diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,7033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.20990764063811923,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00020990764063811922,
+      "grad_norm": 0.8211116790771484,
+      "learning_rate": 2.09643605870021e-07,
+      "loss": 0.9775,
+      "step": 1
+    },
+    {
+      "epoch": 0.00041981528127623844,
+      "grad_norm": 0.7017123103141785,
+      "learning_rate": 4.19287211740042e-07,
+      "loss": 0.868,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006297229219143577,
+      "grad_norm": 0.6900752782821655,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 0.8862,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008396305625524769,
+      "grad_norm": 0.7597951292991638,
+      "learning_rate": 8.38574423480084e-07,
+      "loss": 0.923,
+      "step": 4
+    },
+    {
+      "epoch": 0.0010495382031905961,
+      "grad_norm": 0.8734421730041504,
+      "learning_rate": 1.048218029350105e-06,
+      "loss": 1.0745,
+      "step": 5
+    },
+    {
+      "epoch": 0.0012594458438287153,
+      "grad_norm": 0.8313158750534058,
+      "learning_rate": 1.257861635220126e-06,
+      "loss": 0.9817,
+      "step": 6
+    },
+    {
+      "epoch": 0.0014693534844668346,
+      "grad_norm": 0.656913697719574,
+      "learning_rate": 1.467505241090147e-06,
+      "loss": 0.8544,
+      "step": 7
+    },
+    {
+      "epoch": 0.0016792611251049538,
+      "grad_norm": 0.7147523164749146,
+      "learning_rate": 1.677148846960168e-06,
+      "loss": 0.8955,
+      "step": 8
+    },
+    {
+      "epoch": 0.001889168765743073,
+      "grad_norm": 0.753795325756073,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 0.881,
+      "step": 9
+    },
+    {
+      "epoch": 0.0020990764063811922,
+      "grad_norm": 0.7141973972320557,
+      "learning_rate": 2.09643605870021e-06,
+      "loss": 0.9229,
+      "step": 10
+    },
+    {
+      "epoch": 0.0023089840470193117,
+      "grad_norm": 0.8047632575035095,
+      "learning_rate": 2.306079664570231e-06,
+      "loss": 0.9438,
+      "step": 11
+    },
+    {
+      "epoch": 0.0025188916876574307,
+      "grad_norm": 0.8492773175239563,
+      "learning_rate": 2.515723270440252e-06,
+      "loss": 0.9977,
+      "step": 12
+    },
+    {
+      "epoch": 0.00272879932829555,
+      "grad_norm": 0.744103729724884,
+      "learning_rate": 2.7253668763102727e-06,
+      "loss": 0.9049,
+      "step": 13
+    },
+    {
+      "epoch": 0.002938706968933669,
+      "grad_norm": 0.7073290944099426,
+      "learning_rate": 2.935010482180294e-06,
+      "loss": 0.8824,
+      "step": 14
+    },
+    {
+      "epoch": 0.0031486146095717885,
+      "grad_norm": 0.6808626651763916,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 0.8854,
+      "step": 15
+    },
+    {
+      "epoch": 0.0033585222502099076,
+      "grad_norm": 0.6526097655296326,
+      "learning_rate": 3.354297693920336e-06,
+      "loss": 0.8219,
+      "step": 16
+    },
+    {
+      "epoch": 0.003568429890848027,
+      "grad_norm": 0.8246333003044128,
+      "learning_rate": 3.563941299790356e-06,
+      "loss": 0.9948,
+      "step": 17
+    },
+    {
+      "epoch": 0.003778337531486146,
+      "grad_norm": 0.8494473099708557,
+      "learning_rate": 3.7735849056603773e-06,
+      "loss": 1.0187,
+      "step": 18
+    },
+    {
+      "epoch": 0.003988245172124265,
+      "grad_norm": 0.9259452223777771,
+      "learning_rate": 3.9832285115303985e-06,
+      "loss": 1.0864,
+      "step": 19
+    },
+    {
+      "epoch": 0.0041981528127623844,
+      "grad_norm": 0.8235483765602112,
+      "learning_rate": 4.19287211740042e-06,
+      "loss": 1.0255,
+      "step": 20
+    },
+    {
+      "epoch": 0.004408060453400504,
+      "grad_norm": 0.8101536631584167,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 0.9892,
+      "step": 21
+    },
+    {
+      "epoch": 0.004617968094038623,
+      "grad_norm": 0.7406070232391357,
+      "learning_rate": 4.612159329140462e-06,
+      "loss": 0.9266,
+      "step": 22
+    },
+    {
+      "epoch": 0.004827875734676742,
+      "grad_norm": 0.7151230573654175,
+      "learning_rate": 4.821802935010482e-06,
+      "loss": 0.902,
+      "step": 23
+    },
+    {
+      "epoch": 0.005037783375314861,
+      "grad_norm": 0.6921148896217346,
+      "learning_rate": 5.031446540880504e-06,
+      "loss": 0.8621,
+      "step": 24
+    },
+    {
+      "epoch": 0.005247691015952981,
+      "grad_norm": 0.6169761419296265,
+      "learning_rate": 5.241090146750524e-06,
+      "loss": 0.8296,
+      "step": 25
+    },
+    {
+      "epoch": 0.0054575986565911,
+      "grad_norm": 0.679093599319458,
+      "learning_rate": 5.4507337526205454e-06,
+      "loss": 0.8593,
+      "step": 26
+    },
+    {
+      "epoch": 0.005667506297229219,
+      "grad_norm": 0.8058671951293945,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.9504,
+      "step": 27
+    },
+    {
+      "epoch": 0.005877413937867338,
+      "grad_norm": 0.8705667853355408,
+      "learning_rate": 5.870020964360588e-06,
+      "loss": 1.0245,
+      "step": 28
+    },
+    {
+      "epoch": 0.006087321578505458,
+      "grad_norm": 0.7334048748016357,
+      "learning_rate": 6.079664570230608e-06,
+      "loss": 0.9172,
+      "step": 29
+    },
+    {
+      "epoch": 0.006297229219143577,
+      "grad_norm": 0.7490406632423401,
+      "learning_rate": 6.289308176100629e-06,
+      "loss": 0.9329,
+      "step": 30
+    },
+    {
+      "epoch": 0.006507136859781696,
+      "grad_norm": 0.8739404678344727,
+      "learning_rate": 6.49895178197065e-06,
+      "loss": 0.9584,
+      "step": 31
+    },
+    {
+      "epoch": 0.006717044500419815,
+      "grad_norm": 0.7686614990234375,
+      "learning_rate": 6.708595387840672e-06,
+      "loss": 0.9493,
+      "step": 32
+    },
+    {
+      "epoch": 0.0069269521410579345,
+      "grad_norm": 0.9184572100639343,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.0678,
+      "step": 33
+    },
+    {
+      "epoch": 0.007136859781696054,
+      "grad_norm": 0.8303453326225281,
+      "learning_rate": 7.127882599580712e-06,
+      "loss": 0.9834,
+      "step": 34
+    },
+    {
+      "epoch": 0.0073467674223341725,
+      "grad_norm": 0.6670796871185303,
+      "learning_rate": 7.337526205450735e-06,
+      "loss": 0.8434,
+      "step": 35
+    },
+    {
+      "epoch": 0.007556675062972292,
+      "grad_norm": 0.6518784761428833,
+      "learning_rate": 7.547169811320755e-06,
+      "loss": 0.8445,
+      "step": 36
+    },
+    {
+      "epoch": 0.007766582703610411,
+      "grad_norm": 0.8396414518356323,
+      "learning_rate": 7.756813417190776e-06,
+      "loss": 0.9229,
+      "step": 37
+    },
+    {
+      "epoch": 0.00797649034424853,
+      "grad_norm": 0.9019722938537598,
+      "learning_rate": 7.966457023060797e-06,
+      "loss": 0.9545,
+      "step": 38
+    },
+    {
+      "epoch": 0.00818639798488665,
+      "grad_norm": 0.6634522676467896,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 0.8279,
+      "step": 39
+    },
+    {
+      "epoch": 0.008396305625524769,
+      "grad_norm": 0.86155104637146,
+      "learning_rate": 8.38574423480084e-06,
+      "loss": 0.9326,
+      "step": 40
+    },
+    {
+      "epoch": 0.008606213266162888,
+      "grad_norm": 0.6448208689689636,
+      "learning_rate": 8.59538784067086e-06,
+      "loss": 0.8261,
+      "step": 41
+    },
+    {
+      "epoch": 0.008816120906801008,
+      "grad_norm": 0.7442598938941956,
+      "learning_rate": 8.80503144654088e-06,
+      "loss": 0.9352,
+      "step": 42
+    },
+    {
+      "epoch": 0.009026028547439127,
+      "grad_norm": 0.8854546546936035,
+      "learning_rate": 9.014675052410902e-06,
+      "loss": 1.017,
+      "step": 43
+    },
+    {
+      "epoch": 0.009235936188077247,
+      "grad_norm": 0.66485196352005,
+      "learning_rate": 9.224318658280923e-06,
+      "loss": 0.8096,
+      "step": 44
+    },
+    {
+      "epoch": 0.009445843828715366,
+      "grad_norm": 0.5998132824897766,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.7932,
+      "step": 45
+    },
+    {
+      "epoch": 0.009655751469353484,
+      "grad_norm": 0.6402536034584045,
+      "learning_rate": 9.643605870020965e-06,
+      "loss": 0.8146,
+      "step": 46
+    },
+    {
+      "epoch": 0.009865659109991603,
+      "grad_norm": 0.6665769219398499,
+      "learning_rate": 9.853249475890985e-06,
+      "loss": 0.8435,
+      "step": 47
+    },
+    {
+      "epoch": 0.010075566750629723,
+      "grad_norm": 0.6947203874588013,
+      "learning_rate": 1.0062893081761008e-05,
+      "loss": 0.8712,
+      "step": 48
+    },
+    {
+      "epoch": 0.010285474391267842,
+      "grad_norm": 0.6707759499549866,
+      "learning_rate": 1.0272536687631027e-05,
+      "loss": 0.8382,
+      "step": 49
+    },
+    {
+      "epoch": 0.010495382031905962,
+      "grad_norm": 0.6716253161430359,
+      "learning_rate": 1.0482180293501048e-05,
+      "loss": 0.8887,
+      "step": 50
+    },
+    {
+      "epoch": 0.010705289672544081,
+      "grad_norm": 0.6954374313354492,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 0.895,
+      "step": 51
+    },
+    {
+      "epoch": 0.0109151973131822,
+      "grad_norm": 0.5633372068405151,
+      "learning_rate": 1.0901467505241091e-05,
+      "loss": 0.7971,
+      "step": 52
+    },
+    {
+      "epoch": 0.01112510495382032,
+      "grad_norm": 0.5513401031494141,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.7848,
+      "step": 53
+    },
+    {
+      "epoch": 0.011335012594458438,
+      "grad_norm": 0.6735419034957886,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.8558,
+      "step": 54
+    },
+    {
+      "epoch": 0.011544920235096557,
+      "grad_norm": 0.5273690223693848,
+      "learning_rate": 1.1530398322851153e-05,
+      "loss": 0.7531,
+      "step": 55
+    },
+    {
+      "epoch": 0.011754827875734676,
+      "grad_norm": 0.612610936164856,
+      "learning_rate": 1.1740041928721176e-05,
+      "loss": 0.8009,
+      "step": 56
+    },
+    {
+      "epoch": 0.011964735516372796,
+      "grad_norm": 0.5250133275985718,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 0.7411,
+      "step": 57
+    },
+    {
+      "epoch": 0.012174643157010915,
+      "grad_norm": 0.6210602521896362,
+      "learning_rate": 1.2159329140461215e-05,
+      "loss": 0.814,
+      "step": 58
+    },
+    {
+      "epoch": 0.012384550797649035,
+      "grad_norm": 0.5767892003059387,
+      "learning_rate": 1.2368972746331238e-05,
+      "loss": 0.7512,
+      "step": 59
+    },
+    {
+      "epoch": 0.012594458438287154,
+      "grad_norm": 0.5112836360931396,
+      "learning_rate": 1.2578616352201259e-05,
+      "loss": 0.7405,
+      "step": 60
+    },
+    {
+      "epoch": 0.012804366078925274,
+      "grad_norm": 0.5214811563491821,
+      "learning_rate": 1.2788259958071281e-05,
+      "loss": 0.7152,
+      "step": 61
+    },
+    {
+      "epoch": 0.013014273719563391,
+      "grad_norm": 0.5820367932319641,
+      "learning_rate": 1.29979035639413e-05,
+      "loss": 0.7359,
+      "step": 62
+    },
+    {
+      "epoch": 0.01322418136020151,
+      "grad_norm": 0.5728296637535095,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.7096,
+      "step": 63
+    },
+    {
+      "epoch": 0.01343408900083963,
+      "grad_norm": 0.514997124671936,
+      "learning_rate": 1.3417190775681343e-05,
+      "loss": 0.6945,
+      "step": 64
+    },
+    {
+      "epoch": 0.01364399664147775,
+      "grad_norm": 0.5707572102546692,
+      "learning_rate": 1.3626834381551362e-05,
+      "loss": 0.7065,
+      "step": 65
+    },
+    {
+      "epoch": 0.013853904282115869,
+      "grad_norm": 0.5674712657928467,
+      "learning_rate": 1.3836477987421385e-05,
+      "loss": 0.6759,
+      "step": 66
+    },
+    {
+      "epoch": 0.014063811922753989,
+      "grad_norm": 0.5445975661277771,
+      "learning_rate": 1.4046121593291406e-05,
+      "loss": 0.6487,
+      "step": 67
+    },
+    {
+      "epoch": 0.014273719563392108,
+      "grad_norm": 0.5629355311393738,
+      "learning_rate": 1.4255765199161425e-05,
+      "loss": 0.6635,
+      "step": 68
+    },
+    {
+      "epoch": 0.014483627204030227,
+      "grad_norm": 0.47151610255241394,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.5976,
+      "step": 69
+    },
+    {
+      "epoch": 0.014693534844668345,
+      "grad_norm": 0.44633767008781433,
+      "learning_rate": 1.467505241090147e-05,
+      "loss": 0.5738,
+      "step": 70
