05,
+      "loss": 0.4763,
+      "step": 19588
+    },
+    {
+      "epoch": 25.07392,
+      "grad_norm": 1.1356542110443115,
+      "learning_rate": 1.0838335334133654e-05,
+      "loss": 0.4435,
+      "step": 19589
+    },
+    {
+      "epoch": 25.0752,
+      "grad_norm": 1.1349308490753174,
+      "learning_rate": 1.0836334533813526e-05,
+      "loss": 0.4455,
+      "step": 19590
+    },
+    {
+      "epoch": 25.07648,
+      "grad_norm": 1.1213163137435913,
+      "learning_rate": 1.0834333733493398e-05,
+      "loss": 0.4538,
+      "step": 19591
+    },
+    {
+      "epoch": 25.07776,
+      "grad_norm": 1.151708960533142,
+      "learning_rate": 1.083233293317327e-05,
+      "loss": 0.4627,
+      "step": 19592
+    },
+    {
+      "epoch": 25.07904,
+      "grad_norm": 1.1179898977279663,
+      "learning_rate": 1.0830332132853143e-05,
+      "loss": 0.4877,
+      "step": 19593
+    },
+    {
+      "epoch": 25.08032,
+      "grad_norm": 1.1163991689682007,
+      "learning_rate": 1.0828331332533013e-05,
+      "loss": 0.4423,
+      "step": 19594
+    },
+    {
+      "epoch": 25.0816,
+      "grad_norm": 1.1171714067459106,
+      "learning_rate": 1.0826330532212887e-05,
+      "loss": 0.4341,
+      "step": 19595
+    },
+    {
+      "epoch": 25.08288,
+      "grad_norm": 1.216645359992981,
+      "learning_rate": 1.0824329731892757e-05,
+      "loss": 0.4985,
+      "step": 19596
+    },
+    {
+      "epoch": 25.08416,
+      "grad_norm": 1.1512889862060547,
+      "learning_rate": 1.082232893157263e-05,
+      "loss": 0.479,
+      "step": 19597
+    },
+    {
+      "epoch": 25.08544,
+      "grad_norm": 1.210100531578064,
+      "learning_rate": 1.08203281312525e-05,
+      "loss": 0.476,
+      "step": 19598
+    },
+    {
+      "epoch": 25.08672,
+      "grad_norm": 1.1688969135284424,
+      "learning_rate": 1.0818327330932374e-05,
+      "loss": 0.4753,
+      "step": 19599
+    },
+    {
+      "epoch": 25.088,
+      "grad_norm": 1.1013729572296143,
+      "learning_rate": 1.0816326530612246e-05,
+      "loss": 0.4628,
+      "step": 19600
+    },
+    {
+      "epoch": 25.08928,
+      "grad_norm": 1.096455454826355,
+      "learning_rate": 1.0814325730292118e-05,
+      "loss": 0.4244,
+      "step": 19601
+    },
+    {
+      "epoch": 25.09056,
+      "grad_norm": 1.1417826414108276,
+      "learning_rate": 1.081232492997199e-05,
+      "loss": 0.4263,
+      "step": 19602
+    },
+    {
+      "epoch": 25.09184,
+      "grad_norm": 1.2026113271713257,
+      "learning_rate": 1.0810324129651862e-05,
+      "loss": 0.4583,
+      "step": 19603
+    },
+    {
+      "epoch": 25.09312,
+      "grad_norm": 1.1624215841293335,
+      "learning_rate": 1.0808323329331733e-05,
+      "loss": 0.464,
+      "step": 19604
+    },
+    {
+      "epoch": 25.0944,
+      "grad_norm": 1.2777810096740723,
+      "learning_rate": 1.0806322529011605e-05,
+      "loss": 0.486,
+      "step": 19605
+    },
+    {
+      "epoch": 25.09568,
+      "grad_norm": 1.1731245517730713,
+      "learning_rate": 1.0804321728691477e-05,
+      "loss": 0.4465,
+      "step": 19606
+    },
+    {
+      "epoch": 25.09696,
+      "grad_norm": 1.1680535078048706,
+      "learning_rate": 1.0802320928371349e-05,
+      "loss": 0.4408,
+      "step": 19607
+    },
+    {
+      "epoch": 25.09824,
+      "grad_norm": 1.1202332973480225,
+      "learning_rate": 1.080032012805122e-05,
+      "loss": 0.4328,
+      "step": 19608
+    },
+    {
+      "epoch": 25.09952,
+      "grad_norm": 1.1760560274124146,
+      "learning_rate": 1.0798319327731093e-05,
+      "loss": 0.4216,
+      "step": 19609
+    },
+    {
+      "epoch": 25.1008,
+      "grad_norm": 1.2153133153915405,
+      "learning_rate": 1.0796318527410965e-05,
+      "loss": 0.5195,
+      "step": 19610
+    },
+    {
+      "epoch": 25.10208,
+      "grad_norm": 1.2274070978164673,
+      "learning_rate": 1.0794317727090836e-05,
+      "loss": 0.4786,
+      "step": 19611
+    },
+    {
+      "epoch": 25.10336,
+      "grad_norm": 1.141127109527588,
+      "learning_rate": 1.0792316926770708e-05,
+      "loss": 0.4644,
+      "step": 19612
+    },
+    {
+      "epoch": 25.10464,
+      "grad_norm": 1.1838977336883545,
+      "learning_rate": 1.079031612645058e-05,
+      "loss": 0.494,
+      "step": 19613
+    },
+    {
+      "epoch": 25.10592,
+      "grad_norm": 1.1999531984329224,
+      "learning_rate": 1.0788315326130452e-05,
+      "loss": 0.52,
+      "step": 19614
+    },
+    {
+      "epoch": 25.1072,
+      "grad_norm": 1.1070973873138428,
+      "learning_rate": 1.0786314525810324e-05,
+      "loss": 0.4395,
+      "step": 19615
+    },
+    {
+      "epoch": 25.10848,
+      "grad_norm": 1.1713796854019165,
+      "learning_rate": 1.0784313725490197e-05,
+      "loss": 0.4757,
+      "step": 19616
+    },
+    {
+      "epoch": 25.10976,
+      "grad_norm": 1.1294423341751099,
+      "learning_rate": 1.0782312925170068e-05,
+      "loss": 0.4865,
+      "step": 19617
+    },
+    {
+      "epoch": 25.11104,
+      "grad_norm": 1.1077033281326294,
+      "learning_rate": 1.0780312124849941e-05,
+      "loss": 0.4506,
+      "step": 19618
+    },
+    {
+      "epoch": 25.11232,
+      "grad_norm": 1.2054367065429688,
+      "learning_rate": 1.0778311324529811e-05,
+      "loss": 0.4717,
+      "step": 19619
+    },
+    {
+      "epoch": 25.1136,
+      "grad_norm": 1.1735788583755493,
+      "learning_rate": 1.0776310524209685e-05,
+      "loss": 0.479,
+      "step": 19620
+    },
+    {
+      "epoch": 25.11488,
+      "grad_norm": 1.128080129623413,
+      "learning_rate": 1.0774309723889555e-05,
+      "loss": 0.4248,
+      "step": 19621
+    },
+    {
+      "epoch": 25.11616,
+      "grad_norm": 1.1419153213500977,
+      "learning_rate": 1.0772308923569429e-05,
+      "loss": 0.4552,
+      "step": 19622
+    },
+    {
+      "epoch": 25.11744,
+      "grad_norm": 1.2297865152359009,
+      "learning_rate": 1.07703081232493e-05,
+      "loss": 0.4826,
+      "step": 19623
+    },
+    {
+      "epoch": 25.11872,
+      "grad_norm": 1.1833217144012451,
+      "learning_rate": 1.0768307322929172e-05,
+      "loss": 0.521,
+      "step": 19624
+    },
+    {
+      "epoch": 25.12,
+      "grad_norm": 1.152848243713379,
+      "learning_rate": 1.0766306522609044e-05,
+      "loss": 0.5335,
+      "step": 19625
+    },
+    {
+      "epoch": 25.12128,
+      "grad_norm": 1.1874018907546997,
+      "learning_rate": 1.0764305722288916e-05,
+      "loss": 0.4737,
+      "step": 19626
+    },
+    {
+      "epoch": 25.12256,
+      "grad_norm": 1.2371412515640259,
+      "learning_rate": 1.0762304921968788e-05,
+      "loss": 0.4999,
+      "step": 19627
+    },
+    {
+      "epoch": 25.12384,
+      "grad_norm": 1.167389988899231,
+      "learning_rate": 1.076030412164866e-05,
+      "loss": 0.4559,
+      "step": 19628
+    },
+    {
+      "epoch": 25.12512,
+      "grad_norm": 1.1866403818130493,
+      "learning_rate": 1.0758303321328532e-05,
+      "loss": 0.4684,
+      "step": 19629
+    },
+    {
+      "epoch": 25.1264,
+      "grad_norm": 1.1500517129898071,
+      "learning_rate": 1.0756302521008405e-05,
+      "loss": 0.4702,
+      "step": 19630
+    },
+    {
+      "epoch": 25.12768,
+      "grad_norm": 1.1612963676452637,
+      "learning_rate": 1.0754301720688275e-05,
+      "loss": 0.4909,
+      "step": 19631
+    },
+    {
+      "epoch": 25.12896,
+      "grad_norm": 1.1947745084762573,
+      "learning_rate": 1.0752300920368149e-05,
+      "loss": 0.4601,
+      "step": 19632
+    },
+    {
+      "epoch": 25.13024,
+      "grad_norm": 1.2526357173919678,
+      "learning_rate": 1.0750300120048019e-05,
+      "loss": 0.4815,
+      "step": 19633
+    },
+    {
+      "epoch": 25.13152,
+      "grad_norm": 1.1006672382354736,
+      "learning_rate": 1.0748299319727893e-05,
+      "loss": 0.4635,
+      "step": 19634
+    },
+    {
+      "epoch": 25.1328,
+      "grad_norm": 1.0949301719665527,
+      "learning_rate": 1.0746298519407763e-05,
+      "loss": 0.4385,
+      "step": 19635
+    },
+    {
+      "epoch": 25.13408,
+      "grad_norm": 1.2405381202697754,
+      "learning_rate": 1.0744297719087636e-05,
+      "loss": 0.5009,
+      "step": 19636
+    },
+    {
+      "epoch": 25.13536,
+      "grad_norm": 1.1082509756088257,
+      "learning_rate": 1.0742296918767508e-05,
+      "loss": 0.4463,
+      "step": 19637
+    },
+    {
+      "epoch": 25.13664,
+      "grad_norm": 1.1167351007461548,
+      "learning_rate": 1.074029611844738e-05,
+      "loss": 0.4892,
+      "step": 19638
+    },
+    {
+      "epoch": 25.13792,
+      "grad_norm": 1.158218264579773,
+      "learning_rate": 1.0738295318127252e-05,
+      "loss": 0.4539,
+      "step": 19639
+    },
+    {
+      "epoch": 25.1392,
+      "grad_norm": 1.2034755945205688,