+    },
+    {
+      "epoch": 0.014903442485306465,
+      "grad_norm": 0.48507657647132874,
+      "learning_rate": 1.488469601677149e-05,
+      "loss": 0.5859,
+      "step": 71
+    },
+    {
+      "epoch": 0.015113350125944584,
+      "grad_norm": 0.4147733151912689,
+      "learning_rate": 1.509433962264151e-05,
+      "loss": 0.5386,
+      "step": 72
+    },
+    {
+      "epoch": 0.015323257766582703,
+      "grad_norm": 0.635608434677124,
+      "learning_rate": 1.530398322851153e-05,
+      "loss": 0.6529,
+      "step": 73
+    },
+    {
+      "epoch": 0.015533165407220823,
+      "grad_norm": 0.5556919574737549,
+      "learning_rate": 1.5513626834381552e-05,
+      "loss": 0.5935,
+      "step": 74
+    },
+    {
+      "epoch": 0.015743073047858942,
+      "grad_norm": 0.5627433657646179,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 0.5677,
+      "step": 75
+    },
+    {
+      "epoch": 0.01595298068849706,
+      "grad_norm": 0.5727344155311584,
+      "learning_rate": 1.5932914046121594e-05,
+      "loss": 0.5701,
+      "step": 76
+    },
+    {
+      "epoch": 0.01616288832913518,
+      "grad_norm": 0.5192092657089233,
+      "learning_rate": 1.6142557651991616e-05,
+      "loss": 0.537,
+      "step": 77
+    },
+    {
+      "epoch": 0.0163727959697733,
+      "grad_norm": 0.6583610773086548,
+      "learning_rate": 1.6352201257861635e-05,
+      "loss": 0.5718,
+      "step": 78
+    },
+    {
+      "epoch": 0.01658270361041142,
+      "grad_norm": 0.4762994050979614,
+      "learning_rate": 1.6561844863731658e-05,
+      "loss": 0.5016,
+      "step": 79
+    },
+    {
+      "epoch": 0.016792611251049538,
+      "grad_norm": 0.7903013825416565,
+      "learning_rate": 1.677148846960168e-05,
+      "loss": 0.6123,
+      "step": 80
+    },
+    {
+      "epoch": 0.01700251889168766,
+      "grad_norm": 0.6027877330780029,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.5085,
+      "step": 81
+    },
+    {
+      "epoch": 0.017212426532325777,
+      "grad_norm": 0.6400225162506104,
+      "learning_rate": 1.719077568134172e-05,
+      "loss": 0.5325,
+      "step": 82
+    },
+    {
+      "epoch": 0.017422334172963894,
+      "grad_norm": 0.5193424224853516,
+      "learning_rate": 1.740041928721174e-05,
+      "loss": 0.476,
+      "step": 83
+    },
+    {
+      "epoch": 0.017632241813602016,
+      "grad_norm": 0.5318325757980347,
+      "learning_rate": 1.761006289308176e-05,
+      "loss": 0.4574,
+      "step": 84
+    },
+    {
+      "epoch": 0.017842149454240133,
+      "grad_norm": 0.5530166029930115,
+      "learning_rate": 1.7819706498951782e-05,
+      "loss": 0.4331,
+      "step": 85
+    },
+    {
+      "epoch": 0.018052057094878254,
+      "grad_norm": 0.5483909845352173,
+      "learning_rate": 1.8029350104821805e-05,
+      "loss": 0.4254,
+      "step": 86
+    },
+    {
+      "epoch": 0.018261964735516372,
+      "grad_norm": 0.4599871039390564,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.3828,
+      "step": 87
+    },
+    {
+      "epoch": 0.018471872376154493,
+      "grad_norm": 0.47489672899246216,
+      "learning_rate": 1.8448637316561846e-05,
+      "loss": 0.3989,
+      "step": 88
+    },
+    {
+      "epoch": 0.01868178001679261,
+      "grad_norm": 0.6532279253005981,
+      "learning_rate": 1.865828092243187e-05,
+      "loss": 0.3968,
+      "step": 89
+    },
+    {
+      "epoch": 0.018891687657430732,
+      "grad_norm": 0.707245945930481,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.4099,
+      "step": 90
+    },
+    {
+      "epoch": 0.01910159529806885,
+      "grad_norm": 0.4290582835674286,
+      "learning_rate": 1.9077568134171907e-05,
+      "loss": 0.3546,
+      "step": 91
+    },
+    {
+      "epoch": 0.019311502938706968,
+      "grad_norm": 0.4370197057723999,
+      "learning_rate": 1.928721174004193e-05,
+      "loss": 0.3519,
+      "step": 92
+    },
+    {
+      "epoch": 0.01952141057934509,
+      "grad_norm": 0.4862878918647766,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.3494,
+      "step": 93
+    },
+    {
+      "epoch": 0.019731318219983206,
+      "grad_norm": 0.514576256275177,
+      "learning_rate": 1.970649895178197e-05,
+      "loss": 0.326,
+      "step": 94
+    },
+    {
+      "epoch": 0.019941225860621328,
+      "grad_norm": 0.3785364031791687,
+      "learning_rate": 1.9916142557651993e-05,
+      "loss": 0.3196,
+      "step": 95
+    },
+    {
+      "epoch": 0.020151133501259445,
+      "grad_norm": 0.4459572732448578,
+      "learning_rate": 2.0125786163522016e-05,
+      "loss": 0.3245,
+      "step": 96
+    },
+    {
+      "epoch": 0.020361041141897566,
+      "grad_norm": 0.3634876310825348,
+      "learning_rate": 2.0335429769392035e-05,
+      "loss": 0.2999,
+      "step": 97
+    },
+    {
+      "epoch": 0.020570948782535684,
+      "grad_norm": 0.39789989590644836,
+      "learning_rate": 2.0545073375262054e-05,
+      "loss": 0.2908,
+      "step": 98
+    },
+    {
+      "epoch": 0.020780856423173802,
+      "grad_norm": 0.3628767430782318,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.278,
+      "step": 99
+    },
+    {
+      "epoch": 0.020990764063811923,
+      "grad_norm": 0.3945654332637787,
+      "learning_rate": 2.0964360587002095e-05,
+      "loss": 0.2902,
+      "step": 100
+    },
+    {
+      "epoch": 0.02120067170445004,
+      "grad_norm": 0.2995467185974121,
+      "learning_rate": 2.1174004192872118e-05,
+      "loss": 0.2748,
+      "step": 101
+    },
+    {
+      "epoch": 0.021410579345088162,
+      "grad_norm": 0.2776371240615845,
+      "learning_rate": 2.138364779874214e-05,
+      "loss": 0.2772,
+      "step": 102
+    },
+    {
+      "epoch": 0.02162048698572628,
+      "grad_norm": 0.292316734790802,
+      "learning_rate": 2.159329140461216e-05,
+      "loss": 0.2732,
+      "step": 103
+    },
+    {
+      "epoch": 0.0218303946263644,
+      "grad_norm": 0.26235565543174744,
+      "learning_rate": 2.1802935010482182e-05,
+      "loss": 0.2592,
+      "step": 104
+    },
+    {
+      "epoch": 0.02204030226700252,
+      "grad_norm": 0.2782291769981384,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.2586,
+      "step": 105
+    },
+    {
+      "epoch": 0.02225020990764064,
+      "grad_norm": 0.24781855940818787,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.2571,
+      "step": 106
+    },
+    {
+      "epoch": 0.022460117548278757,
+      "grad_norm": 0.3645837604999542,
+      "learning_rate": 2.2431865828092242e-05,
+      "loss": 0.247,
+      "step": 107
+    },
+    {
+      "epoch": 0.022670025188916875,
+      "grad_norm": 0.4096992313861847,
+      "learning_rate": 2.2641509433962265e-05,
+      "loss": 0.2648,
+      "step": 108
+    },
+    {
+      "epoch": 0.022879932829554996,
+      "grad_norm": 0.36801740527153015,
+      "learning_rate": 2.2851153039832284e-05,
+      "loss": 0.2677,
+      "step": 109
+    },
+    {
+      "epoch": 0.023089840470193114,
+      "grad_norm": 0.221563920378685,
+      "learning_rate": 2.3060796645702306e-05,
+      "loss": 0.2544,
+      "step": 110
+    },
+    {
+      "epoch": 0.023299748110831235,
+      "grad_norm": 0.17734010517597198,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.2458,
+      "step": 111
+    },
+    {
+      "epoch": 0.023509655751469353,
+      "grad_norm": 0.23815220594406128,
+      "learning_rate": 2.348008385744235e-05,
+      "loss": 0.2249,
+      "step": 112
+    },
+    {
+      "epoch": 0.023719563392107474,
+      "grad_norm": 0.24534103274345398,
+      "learning_rate": 2.368972746331237e-05,
+      "loss": 0.2283,
+      "step": 113
+    },
+    {
+      "epoch": 0.02392947103274559,
+      "grad_norm": 0.17044654488563538,
+      "learning_rate": 2.3899371069182393e-05,
+      "loss": 0.2543,
+      "step": 114
+    },
+    {
+      "epoch": 0.02413937867338371,
+      "grad_norm": 0.22411420941352844,
+      "learning_rate": 2.4109014675052412e-05,
+      "loss": 0.2537,
+      "step": 115
+    },
+    {
+      "epoch": 0.02434928631402183,
+      "grad_norm": 0.17107880115509033,
+      "learning_rate": 2.431865828092243e-05,
+      "loss": 0.2477,
+      "step": 116
+    },
+    {
+      "epoch": 0.02455919395465995,
+      "grad_norm": 0.19663883745670319,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 0.2503,
+      "step": 117
+    },
+    {
+      "epoch": 0.02476910159529807,
+      "grad_norm": 0.2755718231201172,
+      "learning_rate": 2.4737945492662476e-05,
+      "loss": 0.201,
+      "step": 118
+    },
+    {
+      "epoch": 0.024979009235936187,
+      "grad_norm": 0.16535572707653046,
+      "learning_rate": 2.4947589098532495e-05,
+      "loss": 0.2364,
+      "step": 119
+    },
+    {
+      "epoch": 0.02518891687657431,
+      "grad_norm": 0.15798090398311615,
+      "learning_rate": 2.5157232704402517e-05,
+      "loss": 0.214,
+      "step": 120
+    },
+    {
+      "epoch": 0.025398824517212426,
+      "grad_norm": 0.1900860220193863,
+      "learning_rate": 2.5366876310272536e-05,
+      "loss": 0.2182,
+      "step": 121
+    },
+    {
+      "epoch": 0.025608732157850547,
+      "grad_norm": 0.18855144083499908,
+      "learning_rate": 2.5576519916142562e-05,
+      "loss": 0.2123,
+      "step": 122
+    },
+    {
+      "epoch": 0.025818639798488665,
+      "grad_norm": 0.1780049353837967,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.2197,
+      "step": 123
+    },
+    {
+      "epoch": 0.026028547439126783,
+      "grad_norm": 0.27040061354637146,
+      "learning_rate": 2.59958071278826e-05,
+      "loss": 0.2395,
+      "step": 124
+    },
+    {
+      "epoch": 0.026238455079764904,
+      "grad_norm": 0.1983417570590973,
+      "learning_rate": 2.6205450733752623e-05,
+      "loss": 0.2197,
+      "step": 125
+    },
+    {
+      "epoch": 0.02644836272040302,
+      "grad_norm": 0.1723383665084839,
+      "learning_rate": 2.641509433962264e-05,
+      "loss": 0.2183,
+      "step": 126
+    },
+    {
+      "epoch": 0.026658270361041143,
+      "grad_norm": 0.24477605521678925,
+      "learning_rate": 2.662473794549266e-05,
+      "loss": 0.2293,
+      "step": 127
+    },
+    {
+      "epoch": 0.02686817800167926,
+      "grad_norm": 0.15110079944133759,
+      "learning_rate": 2.6834381551362687e-05,
+      "loss": 0.2245,
+      "step": 128
+    },
+    {
+      "epoch": 0.02707808564231738,
+      "grad_norm": 0.22366492450237274,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.2031,
+      "step": 129
+    },
+    {
+      "epoch": 0.0272879932829555,
+      "grad_norm": 0.32891571521759033,
+      "learning_rate": 2.7253668763102725e-05,
+      "loss": 0.1821,
+      "step": 130
+    },
+    {
+      "epoch": 0.02749790092359362,
+      "grad_norm": 0.20027081668376923,
+      "learning_rate": 2.746331236897275e-05,
+      "loss": 0.223,
+      "step": 131
+    },
+    {
+      "epoch": 0.027707808564231738,
+      "grad_norm": 0.2269366830587387,
+      "learning_rate": 2.767295597484277e-05,
+      "loss": 0.2246,
+      "step": 132
+    },
+    {
+      "epoch": 0.027917716204869856,
+      "grad_norm": 0.1355280727148056,
+      "learning_rate": 2.788259958071279e-05,
+      "loss": 0.2183,
+      "step": 133
+    },
+    {
+      "epoch": 0.028127623845507977,
+      "grad_norm": 0.29291626811027527,
+      "learning_rate": 2.809224318658281e-05,
+      "loss": 0.2049,
+      "step": 134
+    },
+    {
+      "epoch": 0.028337531486146095,
+      "grad_norm": 0.1778186410665512,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.2095,
+      "step": 135
+    },
+    {
+      "epoch": 0.028547439126784216,
+      "grad_norm": 0.23263931274414062,
+      "learning_rate": 2.851153039832285e-05,
+      "loss": 0.2372,
+      "step": 136
+    },
+    {
+      "epoch": 0.028757346767422334,
+      "grad_norm": 0.2121749222278595,
+      "learning_rate": 2.8721174004192875e-05,
+      "loss": 0.2107,
+      "step": 137
+    },
+    {
+      "epoch": 0.028967254408060455,
+      "grad_norm": 0.19954991340637207,
+      "learning_rate": 2.8930817610062894e-05,
+      "loss": 0.2181,
+      "step": 138
+    },
+    {
+      "epoch": 0.029177162048698572,
+      "grad_norm": 0.15431921184062958,
+      "learning_rate": 2.9140461215932913e-05,
+      "loss": 0.2198,
+      "step": 139
+    },
+    {
+      "epoch": 0.02938706968933669,
+      "grad_norm": 0.17603729665279388,
+      "learning_rate": 2.935010482180294e-05,
+      "loss": 0.2339,
+      "step": 140
+    },
+    {
+      "epoch": 0.02959697732997481,