+      "learning_rate": 1.0736294517807124e-05,
+      "loss": 0.4988,
+      "step": 19640
+    },
+    {
+      "epoch": 25.14048,
+      "grad_norm": 1.1195570230484009,
+      "learning_rate": 1.0734293717486996e-05,
+      "loss": 0.4505,
+      "step": 19641
+    },
+    {
+      "epoch": 25.14176,
+      "grad_norm": 1.1278491020202637,
+      "learning_rate": 1.0732292917166867e-05,
+      "loss": 0.4631,
+      "step": 19642
+    },
+    {
+      "epoch": 25.14304,
+      "grad_norm": 1.1615363359451294,
+      "learning_rate": 1.073029211684674e-05,
+      "loss": 0.4845,
+      "step": 19643
+    },
+    {
+      "epoch": 25.14432,
+      "grad_norm": 1.1552554368972778,
+      "learning_rate": 1.0728291316526611e-05,
+      "loss": 0.4543,
+      "step": 19644
+    },
+    {
+      "epoch": 25.1456,
+      "grad_norm": 1.1733779907226562,
+      "learning_rate": 1.0726290516206483e-05,
+      "loss": 0.5251,
+      "step": 19645
+    },
+    {
+      "epoch": 25.14688,
+      "grad_norm": 1.2017656564712524,
+      "learning_rate": 1.0724289715886355e-05,
+      "loss": 0.5293,
+      "step": 19646
+    },
+    {
+      "epoch": 25.14816,
+      "grad_norm": 1.1338574886322021,
+      "learning_rate": 1.0722288915566227e-05,
+      "loss": 0.4493,
+      "step": 19647
+    },
+    {
+      "epoch": 25.14944,
+      "grad_norm": 1.1379152536392212,
+      "learning_rate": 1.0720288115246099e-05,
+      "loss": 0.4381,
+      "step": 19648
+    },
+    {
+      "epoch": 25.15072,
+      "grad_norm": 1.1826634407043457,
+      "learning_rate": 1.071828731492597e-05,
+      "loss": 0.4921,
+      "step": 19649
+    },
+    {
+      "epoch": 25.152,
+      "grad_norm": 1.1363234519958496,
+      "learning_rate": 1.0716286514605842e-05,
+      "loss": 0.4593,
+      "step": 19650
+    },
+    {
+      "epoch": 25.15328,
+      "grad_norm": 1.157832384109497,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.456,
+      "step": 19651
+    },
+    {
+      "epoch": 25.15456,
+      "grad_norm": 1.1512408256530762,
+      "learning_rate": 1.0712284913965586e-05,
+      "loss": 0.4946,
+      "step": 19652
+    },
+    {
+      "epoch": 25.15584,
+      "grad_norm": 1.2528743743896484,
+      "learning_rate": 1.071028411364546e-05,
+      "loss": 0.5087,
+      "step": 19653
+    },
+    {
+      "epoch": 25.15712,
+      "grad_norm": 1.1764575242996216,
+      "learning_rate": 1.070828331332533e-05,
+      "loss": 0.4661,
+      "step": 19654
+    },
+    {
+      "epoch": 25.1584,
+      "grad_norm": 1.1859891414642334,
+      "learning_rate": 1.0706282513005203e-05,
+      "loss": 0.4623,
+      "step": 19655
+    },
+    {
+      "epoch": 25.15968,
+      "grad_norm": 1.1567497253417969,
+      "learning_rate": 1.0704281712685073e-05,
+      "loss": 0.4522,
+      "step": 19656
+    },
+    {
+      "epoch": 25.16096,
+      "grad_norm": 1.191056489944458,
+      "learning_rate": 1.0702280912364947e-05,
+      "loss": 0.4735,
+      "step": 19657
+    },
+    {
+      "epoch": 25.16224,
+      "grad_norm": 1.1542645692825317,
+      "learning_rate": 1.0700280112044817e-05,
+      "loss": 0.4693,
+      "step": 19658
+    },
+    {
+      "epoch": 25.16352,
+      "grad_norm": 1.1220331192016602,
+      "learning_rate": 1.069827931172469e-05,
+      "loss": 0.4753,
+      "step": 19659
+    },
+    {
+      "epoch": 25.1648,
+      "grad_norm": 1.1264070272445679,
+      "learning_rate": 1.0696278511404562e-05,
+      "loss": 0.4709,
+      "step": 19660
+    },
+    {
+      "epoch": 25.16608,
+      "grad_norm": 1.16756272315979,
+      "learning_rate": 1.0694277711084434e-05,
+      "loss": 0.4573,
+      "step": 19661
+    },
+    {
+      "epoch": 25.16736,
+      "grad_norm": 1.1602298021316528,
+      "learning_rate": 1.0692276910764306e-05,
+      "loss": 0.4363,
+      "step": 19662
+    },
+    {
+      "epoch": 25.16864,
+      "grad_norm": 1.1670148372650146,
+      "learning_rate": 1.0690276110444178e-05,
+      "loss": 0.4925,
+      "step": 19663
+    },
+    {
+      "epoch": 25.16992,
+      "grad_norm": 1.2356739044189453,
+      "learning_rate": 1.068827531012405e-05,
+      "loss": 0.4923,
+      "step": 19664
+    },
+    {
+      "epoch": 25.1712,
+      "grad_norm": 1.2018803358078003,
+      "learning_rate": 1.0686274509803922e-05,
+      "loss": 0.4857,
+      "step": 19665
+    },
+    {
+      "epoch": 25.17248,
+      "grad_norm": 1.1678217649459839,
+      "learning_rate": 1.0684273709483794e-05,
+      "loss": 0.4493,
+      "step": 19666
+    },
+    {
+      "epoch": 25.17376,
+      "grad_norm": 1.1933562755584717,
+      "learning_rate": 1.0682272909163665e-05,
+      "loss": 0.4712,
+      "step": 19667
+    },
+    {
+      "epoch": 25.17504,
+      "grad_norm": 1.1472783088684082,
+      "learning_rate": 1.0680272108843537e-05,
+      "loss": 0.4889,
+      "step": 19668
+    },
+    {
+      "epoch": 25.17632,
+      "grad_norm": 1.113418698310852,
+      "learning_rate": 1.0678271308523411e-05,
+      "loss": 0.4455,
+      "step": 19669
+    },
+    {
+      "epoch": 25.1776,
+      "grad_norm": 1.2086471319198608,
+      "learning_rate": 1.0676270508203281e-05,
+      "loss": 0.4709,
+      "step": 19670
+    },
+    {
+      "epoch": 25.17888,
+      "grad_norm": 1.0935802459716797,
+      "learning_rate": 1.0674269707883155e-05,
+      "loss": 0.4329,
+      "step": 19671
+    },
+    {
+      "epoch": 25.18016,
+      "grad_norm": 1.1935173273086548,
+      "learning_rate": 1.0672268907563025e-05,
+      "loss": 0.4988,
+      "step": 19672
+    },
+    {
+      "epoch": 25.18144,
+      "grad_norm": 1.189258337020874,
+      "learning_rate": 1.0670268107242898e-05,
+      "loss": 0.4994,
+      "step": 19673
+    },
+    {
+      "epoch": 25.18272,
+      "grad_norm": 1.2858033180236816,
+      "learning_rate": 1.0668267306922768e-05,
+      "loss": 0.5511,
+      "step": 19674
+    },
+    {
+      "epoch": 25.184,
+      "grad_norm": 1.2078543901443481,
+      "learning_rate": 1.0666266506602642e-05,
+      "loss": 0.4853,
+      "step": 19675
+    },
+    {
+      "epoch": 25.18528,
+      "grad_norm": 1.1853865385055542,
+      "learning_rate": 1.0664265706282514e-05,
+      "loss": 0.475,
+      "step": 19676
+    },
+    {
+      "epoch": 25.18656,
+      "grad_norm": 1.1798455715179443,
+      "learning_rate": 1.0662264905962386e-05,
+      "loss": 0.5103,
+      "step": 19677
+    },
+    {
+      "epoch": 25.18784,
+      "grad_norm": 1.1571396589279175,
+      "learning_rate": 1.0660264105642258e-05,
+      "loss": 0.4582,
+      "step": 19678
+    },
+    {
+      "epoch": 25.18912,
+      "grad_norm": 1.178780198097229,
+      "learning_rate": 1.065826330532213e-05,
+      "loss": 0.5142,
+      "step": 19679
+    },
+    {
+      "epoch": 25.1904,
+      "grad_norm": 1.1873482465744019,
+      "learning_rate": 1.0656262505002001e-05,
+      "loss": 0.4801,
+      "step": 19680
+    },
+    {
+      "epoch": 25.19168,
+      "grad_norm": 1.1041178703308105,
+      "learning_rate": 1.0654261704681873e-05,
+      "loss": 0.4445,
+      "step": 19681
+    },
+    {
+      "epoch": 25.19296,
+      "grad_norm": 1.1332768201828003,
+      "learning_rate": 1.0652260904361745e-05,
+      "loss": 0.444,
+      "step": 19682
+    },
+    {
+      "epoch": 25.19424,
+      "grad_norm": 1.1968101263046265,
+      "learning_rate": 1.0650260104041619e-05,
+      "loss": 0.5094,
+      "step": 19683
+    },
+    {
+      "epoch": 25.19552,
+      "grad_norm": 1.158816933631897,
+      "learning_rate": 1.0648259303721489e-05,
+      "loss": 0.4494,
+      "step": 19684
+    },
+    {
+      "epoch": 25.1968,
+      "grad_norm": 1.192723274230957,
+      "learning_rate": 1.0646258503401362e-05,
+      "loss": 0.4797,
+      "step": 19685
+    },
+    {
+      "epoch": 25.19808,
+      "grad_norm": 1.202081561088562,
+      "learning_rate": 1.0644257703081232e-05,
+      "loss": 0.441,
+      "step": 19686
+    },
+    {
+      "epoch": 25.19936,
+      "grad_norm": 1.2185006141662598,
+      "learning_rate": 1.0642256902761106e-05,
+      "loss": 0.4857,
+      "step": 19687
+    },
+    {
+      "epoch": 25.20064,
+      "grad_norm": 1.2291009426116943,
+      "learning_rate": 1.0640256102440976e-05,
+      "loss": 0.4885,
+      "step": 19688
+    },
+    {
+      "epoch": 25.20192,
+      "grad_norm": 1.1647063493728638,
+      "learning_rate": 1.063825530212085e-05,
+      "loss": 0.4644,
+      "step": 19689
+    },
+    {
+      "epoch": 25.2032,
+      "grad_norm": 1.197346806526184,
+      "learning_rate": 1.0636254501800722e-05,
+      "loss": 0.5212,
+      "step": 19690
+    },
+    {
+      "epoch": 25.20448,
+      "grad_norm": 1.0721224546432495,
+      "learning_rate": 1.0634253701480593e-05,
+      "loss": 0.4732,
+      "step": 19691
+    },
+    {
+      "epoch": 25.20576,
+      "grad_norm": 1.1543662548065186,
+      "learning_rate": 1.0632252901160465e-05,
+      "loss": 0.4853,
+      "step": 19692
+    },
+    {
+      "epoch": 25.20704,
+      "grad_norm": 1.0685230493545532,
+      "learning_rate": 1.0630252100840337e-05,
+      "loss": 0.4494,
+      "step": 19693
+    },
+    {
+      "epoch": 25.20832,
+      "grad_norm": 1.1875240802764893,
+      "learning_rate": 1.0628251300520209e-05,
+      "loss": 0.4819,
+      "step": 19694
+    },
+    {
+      "epoch": 25.209600000000002,
+      "grad_norm": 1.1269207000732422,
+      "learning_rate": 1.0626250500200081e-05,
+      "loss": 0.4352,
+      "step": 19695
+    },
+    {
+      "epoch": 25.21088,
+      "grad_norm": 1.164787769317627,
+      "learning_rate": 1.0624249699879953e-05,
+      "loss": 0.4917,
+      "step": 19696
+    },
+    {
+      "epoch": 25.21216,
+      "grad_norm": 1.168983817100525,
+      "learning_rate": 1.0622248899559825e-05,
+      "loss": 0.454,
+      "step": 19697
+    },
+    {
+      "epoch": 25.21344,
+      "grad_norm": 1.1204594373703003,
+      "learning_rate": 1.0620248099239696e-05,
+      "loss": 0.428,
+      "step": 19698
+    },
+    {
+      "epoch": 25.21472,
+      "grad_norm": 1.1424245834350586,
+      "learning_rate": 1.0618247298919568e-05,
+      "loss": 0.441,
+      "step": 19699
+    },
+    {
+      "epoch": 25.216,
+      "grad_norm": 1.2119210958480835,
+      "learning_rate": 1.061624649859944e-05,
+      "loss": 0.4883,
+      "step": 19700
+    },
+    {
+      "epoch": 25.21728,
+      "grad_norm": 1.1396766901016235,
+      "learning_rate": 1.0614245698279312e-05,
+      "loss": 0.4449,
+      "step": 19701
+    },
+    {
+      "epoch": 25.21856,
+      "grad_norm": 1.164551019668579,
+      "learning_rate": 1.0612244897959184e-05,
+      "loss": 0.466,
+      "step": 19702
+    },
+    {
+      "epoch": 25.21984,
+      "grad_norm": 1.1590672731399536,
+      "learning_rate": 1.0610244097639056e-05,
+      "loss": 0.4948,
+      "step": 19703
+    },
+    {
+      "epoch": 25.22112,
+      "grad_norm": 1.2192983627319336,
+      "learning_rate": 1.0608243297318928e-05,
+      "loss": 0.463,
+      "step": 19704
+    },
+    {
+      "epoch": 25.2224,
+      "grad_norm": 1.1451095342636108,
+      "learning_rate": 1.06062424969988e-05,
+      "loss": 0.4541,
+      "step": 19705
+    },
+    {
+      "epoch": 25.22368,
+      "grad_norm": 1.203220009803772,
+      "learning_rate": 1.0604241696678673e-05,
+      "loss": 0.4684,
+      "step": 19706
+    },
+    {
+      "epoch": 25.22496,
+      "grad_norm": 1.1325691938400269,
+      "learning_rate": 1.0602240896358543e-05,
+      "loss": 0.5005,
+      "step": 19707
+    },
+    {
+      "epoch": 25.22624,
+      "grad_norm": 1.106677532196045,
+      "learning_rate": 1.0600240096038417e-05,
+      "loss": 0.5002,
+      "step": 19708
+    },
+    {
+      "epoch": 25.22752,
+      "grad_norm": 1.176331639289856,
+      "learning_rate": 1.0598239295718287e-05,
+      "loss": 0.4872,
+      "step": 19709
+    },
+    {
+      "epoch": 25.2288,
+      "grad_norm": 1.1356264352798462,
+      "learning_rate": 1.059623849539816e-05,
+      "loss": 0.4681,
+      "step": 19710
+    },
+    {
+      "epoch": 25.23008,
+      "grad_norm": 1.0752949714660645,
+      "learning_rate": 1.059423769507803e-05,
+      "loss": 0.4497,
+      "step": 19711
+    },
+    {
+      "epoch": 25.23136,
+      "grad_norm": 1.1183545589447021,
+      "learning_rate": 1.0592236894757904e-05,
+      "loss": 0.4709,
+      "step": 19712
+    },
+    {
+      "epoch": 25.23264,
+      "grad_norm": 1.1348707675933838,
+      "learning_rate": 1.0590236094437776e-05,
+      "loss": 0.4373,
+      "step": 19713
+    },
+    {
+      "epoch": 25.23392,
+      "grad_norm": 1.201171636581421,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.4347,
+      "step": 19714
+    },
+    {
+      "epoch": 25.2352,
+      "grad_norm": 1.2042893171310425,
+      "learning_rate": 1.058623449379752e-05,
+      "loss": 0.4988,
+      "step": 19715
+    },
+    {
+      "epoch": 25.23648,
+      "grad_norm": 1.172882318496704,
+      "learning_rate": 1.0584233693477392e-05,
+      "loss": 0.4931,
+      "step": 19716
+    },
+    {
+      "epoch": 25.23776,
+      "grad_norm": 1.1809163093566895,
+      "learning_rate": 1.0582232893157263e-05,
+      "loss": 0.463,
+      "step": 19717
+    },
+    {
+      "epoch": 25.23904,
+      "grad_norm": 1.2052890062332153,
+      "learning_rate": 1.0580232092837135e-05,
+      "loss": 0.487,
+      "step": 19718
+    },
+    {
+      "epoch": 25.24032,
+      "grad_norm": 1.1632746458053589,
+      "learning_rate": 1.0578231292517007e-05,
+      "loss": 0.4584,
+      "step": 19719
+    },
+    {
+      "epoch": 25.2416,
+      "grad_norm": 1.1408315896987915,
+      "learning_rate": 1.057623049219688e-05,
+      "loss": 0.4565,
+      "step": 19720
+    },
+    {
+      "epoch": 25.24288,
+      "grad_norm": 1.1668243408203125,
+      "learning_rate": 1.057422969187675e-05,
+      "loss": 0.4432,
+      "step": 19721
+    },
+    {
+      "epoch": 25.24416,
+      "grad_norm": 1.1259456872940063,
+      "learning_rate": 1.0572228891556624e-05,
+      "loss": 0.456,
+      "step": 19722
+    },
+    {
+      "epoch": 25.24544,
+      "grad_norm": 1.109616994857788,
+      "learning_rate": 1.0570228091236495e-05,
+      "loss": 0.4352,
+      "step": 19723
+    },
+    {
+      "epoch": 25.24672,
+      "grad_norm": 1.1580768823623657,
+      "learning_rate": 1.0568227290916368e-05,
+      "loss": 0.451,
+      "step": 19724
+    },
+    {
+      "epoch": 25.248,
+      "grad_norm": 1.1380645036697388,
+      "learning_rate": 1.0566226490596238e-05,
+      "loss": 0.4576,
+      "step": 19725
+    },
+    {
+      "epoch": 25.24928,
+      "grad_norm": 1.1373761892318726,
+      "learning_rate": 1.0564225690276112e-05,
+      "loss": 0.4766,
+      "step": 19726
+    },
+    {
+      "epoch": 25.25056,
+      "grad_norm": 1.169333577156067,
+      "learning_rate": 1.0562224889955982e-05,
+      "loss": 0.5047,
+      "step": 19727
+    },
+    {
+      "epoch": 25.25184,
+      "grad_norm": 1.1574784517288208,
+      "learning_rate": 1.0560224089635856e-05,
+      "loss": 0.4608,
+      "step": 19728
+    },
+    {
+      "epoch": 25.25312,
+      "grad_norm": 1.2007426023483276,
+      "learning_rate": 1.0558223289315727e-05,
+      "loss": 0.4717,
+      "step": 19729
+    },
+    {
+      "epoch": 25.2544,
+      "grad_norm": 1.2134323120117188,
+      "learning_rate": 1.05562224889956e-05,
+      "loss": 0.4757,
+      "step": 19730
+    },
+    {
+      "epoch": 25.25568,
+      "grad_norm": 1.1803364753723145,
+      "learning_rate": 1.0554221688675471e-05,
+      "loss": 0.462,
+      "step": 19731
+    },
+    {
+      "epoch": 25.25696,
+      "grad_norm": 1.1004109382629395,
+      "learning_rate": 1.0552220888355343e-05,
+      "loss": 0.4594,
+      "step": 19732
+    },
+    {
+      "epoch": 25.25824,
+      "grad_norm": 1.1959086656570435,
+      "learning_rate": 1.0550220088035215e-05,
+      "loss": 0.4424,
+      "step": 19733
+    },
+    {
+      "epoch": 25.25952,
+      "grad_norm": 1.1749048233032227,
+      "learning_rate": 1.0548219287715087e-05,
+      "loss": 0.4687,
+      "step": 19734
+    },
+    {
+      "epoch": 25.2608,
+      "grad_norm": 1.2056117057800293,
+      "learning_rate": 1.0546218487394959e-05,
+      "loss": 0.4574,
+      "step": 19735
+    },
+    {
+      "epoch": 25.26208,
+      "grad_norm": 1.1588826179504395,
+      "learning_rate": 1.054421768707483e-05,
+      "loss": 0.4271,
+      "step": 19736
+    },
+    {
+      "epoch": 25.26336,
+      "grad_norm": 1.1901261806488037,
+      "learning_rate": 1.0542216886754702e-05,
+      "loss": 0.4621,
+      "step": 19737
+    },
+    {
+      "epoch": 25.26464,
+      "grad_norm": 1.1712379455566406,
+      "learning_rate": 1.0540216086434574e-05,
+      "loss": 0.4774,
+      "step": 19738
+    },
+    {
+      "epoch": 25.26592,
+      "grad_norm": 1.1799527406692505,
+      "learning_rate": 1.0538215286114446e-05,
+      "loss": 0.4345,
+      "step": 19739
+    },
+    {
+      "epoch": 25.2672,
+      "grad_norm": 1.1726861000061035,
+      "learning_rate": 1.0536214485794318e-05,
+      "loss": 0.5192,
+      "step": 19740
+    },
+    {
+      "epoch": 25.26848,
+      "grad_norm": 1.1405744552612305,
+      "learning_rate": 1.053421368547419e-05,
+      "loss": 0.4844,
+      "step": 19741
+    },
+    {
+      "epoch": 25.26976,
+      "grad_norm": 1.1705641746520996,
+      "learning_rate": 1.0532212885154062e-05,
+      "loss": 0.4587,
+      "step": 19742
+    },
+    {
+      "epoch": 25.27104,
+      "grad_norm": 1.0993807315826416,
+      "learning_rate": 1.0530212084833935e-05,
+      "loss": 0.4864,