+      "grad_norm": 0.1604471355676651,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.2064,
+      "step": 141
+    },
+    {
+      "epoch": 0.02980688497061293,
+      "grad_norm": 0.17169184982776642,
+      "learning_rate": 2.976939203354298e-05,
+      "loss": 0.1987,
+      "step": 142
+    },
+    {
+      "epoch": 0.03001679261125105,
+      "grad_norm": 0.1285230815410614,
+      "learning_rate": 2.9979035639413e-05,
+      "loss": 0.2342,
+      "step": 143
+    },
+    {
+      "epoch": 0.030226700251889168,
+      "grad_norm": 0.1755395084619522,
+      "learning_rate": 3.018867924528302e-05,
+      "loss": 0.222,
+      "step": 144
+    },
+    {
+      "epoch": 0.03043660789252729,
+      "grad_norm": 0.15474987030029297,
+      "learning_rate": 3.0398322851153044e-05,
+      "loss": 0.2159,
+      "step": 145
+    },
+    {
+      "epoch": 0.030646515533165407,
+      "grad_norm": 0.12986472249031067,
+      "learning_rate": 3.060796645702306e-05,
+      "loss": 0.2124,
+      "step": 146
+    },
+    {
+      "epoch": 0.030856423173803528,
+      "grad_norm": 0.1458188146352768,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.214,
+      "step": 147
+    },
+    {
+      "epoch": 0.031066330814441646,
+      "grad_norm": 0.1323792040348053,
+      "learning_rate": 3.1027253668763105e-05,
+      "loss": 0.2153,
+      "step": 148
+    },
+    {
+      "epoch": 0.03127623845507976,
+      "grad_norm": 0.16542711853981018,
+      "learning_rate": 3.1236897274633124e-05,
+      "loss": 0.2154,
+      "step": 149
+    },
+    {
+      "epoch": 0.031486146095717885,
+      "grad_norm": 0.17730407416820526,
+      "learning_rate": 3.144654088050314e-05,
+      "loss": 0.2202,
+      "step": 150
+    },
+    {
+      "epoch": 0.031696053736356006,
+      "grad_norm": 0.15039502084255219,
+      "learning_rate": 3.165618448637317e-05,
+      "loss": 0.2056,
+      "step": 151
+    },
+    {
+      "epoch": 0.03190596137699412,
+      "grad_norm": 0.20309150218963623,
+      "learning_rate": 3.186582809224319e-05,
+      "loss": 0.2175,
+      "step": 152
+    },
+    {
+      "epoch": 0.03211586901763224,
+      "grad_norm": 0.16652604937553406,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.2076,
+      "step": 153
+    },
+    {
+      "epoch": 0.03232577665827036,
+      "grad_norm": 0.14530467987060547,
+      "learning_rate": 3.228511530398323e-05,
+      "loss": 0.2167,
+      "step": 154
+    },
+    {
+      "epoch": 0.032535684298908484,
+      "grad_norm": 0.13003528118133545,
+      "learning_rate": 3.249475890985325e-05,
+      "loss": 0.2089,
+      "step": 155
+    },
+    {
+      "epoch": 0.0327455919395466,
+      "grad_norm": 0.16985855996608734,
+      "learning_rate": 3.270440251572327e-05,
+      "loss": 0.1994,
+      "step": 156
+    },
+    {
+      "epoch": 0.03295549958018472,
+      "grad_norm": 0.18479777872562408,
+      "learning_rate": 3.29140461215933e-05,
+      "loss": 0.2127,
+      "step": 157
+    },
+    {
+      "epoch": 0.03316540722082284,
+      "grad_norm": 0.1541491150856018,
+      "learning_rate": 3.3123689727463316e-05,
+      "loss": 0.2208,
+      "step": 158
+    },
+    {
+      "epoch": 0.033375314861460954,
+      "grad_norm": 0.13511165976524353,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.218,
+      "step": 159
+    },
+    {
+      "epoch": 0.033585222502099076,
+      "grad_norm": 0.1392865628004074,
+      "learning_rate": 3.354297693920336e-05,
+      "loss": 0.226,
+      "step": 160
+    },
+    {
+      "epoch": 0.0337951301427372,
+      "grad_norm": 0.1614847183227539,
+      "learning_rate": 3.375262054507338e-05,
+      "loss": 0.2072,
+      "step": 161
+    },
+    {
+      "epoch": 0.03400503778337532,
+      "grad_norm": 0.12186679244041443,
+      "learning_rate": 3.39622641509434e-05,
+      "loss": 0.2193,
+      "step": 162
+    },
+    {
+      "epoch": 0.03421494542401343,
+      "grad_norm": 0.1777278333902359,
+      "learning_rate": 3.417190775681342e-05,
+      "loss": 0.202,
+      "step": 163
+    },
+    {
+      "epoch": 0.03442485306465155,
+      "grad_norm": 0.13323499262332916,
+      "learning_rate": 3.438155136268344e-05,
+      "loss": 0.1983,
+      "step": 164
+    },
+    {
+      "epoch": 0.034634760705289674,
+      "grad_norm": 0.22301942110061646,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.2252,
+      "step": 165
+    },
+    {
+      "epoch": 0.03484466834592779,
+      "grad_norm": 0.1538372039794922,
+      "learning_rate": 3.480083857442348e-05,
+      "loss": 0.2071,
+      "step": 166
+    },
+    {
+      "epoch": 0.03505457598656591,
+      "grad_norm": 0.128251314163208,
+      "learning_rate": 3.50104821802935e-05,
+      "loss": 0.214,
+      "step": 167
+    },
+    {
+      "epoch": 0.03526448362720403,
+      "grad_norm": 0.26556313037872314,
+      "learning_rate": 3.522012578616352e-05,
+      "loss": 0.1863,
+      "step": 168
+    },
+    {
+      "epoch": 0.03547439126784215,
+      "grad_norm": 0.14938302338123322,
+      "learning_rate": 3.5429769392033546e-05,
+      "loss": 0.2152,
+      "step": 169
+    },
+    {
+      "epoch": 0.035684298908480266,
+      "grad_norm": 0.19645771384239197,
+      "learning_rate": 3.5639412997903565e-05,
+      "loss": 0.1977,
+      "step": 170
+    },
+    {
+      "epoch": 0.03589420654911839,
+      "grad_norm": 0.1725340038537979,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.1981,
+      "step": 171
+    },
+    {
+      "epoch": 0.03610411418975651,
+      "grad_norm": 0.16048069298267365,
+      "learning_rate": 3.605870020964361e-05,
+      "loss": 0.1983,
+      "step": 172
+    },
+    {
+      "epoch": 0.03631402183039462,
+      "grad_norm": 0.21855585277080536,
+      "learning_rate": 3.626834381551363e-05,
+      "loss": 0.2321,
+      "step": 173
+    },
+    {
+      "epoch": 0.036523929471032744,
+      "grad_norm": 0.1136787012219429,
+      "learning_rate": 3.647798742138365e-05,
+      "loss": 0.1973,
+      "step": 174
+    },
+    {
+      "epoch": 0.036733837111670865,
+      "grad_norm": 0.15145623683929443,
+      "learning_rate": 3.6687631027253674e-05,
+      "loss": 0.1947,
+      "step": 175
+    },
+    {
+      "epoch": 0.036943744752308987,
+      "grad_norm": 0.21631890535354614,
+      "learning_rate": 3.689727463312369e-05,
+      "loss": 0.2151,
+      "step": 176
+    },
+    {
+      "epoch": 0.0371536523929471,
+      "grad_norm": 0.2623152434825897,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.1814,
+      "step": 177
+    },
+    {
+      "epoch": 0.03736356003358522,
+      "grad_norm": 0.1753605455160141,
+      "learning_rate": 3.731656184486374e-05,
+      "loss": 0.203,
+      "step": 178
+    },
+    {
+      "epoch": 0.03757346767422334,
+      "grad_norm": 0.10878176242113113,
+      "learning_rate": 3.752620545073376e-05,
+      "loss": 0.2052,
+      "step": 179
+    },
+    {
+      "epoch": 0.037783375314861464,
+      "grad_norm": 0.13699688017368317,
+      "learning_rate": 3.7735849056603776e-05,
+      "loss": 0.2004,
+      "step": 180
+    },
+    {
+      "epoch": 0.03799328295549958,
+      "grad_norm": 0.14288806915283203,
+      "learning_rate": 3.7945492662473795e-05,
+      "loss": 0.1745,
+      "step": 181
+    },
+    {
+      "epoch": 0.0382031905961377,
+      "grad_norm": 0.12457548081874847,
+      "learning_rate": 3.8155136268343814e-05,
+      "loss": 0.194,
+      "step": 182
+    },
+    {
+      "epoch": 0.03841309823677582,
+      "grad_norm": 0.167145237326622,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.2027,
+      "step": 183
+    },
+    {
+      "epoch": 0.038623005877413935,
+      "grad_norm": 0.12857979536056519,
+      "learning_rate": 3.857442348008386e-05,
+      "loss": 0.2126,
+      "step": 184
+    },
+    {
+      "epoch": 0.038832913518052056,
+      "grad_norm": 0.16190126538276672,
+      "learning_rate": 3.878406708595388e-05,
+      "loss": 0.2037,
+      "step": 185
+    },
+    {
+      "epoch": 0.03904282115869018,
+      "grad_norm": 0.168744757771492,
+      "learning_rate": 3.8993710691823904e-05,
+      "loss": 0.2108,
+      "step": 186
+    },
+    {
+      "epoch": 0.0392527287993283,
+      "grad_norm": 0.1676539033651352,
+      "learning_rate": 3.920335429769392e-05,
+      "loss": 0.216,
+      "step": 187
+    },
+    {
+      "epoch": 0.03946263643996641,
+      "grad_norm": 0.13556820154190063,
+      "learning_rate": 3.941299790356394e-05,
+      "loss": 0.2085,
+      "step": 188
+    },
+    {
+      "epoch": 0.039672544080604534,
+      "grad_norm": 0.1797979772090912,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.1933,
+      "step": 189
+    },
+    {
+      "epoch": 0.039882451721242655,
+      "grad_norm": 0.20826327800750732,
+      "learning_rate": 3.983228511530399e-05,
+      "loss": 0.2214,
+      "step": 190
+    },
+    {
+      "epoch": 0.04009235936188077,
+      "grad_norm": 0.19972363114356995,
+      "learning_rate": 4.0041928721174006e-05,
+      "loss": 0.1941,
+      "step": 191
+    },
+    {
+      "epoch": 0.04030226700251889,
+      "grad_norm": 0.149556502699852,
+      "learning_rate": 4.025157232704403e-05,
+      "loss": 0.1956,
+      "step": 192
+    },
+    {
+      "epoch": 0.04051217464315701,
+      "grad_norm": 0.22496013343334198,
+      "learning_rate": 4.046121593291405e-05,
+      "loss": 0.1973,
+      "step": 193
+    },
+    {
+      "epoch": 0.04072208228379513,
+      "grad_norm": 0.16132576763629913,
+      "learning_rate": 4.067085953878407e-05,
+      "loss": 0.203,
+      "step": 194
+    },
+    {
+      "epoch": 0.04093198992443325,
+      "grad_norm": 0.17156128585338593,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.1942,
+      "step": 195
+    },
+    {
+      "epoch": 0.04114189756507137,
+      "grad_norm": 0.14846180379390717,
+      "learning_rate": 4.109014675052411e-05,
+      "loss": 0.2005,
+      "step": 196
+    },
+    {
+      "epoch": 0.04135180520570949,
+      "grad_norm": 0.20252752304077148,
+      "learning_rate": 4.129979035639413e-05,
+      "loss": 0.186,
+      "step": 197
+    },
+    {
+      "epoch": 0.041561712846347604,
+      "grad_norm": 0.16286462545394897,
+      "learning_rate": 4.150943396226415e-05,
+      "loss": 0.1997,
+      "step": 198
+    },
+    {
+      "epoch": 0.041771620486985725,
+      "grad_norm": 0.13008786737918854,
+      "learning_rate": 4.171907756813417e-05,
+      "loss": 0.2052,
+      "step": 199
+    },
+    {
+      "epoch": 0.041981528127623846,
+      "grad_norm": 0.13853180408477783,
+      "learning_rate": 4.192872117400419e-05,
+      "loss": 0.1907,
+      "step": 200
+    },
+    {
+      "epoch": 0.04219143576826197,
+      "grad_norm": 0.20382314920425415,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.1931,
+      "step": 201
+    },
+    {
+      "epoch": 0.04240134340890008,
+      "grad_norm": 0.23206844925880432,
+      "learning_rate": 4.2348008385744236e-05,
+      "loss": 0.1849,
+      "step": 202
+    },
+    {
+      "epoch": 0.0426112510495382,
+      "grad_norm": 0.2456827312707901,
+      "learning_rate": 4.2557651991614255e-05,
+      "loss": 0.2088,
+      "step": 203
+    },
+    {
+      "epoch": 0.042821158690176324,
+      "grad_norm": 0.15247821807861328,
+      "learning_rate": 4.276729559748428e-05,
+      "loss": 0.1944,
+      "step": 204
+    },
+    {
+      "epoch": 0.043031066330814445,
+      "grad_norm": 0.174981027841568,
+      "learning_rate": 4.29769392033543e-05,
+      "loss": 0.2161,
+      "step": 205
+    },
+    {
+      "epoch": 0.04324097397145256,
+      "grad_norm": 0.20193031430244446,
+      "learning_rate": 4.318658280922432e-05,
+      "loss": 0.2084,
+      "step": 206
+    },
+    {
+      "epoch": 0.04345088161209068,
+      "grad_norm": 0.20125791430473328,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.189,
+      "step": 207
+    },
+    {
+      "epoch": 0.0436607892527288,
+      "grad_norm": 0.16958673298358917,
+      "learning_rate": 4.3605870020964364e-05,
+      "loss": 0.1952,
+      "step": 208
+    },
+    {
+      "epoch": 0.043870696893366916,
+      "grad_norm": 0.20714177191257477,
+      "learning_rate": 4.381551362683438e-05,
+      "loss": 0.1832,
+      "step": 209
+    },
+    {
+      "epoch": 0.04408060453400504,
+      "grad_norm": 0.14229562878608704,
+      "learning_rate": 4.402515723270441e-05,
+      "loss": 0.1806,
+      "step": 210
+    },
+    {
+      "epoch": 0.04429051217464316,
+      "grad_norm": 0.1985626220703125,
+      "learning_rate": 4.423480083857443e-05,
+      "loss": 0.1973,
+      "step": 211
+    },
+    {
+      "epoch": 0.04450041981528128,
+      "grad_norm": 0.13714846968650818,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.2018,
+      "step": 212
+    },
+    {
+      "epoch": 0.044710327455919394,
+      "grad_norm": 0.25591611862182617,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.2001,
+      "step": 213
+    },
+    {
+      "epoch": 0.044920235096557515,
+      "grad_norm": 0.29017260670661926,
+      "learning_rate": 4.4863731656184485e-05,
+      "loss": 0.2278,
+      "step": 214
+    },
+    {
+      "epoch": 0.045130142737195636,
+      "grad_norm": 0.12594066560268402,
+      "learning_rate": 4.5073375262054504e-05,
+      "loss": 0.2023,
+      "step": 215
+    },
+    {
+      "epoch": 0.04534005037783375,
+      "grad_norm": 0.1602821797132492,
+      "learning_rate": 4.528301886792453e-05,
+      "loss": 0.2101,
+      "step": 216
+    },
+    {
+      "epoch": 0.04554995801847187,
+      "grad_norm": 0.36664336919784546,
+      "learning_rate": 4.549266247379455e-05,
+      "loss": 0.1665,
+      "step": 217
+    },
+    {
+      "epoch": 0.04575986565910999,
+      "grad_norm": 0.20665952563285828,
+      "learning_rate": 4.570230607966457e-05,
+      "loss": 0.1965,
+      "step": 218
+    },
+    {
+      "epoch": 0.045969773299748114,
+      "grad_norm": 0.15664128959178925,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.2104,
+      "step": 219
+    },
+    {
+      "epoch": 0.04617968094038623,
+      "grad_norm": 0.14733830094337463,
+      "learning_rate": 4.612159329140461e-05,
+      "loss": 0.1955,
+      "step": 220
+    },
+    {
+      "epoch": 0.04638958858102435,
+      "grad_norm": 0.19135522842407227,
+      "learning_rate": 4.633123689727464e-05,
+      "loss": 0.1947,
+      "step": 221
+    },
+    {
+      "epoch": 0.04659949622166247,
+      "grad_norm": 0.24508413672447205,
+      "learning_rate": 4.654088050314466e-05,
+      "loss": 0.2005,
+      "step": 222
+    },
+    {
+      "epoch": 0.046809403862300585,
+      "grad_norm": 0.16794531047344208,
+      "learning_rate": 4.6750524109014677e-05,
+      "loss": 0.1958,
+      "step": 223
+    },
+    {
+      "epoch": 0.047019311502938706,
+      "grad_norm": 0.1870536506175995,
+      "learning_rate": 4.69601677148847e-05,
+      "loss": 0.1868,
+      "step": 224
+    },
+    {
+      "epoch": 0.04722921914357683,
+      "grad_norm": 0.16214439272880554,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.2003,
+      "step": 225
+    },
+    {
+      "epoch": 0.04743912678421495,
+      "grad_norm": 0.24978841841220856,
+      "learning_rate": 4.737945492662474e-05,
+      "loss": 0.1909,
+      "step": 226
+    },
+    {
+      "epoch": 0.04764903442485306,
+      "grad_norm": 0.163265198469162,
+      "learning_rate": 4.7589098532494766e-05,
+      "loss": 0.2056,
+      "step": 227
+    },
+    {
+      "epoch": 0.04785894206549118,
+      "grad_norm": 0.24885287880897522,
+      "learning_rate": 4.7798742138364785e-05,
+      "loss": 0.2035,
+      "step": 228
+    },
+    {
+      "epoch": 0.048068849706129305,
+      "grad_norm": 0.1393681764602661,
+      "learning_rate": 4.8008385744234804e-05,
+      "loss": 0.1976,
+      "step": 229
+    },
+    {
+      "epoch": 0.04827875734676742,
+      "grad_norm": 0.17042241990566254,
+      "learning_rate": 4.8218029350104823e-05,
+      "loss": 0.1824,
+      "step": 230
+    },
+    {
+      "epoch": 0.04848866498740554,
+      "grad_norm": 0.1625502109527588,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.1904,
+      "step": 231
+    },
+    {
+      "epoch": 0.04869857262804366,
+      "grad_norm": 0.14994169771671295,
+      "learning_rate": 4.863731656184486e-05,
+      "loss": 0.1926,
+      "step": 232
+    },
+    {
+      "epoch": 0.04890848026868178,
+      "grad_norm": 0.15602821111679077,
+      "learning_rate": 4.884696016771489e-05,
+      "loss": 0.1828,
+      "step": 233
+    },
+    {
+      "epoch": 0.0491183879093199,
+      "grad_norm": 0.13405688107013702,
+      "learning_rate": 4.9056603773584906e-05,
+      "loss": 0.1989,
+      "step": 234
+    },
+    {
+      "epoch": 0.04932829554995802,
+      "grad_norm": 0.18559689819812775,
+      "learning_rate": 4.9266247379454926e-05,
+      "loss": 0.2131,
+      "step": 235
+    },
+    {
+      "epoch": 0.04953820319059614,
+      "grad_norm": 0.1557319462299347,
+      "learning_rate": 4.947589098532495e-05,
+      "loss": 0.19,
+      "step": 236
+    },
+    {
+      "epoch": 0.04974811083123426,
+      "grad_norm": 0.2162303477525711,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.196,
+      "step": 237
+    },
+    {
+      "epoch": 0.049958018471872374,
+      "grad_norm": 0.17403477430343628,
+      "learning_rate": 4.989517819706499e-05,
+      "loss": 0.1997,
+      "step": 238
+    },
+    {
+      "epoch": 0.050167926112510496,
+      "grad_norm": 0.11738390475511551,
+      "learning_rate": 5.010482180293501e-05,
+      "loss": 0.1955,
+      "step": 239
+    },
+    {
+      "epoch": 0.05037783375314862,
+      "grad_norm": 0.15942999720573425,
+      "learning_rate": 5.0314465408805034e-05,
+      "loss": 0.1899,
+      "step": 240
+    },
+    {
+      "epoch": 0.05058774139378673,
+      "grad_norm": 0.14695511758327484,
+      "learning_rate": 5.052410901467506e-05,
+      "loss": 0.2032,
+      "step": 241
+    },
+    {
+      "epoch": 0.05079764903442485,
+      "grad_norm": 0.16291062533855438,
+      "learning_rate": 5.073375262054507e-05,
+      "loss": 0.1929,
+      "step": 242
+    },
+    {
+      "epoch": 0.05100755667506297,
+      "grad_norm": 0.18514905869960785,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.1899,
+      "step": 243
+    },
+    {
+      "epoch": 0.051217464315701094,
+      "grad_norm": 0.2196233868598938,
+      "learning_rate": 5.1153039832285124e-05,
+      "loss": 0.1741,
+      "step": 244
+    },
+    {
+      "epoch": 0.05142737195633921,
+      "grad_norm": 0.19183433055877686,
+      "learning_rate": 5.1362683438155136e-05,
+      "loss": 0.1985,
+      "step": 245
+    },
+    {
+      "epoch": 0.05163727959697733,
+      "grad_norm": 0.1604142189025879,
+      "learning_rate": 5.157232704402516e-05,
+      "loss": 0.1865,
+      "step": 246
+    },
+    {
+      "epoch": 0.05184718723761545,
+      "grad_norm": 0.18311725556850433,
+      "learning_rate": 5.178197064989518e-05,
+      "loss": 0.2005,
+      "step": 247
+    },
+    {
+      "epoch": 0.052057094878253565,
+      "grad_norm": 0.21732251346111298,
+      "learning_rate": 5.19916142557652e-05,
+      "loss": 0.2032,
+      "step": 248
+    },
+    {
+      "epoch": 0.052267002518891686,
+      "grad_norm": 0.2600694000720978,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.1838,
+      "step": 249
+    },
+    {
+      "epoch": 0.05247691015952981,
+      "grad_norm": 0.18634290993213654,
+      "learning_rate": 5.2410901467505245e-05,
+      "loss": 0.1809,
+      "step": 250
+    },
+    {
+      "epoch": 0.05268681780016793,
+      "grad_norm": 0.14735780656337738,
+      "learning_rate": 5.262054507337526e-05,
+      "loss": 0.1981,
+      "step": 251
+    },
+    {
+      "epoch": 0.05289672544080604,
+      "grad_norm": 0.16118381917476654,
+      "learning_rate": 5.283018867924528e-05,
+      "loss": 0.1836,
+      "step": 252
+    },
+    {
+      "epoch": 0.053106633081444164,
+      "grad_norm": 0.17707999050617218,
+      "learning_rate": 5.303983228511531e-05,
+      "loss": 0.2044,
+      "step": 253
+    },
+    {
+      "epoch": 0.053316540722082285,
+      "grad_norm": 0.25897523760795593,
+      "learning_rate": 5.324947589098532e-05,
+      "loss": 0.1753,
+      "step": 254
+    },
+    {
+      "epoch": 0.0535264483627204,
+      "grad_norm": 0.1371389776468277,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.1839,
+      "step": 255
+    },
+    {
+      "epoch": 0.05373635600335852,
+      "grad_norm": 0.21849682927131653,
+      "learning_rate": 5.366876310272537e-05,
+      "loss": 0.1938,
+      "step": 256
+    },
+    {
+      "epoch": 0.05394626364399664,
+      "grad_norm": 0.16861748695373535,
+      "learning_rate": 5.3878406708595385e-05,
+      "loss": 0.1751,
+      "step": 257
+    },
+    {
+      "epoch": 0.05415617128463476,
+      "grad_norm": 0.19400931894779205,
+      "learning_rate": 5.408805031446541e-05,
+      "loss": 0.2117,
+      "step": 258
+    },
+    {
+      "epoch": 0.05436607892527288,
+      "grad_norm": 0.18074113130569458,
+      "learning_rate": 5.429769392033544e-05,
+      "loss": 0.1949,
+      "step": 259
+    },
+    {
+      "epoch": 0.054575986565911,
+      "grad_norm": 0.17707990109920502,
+      "learning_rate": 5.450733752620545e-05,
+      "loss": 0.199,
+      "step": 260
+    },
+    {
+      "epoch": 0.05478589420654912,
+      "grad_norm": 0.16568966209888458,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.2027,
+      "step": 261
+    },
+    {
+      "epoch": 0.05499580184718724,
+      "grad_norm": 0.24486149847507477,
+      "learning_rate": 5.49266247379455e-05,
+      "loss": 0.2042,
+      "step": 262
+    },
+    {
+      "epoch": 0.055205709487825355,
+      "grad_norm": 0.20431551337242126,
+      "learning_rate": 5.513626834381551e-05,
+      "loss": 0.2013,
+      "step": 263
+    },
+    {
+      "epoch": 0.055415617128463476,
+      "grad_norm": 0.21347559988498688,
+      "learning_rate": 5.534591194968554e-05,
+      "loss": 0.1687,
+      "step": 264
+    },
+    {
+      "epoch": 0.0556255247691016,
+      "grad_norm": 0.22354485094547272,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.1819,
+      "step": 265
+    },
+    {
+      "epoch": 0.05583543240973971,
+      "grad_norm": 0.2891826033592224,
+      "learning_rate": 5.576519916142558e-05,
+      "loss": 0.1927,
+      "step": 266
+    },
+    {
+      "epoch": 0.05604534005037783,
+      "grad_norm": 0.2465565800666809,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.1852,
+      "step": 267
+    },
+    {
+      "epoch": 0.056255247691015954,
+      "grad_norm": 0.1743604689836502,
+      "learning_rate": 5.618448637316562e-05,
+      "loss": 0.1722,
+      "step": 268
+    },
+    {
+      "epoch": 0.056465155331654075,
+      "grad_norm": 0.2461417019367218,
+      "learning_rate": 5.6394129979035634e-05,
+      "loss": 0.1816,
+      "step": 269
+    },
+    {
+      "epoch": 0.05667506297229219,
+      "grad_norm": 0.21032604575157166,
+      "learning_rate": 5.660377358490566e-05,
+      "loss": 0.182,
+      "step": 270
+    },
+    {
+      "epoch": 0.05688497061293031,
+      "grad_norm": 0.2683754861354828,
+      "learning_rate": 5.6813417190775686e-05,
+      "loss": 0.1718,
+      "step": 271
+    },
+    {
+      "epoch": 0.05709487825356843,
+      "grad_norm": 0.18455228209495544,
+      "learning_rate": 5.70230607966457e-05,
+      "loss": 0.1954,
+      "step": 272
+    },
+    {
+      "epoch": 0.057304785894206546,
+      "grad_norm": 0.22255247831344604,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.1922,
+      "step": 273
+    },
+    {
+      "epoch": 0.05751469353484467,
+      "grad_norm": 0.22789119184017181,
+      "learning_rate": 5.744234800838575e-05,
+      "loss": 0.1882,
+      "step": 274
+    },
+    {
+      "epoch": 0.05772460117548279,
+      "grad_norm": 0.20874802768230438,
+      "learning_rate": 5.765199161425576e-05,
+      "loss": 0.1859,
+      "step": 275
+    },
+    {
+      "epoch": 0.05793450881612091,
+      "grad_norm": 0.17554089426994324,
+      "learning_rate": 5.786163522012579e-05,
+      "loss": 0.1793,
+      "step": 276
+    },
+    {
+      "epoch": 0.058144416456759024,
+      "grad_norm": 0.2011173665523529,
+      "learning_rate": 5.8071278825995814e-05,
+      "loss": 0.2009,
+      "step": 277
+    },
+    {
+      "epoch": 0.058354324097397145,
+      "grad_norm": 0.28261420130729675,
+      "learning_rate": 5.8280922431865826e-05,
+      "loss": 0.2002,
+      "step": 278
+    },
+    {
+      "epoch": 0.058564231738035266,
+      "grad_norm": 0.2356766015291214,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.2095,
+      "step": 279
+    },
+    {
+      "epoch": 0.05877413937867338,
+      "grad_norm": 0.15072722733020782,
+      "learning_rate": 5.870020964360588e-05,
+      "loss": 0.1781,
+      "step": 280
+    },
+    {
+      "epoch": 0.0589840470193115,
+      "grad_norm": 0.1446981132030487,
+      "learning_rate": 5.89098532494759e-05,
+      "loss": 0.1813,
+      "step": 281
+    },
+    {
+      "epoch": 0.05919395465994962,
+      "grad_norm": 0.16516901552677155,