+      "step": 19743
+    },
+    {
+      "epoch": 25.27232,
+      "grad_norm": 1.1683118343353271,
+      "learning_rate": 1.0528211284513805e-05,
+      "loss": 0.4339,
+      "step": 19744
+    },
+    {
+      "epoch": 25.2736,
+      "grad_norm": 1.18573796749115,
+      "learning_rate": 1.0526210484193679e-05,
+      "loss": 0.5003,
+      "step": 19745
+    },
+    {
+      "epoch": 25.27488,
+      "grad_norm": 1.1361162662506104,
+      "learning_rate": 1.0524209683873549e-05,
+      "loss": 0.4511,
+      "step": 19746
+    },
+    {
+      "epoch": 25.27616,
+      "grad_norm": 1.144995927810669,
+      "learning_rate": 1.0522208883553422e-05,
+      "loss": 0.4936,
+      "step": 19747
+    },
+    {
+      "epoch": 25.27744,
+      "grad_norm": 1.0969334840774536,
+      "learning_rate": 1.0520208083233293e-05,
+      "loss": 0.4298,
+      "step": 19748
+    },
+    {
+      "epoch": 25.27872,
+      "grad_norm": 1.1504309177398682,
+      "learning_rate": 1.0518207282913166e-05,
+      "loss": 0.4984,
+      "step": 19749
+    },
+    {
+      "epoch": 25.28,
+      "grad_norm": 1.1646525859832764,
+      "learning_rate": 1.0516206482593036e-05,
+      "loss": 0.4784,
+      "step": 19750
+    },
+    {
+      "epoch": 25.28128,
+      "grad_norm": 1.2210896015167236,
+      "learning_rate": 1.051420568227291e-05,
+      "loss": 0.5125,
+      "step": 19751
+    },
+    {
+      "epoch": 25.28256,
+      "grad_norm": 1.1825578212738037,
+      "learning_rate": 1.0512204881952782e-05,
+      "loss": 0.4775,
+      "step": 19752
+    },
+    {
+      "epoch": 25.28384,
+      "grad_norm": 1.098472237586975,
+      "learning_rate": 1.0510204081632654e-05,
+      "loss": 0.4602,
+      "step": 19753
+    },
+    {
+      "epoch": 25.28512,
+      "grad_norm": 1.095916748046875,
+      "learning_rate": 1.0508203281312525e-05,
+      "loss": 0.4521,
+      "step": 19754
+    },
+    {
+      "epoch": 25.2864,
+      "grad_norm": 1.1654385328292847,
+      "learning_rate": 1.0506202480992397e-05,
+      "loss": 0.4875,
+      "step": 19755
+    },
+    {
+      "epoch": 25.28768,
+      "grad_norm": 1.1669584512710571,
+      "learning_rate": 1.050420168067227e-05,
+      "loss": 0.4939,
+      "step": 19756
+    },
+    {
+      "epoch": 25.28896,
+      "grad_norm": 1.1289222240447998,
+      "learning_rate": 1.0502200880352141e-05,
+      "loss": 0.4583,
+      "step": 19757
+    },
+    {
+      "epoch": 25.29024,
+      "grad_norm": 1.213550090789795,
+      "learning_rate": 1.0500200080032013e-05,
+      "loss": 0.4899,
+      "step": 19758
+    },
+    {
+      "epoch": 25.29152,
+      "grad_norm": 1.2919397354125977,
+      "learning_rate": 1.0498199279711886e-05,
+      "loss": 0.5269,
+      "step": 19759
+    },
+    {
+      "epoch": 25.2928,
+      "grad_norm": 1.1487518548965454,
+      "learning_rate": 1.0496198479391757e-05,
+      "loss": 0.4528,
+      "step": 19760
+    },
+    {
+      "epoch": 25.29408,
+      "grad_norm": 1.0663877725601196,
+      "learning_rate": 1.049419767907163e-05,
+      "loss": 0.4368,
+      "step": 19761
+    },
+    {
+      "epoch": 25.29536,
+      "grad_norm": 1.125251293182373,
+      "learning_rate": 1.04921968787515e-05,
+      "loss": 0.4331,
+      "step": 19762
+    },
+    {
+      "epoch": 25.29664,
+      "grad_norm": 1.1998741626739502,
+      "learning_rate": 1.0490196078431374e-05,
+      "loss": 0.4779,
+      "step": 19763
+    },
+    {
+      "epoch": 25.29792,
+      "grad_norm": 1.1586112976074219,
+      "learning_rate": 1.0488195278111244e-05,
+      "loss": 0.4594,
+      "step": 19764
+    },
+    {
+      "epoch": 25.2992,
+      "grad_norm": 1.1433591842651367,
+      "learning_rate": 1.0486194477791118e-05,
+      "loss": 0.4665,
+      "step": 19765
+    },
+    {
+      "epoch": 25.30048,
+      "grad_norm": 1.2368263006210327,
+      "learning_rate": 1.048419367747099e-05,
+      "loss": 0.5092,
+      "step": 19766
+    },
+    {
+      "epoch": 25.30176,
+      "grad_norm": 1.1645898818969727,
+      "learning_rate": 1.0482192877150861e-05,
+      "loss": 0.4237,
+      "step": 19767
+    },
+    {
+      "epoch": 25.30304,
+      "grad_norm": 1.1961017847061157,
+      "learning_rate": 1.0480192076830733e-05,
+      "loss": 0.4801,
+      "step": 19768
+    },
+    {
+      "epoch": 25.30432,
+      "grad_norm": 1.1524591445922852,
+      "learning_rate": 1.0478191276510605e-05,
+      "loss": 0.4472,
+      "step": 19769
+    },
+    {
+      "epoch": 25.3056,
+      "grad_norm": 1.2043436765670776,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 0.4915,
+      "step": 19770
+    },
+    {
+      "epoch": 25.30688,
+      "grad_norm": 1.173971176147461,
+      "learning_rate": 1.0474189675870349e-05,
+      "loss": 0.464,
+      "step": 19771
+    },
+    {
+      "epoch": 25.30816,
+      "grad_norm": 1.1863977909088135,
+      "learning_rate": 1.047218887555022e-05,
+      "loss": 0.4996,
+      "step": 19772
+    },
+    {
+      "epoch": 25.30944,
+      "grad_norm": 1.2210062742233276,
+      "learning_rate": 1.0470188075230092e-05,
+      "loss": 0.4825,
+      "step": 19773
+    },
+    {
+      "epoch": 25.31072,
+      "grad_norm": 1.209662914276123,
+      "learning_rate": 1.0468187274909964e-05,
+      "loss": 0.4904,
+      "step": 19774
+    },
+    {
+      "epoch": 25.312,
+      "grad_norm": 1.1289560794830322,
+      "learning_rate": 1.0466186474589836e-05,
+      "loss": 0.4321,
+      "step": 19775
+    },
+    {
+      "epoch": 25.31328,
+      "grad_norm": 1.1407843828201294,
+      "learning_rate": 1.0464185674269708e-05,
+      "loss": 0.4982,
+      "step": 19776
+    },
+    {
+      "epoch": 25.31456,
+      "grad_norm": 1.0865232944488525,
+      "learning_rate": 1.046218487394958e-05,
+      "loss": 0.4361,
+      "step": 19777
+    },
+    {
+      "epoch": 25.31584,
+      "grad_norm": 1.192724585533142,
+      "learning_rate": 1.0460184073629452e-05,
+      "loss": 0.5161,
+      "step": 19778
+    },
+    {
+      "epoch": 25.31712,
+      "grad_norm": 1.1117703914642334,
+      "learning_rate": 1.0458183273309324e-05,
+      "loss": 0.4656,
+      "step": 19779
+    },
+    {
+      "epoch": 25.3184,
+      "grad_norm": 1.1453944444656372,
+      "learning_rate": 1.0456182472989195e-05,
+      "loss": 0.4729,
+      "step": 19780
+    },
+    {
+      "epoch": 25.31968,
+      "grad_norm": 1.1950432062149048,
+      "learning_rate": 1.0454181672669067e-05,
+      "loss": 0.4684,
+      "step": 19781
+    },
+    {
+      "epoch": 25.32096,
+      "grad_norm": 1.17280912399292,
+      "learning_rate": 1.0452180872348941e-05,
+      "loss": 0.4754,
+      "step": 19782
+    },
+    {
+      "epoch": 25.32224,
+      "grad_norm": 1.2976539134979248,
+      "learning_rate": 1.0450180072028811e-05,
+      "loss": 0.5065,
+      "step": 19783
+    },
+    {
+      "epoch": 25.32352,
+      "grad_norm": 1.2335752248764038,
+      "learning_rate": 1.0448179271708685e-05,
+      "loss": 0.4805,
+      "step": 19784
+    },
+    {
+      "epoch": 25.3248,
+      "grad_norm": 1.1491906642913818,
+      "learning_rate": 1.0446178471388555e-05,
+      "loss": 0.424,
+      "step": 19785
+    },
+    {
+      "epoch": 25.32608,
+      "grad_norm": 1.1486718654632568,
+      "learning_rate": 1.0444177671068428e-05,
+      "loss": 0.4615,
+      "step": 19786
+    },
+    {
+      "epoch": 25.32736,
+      "grad_norm": 1.197849988937378,
+      "learning_rate": 1.0442176870748298e-05,
+      "loss": 0.4458,
+      "step": 19787
+    },
+    {
+      "epoch": 25.32864,
+      "grad_norm": 1.2207767963409424,
+      "learning_rate": 1.0440176070428172e-05,
+      "loss": 0.5007,
+      "step": 19788
+    },
+    {
+      "epoch": 25.32992,
+      "grad_norm": 1.2052980661392212,
+      "learning_rate": 1.0438175270108044e-05,
+      "loss": 0.4814,
+      "step": 19789
+    },
+    {
+      "epoch": 25.3312,
+      "grad_norm": 1.1656453609466553,
+      "learning_rate": 1.0436174469787916e-05,
+      "loss": 0.4519,
+      "step": 19790
+    },
+    {
+      "epoch": 25.33248,
+      "grad_norm": 1.150758981704712,
+      "learning_rate": 1.0434173669467788e-05,
+      "loss": 0.4832,
+      "step": 19791
+    },
+    {
+      "epoch": 25.33376,
+      "grad_norm": 1.1945995092391968,
+      "learning_rate": 1.043217286914766e-05,
+      "loss": 0.469,
+      "step": 19792
+    },
+    {
+      "epoch": 25.33504,
+      "grad_norm": 1.1622135639190674,
+      "learning_rate": 1.0430172068827531e-05,
+      "loss": 0.4425,
+      "step": 19793
+    },
+    {
+      "epoch": 25.33632,
+      "grad_norm": 1.1778172254562378,
+      "learning_rate": 1.0428171268507403e-05,
+      "loss": 0.4779,
+      "step": 19794
+    },
+    {
+      "epoch": 25.3376,
+      "grad_norm": 1.2094917297363281,
+      "learning_rate": 1.0426170468187275e-05,