+      "learning_rate": 5.9119496855345916e-05,
+      "loss": 0.1946,
+      "step": 282
+    },
+    {
+      "epoch": 0.059403862300587744,
+      "grad_norm": 0.16732774674892426,
+      "learning_rate": 5.9329140461215935e-05,
+      "loss": 0.1813,
+      "step": 283
+    },
+    {
+      "epoch": 0.05961376994122586,
+      "grad_norm": 0.2000836730003357,
+      "learning_rate": 5.953878406708596e-05,
+      "loss": 0.1918,
+      "step": 284
+    },
+    {
+      "epoch": 0.05982367758186398,
+      "grad_norm": 0.15576116740703583,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.1814,
+      "step": 285
+    },
+    {
+      "epoch": 0.0600335852225021,
+      "grad_norm": 0.18421867489814758,
+      "learning_rate": 5.9958071278826e-05,
+      "loss": 0.1745,
+      "step": 286
+    },
+    {
+      "epoch": 0.06024349286314022,
+      "grad_norm": 0.2663988471031189,
+      "learning_rate": 6.0167714884696025e-05,
+      "loss": 0.1889,
+      "step": 287
+    },
+    {
+      "epoch": 0.060453400503778336,
+      "grad_norm": 0.18096649646759033,
+      "learning_rate": 6.037735849056604e-05,
+      "loss": 0.191,
+      "step": 288
+    },
+    {
+      "epoch": 0.06066330814441646,
+      "grad_norm": 0.24025796353816986,
+      "learning_rate": 6.058700209643606e-05,
+      "loss": 0.1907,
+      "step": 289
+    },
+    {
+      "epoch": 0.06087321578505458,
+      "grad_norm": 0.1682557910680771,
+      "learning_rate": 6.079664570230609e-05,
+      "loss": 0.1747,
+      "step": 290
+    },
+    {
+      "epoch": 0.06108312342569269,
+      "grad_norm": 0.2801767587661743,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.1813,
+      "step": 291
+    },
+    {
+      "epoch": 0.061293031066330814,
+      "grad_norm": 0.2288123071193695,
+      "learning_rate": 6.121593291404612e-05,
+      "loss": 0.1681,
+      "step": 292
+    },
+    {
+      "epoch": 0.061502938706968935,
+      "grad_norm": 0.14664186537265778,
+      "learning_rate": 6.142557651991615e-05,
+      "loss": 0.1723,
+      "step": 293
+    },
+    {
+      "epoch": 0.061712846347607056,
+      "grad_norm": 0.15858514606952667,
+      "learning_rate": 6.163522012578616e-05,
+      "loss": 0.198,
+      "step": 294
+    },
+    {
+      "epoch": 0.06192275398824517,
+      "grad_norm": 0.24689258635044098,
+      "learning_rate": 6.184486373165618e-05,
+      "loss": 0.1865,
+      "step": 295
+    },
+    {
+      "epoch": 0.06213266162888329,
+      "grad_norm": 0.19141900539398193,
+      "learning_rate": 6.205450733752621e-05,
+      "loss": 0.2103,
+      "step": 296
+    },
+    {
+      "epoch": 0.06234256926952141,
+      "grad_norm": 0.22445173561573029,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.1778,
+      "step": 297
+    },
+    {
+      "epoch": 0.06255247691015953,
+      "grad_norm": 0.1805533766746521,
+      "learning_rate": 6.247379454926625e-05,
+      "loss": 0.1752,
+      "step": 298
+    },
+    {
+      "epoch": 0.06276238455079765,
+      "grad_norm": 0.17849349975585938,
+      "learning_rate": 6.268343815513627e-05,
+      "loss": 0.1735,
+      "step": 299
+    },
+    {
+      "epoch": 0.06297229219143577,
+      "grad_norm": 0.2601464092731476,
+      "learning_rate": 6.289308176100629e-05,
+      "loss": 0.1868,
+      "step": 300
+    },
+    {
+      "epoch": 0.06318219983207389,
+      "grad_norm": 0.2865089178085327,
+      "learning_rate": 6.310272536687631e-05,
+      "loss": 0.2058,
+      "step": 301
+    },
+    {
+      "epoch": 0.06339210747271201,
+      "grad_norm": 0.1764407902956009,
+      "learning_rate": 6.331236897274634e-05,
+      "loss": 0.1828,
+      "step": 302
+    },
+    {
+      "epoch": 0.06360201511335013,
+      "grad_norm": 0.136027991771698,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.1841,
+      "step": 303
+    },
+    {
+      "epoch": 0.06381192275398824,
+      "grad_norm": 0.26960527896881104,
+      "learning_rate": 6.373165618448638e-05,
+      "loss": 0.1806,
+      "step": 304
+    },
+    {
+      "epoch": 0.06402183039462636,
+      "grad_norm": 0.2371356189250946,
+      "learning_rate": 6.39412997903564e-05,
+      "loss": 0.1755,
+      "step": 305
+    },
+    {
+      "epoch": 0.06423173803526448,
+      "grad_norm": 0.16067345440387726,
+      "learning_rate": 6.415094339622641e-05,
+      "loss": 0.1899,
+      "step": 306
+    },
+    {
+      "epoch": 0.0644416456759026,
+      "grad_norm": 0.1733190417289734,
+      "learning_rate": 6.436058700209644e-05,
+      "loss": 0.1829,
+      "step": 307
+    },
+    {
+      "epoch": 0.06465155331654072,
+      "grad_norm": 0.19170600175857544,
+      "learning_rate": 6.457023060796647e-05,
+      "loss": 0.1777,
+      "step": 308
+    },
+    {
+      "epoch": 0.06486146095717885,
+      "grad_norm": 0.17290905117988586,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.187,
+      "step": 309
+    },
+    {
+      "epoch": 0.06507136859781697,
+      "grad_norm": 0.3149113059043884,
+      "learning_rate": 6.49895178197065e-05,
+      "loss": 0.2037,
+      "step": 310
+    },
+    {
+      "epoch": 0.06528127623845507,
+      "grad_norm": 0.15404744446277618,
+      "learning_rate": 6.519916142557653e-05,
+      "loss": 0.1909,
+      "step": 311
+    },
+    {
+      "epoch": 0.0654911838790932,
+      "grad_norm": 0.157347172498703,
+      "learning_rate": 6.540880503144654e-05,
+      "loss": 0.191,
+      "step": 312
+    },
+    {
+      "epoch": 0.06570109151973132,
+      "grad_norm": 0.23342733085155487,
+      "learning_rate": 6.561844863731657e-05,
+      "loss": 0.1867,
+      "step": 313
+    },
+    {
+      "epoch": 0.06591099916036944,
+      "grad_norm": 0.2597595155239105,
+      "learning_rate": 6.58280922431866e-05,
+      "loss": 0.1949,
+      "step": 314
+    },
+    {
+      "epoch": 0.06612090680100756,
+      "grad_norm": 0.23665842413902283,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.1898,
+      "step": 315
+    },
+    {
+      "epoch": 0.06633081444164568,
+      "grad_norm": 0.18460237979888916,
+      "learning_rate": 6.624737945492663e-05,
+      "loss": 0.1824,
+      "step": 316
+    },
+    {
+      "epoch": 0.0665407220822838,
+      "grad_norm": 0.1677280068397522,
+      "learning_rate": 6.645702306079666e-05,
+      "loss": 0.1871,
+      "step": 317
+    },
+    {
+      "epoch": 0.06675062972292191,
+      "grad_norm": 0.1769377440214157,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.1779,
+      "step": 318
+    },
+    {
+      "epoch": 0.06696053736356003,
+      "grad_norm": 0.1881011724472046,
+      "learning_rate": 6.68763102725367e-05,
+      "loss": 0.1711,
+      "step": 319
+    },
+    {
+      "epoch": 0.06717044500419815,
+      "grad_norm": 0.19110549986362457,
+      "learning_rate": 6.708595387840672e-05,
+      "loss": 0.1823,
+      "step": 320
+    },
+    {
+      "epoch": 0.06738035264483627,
+      "grad_norm": 0.26796162128448486,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.1863,
+      "step": 321
+    },
+    {
+      "epoch": 0.0675902602854744,
+      "grad_norm": 0.17290090024471283,
+      "learning_rate": 6.750524109014676e-05,
+      "loss": 0.1816,
+      "step": 322
+    },
+    {
+      "epoch": 0.06780016792611251,
+      "grad_norm": 0.2324109524488449,
+      "learning_rate": 6.771488469601677e-05,
+      "loss": 0.1944,
+      "step": 323
+    },
+    {
+      "epoch": 0.06801007556675064,
+      "grad_norm": 0.24944299459457397,
+      "learning_rate": 6.79245283018868e-05,
+      "loss": 0.1965,
+      "step": 324
+    },
+    {
+      "epoch": 0.06821998320738874,
+      "grad_norm": 0.2102229744195938,
+      "learning_rate": 6.813417190775681e-05,
+      "loss": 0.221,
+      "step": 325
+    },
+    {
+      "epoch": 0.06842989084802686,
+      "grad_norm": 0.22497773170471191,
+      "learning_rate": 6.834381551362684e-05,
+      "loss": 0.1818,
+      "step": 326
+    },
+    {
+      "epoch": 0.06863979848866499,
+      "grad_norm": 0.19047041237354279,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.1817,
+      "step": 327
+    },
+    {
+      "epoch": 0.0688497061293031,
+      "grad_norm": 0.19890040159225464,
+      "learning_rate": 6.876310272536687e-05,
+      "loss": 0.2054,
+      "step": 328
+    },
+    {
+      "epoch": 0.06905961376994123,
+      "grad_norm": 0.18274420499801636,
+      "learning_rate": 6.89727463312369e-05,
+      "loss": 0.1732,
+      "step": 329
+    },
+    {
+      "epoch": 0.06926952141057935,
+      "grad_norm": 0.20556879043579102,
+      "learning_rate": 6.918238993710691e-05,
+      "loss": 0.1773,
+      "step": 330
+    },
+    {
+      "epoch": 0.06947942905121747,
+      "grad_norm": 0.19632075726985931,
+      "learning_rate": 6.939203354297694e-05,
+      "loss": 0.1841,
+      "step": 331
+    },
+    {
+      "epoch": 0.06968933669185558,
+      "grad_norm": 0.25757917761802673,
+      "learning_rate": 6.960167714884696e-05,
+      "loss": 0.1861,
+      "step": 332
+    },
+    {
+      "epoch": 0.0698992443324937,
+      "grad_norm": 0.1654757410287857,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.1797,
+      "step": 333
+    },
+    {
+      "epoch": 0.07010915197313182,
+      "grad_norm": 0.1424175500869751,
+      "learning_rate": 7.0020964360587e-05,
+      "loss": 0.1659,
+      "step": 334
+    },
+    {
+      "epoch": 0.07031905961376994,
+      "grad_norm": 0.21559248864650726,
+      "learning_rate": 7.023060796645703e-05,
+      "loss": 0.175,
+      "step": 335
+    },
+    {
+      "epoch": 0.07052896725440806,
+      "grad_norm": 0.3100188076496124,
+      "learning_rate": 7.044025157232704e-05,
+      "loss": 0.1644,
+      "step": 336
+    },
+    {
+      "epoch": 0.07073887489504618,
+      "grad_norm": 0.2530849277973175,
+      "learning_rate": 7.064989517819707e-05,
+      "loss": 0.1697,
+      "step": 337
+    },
+    {
+      "epoch": 0.0709487825356843,
+      "grad_norm": 0.2111438810825348,
+      "learning_rate": 7.085953878406709e-05,
+      "loss": 0.202,
+      "step": 338
+    },
+    {
+      "epoch": 0.07115869017632241,
+      "grad_norm": 0.21798165142536163,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.1832,
+      "step": 339
+    },
+    {
+      "epoch": 0.07136859781696053,
+      "grad_norm": 0.24779516458511353,
+      "learning_rate": 7.127882599580713e-05,
+      "loss": 0.1934,
+      "step": 340
+    },
+    {
+      "epoch": 0.07157850545759865,
+      "grad_norm": 0.21718356013298035,
+      "learning_rate": 7.148846960167716e-05,
+      "loss": 0.1951,
+      "step": 341
+    },
+    {
+      "epoch": 0.07178841309823678,
+      "grad_norm": 0.22320568561553955,
+      "learning_rate": 7.169811320754717e-05,
+      "loss": 0.1619,
+      "step": 342
+    },
+    {
+      "epoch": 0.0719983207388749,
+      "grad_norm": 0.19393590092658997,
+      "learning_rate": 7.19077568134172e-05,
+      "loss": 0.201,
+      "step": 343
+    },
+    {
+      "epoch": 0.07220822837951302,
+      "grad_norm": 0.1626208871603012,
+      "learning_rate": 7.211740041928722e-05,
+      "loss": 0.1848,
+      "step": 344
+    },
+    {
+      "epoch": 0.07241813602015114,
+      "grad_norm": 0.2256711721420288,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.192,
+      "step": 345
+    },
+    {
+      "epoch": 0.07262804366078925,
+      "grad_norm": 0.2225414514541626,
+      "learning_rate": 7.253668763102726e-05,
+      "loss": 0.1907,
+      "step": 346
+    },
+    {
+      "epoch": 0.07283795130142737,
+      "grad_norm": 0.1786690205335617,
+      "learning_rate": 7.274633123689728e-05,
+      "loss": 0.1756,
+      "step": 347
+    },
+    {
+      "epoch": 0.07304785894206549,
+      "grad_norm": 0.2155577391386032,
+      "learning_rate": 7.29559748427673e-05,
+      "loss": 0.1739,
+      "step": 348
+    },
+    {
+      "epoch": 0.07325776658270361,
+      "grad_norm": 0.32533329725265503,
+      "learning_rate": 7.316561844863732e-05,
+      "loss": 0.1988,
+      "step": 349
+    },
+    {
+      "epoch": 0.07346767422334173,
+      "grad_norm": 0.1870083510875702,
+      "learning_rate": 7.337526205450735e-05,
+      "loss": 0.1615,
+      "step": 350
+    },
+    {
+      "epoch": 0.07367758186397985,
+      "grad_norm": 0.2160840779542923,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.1901,
+      "step": 351
+    },
+    {
+      "epoch": 0.07388748950461797,
+      "grad_norm": 0.19049416482448578,
+      "learning_rate": 7.379454926624739e-05,
+      "loss": 0.1763,
+      "step": 352
+    },
+    {
+      "epoch": 0.0740973971452561,