+      "loss": 0.4715,
+      "step": 19795
+    },
+    {
+      "epoch": 25.33888,
+      "grad_norm": 1.206709861755371,
+      "learning_rate": 1.0424169667867149e-05,
+      "loss": 0.5091,
+      "step": 19796
+    },
+    {
+      "epoch": 25.34016,
+      "grad_norm": 1.2187248468399048,
+      "learning_rate": 1.0422168867547019e-05,
+      "loss": 0.5013,
+      "step": 19797
+    },
+    {
+      "epoch": 25.34144,
+      "grad_norm": 1.3126270771026611,
+      "learning_rate": 1.0420168067226892e-05,
+      "loss": 0.4873,
+      "step": 19798
+    },
+    {
+      "epoch": 25.34272,
+      "grad_norm": 1.1935737133026123,
+      "learning_rate": 1.0418167266906762e-05,
+      "loss": 0.4794,
+      "step": 19799
+    },
+    {
+      "epoch": 25.344,
+      "grad_norm": 1.2357202768325806,
+      "learning_rate": 1.0416166466586636e-05,
+      "loss": 0.4987,
+      "step": 19800
+    },
+    {
+      "epoch": 25.34528,
+      "grad_norm": 1.1685526371002197,
+      "learning_rate": 1.0414165666266506e-05,
+      "loss": 0.4617,
+      "step": 19801
+    },
+    {
+      "epoch": 25.34656,
+      "grad_norm": 1.1791237592697144,
+      "learning_rate": 1.041216486594638e-05,
+      "loss": 0.4917,
+      "step": 19802
+    },
+    {
+      "epoch": 25.34784,
+      "grad_norm": 1.1622064113616943,
+      "learning_rate": 1.0410164065626252e-05,
+      "loss": 0.4591,
+      "step": 19803
+    },
+    {
+      "epoch": 25.34912,
+      "grad_norm": 1.1878505945205688,
+      "learning_rate": 1.0408163265306123e-05,
+      "loss": 0.483,
+      "step": 19804
+    },
+    {
+      "epoch": 25.3504,
+      "grad_norm": 1.2232707738876343,
+      "learning_rate": 1.0406162464985995e-05,
+      "loss": 0.4873,
+      "step": 19805
+    },
+    {
+      "epoch": 25.35168,
+      "grad_norm": 1.1818996667861938,
+      "learning_rate": 1.0404161664665867e-05,
+      "loss": 0.4899,
+      "step": 19806
+    },
+    {
+      "epoch": 25.35296,
+      "grad_norm": 1.1730505228042603,
+      "learning_rate": 1.0402160864345739e-05,
+      "loss": 0.4547,
+      "step": 19807
+    },
+    {
+      "epoch": 25.35424,
+      "grad_norm": 1.2024863958358765,
+      "learning_rate": 1.040016006402561e-05,
+      "loss": 0.4791,
+      "step": 19808
+    },
+    {
+      "epoch": 25.35552,
+      "grad_norm": 1.1775726079940796,
+      "learning_rate": 1.0398159263705483e-05,
+      "loss": 0.4802,
+      "step": 19809
+    },
+    {
+      "epoch": 25.3568,
+      "grad_norm": 1.1456516981124878,
+      "learning_rate": 1.0396158463385355e-05,
+      "loss": 0.4666,
+      "step": 19810
+    },
+    {
+      "epoch": 25.35808,
+      "grad_norm": 1.1130499839782715,
+      "learning_rate": 1.0394157663065226e-05,
+      "loss": 0.4674,
+      "step": 19811
+    },
+    {
+      "epoch": 25.35936,
+      "grad_norm": 1.2441742420196533,
+      "learning_rate": 1.0392156862745098e-05,
+      "loss": 0.5162,
+      "step": 19812
+    },
+    {
+      "epoch": 25.36064,
+      "grad_norm": 1.139827847480774,
+      "learning_rate": 1.039015606242497e-05,
+      "loss": 0.4504,
+      "step": 19813
+    },
+    {
+      "epoch": 25.36192,
+      "grad_norm": 1.1869721412658691,
+      "learning_rate": 1.0388155262104842e-05,
+      "loss": 0.5,
+      "step": 19814
+    },
+    {
+      "epoch": 25.3632,
+      "grad_norm": 1.1606616973876953,
+      "learning_rate": 1.0386154461784714e-05,
+      "loss": 0.5037,
+      "step": 19815
+    },
+    {
+      "epoch": 25.36448,
+      "grad_norm": 1.1421772241592407,
+      "learning_rate": 1.0384153661464586e-05,
+      "loss": 0.4588,
+      "step": 19816
+    },
+    {
+      "epoch": 25.36576,
+      "grad_norm": 1.224739909172058,
+      "learning_rate": 1.0382152861144458e-05,
+      "loss": 0.4998,
+      "step": 19817
+    },
+    {
+      "epoch": 25.36704,
+      "grad_norm": 1.2049306631088257,
+      "learning_rate": 1.038015206082433e-05,
+      "loss": 0.5418,
+      "step": 19818
+    },
+    {
+      "epoch": 25.36832,
+      "grad_norm": 1.1043879985809326,
+      "learning_rate": 1.0378151260504203e-05,
+      "loss": 0.4709,
+      "step": 19819
+    },
+    {
+      "epoch": 25.3696,
+      "grad_norm": 1.1660434007644653,
+      "learning_rate": 1.0376150460184073e-05,
+      "loss": 0.48,
+      "step": 19820
+    },
+    {
+      "epoch": 25.37088,
+      "grad_norm": 1.2284889221191406,
+      "learning_rate": 1.0374149659863947e-05,
+      "loss": 0.5137,
+      "step": 19821
+    },
+    {
+      "epoch": 25.37216,
+      "grad_norm": 1.163232684135437,
+      "learning_rate": 1.0372148859543817e-05,
+      "loss": 0.4841,
+      "step": 19822
+    },
+    {
+      "epoch": 25.37344,
+      "grad_norm": 1.2111849784851074,
+      "learning_rate": 1.037014805922369e-05,
+      "loss": 0.4561,
+      "step": 19823
+    },
+    {
+      "epoch": 25.37472,
+      "grad_norm": 1.1000887155532837,
+      "learning_rate": 1.036814725890356e-05,
+      "loss": 0.4191,
+      "step": 19824
+    },
+    {
+      "epoch": 25.376,
+      "grad_norm": 1.1611400842666626,
+      "learning_rate": 1.0366146458583434e-05,
+      "loss": 0.5186,
+      "step": 19825
+    },
+    {
+      "epoch": 25.37728,
+      "grad_norm": 1.2061467170715332,
+      "learning_rate": 1.0364145658263306e-05,
+      "loss": 0.4728,
+      "step": 19826
+    },
+    {
+      "epoch": 25.37856,
+      "grad_norm": 1.147557020187378,
+      "learning_rate": 1.0362144857943178e-05,
+      "loss": 0.4578,
+      "step": 19827
+    },
+    {
+      "epoch": 25.37984,
+      "grad_norm": 1.1873794794082642,
+      "learning_rate": 1.036014405762305e-05,
+      "loss": 0.4687,
+      "step": 19828
+    },
+    {
+      "epoch": 25.38112,
+      "grad_norm": 1.1945492029190063,
+      "learning_rate": 1.0358143257302922e-05,
+      "loss": 0.4752,
+      "step": 19829
+    },
+    {
+      "epoch": 25.3824,
+      "grad_norm": 1.1753865480422974,
+      "learning_rate": 1.0356142456982793e-05,
+      "loss": 0.4974,
+      "step": 19830
+    },
+    {
+      "epoch": 25.38368,
+      "grad_norm": 1.1818007230758667,
+      "learning_rate": 1.0354141656662665e-05,
+      "loss": 0.4648,
+      "step": 19831
+    },
+    {
+      "epoch": 25.38496,
+      "grad_norm": 1.2011675834655762,
+      "learning_rate": 1.0352140856342537e-05,
+      "loss": 0.4546,
+      "step": 19832
+    },
+    {
+      "epoch": 25.38624,
+      "grad_norm": 1.203870415687561,
+      "learning_rate": 1.0350140056022409e-05,
+      "loss": 0.4703,
+      "step": 19833
+    },
+    {
+      "epoch": 25.38752,
+      "grad_norm": 1.1994456052780151,
+      "learning_rate": 1.034813925570228e-05,
+      "loss": 0.4695,
+      "step": 19834
+    },
+    {
+      "epoch": 25.3888,
+      "grad_norm": 1.1361202001571655,
+      "learning_rate": 1.0346138455382154e-05,
+      "loss": 0.4831,
+      "step": 19835
+    },
+    {
+      "epoch": 25.39008,
+      "grad_norm": 1.1568169593811035,
+      "learning_rate": 1.0344137655062025e-05,
+      "loss": 0.4666,
+      "step": 19836
+    },
+    {
+      "epoch": 25.39136,
+      "grad_norm": 1.1744780540466309,
+      "learning_rate": 1.0342136854741898e-05,
+      "loss": 0.4628,
+      "step": 19837
+    },
+    {
+      "epoch": 25.39264,
+      "grad_norm": 1.1357437372207642,
+      "learning_rate": 1.0340136054421768e-05,
+      "loss": 0.5288,
+      "step": 19838
+    },
+    {
+      "epoch": 25.39392,
+      "grad_norm": 1.2219725847244263,
+      "learning_rate": 1.0338135254101642e-05,
+      "loss": 0.4923,
+      "step": 19839
+    },
+    {
+      "epoch": 25.3952,
+      "grad_norm": 1.1251201629638672,
+      "learning_rate": 1.0336134453781512e-05,
+      "loss": 0.4548,
+      "step": 19840
+    },
+    {
+      "epoch": 25.39648,
+      "grad_norm": 1.1264392137527466,
+      "learning_rate": 1.0334133653461385e-05,
+      "loss": 0.4398,
+      "step": 19841
+    },
+    {
+      "epoch": 25.39776,
+      "grad_norm": 1.1508971452713013,
+      "learning_rate": 1.0332132853141257e-05,
+      "loss": 0.4722,
+      "step": 19842
+    },
+    {
+      "epoch": 25.39904,
+      "grad_norm": 1.1928080320358276,
+      "learning_rate": 1.033013205282113e-05,
+      "loss": 0.4667,
+      "step": 19843
+    },
+    {
+      "epoch": 25.40032,
+      "grad_norm": 1.1580041646957397,
+      "learning_rate": 1.0328131252501001e-05,
+      "loss": 0.4179,
+      "step": 19844
+    },
+    {
+      "epoch": 25.4016,
+      "grad_norm": 1.135171890258789,
+      "learning_rate": 1.0326130452180873e-05,
+      "loss": 0.4607,
+      "step": 19845
+    },
+    {
+      "epoch": 25.40288,
+      "grad_norm": 1.2124780416488647,
+      "learning_rate": 1.0324129651860745e-05,
+      "loss": 0.502,
+      "step": 19846
+    },
+    {
+      "epoch": 25.40416,