+      "grad_norm": 0.15733250975608826,
+      "learning_rate": 7.400419287211741e-05,
+      "loss": 0.1812,
+      "step": 353
+    },
+    {
+      "epoch": 0.0743073047858942,
+      "grad_norm": 0.22470858693122864,
+      "learning_rate": 7.421383647798742e-05,
+      "loss": 0.194,
+      "step": 354
+    },
+    {
+      "epoch": 0.07451721242653232,
+      "grad_norm": 0.3271860182285309,
+      "learning_rate": 7.442348008385745e-05,
+      "loss": 0.1708,
+      "step": 355
+    },
+    {
+      "epoch": 0.07472712006717044,
+      "grad_norm": 0.17839424312114716,
+      "learning_rate": 7.463312368972748e-05,
+      "loss": 0.182,
+      "step": 356
+    },
+    {
+      "epoch": 0.07493702770780857,
+      "grad_norm": 0.1907908171415329,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.163,
+      "step": 357
+    },
+    {
+      "epoch": 0.07514693534844669,
+      "grad_norm": 0.20342503488063812,
+      "learning_rate": 7.505241090146751e-05,
+      "loss": 0.2029,
+      "step": 358
+    },
+    {
+      "epoch": 0.07535684298908481,
+      "grad_norm": 0.21872438490390778,
+      "learning_rate": 7.526205450733753e-05,
+      "loss": 0.192,
+      "step": 359
+    },
+    {
+      "epoch": 0.07556675062972293,
+      "grad_norm": 0.22313977777957916,
+      "learning_rate": 7.547169811320755e-05,
+      "loss": 0.1823,
+      "step": 360
+    },
+    {
+      "epoch": 0.07577665827036104,
+      "grad_norm": 0.1931924819946289,
+      "learning_rate": 7.568134171907756e-05,
+      "loss": 0.1992,
+      "step": 361
+    },
+    {
+      "epoch": 0.07598656591099916,
+      "grad_norm": 0.2859954535961151,
+      "learning_rate": 7.589098532494759e-05,
+      "loss": 0.1684,
+      "step": 362
+    },
+    {
+      "epoch": 0.07619647355163728,
+      "grad_norm": 0.18601499497890472,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.1871,
+      "step": 363
+    },
+    {
+      "epoch": 0.0764063811922754,
+      "grad_norm": 0.26345667243003845,
+      "learning_rate": 7.631027253668763e-05,
+      "loss": 0.1883,
+      "step": 364
+    },
+    {
+      "epoch": 0.07661628883291352,
+      "grad_norm": 0.24455974996089935,
+      "learning_rate": 7.651991614255765e-05,
+      "loss": 0.1991,
+      "step": 365
+    },
+    {
+      "epoch": 0.07682619647355164,
+      "grad_norm": 0.1787412464618683,
+      "learning_rate": 7.672955974842768e-05,
+      "loss": 0.1895,
+      "step": 366
+    },
+    {
+      "epoch": 0.07703610411418976,
+      "grad_norm": 0.2711624205112457,
+      "learning_rate": 7.693920335429769e-05,
+      "loss": 0.1889,
+      "step": 367
+    },
+    {
+      "epoch": 0.07724601175482787,
+      "grad_norm": 0.2764052152633667,
+      "learning_rate": 7.714884696016772e-05,
+      "loss": 0.1895,
+      "step": 368
+    },
+    {
+      "epoch": 0.07745591939546599,
+      "grad_norm": 0.15490169823169708,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.1812,
+      "step": 369
+    },
+    {
+      "epoch": 0.07766582703610411,
+      "grad_norm": 0.18659183382987976,
+      "learning_rate": 7.756813417190776e-05,
+      "loss": 0.1871,
+      "step": 370
+    },
+    {
+      "epoch": 0.07787573467674223,
+      "grad_norm": 0.19188903272151947,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.1716,
+      "step": 371
+    },
+    {
+      "epoch": 0.07808564231738035,
+      "grad_norm": 0.22174161672592163,
+      "learning_rate": 7.798742138364781e-05,
+      "loss": 0.1795,
+      "step": 372
+    },
+    {
+      "epoch": 0.07829554995801848,
+      "grad_norm": 0.1882723569869995,
+      "learning_rate": 7.819706498951782e-05,
+      "loss": 0.1874,
+      "step": 373
+    },
+    {
+      "epoch": 0.0785054575986566,
+      "grad_norm": 0.1762145459651947,
+      "learning_rate": 7.840670859538785e-05,
+      "loss": 0.1852,
+      "step": 374
+    },
+    {
+      "epoch": 0.0787153652392947,
+      "grad_norm": 0.17931701242923737,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.1817,
+      "step": 375
+    },
+    {
+      "epoch": 0.07892527287993283,
+      "grad_norm": 0.1833990216255188,
+      "learning_rate": 7.882599580712788e-05,
+      "loss": 0.1852,
+      "step": 376
+    },
+    {
+      "epoch": 0.07913518052057095,
+      "grad_norm": 0.2758026421070099,
+      "learning_rate": 7.903563941299791e-05,
+      "loss": 0.1868,
+      "step": 377
+    },
+    {
+      "epoch": 0.07934508816120907,
+      "grad_norm": 0.24222204089164734,
+      "learning_rate": 7.924528301886794e-05,
+      "loss": 0.186,
+      "step": 378
+    },
+    {
+      "epoch": 0.07955499580184719,
+      "grad_norm": 0.17609156668186188,
+      "learning_rate": 7.945492662473795e-05,
+      "loss": 0.1807,
+      "step": 379
+    },
+    {
+      "epoch": 0.07976490344248531,
+      "grad_norm": 0.23695167899131775,
+      "learning_rate": 7.966457023060797e-05,
+      "loss": 0.1799,
+      "step": 380
+    },
+    {
+      "epoch": 0.07997481108312343,
+      "grad_norm": 0.25356245040893555,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.1899,
+      "step": 381
+    },
+    {
+      "epoch": 0.08018471872376154,
+      "grad_norm": 0.23144365847110748,
+      "learning_rate": 8.008385744234801e-05,
+      "loss": 0.192,
+      "step": 382
+    },
+    {
+      "epoch": 0.08039462636439966,
+      "grad_norm": 0.1521812379360199,
+      "learning_rate": 8.029350104821804e-05,
+      "loss": 0.1864,
+      "step": 383
+    },
+    {
+      "epoch": 0.08060453400503778,
+      "grad_norm": 0.16725748777389526,
+      "learning_rate": 8.050314465408806e-05,
+      "loss": 0.202,
+      "step": 384
+    },
+    {
+      "epoch": 0.0808144416456759,
+      "grad_norm": 0.21173058450222015,
+      "learning_rate": 8.071278825995808e-05,
+      "loss": 0.1751,
+      "step": 385
+    },
+    {
+      "epoch": 0.08102434928631402,
+      "grad_norm": 0.15676653385162354,
+      "learning_rate": 8.09224318658281e-05,
+      "loss": 0.1946,
+      "step": 386
+    },
+    {
+      "epoch": 0.08123425692695214,
+      "grad_norm": 0.21838362514972687,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.1889,
+      "step": 387
+    },
+    {
+      "epoch": 0.08144416456759027,
+      "grad_norm": 0.19586238265037537,
+      "learning_rate": 8.134171907756814e-05,
+      "loss": 0.1884,
+      "step": 388
+    },
+    {
+      "epoch": 0.08165407220822837,
+      "grad_norm": 0.21012739837169647,
+      "learning_rate": 8.155136268343817e-05,
+      "loss": 0.1822,
+      "step": 389
+    },
+    {
+      "epoch": 0.0818639798488665,
+      "grad_norm": 0.2092917114496231,
+      "learning_rate": 8.176100628930818e-05,
+      "loss": 0.1783,
+      "step": 390
+    },
+    {
+      "epoch": 0.08207388748950462,
+      "grad_norm": 0.3745954930782318,
+      "learning_rate": 8.19706498951782e-05,
+      "loss": 0.1976,
+      "step": 391
+    },
+    {
+      "epoch": 0.08228379513014274,
+      "grad_norm": 0.2579379081726074,
+      "learning_rate": 8.218029350104822e-05,
+      "loss": 0.197,
+      "step": 392
+    },
+    {
+      "epoch": 0.08249370277078086,
+      "grad_norm": 0.18806852400302887,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.1873,
+      "step": 393
+    },
+    {
+      "epoch": 0.08270361041141898,
+      "grad_norm": 0.24592849612236023,
+      "learning_rate": 8.259958071278825e-05,
+      "loss": 0.2083,
+      "step": 394
+    },
+    {
+      "epoch": 0.0829135180520571,
+      "grad_norm": 0.2678208649158478,
+      "learning_rate": 8.280922431865828e-05,
+      "loss": 0.1878,
+      "step": 395
+    },
+    {
+      "epoch": 0.08312342569269521,
+      "grad_norm": 0.2023075520992279,
+      "learning_rate": 8.30188679245283e-05,
+      "loss": 0.1861,
+      "step": 396
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 0.2390558272600174,
+      "learning_rate": 8.322851153039832e-05,
+      "loss": 0.1802,
+      "step": 397
+    },
+    {
+      "epoch": 0.08354324097397145,
+      "grad_norm": 0.175222247838974,
+      "learning_rate": 8.343815513626834e-05,
+      "loss": 0.1674,
+      "step": 398
+    },
+    {
+      "epoch": 0.08375314861460957,
+      "grad_norm": 0.2340380698442459,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.17,
+      "step": 399
+    },
+    {
+      "epoch": 0.08396305625524769,
+      "grad_norm": 0.20814655721187592,
+      "learning_rate": 8.385744234800838e-05,
+      "loss": 0.1513,
+      "step": 400
+    },
+    {
+      "epoch": 0.08417296389588581,
+      "grad_norm": 0.34867948293685913,
+      "learning_rate": 8.406708595387841e-05,
+      "loss": 0.1644,
+      "step": 401
+    },
+    {
+      "epoch": 0.08438287153652393,
+      "grad_norm": 0.21573619544506073,
+      "learning_rate": 8.427672955974843e-05,
+      "loss": 0.1785,
+      "step": 402
+    },
+    {
+      "epoch": 0.08459277917716204,
+      "grad_norm": 0.21437713503837585,
+      "learning_rate": 8.448637316561845e-05,
+      "loss": 0.1523,
+      "step": 403
+    },
+    {
+      "epoch": 0.08480268681780016,
+      "grad_norm": 0.2250152826309204,
+      "learning_rate": 8.469601677148847e-05,
+      "loss": 0.1728,
+      "step": 404
+    },
+    {
+      "epoch": 0.08501259445843828,
+      "grad_norm": 0.2514733672142029,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.1755,
+      "step": 405
+    },
+    {
+      "epoch": 0.0852225020990764,
+      "grad_norm": 0.16003377735614777,
+      "learning_rate": 8.511530398322851e-05,
+      "loss": 0.1935,
+      "step": 406
+    },
+    {
+      "epoch": 0.08543240973971453,
+      "grad_norm": 0.1792365163564682,
+      "learning_rate": 8.532494758909854e-05,
+      "loss": 0.1991,
+      "step": 407
+    },
+    {
+      "epoch": 0.08564231738035265,
+      "grad_norm": 0.16854703426361084,
+      "learning_rate": 8.553459119496856e-05,
+      "loss": 0.1904,
+      "step": 408
+    },
+    {
+      "epoch": 0.08585222502099077,
+      "grad_norm": 0.19401603937149048,
+      "learning_rate": 8.574423480083857e-05,
+      "loss": 0.1681,
+      "step": 409
+    },
+    {
+      "epoch": 0.08606213266162889,
+      "grad_norm": 0.15639828145503998,
+      "learning_rate": 8.59538784067086e-05,
+      "loss": 0.1628,
+      "step": 410
+    },
+    {
+      "epoch": 0.086272040302267,
+      "grad_norm": 0.15357258915901184,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.1905,
+      "step": 411
+    },
+    {
+      "epoch": 0.08648194794290512,
+      "grad_norm": 0.256944477558136,
+      "learning_rate": 8.637316561844864e-05,
+      "loss": 0.2015,
+      "step": 412
+    },
+    {
+      "epoch": 0.08669185558354324,
+      "grad_norm": 0.13482192158699036,
+      "learning_rate": 8.658280922431866e-05,
+      "loss": 0.1718,
+      "step": 413
+    },
+    {
+      "epoch": 0.08690176322418136,
+      "grad_norm": 0.16525831818580627,
+      "learning_rate": 8.679245283018869e-05,
+      "loss": 0.1778,
+      "step": 414
+    },
+    {
+      "epoch": 0.08711167086481948,
+      "grad_norm": 0.2145531326532364,
+      "learning_rate": 8.70020964360587e-05,
+      "loss": 0.1714,
+      "step": 415
+    },
+    {
+      "epoch": 0.0873215785054576,
+      "grad_norm": 0.20481255650520325,
+      "learning_rate": 8.721174004192873e-05,
+      "loss": 0.1832,
+      "step": 416
+    },
+    {
+      "epoch": 0.08753148614609572,
+      "grad_norm": 0.14865756034851074,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.1659,
+      "step": 417
+    },
+    {
+      "epoch": 0.08774139378673383,
+      "grad_norm": 0.14713706076145172,
+      "learning_rate": 8.763102725366877e-05,
+      "loss": 0.1674,
+      "step": 418
+    },
+    {
+      "epoch": 0.08795130142737195,
+      "grad_norm": 0.1728464812040329,
+      "learning_rate": 8.784067085953879e-05,
+      "loss": 0.1727,
+      "step": 419
+    },
+    {
+      "epoch": 0.08816120906801007,
+      "grad_norm": 0.2646033465862274,
+      "learning_rate": 8.805031446540882e-05,
+      "loss": 0.1973,
+      "step": 420
+    },
+    {
+      "epoch": 0.0883711167086482,
+      "grad_norm": 0.2262433022260666,
+      "learning_rate": 8.825995807127883e-05,
+      "loss": 0.1817,
+      "step": 421
+    },
+    {
+      "epoch": 0.08858102434928632,
+      "grad_norm": 0.16398945450782776,
+      "learning_rate": 8.846960167714886e-05,
+      "loss": 0.1628,
+      "step": 422
+    },
+    {
+      "epoch": 0.08879093198992444,
+      "grad_norm": 0.15976634621620178,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.1678,
+      "step": 423
+    },