+      "grad_norm": 1.1926053762435913,
+      "learning_rate": 1.0322128851540617e-05,
+      "loss": 0.4687,
+      "step": 19847
+    },
+    {
+      "epoch": 25.40544,
+      "grad_norm": 1.2334791421890259,
+      "learning_rate": 1.0320128051220488e-05,
+      "loss": 0.5041,
+      "step": 19848
+    },
+    {
+      "epoch": 25.40672,
+      "grad_norm": 1.1845077276229858,
+      "learning_rate": 1.0318127250900362e-05,
+      "loss": 0.4649,
+      "step": 19849
+    },
+    {
+      "epoch": 25.408,
+      "grad_norm": 1.1910468339920044,
+      "learning_rate": 1.0316126450580232e-05,
+      "loss": 0.4352,
+      "step": 19850
+    },
+    {
+      "epoch": 25.40928,
+      "grad_norm": 1.1773654222488403,
+      "learning_rate": 1.0314125650260106e-05,
+      "loss": 0.4306,
+      "step": 19851
+    },
+    {
+      "epoch": 25.41056,
+      "grad_norm": 1.211543083190918,
+      "learning_rate": 1.0312124849939976e-05,
+      "loss": 0.4869,
+      "step": 19852
+    },
+    {
+      "epoch": 25.41184,
+      "grad_norm": 1.2373056411743164,
+      "learning_rate": 1.031012404961985e-05,
+      "loss": 0.4914,
+      "step": 19853
+    },
+    {
+      "epoch": 25.41312,
+      "grad_norm": 1.2094565629959106,
+      "learning_rate": 1.030812324929972e-05,
+      "loss": 0.4829,
+      "step": 19854
+    },
+    {
+      "epoch": 25.4144,
+      "grad_norm": 1.1150031089782715,
+      "learning_rate": 1.0306122448979593e-05,
+      "loss": 0.4435,
+      "step": 19855
+    },
+    {
+      "epoch": 25.41568,
+      "grad_norm": 1.189208984375,
+      "learning_rate": 1.0304121648659465e-05,
+      "loss": 0.4767,
+      "step": 19856
+    },
+    {
+      "epoch": 25.41696,
+      "grad_norm": 1.1635322570800781,
+      "learning_rate": 1.0302120848339337e-05,
+      "loss": 0.4754,
+      "step": 19857
+    },
+    {
+      "epoch": 25.41824,
+      "grad_norm": 1.2085617780685425,
+      "learning_rate": 1.0300120048019209e-05,
+      "loss": 0.5219,
+      "step": 19858
+    },
+    {
+      "epoch": 25.41952,
+      "grad_norm": 1.150395393371582,
+      "learning_rate": 1.029811924769908e-05,
+      "loss": 0.4459,
+      "step": 19859
+    },
+    {
+      "epoch": 25.4208,
+      "grad_norm": 1.1318018436431885,
+      "learning_rate": 1.0296118447378952e-05,
+      "loss": 0.453,
+      "step": 19860
+    },
+    {
+      "epoch": 25.42208,
+      "grad_norm": 1.1467453241348267,
+      "learning_rate": 1.0294117647058824e-05,
+      "loss": 0.4603,
+      "step": 19861
+    },
+    {
+      "epoch": 25.42336,
+      "grad_norm": 1.1553596258163452,
+      "learning_rate": 1.0292116846738696e-05,
+      "loss": 0.4754,
+      "step": 19862
+    },
+    {
+      "epoch": 25.42464,
+      "grad_norm": 1.1549872159957886,
+      "learning_rate": 1.0290116046418568e-05,
+      "loss": 0.4961,
+      "step": 19863
+    },
+    {
+      "epoch": 25.42592,
+      "grad_norm": 1.1191171407699585,
+      "learning_rate": 1.028811524609844e-05,
+      "loss": 0.4749,
+      "step": 19864
+    },
+    {
+      "epoch": 25.4272,
+      "grad_norm": 1.1042835712432861,
+      "learning_rate": 1.0286114445778312e-05,
+      "loss": 0.4217,
+      "step": 19865
+    },
+    {
+      "epoch": 25.42848,
+      "grad_norm": 1.2307777404785156,
+      "learning_rate": 1.0284113645458184e-05,
+      "loss": 0.4629,
+      "step": 19866
+    },
+    {
+      "epoch": 25.42976,
+      "grad_norm": 1.1764090061187744,
+      "learning_rate": 1.0282112845138055e-05,
+      "loss": 0.4898,
+      "step": 19867
+    },
+    {
+      "epoch": 25.43104,
+      "grad_norm": 1.189038634300232,
+      "learning_rate": 1.0280112044817927e-05,
+      "loss": 0.4423,
+      "step": 19868
+    },
+    {
+      "epoch": 25.43232,
+      "grad_norm": 1.1858000755310059,
+      "learning_rate": 1.02781112444978e-05,
+      "loss": 0.4757,
+      "step": 19869
+    },
+    {
+      "epoch": 25.4336,
+      "grad_norm": 1.191748023033142,
+      "learning_rate": 1.0276110444177671e-05,
+      "loss": 0.4882,
+      "step": 19870
+    },
+    {
+      "epoch": 25.43488,
+      "grad_norm": 1.2290452718734741,
+      "learning_rate": 1.0274109643857543e-05,
+      "loss": 0.4923,
+      "step": 19871
+    },
+    {
+      "epoch": 25.43616,
+      "grad_norm": 1.1514322757720947,
+      "learning_rate": 1.0272108843537416e-05,
+      "loss": 0.4417,
+      "step": 19872
+    },
+    {
+      "epoch": 25.43744,
+      "grad_norm": 1.162440299987793,
+      "learning_rate": 1.0270108043217287e-05,
+      "loss": 0.4663,
+      "step": 19873
+    },
+    {
+      "epoch": 25.43872,
+      "grad_norm": 1.1626965999603271,
+      "learning_rate": 1.026810724289716e-05,
+      "loss": 0.4808,
+      "step": 19874
+    },
+    {
+      "epoch": 25.44,
+      "grad_norm": 1.1720807552337646,
+      "learning_rate": 1.026610644257703e-05,
+      "loss": 0.4818,
+      "step": 19875
+    },
+    {
+      "epoch": 25.44128,
+      "grad_norm": 1.1925668716430664,
+      "learning_rate": 1.0264105642256904e-05,
+      "loss": 0.4541,
+      "step": 19876
+    },
+    {
+      "epoch": 25.44256,
+      "grad_norm": 1.2147929668426514,
+      "learning_rate": 1.0262104841936774e-05,
+      "loss": 0.465,
+      "step": 19877
+    },
+    {
+      "epoch": 25.44384,
+      "grad_norm": 1.1247954368591309,
+      "learning_rate": 1.0260104041616648e-05,
+      "loss": 0.4811,
+      "step": 19878
+    },
+    {
+      "epoch": 25.44512,
+      "grad_norm": 1.1577873229980469,
+      "learning_rate": 1.025810324129652e-05,
+      "loss": 0.4534,
+      "step": 19879
+    },
+    {
+      "epoch": 25.4464,
+      "grad_norm": 1.1290161609649658,
+      "learning_rate": 1.0256102440976391e-05,
+      "loss": 0.4636,
+      "step": 19880
+    },
+    {
+      "epoch": 25.44768,
+      "grad_norm": 1.1666961908340454,
+      "learning_rate": 1.0254101640656263e-05,
+      "loss": 0.4773,
+      "step": 19881
+    },
+    {
+      "epoch": 25.44896,
+      "grad_norm": 1.210201382637024,
+      "learning_rate": 1.0252100840336135e-05,
+      "loss": 0.5152,
+      "step": 19882
+    },
+    {
+      "epoch": 25.45024,
+      "grad_norm": 1.1907308101654053,
+      "learning_rate": 1.0250100040016007e-05,
+      "loss": 0.4904,
+      "step": 19883
+    },
+    {
+      "epoch": 25.45152,
+      "grad_norm": 1.2196985483169556,
+      "learning_rate": 1.0248099239695879e-05,
+      "loss": 0.499,
+      "step": 19884
+    },
+    {
+      "epoch": 25.4528,
+      "grad_norm": 1.1352413892745972,
+      "learning_rate": 1.024609843937575e-05,
+      "loss": 0.4525,
+      "step": 19885
+    },
+    {
+      "epoch": 25.45408,
+      "grad_norm": 1.1208443641662598,
+      "learning_rate": 1.0244097639055624e-05,
+      "loss": 0.4592,
+      "step": 19886
+    },
+    {
+      "epoch": 25.45536,
+      "grad_norm": 1.152306079864502,
+      "learning_rate": 1.0242096838735494e-05,
+      "loss": 0.4724,
+      "step": 19887
+    },
+    {
+      "epoch": 25.45664,
+      "grad_norm": 1.1952404975891113,
+      "learning_rate": 1.0240096038415368e-05,
+      "loss": 0.507,
+      "step": 19888
+    },
+    {
+      "epoch": 25.45792,
+      "grad_norm": 1.207161545753479,
+      "learning_rate": 1.0238095238095238e-05,
+      "loss": 0.4736,
+      "step": 19889
+    },
+    {
+      "epoch": 25.4592,
+      "grad_norm": 1.2183301448822021,
+      "learning_rate": 1.0236094437775112e-05,
+      "loss": 0.4445,
+      "step": 19890
+    },
+    {
+      "epoch": 25.46048,
+      "grad_norm": 1.1657265424728394,
+      "learning_rate": 1.0234093637454982e-05,
+      "loss": 0.4539,
+      "step": 19891
+    },
+    {
+      "epoch": 25.46176,
+      "grad_norm": 1.19996976852417,
+      "learning_rate": 1.0232092837134855e-05,
+      "loss": 0.4312,
+      "step": 19892
+    },
+    {
+      "epoch": 25.46304,
+      "grad_norm": 1.202269434928894,
+      "learning_rate": 1.0230092036814725e-05,
+      "loss": 0.4801,
+      "step": 19893
+    },
+    {
+      "epoch": 25.46432,
+      "grad_norm": 1.1666241884231567,
+      "learning_rate": 1.0228091236494599e-05,
+      "loss": 0.4813,
+      "step": 19894
+    },
+    {
+      "epoch": 25.4656,
+      "grad_norm": 1.2247138023376465,
+      "learning_rate": 1.022609043617447e-05,
+      "loss": 0.4749,
+      "step": 19895
+    },
+    {
+      "epoch": 25.46688,
+      "grad_norm": 1.2047659158706665,
+      "learning_rate": 1.0224089635854343e-05,
+      "loss": 0.4895,
+      "step": 19896
+    },
+    {
+      "epoch": 25.46816,
+      "grad_norm": 1.2424787282943726,
+      "learning_rate": 1.0222088835534215e-05,
+      "loss": 0.4762,
+      "step": 19897
+    },
+    {
+      "epoch": 25.46944,
+      "grad_norm": 1.190301775932312,
+      "learning_rate": 1.0220088035214086e-05,
+      "loss": 0.4786,
+      "step": 19898