+    {
+      "epoch": 0.08900083963056256,
+      "grad_norm": 0.19314904510974884,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.1753,
+      "step": 424
+    },
+    {
+      "epoch": 0.08921074727120067,
+      "grad_norm": 0.21701818704605103,
+      "learning_rate": 8.909853249475892e-05,
+      "loss": 0.19,
+      "step": 425
+    },
+    {
+      "epoch": 0.08942065491183879,
+      "grad_norm": 0.22768795490264893,
+      "learning_rate": 8.930817610062893e-05,
+      "loss": 0.1613,
+      "step": 426
+    },
+    {
+      "epoch": 0.08963056255247691,
+      "grad_norm": 0.13172288239002228,
+      "learning_rate": 8.951781970649896e-05,
+      "loss": 0.1749,
+      "step": 427
+    },
+    {
+      "epoch": 0.08984047019311503,
+      "grad_norm": 0.2015480250120163,
+      "learning_rate": 8.972746331236897e-05,
+      "loss": 0.1857,
+      "step": 428
+    },
+    {
+      "epoch": 0.09005037783375315,
+      "grad_norm": 0.20239531993865967,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.1804,
+      "step": 429
+    },
+    {
+      "epoch": 0.09026028547439127,
+      "grad_norm": 0.17528317868709564,
+      "learning_rate": 9.014675052410901e-05,
+      "loss": 0.1729,
+      "step": 430
+    },
+    {
+      "epoch": 0.0904701931150294,
+      "grad_norm": 0.16663801670074463,
+      "learning_rate": 9.035639412997903e-05,
+      "loss": 0.1896,
+      "step": 431
+    },
+    {
+      "epoch": 0.0906801007556675,
+      "grad_norm": 0.18777558207511902,
+      "learning_rate": 9.056603773584906e-05,
+      "loss": 0.1801,
+      "step": 432
+    },
+    {
+      "epoch": 0.09089000839630562,
+      "grad_norm": 0.1440989226102829,
+      "learning_rate": 9.077568134171907e-05,
+      "loss": 0.179,
+      "step": 433
+    },
+    {
+      "epoch": 0.09109991603694374,
+      "grad_norm": 0.1786854863166809,
+      "learning_rate": 9.09853249475891e-05,
+      "loss": 0.1799,
+      "step": 434
+    },
+    {
+      "epoch": 0.09130982367758186,
+      "grad_norm": 0.20794442296028137,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.1731,
+      "step": 435
+    },
+    {
+      "epoch": 0.09151973131821999,
+      "grad_norm": 0.16377133131027222,
+      "learning_rate": 9.140461215932914e-05,
+      "loss": 0.1815,
+      "step": 436
+    },
+    {
+      "epoch": 0.0917296389588581,
+      "grad_norm": 0.174666628241539,
+      "learning_rate": 9.161425576519916e-05,
+      "loss": 0.1798,
+      "step": 437
+    },
+    {
+      "epoch": 0.09193954659949623,
+      "grad_norm": 0.2127188891172409,
+      "learning_rate": 9.182389937106919e-05,
+      "loss": 0.1805,
+      "step": 438
+    },
+    {
+      "epoch": 0.09214945424013433,
+      "grad_norm": 0.1936446726322174,
+      "learning_rate": 9.203354297693921e-05,
+      "loss": 0.1723,
+      "step": 439
+    },
+    {
+      "epoch": 0.09235936188077246,
+      "grad_norm": 0.18736332654953003,
+      "learning_rate": 9.224318658280923e-05,
+      "loss": 0.1849,
+      "step": 440
+    },
+    {
+      "epoch": 0.09256926952141058,
+      "grad_norm": 0.16817238926887512,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.1746,
+      "step": 441
+    },
+    {
+      "epoch": 0.0927791771620487,
+      "grad_norm": 0.18249107897281647,
+      "learning_rate": 9.266247379454928e-05,
+      "loss": 0.1628,
+      "step": 442
+    },
+    {
+      "epoch": 0.09298908480268682,
+      "grad_norm": 0.1728898137807846,
+      "learning_rate": 9.287211740041929e-05,
+      "loss": 0.203,
+      "step": 443
+    },
+    {
+      "epoch": 0.09319899244332494,
+      "grad_norm": 0.16144797205924988,
+      "learning_rate": 9.308176100628931e-05,
+      "loss": 0.1843,
+      "step": 444
+    },
+    {
+      "epoch": 0.09340890008396306,
+      "grad_norm": 0.19680747389793396,
+      "learning_rate": 9.329140461215934e-05,
+      "loss": 0.168,
+      "step": 445
+    },
+    {
+      "epoch": 0.09361880772460117,
+      "grad_norm": 0.16198395192623138,
+      "learning_rate": 9.350104821802935e-05,
+      "loss": 0.1598,
+      "step": 446
+    },
+    {
+      "epoch": 0.09382871536523929,
+      "grad_norm": 0.17398878931999207,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.17,
+      "step": 447
+    },
+    {
+      "epoch": 0.09403862300587741,
+      "grad_norm": 0.18602675199508667,
+      "learning_rate": 9.39203354297694e-05,
+      "loss": 0.1716,
+      "step": 448
+    },
+    {
+      "epoch": 0.09424853064651553,
+      "grad_norm": 0.18403322994709015,
+      "learning_rate": 9.412997903563942e-05,
+      "loss": 0.1948,
+      "step": 449
+    },
+    {
+      "epoch": 0.09445843828715365,
+      "grad_norm": 0.18783587217330933,
+      "learning_rate": 9.433962264150944e-05,
+      "loss": 0.1812,
+      "step": 450
+    },
+    {
+      "epoch": 0.09466834592779177,
+      "grad_norm": 0.20252300798892975,
+      "learning_rate": 9.454926624737947e-05,
+      "loss": 0.1666,
+      "step": 451
+    },
+    {
+      "epoch": 0.0948782535684299,
+      "grad_norm": 0.20129899680614471,
+      "learning_rate": 9.475890985324948e-05,
+      "loss": 0.1695,
+      "step": 452
+    },
+    {
+      "epoch": 0.095088161209068,
+      "grad_norm": 0.17035968601703644,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.1678,
+      "step": 453
+    },
+    {
+      "epoch": 0.09529806884970612,
+      "grad_norm": 0.20403030514717102,
+      "learning_rate": 9.517819706498953e-05,
+      "loss": 0.1843,
+      "step": 454
+    },
+    {
+      "epoch": 0.09550797649034425,
+      "grad_norm": 0.17489562928676605,
+      "learning_rate": 9.538784067085954e-05,
+      "loss": 0.1657,
+      "step": 455
+    },
+    {
+      "epoch": 0.09571788413098237,
+      "grad_norm": 0.19699983298778534,
+      "learning_rate": 9.559748427672957e-05,
+      "loss": 0.1634,
+      "step": 456
+    },
+    {
+      "epoch": 0.09592779177162049,
+      "grad_norm": 0.17113354802131653,
+      "learning_rate": 9.58071278825996e-05,
+      "loss": 0.1879,
+      "step": 457
+    },
+    {
+      "epoch": 0.09613769941225861,
+      "grad_norm": 0.2290397435426712,
+      "learning_rate": 9.601677148846961e-05,
+      "loss": 0.1784,
+      "step": 458
+    },
+    {
+      "epoch": 0.09634760705289673,
+      "grad_norm": 0.2173147350549698,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.1673,
+      "step": 459
+    },
+    {
+      "epoch": 0.09655751469353484,
+      "grad_norm": 0.18280835449695587,
+      "learning_rate": 9.643605870020965e-05,
+      "loss": 0.1806,
+      "step": 460
+    },
+    {
+      "epoch": 0.09676742233417296,
+      "grad_norm": 0.14638672769069672,
+      "learning_rate": 9.664570230607967e-05,
+      "loss": 0.1833,
+      "step": 461
+    },
+    {
+      "epoch": 0.09697732997481108,
+      "grad_norm": 0.16228064894676208,
+      "learning_rate": 9.685534591194969e-05,
+      "loss": 0.1762,
+      "step": 462
+    },
+    {
+      "epoch": 0.0971872376154492,
+      "grad_norm": 0.1836690902709961,
+      "learning_rate": 9.706498951781971e-05,
+      "loss": 0.1782,
+      "step": 463
+    },
+    {
+      "epoch": 0.09739714525608732,
+      "grad_norm": 0.19470515847206116,
+      "learning_rate": 9.727463312368972e-05,
+      "loss": 0.185,
+      "step": 464
+    },
+    {
+      "epoch": 0.09760705289672544,
+      "grad_norm": 0.1833791732788086,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.1753,
+      "step": 465
+    },
+    {
+      "epoch": 0.09781696053736356,
+      "grad_norm": 0.22608265280723572,
+      "learning_rate": 9.769392033542977e-05,
+      "loss": 0.1792,
+      "step": 466
+    },
+    {
+      "epoch": 0.09802686817800169,
+      "grad_norm": 0.16552825272083282,
+      "learning_rate": 9.790356394129979e-05,
+      "loss": 0.1808,
+      "step": 467
+    },
+    {
+      "epoch": 0.0982367758186398,
+      "grad_norm": 0.2294851690530777,
+      "learning_rate": 9.811320754716981e-05,
+      "loss": 0.1935,
+      "step": 468
+    },
+    {
+      "epoch": 0.09844668345927791,
+      "grad_norm": 0.26589101552963257,
+      "learning_rate": 9.832285115303984e-05,
+      "loss": 0.1676,
+      "step": 469
+    },
+    {
+      "epoch": 0.09865659109991604,
+      "grad_norm": 0.22315791249275208,
+      "learning_rate": 9.853249475890985e-05,
+      "loss": 0.1806,
+      "step": 470
+    },
+    {
+      "epoch": 0.09886649874055416,
+      "grad_norm": 0.16855137050151825,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.1816,
+      "step": 471
+    },
+    {
+      "epoch": 0.09907640638119228,
+      "grad_norm": 0.19197392463684082,
+      "learning_rate": 9.89517819706499e-05,
+      "loss": 0.1923,
+      "step": 472
+    },
+    {
+      "epoch": 0.0992863140218304,
+      "grad_norm": 0.18722014129161835,
+      "learning_rate": 9.916142557651992e-05,
+      "loss": 0.1823,
+      "step": 473
+    },
+    {
+      "epoch": 0.09949622166246852,
+      "grad_norm": 0.15668706595897675,
+      "learning_rate": 9.937106918238994e-05,
+      "loss": 0.1896,
+      "step": 474
+    },
+    {
+      "epoch": 0.09970612930310663,
+      "grad_norm": 0.17297013103961945,
+      "learning_rate": 9.958071278825997e-05,
+      "loss": 0.1907,
+      "step": 475
+    },
+    {
+      "epoch": 0.09991603694374475,
+      "grad_norm": 0.23546694219112396,
+      "learning_rate": 9.979035639412998e-05,
+      "loss": 0.1647,
+      "step": 476
+    },
+    {
+      "epoch": 0.10012594458438287,
+      "grad_norm": 0.1627054661512375,
+      "learning_rate": 0.0001,
+      "loss": 0.1686,
+      "step": 477
+    },
+    {
+      "epoch": 0.10033585222502099,
+      "grad_norm": 0.21043647825717926,
+      "learning_rate": 9.999998657442895e-05,
+      "loss": 0.1865,
+      "step": 478
+    },
+    {
+      "epoch": 0.10054575986565911,
+      "grad_norm": 0.19615764915943146,
+      "learning_rate": 9.999994629772298e-05,
+      "loss": 0.1683,
+      "step": 479
+    },
+    {
+      "epoch": 0.10075566750629723,
+      "grad_norm": 0.2884671092033386,
+      "learning_rate": 9.999987916990372e-05,
+      "loss": 0.1858,
+      "step": 480
+    },
+    {
+      "epoch": 0.10096557514693535,
+      "grad_norm": 0.2381323128938675,
+      "learning_rate": 9.999978519100723e-05,
+      "loss": 0.1879,
+      "step": 481
+    },
+    {
+      "epoch": 0.10117548278757346,
+      "grad_norm": 0.19187557697296143,
+      "learning_rate": 9.999966436108398e-05,
+      "loss": 0.1808,
+      "step": 482
+    },
+    {
+      "epoch": 0.10138539042821158,
+      "grad_norm": 0.19443491101264954,
+      "learning_rate": 9.999951668019887e-05,
+      "loss": 0.1659,
+      "step": 483
+    },
+    {
+      "epoch": 0.1015952980688497,
+      "grad_norm": 0.20151716470718384,
+      "learning_rate": 9.999934214843116e-05,
+      "loss": 0.1679,
+      "step": 484
+    },
+    {
+      "epoch": 0.10180520570948783,
+      "grad_norm": 0.2896507978439331,
+      "learning_rate": 9.999914076587464e-05,
+      "loss": 0.1734,
+      "step": 485
+    },
+    {
+      "epoch": 0.10201511335012595,
+      "grad_norm": 0.21598441898822784,
+      "learning_rate": 9.999891253263741e-05,
+      "loss": 0.1779,
+      "step": 486
+    },
+    {
+      "epoch": 0.10222502099076407,
+      "grad_norm": 0.196011021733284,
+      "learning_rate": 9.999865744884207e-05,
+      "loss": 0.1815,
+      "step": 487
+    },
+    {
+      "epoch": 0.10243492863140219,
+      "grad_norm": 0.20962318778038025,
+      "learning_rate": 9.999837551462558e-05,
+      "loss": 0.1727,
+      "step": 488
+    },
+    {
+      "epoch": 0.1026448362720403,
+      "grad_norm": 0.18339572846889496,
+      "learning_rate": 9.999806673013935e-05,
+      "loss": 0.1689,
+      "step": 489
+    },
+    {
+      "epoch": 0.10285474391267842,
+      "grad_norm": 0.20994813740253448,
+      "learning_rate": 9.999773109554922e-05,
+      "loss": 0.1768,
+      "step": 490
+    },
+    {
+      "epoch": 0.10306465155331654,
+      "grad_norm": 0.163935124874115,
+      "learning_rate": 9.999736861103541e-05,
+      "loss": 0.1777,
+      "step": 491
+    },
+    {
+      "epoch": 0.10327455919395466,
+      "grad_norm": 0.1465967446565628,
+      "learning_rate": 9.99969792767926e-05,
+      "loss": 0.1846,
+      "step": 492
+    },
+    {
+      "epoch": 0.10348446683459278,
+      "grad_norm": 0.21443922817707062,
+      "learning_rate": 9.999656309302987e-05,
+      "loss": 0.1807,
+      "step": 493
+    },
+    {
+      "epoch": 0.1036943744752309,
+      "grad_norm": 0.18504248559474945,