+    },
+    {
+      "epoch": 25.47072,
+      "grad_norm": 1.2236806154251099,
+      "learning_rate": 1.0218087234893958e-05,
+      "loss": 0.5193,
+      "step": 19899
+    },
+    {
+      "epoch": 25.472,
+      "grad_norm": 1.1594198942184448,
+      "learning_rate": 1.021608643457383e-05,
+      "loss": 0.5058,
+      "step": 19900
+    },
+    {
+      "epoch": 25.47328,
+      "grad_norm": 1.1569578647613525,
+      "learning_rate": 1.0214085634253702e-05,
+      "loss": 0.4701,
+      "step": 19901
+    },
+    {
+      "epoch": 25.47456,
+      "grad_norm": 1.1034061908721924,
+      "learning_rate": 1.0212084833933574e-05,
+      "loss": 0.4354,
+      "step": 19902
+    },
+    {
+      "epoch": 25.47584,
+      "grad_norm": 1.136178970336914,
+      "learning_rate": 1.0210084033613446e-05,
+      "loss": 0.4281,
+      "step": 19903
+    },
+    {
+      "epoch": 25.47712,
+      "grad_norm": 1.1307159662246704,
+      "learning_rate": 1.0208083233293318e-05,
+      "loss": 0.4523,
+      "step": 19904
+    },
+    {
+      "epoch": 25.4784,
+      "grad_norm": 1.1765360832214355,
+      "learning_rate": 1.020608243297319e-05,
+      "loss": 0.4607,
+      "step": 19905
+    },
+    {
+      "epoch": 25.47968,
+      "grad_norm": 1.0838277339935303,
+      "learning_rate": 1.0204081632653061e-05,
+      "loss": 0.4631,
+      "step": 19906
+    },
+    {
+      "epoch": 25.48096,
+      "grad_norm": 1.1757216453552246,
+      "learning_rate": 1.0202080832332933e-05,
+      "loss": 0.4862,
+      "step": 19907
+    },
+    {
+      "epoch": 25.48224,
+      "grad_norm": 1.1511602401733398,
+      "learning_rate": 1.0200080032012805e-05,
+      "loss": 0.456,
+      "step": 19908
+    },
+    {
+      "epoch": 25.48352,
+      "grad_norm": 1.1518641710281372,
+      "learning_rate": 1.0198079231692679e-05,
+      "loss": 0.5169,
+      "step": 19909
+    },
+    {
+      "epoch": 25.4848,
+      "grad_norm": 1.1432472467422485,
+      "learning_rate": 1.0196078431372549e-05,
+      "loss": 0.4253,
+      "step": 19910
+    },
+    {
+      "epoch": 25.48608,
+      "grad_norm": 1.1166365146636963,
+      "learning_rate": 1.0194077631052422e-05,
+      "loss": 0.464,
+      "step": 19911
+    },
+    {
+      "epoch": 25.48736,
+      "grad_norm": 1.0852631330490112,
+      "learning_rate": 1.0192076830732292e-05,
+      "loss": 0.4376,
+      "step": 19912
+    },
+    {
+      "epoch": 25.48864,
+      "grad_norm": 1.1255972385406494,
+      "learning_rate": 1.0190076030412166e-05,
+      "loss": 0.4564,
+      "step": 19913
+    },
+    {
+      "epoch": 25.48992,
+      "grad_norm": 1.1327401399612427,
+      "learning_rate": 1.0188075230092036e-05,
+      "loss": 0.4365,
+      "step": 19914
+    },
+    {
+      "epoch": 25.4912,
+      "grad_norm": 1.1716783046722412,
+      "learning_rate": 1.018607442977191e-05,
+      "loss": 0.5002,
+      "step": 19915
+    },
+    {
+      "epoch": 25.49248,
+      "grad_norm": 1.2162928581237793,
+      "learning_rate": 1.018407362945178e-05,
+      "loss": 0.5105,
+      "step": 19916
+    },
+    {
+      "epoch": 25.49376,
+      "grad_norm": 1.23090660572052,
+      "learning_rate": 1.0182072829131653e-05,
+      "loss": 0.4842,
+      "step": 19917
+    },
+    {
+      "epoch": 25.49504,
+      "grad_norm": 1.1792917251586914,
+      "learning_rate": 1.0180072028811525e-05,
+      "loss": 0.505,
+      "step": 19918
+    },
+    {
+      "epoch": 25.49632,
+      "grad_norm": 1.1648669242858887,
+      "learning_rate": 1.0178071228491397e-05,
+      "loss": 0.4483,
+      "step": 19919
+    },
+    {
+      "epoch": 25.4976,
+      "grad_norm": 1.1607261896133423,
+      "learning_rate": 1.0176070428171269e-05,
+      "loss": 0.4921,
+      "step": 19920
+    },
+    {
+      "epoch": 25.49888,
+      "grad_norm": 1.116874098777771,
+      "learning_rate": 1.017406962785114e-05,
+      "loss": 0.4774,
+      "step": 19921
+    },
+    {
+      "epoch": 25.50016,
+      "grad_norm": 1.107079267501831,
+      "learning_rate": 1.0172068827531013e-05,
+      "loss": 0.4342,
+      "step": 19922
+    },
+    {
+      "epoch": 25.50144,
+      "grad_norm": 1.134993553161621,
+      "learning_rate": 1.0170068027210885e-05,
+      "loss": 0.4498,
+      "step": 19923
+    },
+    {
+      "epoch": 25.50272,
+      "grad_norm": 1.218907356262207,
+      "learning_rate": 1.0168067226890756e-05,
+      "loss": 0.4602,
+      "step": 19924
+    },
+    {
+      "epoch": 25.504,
+      "grad_norm": 1.1606868505477905,
+      "learning_rate": 1.016606642657063e-05,
+      "loss": 0.4432,
+      "step": 19925
+    },
+    {
+      "epoch": 25.50528,
+      "grad_norm": 1.0475679636001587,
+      "learning_rate": 1.01640656262505e-05,
+      "loss": 0.4417,
+      "step": 19926
+    },
+    {
+      "epoch": 25.50656,
+      "grad_norm": 1.0612995624542236,
+      "learning_rate": 1.0162064825930374e-05,
+      "loss": 0.4264,
+      "step": 19927
+    },
+    {
+      "epoch": 25.50784,
+      "grad_norm": 1.1685255765914917,
+      "learning_rate": 1.0160064025610244e-05,
+      "loss": 0.4719,
+      "step": 19928
+    },
+    {
+      "epoch": 25.50912,
+      "grad_norm": 1.1424115896224976,
+      "learning_rate": 1.0158063225290117e-05,
+      "loss": 0.4454,
+      "step": 19929
+    },
+    {
+      "epoch": 25.5104,
+      "grad_norm": 1.1809263229370117,
+      "learning_rate": 1.0156062424969988e-05,
+      "loss": 0.4515,
+      "step": 19930
+    },
+    {
+      "epoch": 25.51168,
+      "grad_norm": 1.25111985206604,
+      "learning_rate": 1.0154061624649861e-05,
+      "loss": 0.5329,
+      "step": 19931
+    },
+    {
+      "epoch": 25.51296,
+      "grad_norm": 1.130300760269165,
+      "learning_rate": 1.0152060824329733e-05,
+      "loss": 0.4795,
+      "step": 19932
+    },
+    {
+      "epoch": 25.51424,
+      "grad_norm": 1.1218140125274658,
+      "learning_rate": 1.0150060024009605e-05,
+      "loss": 0.4573,
+      "step": 19933
+    },
+    {
+      "epoch": 25.51552,
+      "grad_norm": 1.183595895767212,
+      "learning_rate": 1.0148059223689477e-05,
+      "loss": 0.4736,
+      "step": 19934
+    },
+    {
+      "epoch": 25.5168,
+      "grad_norm": 1.1084133386611938,
+      "learning_rate": 1.0146058423369348e-05,
+      "loss": 0.4453,
+      "step": 19935
+    },
+    {
+      "epoch": 25.51808,
+      "grad_norm": 1.1661556959152222,
+      "learning_rate": 1.014405762304922e-05,
+      "loss": 0.4992,
+      "step": 19936
+    },
+    {
+      "epoch": 25.51936,
+      "grad_norm": 1.1977107524871826,
+      "learning_rate": 1.0142056822729092e-05,
+      "loss": 0.4746,
+      "step": 19937
+    },
+    {
+      "epoch": 25.52064,
+      "grad_norm": 1.1810569763183594,
+      "learning_rate": 1.0140056022408964e-05,
+      "loss": 0.4634,
+      "step": 19938
+    },
+    {
+      "epoch": 25.52192,
+      "grad_norm": 1.1430659294128418,
+      "learning_rate": 1.0138055222088836e-05,
+      "loss": 0.5255,
+      "step": 19939
+    },
+    {
+      "epoch": 25.5232,
+      "grad_norm": 1.135312795639038,
+      "learning_rate": 1.0136054421768708e-05,
+      "loss": 0.4624,
+      "step": 19940
+    },
+    {
+      "epoch": 25.52448,
+      "grad_norm": 1.0835295915603638,
+      "learning_rate": 1.013405362144858e-05,
+      "loss": 0.4393,
+      "step": 19941
+    },
+    {
+      "epoch": 25.52576,
+      "grad_norm": 1.1327999830245972,
+      "learning_rate": 1.0132052821128451e-05,
+      "loss": 0.4165,
+      "step": 19942
+    },
+    {
+      "epoch": 25.52704,
+      "grad_norm": 1.2144399881362915,
+      "learning_rate": 1.0130052020808323e-05,
+      "loss": 0.5114,
+      "step": 19943
+    },
+    {
+      "epoch": 25.52832,
+      "grad_norm": 1.2352372407913208,
+      "learning_rate": 1.0128051220488195e-05,
+      "loss": 0.5182,
+      "step": 19944
+    },
+    {
+      "epoch": 25.5296,
+      "grad_norm": 1.109834909439087,
+      "learning_rate": 1.0126050420168067e-05,
+      "loss": 0.4612,
+      "step": 19945
+    },
+    {
+      "epoch": 25.53088,
+      "grad_norm": 1.163332462310791,
+      "learning_rate": 1.0124049619847939e-05,
+      "loss": 0.4823,
+      "step": 19946
+    },
+    {
+      "epoch": 25.53216,
+      "grad_norm": 1.219772458076477,
+      "learning_rate": 1.012204881952781e-05,
+      "loss": 0.5057,
+      "step": 19947
+    },
+    {
+      "epoch": 25.53344,
+      "grad_norm": 1.111663818359375,
+      "learning_rate": 1.0120048019207684e-05,
+      "loss": 0.4297,
+      "step": 19948
+    },
+    {
+      "epoch": 25.53472,
+      "grad_norm": 1.1989513635635376,
+      "learning_rate": 1.0118047218887554e-05,
+      "loss": 0.4661,
+      "step": 19949
+    },
+    {
+      "epoch": 25.536,
+      "grad_norm": 1.1225396394729614,
+      "learning_rate": 1.0116046418567428e-05,
+      "loss": 0.5113,
+      "step": 19950
+    },
+    {
+      "epoch": 25.53728,
+      "grad_norm": 1.134542465209961,
+      "learning_rate": 1.0114045618247298e-05,
+      "loss": 0.4717,
+      "step": 19951
+    },
+    {
+      "epoch": 25.53856,
+      "grad_norm": 1.1759467124938965,
+      "learning_rate": 1.0112044817927172e-05,
+      "loss": 0.4592,
+      "step": 19952
+    },
+    {
+      "epoch": 25.53984,
+      "grad_norm": 1.1189574003219604,
+      "learning_rate": 1.0110044017607042e-05,
+      "loss": 0.4471,
+      "step": 19953
+    },
+    {
+      "epoch": 25.54112,
+      "grad_norm": 1.2080943584442139,
+      "learning_rate": 1.0108043217286915e-05,
+      "loss": 0.5141,
+      "step": 19954
+    },
+    {
+      "epoch": 25.5424,
+      "grad_norm": 1.1960744857788086,
+      "learning_rate": 1.0106042416966787e-05,
+      "loss": 0.4721,
+      "step": 19955
+    },
+    {
+      "epoch": 25.54368,
+      "grad_norm": 1.1887983083724976,
+      "learning_rate": 1.010404161664666e-05,
+      "loss": 0.4531,
+      "step": 19956
+    },
+    {
+      "epoch": 25.54496,
+      "grad_norm": 1.228134036064148,
+      "learning_rate": 1.0102040816326531e-05,
+      "loss": 0.476,
+      "step": 19957
+    },
+    {
+      "epoch": 25.54624,
+      "grad_norm": 1.1741596460342407,
+      "learning_rate": 1.0100040016006403e-05,
+      "loss": 0.4897,
+      "step": 19958
+    },
+    {
+      "epoch": 25.54752,
+      "grad_norm": 1.1398587226867676,
+      "learning_rate": 1.0098039215686275e-05,
+      "loss": 0.4474,
+      "step": 19959
+    },
+    {
+      "epoch": 25.5488,
+      "grad_norm": 1.1780531406402588,
+      "learning_rate": 1.0096038415366147e-05,
+      "loss": 0.4595,
+      "step": 19960
+    },
+    {
+      "epoch": 25.55008,
+      "grad_norm": 1.1379321813583374,
+      "learning_rate": 1.0094037615046018e-05,
+      "loss": 0.4659,
+      "step": 19961
+    },
+    {
+      "epoch": 25.55136,
+      "grad_norm": 1.0986708402633667,
+      "learning_rate": 1.0092036814725892e-05,
+      "loss": 0.4721,
+      "step": 19962
+    },
+    {
+      "epoch": 25.55264,
+      "grad_norm": 1.1313260793685913,
+      "learning_rate": 1.0090036014405762e-05,
+      "loss": 0.4701,
+      "step": 19963
+    },
+    {
+      "epoch": 25.55392,
+      "grad_norm": 1.0673236846923828,
+      "learning_rate": 1.0088035214085636e-05,
+      "loss": 0.4375,
+      "step": 19964
+    },
+    {
+      "epoch": 25.5552,
+      "grad_norm": 1.1325130462646484,
+      "learning_rate": 1.0086034413765506e-05,
+      "loss": 0.4712,
+      "step": 19965
+    },
+    {
+      "epoch": 25.55648,
+      "grad_norm": 1.095177412033081,
+      "learning_rate": 1.008403361344538e-05,
+      "loss": 0.4617,
+      "step": 19966
+    },
+    {
+      "epoch": 25.557760000000002,
+      "grad_norm": 1.203389048576355,
+      "learning_rate": 1.008203281312525e-05,
+      "loss": 0.4962,
+      "step": 19967
+    },
+    {
+      "epoch": 25.55904,
+      "grad_norm": 1.1400015354156494,
+      "learning_rate": 1.0080032012805123e-05,
+      "loss": 0.4451,
+      "step": 19968
+    },
+    {
+      "epoch": 25.56032,
+      "grad_norm": 1.153382420539856,
+      "learning_rate": 1.0078031212484995e-05,
+      "loss": 0.4675,
+      "step": 19969
+    },
+    {
+      "epoch": 25.5616,
+      "grad_norm": 1.207768440246582,
+      "learning_rate": 1.0076030412164867e-05,
+      "loss": 0.5011,
+      "step": 19970
+    },
+    {
+      "epoch": 25.56288,
+      "grad_norm": 1.1720563173294067,
+      "learning_rate": 1.0074029611844739e-05,
+      "loss": 0.4832,
+      "step": 19971
+    },
+    {
+      "epoch": 25.56416,
+      "grad_norm": 1.2000620365142822,
+      "learning_rate": 1.007202881152461e-05,
+      "loss": 0.4259,
+      "step": 19972
+    },
+    {
+      "epoch": 25.56544,
+      "grad_norm": 1.208329200744629,
+      "learning_rate": 1.0070028011204482e-05,
+      "loss": 0.543,
+      "step": 19973
+    },
+    {
+      "epoch": 25.56672,
+      "grad_norm": 1.165389895439148,
+      "learning_rate": 1.0068027210884354e-05,
+      "loss": 0.4622,
+      "step": 19974
+    },
+    {
+      "epoch": 25.568,
+      "grad_norm": 1.1432007551193237,
+      "learning_rate": 1.0066026410564226e-05,
+      "loss": 0.4856,
+      "step": 19975
+    },
+    {
+      "epoch": 25.56928,
+      "grad_norm": 1.1231637001037598,
+      "learning_rate": 1.0064025610244098e-05,
+      "loss": 0.4722,
+      "step": 19976
+    },
+    {
+      "epoch": 25.57056,
+      "grad_norm": 1.2253926992416382,
+      "learning_rate": 1.006202480992397e-05,
+      "loss": 0.4966,
+      "step": 19977
+    },
+    {
+      "epoch": 25.57184,
+      "grad_norm": 1.1498738527297974,
+      "learning_rate": 1.0060024009603842e-05,
+      "loss": 0.4766,
+      "step": 19978
+    },
+    {
+      "epoch": 25.57312,
+      "grad_norm": 1.264626383781433,
+      "learning_rate": 1.0058023209283714e-05,
+      "loss": 0.547,
+      "step": 19979
+    },
+    {
+      "epoch": 25.5744,
+      "grad_norm": 1.1682270765304565,
+      "learning_rate": 1.0056022408963585e-05,
+      "loss": 0.4642,
+      "step": 19980
+    },
+    {
+      "epoch": 25.57568,
+      "grad_norm": 1.1158068180084229,
+      "learning_rate": 1.0054021608643457e-05,
+      "loss": 0.4469,
+      "step": 19981
+    },
+    {
+      "epoch": 25.57696,
+      "grad_norm": 1.223067283630371,
+      "learning_rate": 1.005202080832333e-05,
+      "loss": 0.4862,
+      "step": 19982
+    },
+    {
+      "epoch": 25.57824,
+      "grad_norm": 1.2305022478103638,
+      "learning_rate": 1.0050020008003201e-05,
+      "loss": 0.5224,
+      "step": 19983
+    },
+    {
+      "epoch": 25.57952,
+      "grad_norm": 1.176965594291687,
+      "learning_rate": 1.0048019207683073e-05,
+      "loss": 0.4521,
+      "step": 19984
+    },
+    {
+      "epoch": 25.5808,
+      "grad_norm": 1.1142029762268066,
+      "learning_rate": 1.0046018407362946e-05,
+      "loss": 0.4352,
+      "step": 19985
+    },
+    {
+      "epoch": 25.58208,
+      "grad_norm": 1.123673677444458,
+      "learning_rate": 1.0044017607042817e-05,
+      "loss": 0.4876,
+      "step": 19986
+    },
+    {
+      "epoch": 25.58336,
+      "grad_norm": 1.1586241722106934,
+      "learning_rate": 1.004201680672269e-05,
+      "loss": 0.4535,
+      "step": 19987
+    },
+    {
+      "epoch": 25.58464,
+      "grad_norm": 1.125301480293274,
+      "learning_rate": 1.004001600640256e-05,
+      "loss": 0.4405,
+      "step": 19988
+    },
+    {
+      "epoch": 25.58592,
+      "grad_norm": 1.1621462106704712,
+      "learning_rate": 1.0038015206082434e-05,
+      "loss": 0.4536,
+      "step": 19989
+    },
+    {
+      "epoch": 25.5872,
+      "grad_norm": 1.1989449262619019,
+      "learning_rate": 1.0036014405762304e-05,
+      "loss": 0.4923,
+      "step": 19990
+    },
+    {
+      "epoch": 25.58848,
+      "grad_norm": 1.2146766185760498,
+      "learning_rate": 1.0034013605442178e-05,
+      "loss": 0.4667,
+      "step": 19991
+    },
+    {
+      "epoch": 25.58976,
+      "grad_norm": 1.1815863847732544,
+      "learning_rate": 1.003201280512205e-05,
+      "loss": 0.4751,
+      "step": 19992
+    },
+    {
+      "epoch": 25.59104,
+      "grad_norm": 1.1977012157440186,
+      "learning_rate": 1.0030012004801921e-05,
+      "loss": 0.4829,
+      "step": 19993
+    },
+    {
+      "epoch": 25.59232,
+      "grad_norm": 1.1629470586776733,
+      "learning_rate": 1.0028011204481793e-05,
+      "loss": 0.4888,
+      "step": 19994
+    },
+    {
+      "epoch": 25.5936,
+      "grad_norm": 1.2305023670196533,
+      "learning_rate": 1.0026010404161665e-05,
+      "loss": 0.5316,
+      "step": 19995
+    },
+    {
+      "epoch": 25.59488,
+      "grad_norm": 1.1644953489303589,
+      "learning_rate": 1.0024009603841537e-05,
+      "loss": 0.5375,
+      "step": 19996
+    },
+    {
+      "epoch": 25.59616,
+      "grad_norm": 1.1814461946487427,
+      "learning_rate": 1.0022008803521409e-05,
+      "loss": 0.4741,
+      "step": 19997
+    },
+    {
+      "epoch": 25.59744,
+      "grad_norm": 1.1363074779510498,
+      "learning_rate": 1.002000800320128e-05,
+      "loss": 0.4249,
+      "step": 19998
+    },
+    {
+      "epoch": 25.59872,
+      "grad_norm": 1.2371721267700195,
+      "learning_rate": 1.0018007202881154e-05,
+      "loss": 0.4753,
+      "step": 19999
+    },
+    {
+      "epoch": 25.6,
+      "grad_norm": 1.137081265449524,
+      "learning_rate": 1.0016006402561024e-05,
+      "loss": 0.4507,
+      "step": 20000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 25000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 33,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.571466040761385e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}