+      "learning_rate": 9.999612005997071e-05,
+      "loss": 0.1548,
+      "step": 494
+    },
+    {
+      "epoch": 0.10390428211586902,
+      "grad_norm": 0.15490441024303436,
+      "learning_rate": 9.999565017785305e-05,
+      "loss": 0.1696,
+      "step": 495
+    },
+    {
+      "epoch": 0.10411418975650713,
+      "grad_norm": 0.1881389319896698,
+      "learning_rate": 9.999515344692923e-05,
+      "loss": 0.1852,
+      "step": 496
+    },
+    {
+      "epoch": 0.10432409739714525,
+      "grad_norm": 0.16337451338768005,
+      "learning_rate": 9.999462986746598e-05,
+      "loss": 0.1834,
+      "step": 497
+    },
+    {
+      "epoch": 0.10453400503778337,
+      "grad_norm": 0.16641898453235626,
+      "learning_rate": 9.99940794397445e-05,
+      "loss": 0.187,
+      "step": 498
+    },
+    {
+      "epoch": 0.1047439126784215,
+      "grad_norm": 0.15948446094989777,
+      "learning_rate": 9.999350216406038e-05,
+      "loss": 0.1835,
+      "step": 499
+    },
+    {
+      "epoch": 0.10495382031905962,
+      "grad_norm": 0.1550200879573822,
+      "learning_rate": 9.999289804072363e-05,
+      "loss": 0.1706,
+      "step": 500
+    },
+    {
+      "epoch": 0.10516372795969774,
+      "grad_norm": 0.16597698628902435,
+      "learning_rate": 9.999226707005867e-05,
+      "loss": 0.1811,
+      "step": 501
+    },
+    {
+      "epoch": 0.10537363560033586,
+      "grad_norm": 0.17551501095294952,
+      "learning_rate": 9.999160925240434e-05,
+      "loss": 0.1677,
+      "step": 502
+    },
+    {
+      "epoch": 0.10558354324097396,
+      "grad_norm": 0.15515847504138947,
+      "learning_rate": 9.999092458811393e-05,
+      "loss": 0.1789,
+      "step": 503
+    },
+    {
+      "epoch": 0.10579345088161209,
+      "grad_norm": 0.18121638894081116,
+      "learning_rate": 9.99902130775551e-05,
+      "loss": 0.1684,
+      "step": 504
+    },
+    {
+      "epoch": 0.10600335852225021,
+      "grad_norm": 0.1853945255279541,
+      "learning_rate": 9.998947472110994e-05,
+      "loss": 0.1907,
+      "step": 505
+    },
+    {
+      "epoch": 0.10621326616288833,
+      "grad_norm": 0.21303139626979828,
+      "learning_rate": 9.998870951917496e-05,
+      "loss": 0.1712,
+      "step": 506
+    },
+    {
+      "epoch": 0.10642317380352645,
+      "grad_norm": 0.16773764789104462,
+      "learning_rate": 9.998791747216113e-05,
+      "loss": 0.1756,
+      "step": 507
+    },
+    {
+      "epoch": 0.10663308144416457,
+      "grad_norm": 0.18033501505851746,
+      "learning_rate": 9.998709858049376e-05,
+      "loss": 0.1654,
+      "step": 508
+    },
+    {
+      "epoch": 0.10684298908480269,
+      "grad_norm": 0.14199328422546387,
+      "learning_rate": 9.998625284461263e-05,
+      "loss": 0.1587,
+      "step": 509
+    },
+    {
+      "epoch": 0.1070528967254408,
+      "grad_norm": 0.19968685507774353,
+      "learning_rate": 9.998538026497192e-05,
+      "loss": 0.1796,
+      "step": 510
+    },
+    {
+      "epoch": 0.10726280436607892,
+      "grad_norm": 0.17311611771583557,
+      "learning_rate": 9.998448084204021e-05,
+      "loss": 0.1864,
+      "step": 511
+    },
+    {
+      "epoch": 0.10747271200671704,
+      "grad_norm": 0.20124119520187378,
+      "learning_rate": 9.998355457630053e-05,
+      "loss": 0.1829,
+      "step": 512
+    },
+    {
+      "epoch": 0.10768261964735516,
+      "grad_norm": 0.12473297864198685,
+      "learning_rate": 9.998260146825029e-05,
+      "loss": 0.175,
+      "step": 513
+    },
+    {
+      "epoch": 0.10789252728799328,
+      "grad_norm": 0.1696644425392151,
+      "learning_rate": 9.998162151840135e-05,
+      "loss": 0.1762,
+      "step": 514
+    },
+    {
+      "epoch": 0.1081024349286314,
+      "grad_norm": 0.1781477928161621,
+      "learning_rate": 9.998061472727996e-05,
+      "loss": 0.1679,
+      "step": 515
+    },
+    {
+      "epoch": 0.10831234256926953,
+      "grad_norm": 0.19112960994243622,
+      "learning_rate": 9.997958109542675e-05,
+      "loss": 0.1553,
+      "step": 516
+    },
+    {
+      "epoch": 0.10852225020990765,
+      "grad_norm": 0.1417030394077301,
+      "learning_rate": 9.997852062339685e-05,
+      "loss": 0.1737,
+      "step": 517
+    },
+    {
+      "epoch": 0.10873215785054575,
+      "grad_norm": 0.15080858767032623,
+      "learning_rate": 9.997743331175976e-05,
+      "loss": 0.1595,
+      "step": 518
+    },
+    {
+      "epoch": 0.10894206549118388,
+      "grad_norm": 0.2046668380498886,
+      "learning_rate": 9.997631916109937e-05,
+      "loss": 0.1839,
+      "step": 519
+    },
+    {
+      "epoch": 0.109151973131822,
+      "grad_norm": 0.19941595196723938,
+      "learning_rate": 9.997517817201401e-05,
+      "loss": 0.1718,
+      "step": 520
+    },
+    {
+      "epoch": 0.10936188077246012,
+      "grad_norm": 0.15989692509174347,
+      "learning_rate": 9.997401034511642e-05,
+      "loss": 0.1613,
+      "step": 521
+    },
+    {
+      "epoch": 0.10957178841309824,
+      "grad_norm": 0.1697997897863388,
+      "learning_rate": 9.997281568103374e-05,
+      "loss": 0.1603,
+      "step": 522
+    },
+    {
+      "epoch": 0.10978169605373636,
+      "grad_norm": 0.1840822696685791,
+      "learning_rate": 9.997159418040754e-05,
+      "loss": 0.1735,
+      "step": 523
+    },
+    {
+      "epoch": 0.10999160369437448,
+      "grad_norm": 0.20991730690002441,
+      "learning_rate": 9.99703458438938e-05,
+      "loss": 0.154,
+      "step": 524
+    },
+    {
+      "epoch": 0.11020151133501259,
+      "grad_norm": 0.16802968084812164,
+      "learning_rate": 9.99690706721629e-05,
+      "loss": 0.1761,
+      "step": 525
+    },
+    {
+      "epoch": 0.11041141897565071,
+      "grad_norm": 0.18329255282878876,
+      "learning_rate": 9.996776866589962e-05,
+      "loss": 0.1609,
+      "step": 526
+    },
+    {
+      "epoch": 0.11062132661628883,
+      "grad_norm": 0.18645748496055603,
+      "learning_rate": 9.996643982580318e-05,
+      "loss": 0.1793,
+      "step": 527
+    },
+    {
+      "epoch": 0.11083123425692695,
+      "grad_norm": 0.1966720074415207,
+      "learning_rate": 9.996508415258722e-05,
+      "loss": 0.1714,
+      "step": 528
+    },
+    {
+      "epoch": 0.11104114189756507,
+      "grad_norm": 0.18155452609062195,
+      "learning_rate": 9.996370164697974e-05,
+      "loss": 0.1673,
+      "step": 529
+    },
+    {
+      "epoch": 0.1112510495382032,
+      "grad_norm": 0.2004195600748062,
+      "learning_rate": 9.996229230972317e-05,
+      "loss": 0.1865,
+      "step": 530
+    },
+    {
+      "epoch": 0.11146095717884132,
+      "grad_norm": 0.15521694719791412,
+      "learning_rate": 9.996085614157438e-05,
+      "loss": 0.1757,
+      "step": 531
+    },
+    {
+      "epoch": 0.11167086481947942,
+      "grad_norm": 0.1686578243970871,
+      "learning_rate": 9.995939314330462e-05,
+      "loss": 0.1768,
+      "step": 532
+    },
+    {
+      "epoch": 0.11188077246011754,
+      "grad_norm": 0.20034368336200714,
+      "learning_rate": 9.995790331569954e-05,
+      "loss": 0.1823,
+      "step": 533
+    },
+    {
+      "epoch": 0.11209068010075567,
+      "grad_norm": 0.1494702696800232,
+      "learning_rate": 9.995638665955922e-05,
+      "loss": 0.175,
+      "step": 534
+    },
+    {
+      "epoch": 0.11230058774139379,
+      "grad_norm": 0.16365233063697815,
+      "learning_rate": 9.995484317569814e-05,
+      "loss": 0.1716,
+      "step": 535
+    },
+    {
+      "epoch": 0.11251049538203191,
+      "grad_norm": 0.19227434694766998,
+      "learning_rate": 9.995327286494521e-05,
+      "loss": 0.1605,
+      "step": 536
+    },
+    {
+      "epoch": 0.11272040302267003,
+      "grad_norm": 0.21946166455745697,
+      "learning_rate": 9.995167572814365e-05,
+      "loss": 0.182,
+      "step": 537
+    },
+    {
+      "epoch": 0.11293031066330815,
+      "grad_norm": 0.2211793065071106,
+      "learning_rate": 9.995005176615124e-05,
+      "loss": 0.1783,
+      "step": 538
+    },
+    {
+      "epoch": 0.11314021830394626,
+      "grad_norm": 0.2154102325439453,
+      "learning_rate": 9.994840097984006e-05,
+      "loss": 0.1888,
+      "step": 539
+    },
+    {
+      "epoch": 0.11335012594458438,
+      "grad_norm": 0.20600587129592896,
+      "learning_rate": 9.994672337009658e-05,
+      "loss": 0.1871,
+      "step": 540
+    },
+    {
+      "epoch": 0.1135600335852225,
+      "grad_norm": 0.22028079628944397,
+      "learning_rate": 9.994501893782176e-05,
+      "loss": 0.1855,
+      "step": 541
+    },
+    {
+      "epoch": 0.11376994122586062,
+      "grad_norm": 0.23957398533821106,
+      "learning_rate": 9.99432876839309e-05,
+      "loss": 0.1616,
+      "step": 542
+    },
+    {
+      "epoch": 0.11397984886649874,
+      "grad_norm": 0.14516577124595642,
+      "learning_rate": 9.994152960935375e-05,
+      "loss": 0.1864,
+      "step": 543
+    },
+    {
+      "epoch": 0.11418975650713686,
+      "grad_norm": 0.14327426254749298,
+      "learning_rate": 9.99397447150344e-05,
+      "loss": 0.1596,
+      "step": 544
+    },
+    {
+      "epoch": 0.11439966414777498,
+      "grad_norm": 0.15387804806232452,
+      "learning_rate": 9.99379330019314e-05,
+      "loss": 0.1638,
+      "step": 545
+    },
+    {
+      "epoch": 0.11460957178841309,
+      "grad_norm": 0.14998720586299896,
+      "learning_rate": 9.993609447101767e-05,
+      "loss": 0.1456,
+      "step": 546
+    },
+    {
+      "epoch": 0.11481947942905121,
+      "grad_norm": 0.21649526059627533,
+      "learning_rate": 9.993422912328054e-05,
+      "loss": 0.1656,
+      "step": 547
+    },
+    {
+      "epoch": 0.11502938706968933,
+      "grad_norm": 0.19370022416114807,
+      "learning_rate": 9.993233695972175e-05,
+      "loss": 0.173,
+      "step": 548
+    },
+    {
+      "epoch": 0.11523929471032746,
+      "grad_norm": 0.20101432502269745,
+      "learning_rate": 9.993041798135745e-05,
+      "loss": 0.1914,
+      "step": 549
+    },
+    {
+      "epoch": 0.11544920235096558,
+      "grad_norm": 0.18047718703746796,
+      "learning_rate": 9.992847218921816e-05,
+      "loss": 0.2084,
+      "step": 550
+    },
+    {
+      "epoch": 0.1156591099916037,
+      "grad_norm": 0.18755191564559937,
+      "learning_rate": 9.99264995843488e-05,
+      "loss": 0.1646,
+      "step": 551
+    },
+    {
+      "epoch": 0.11586901763224182,
+      "grad_norm": 0.2605830430984497,
+      "learning_rate": 9.992450016780876e-05,
+      "loss": 0.1767,
+      "step": 552
+    },
+    {
+      "epoch": 0.11607892527287993,
+      "grad_norm": 0.1701487898826599,
+      "learning_rate": 9.99224739406717e-05,
+      "loss": 0.173,
+      "step": 553
+    },
+    {
+      "epoch": 0.11628883291351805,
+      "grad_norm": 0.19497598707675934,
+      "learning_rate": 9.99204209040258e-05,
+      "loss": 0.1857,
+      "step": 554
+    },
+    {
+      "epoch": 0.11649874055415617,
+      "grad_norm": 0.16379636526107788,
+      "learning_rate": 9.991834105897356e-05,
+      "loss": 0.154,
+      "step": 555
+    },
+    {
+      "epoch": 0.11670864819479429,
+      "grad_norm": 0.18630164861679077,
+      "learning_rate": 9.991623440663192e-05,
+      "loss": 0.1871,
+      "step": 556
+    },
+    {
+      "epoch": 0.11691855583543241,
+      "grad_norm": 0.219542995095253,
+      "learning_rate": 9.991410094813221e-05,
+      "loss": 0.1877,
+      "step": 557
+    },
+    {
+      "epoch": 0.11712846347607053,
+      "grad_norm": 0.1855912059545517,
+      "learning_rate": 9.991194068462011e-05,
+      "loss": 0.1771,
+      "step": 558
+    },
+    {
+      "epoch": 0.11733837111670865,
+      "grad_norm": 0.15919265151023865,
+      "learning_rate": 9.990975361725577e-05,
+      "loss": 0.1684,
+      "step": 559
+    },
+    {
+      "epoch": 0.11754827875734676,
+      "grad_norm": 0.14570386707782745,
+      "learning_rate": 9.990753974721366e-05,
+      "loss": 0.1656,
+      "step": 560
+    },
+    {
+      "epoch": 0.11775818639798488,
+      "grad_norm": 0.14153516292572021,
+      "learning_rate": 9.990529907568272e-05,
+      "loss": 0.1642,
+      "step": 561
+    },
+    {
+      "epoch": 0.117968094038623,
+      "grad_norm": 0.14015786349773407,
+      "learning_rate": 9.99030316038662e-05,
+      "loss": 0.1704,
+      "step": 562
+    },
+    {
+      "epoch": 0.11817800167926112,
+      "grad_norm": 0.2615254819393158,
+      "learning_rate": 9.99007373329818e-05,
+      "loss": 0.1601,
+      "step": 563
+    },
+    {
+      "epoch": 0.11838790931989925,
+      "grad_norm": 0.2077956199645996,
+      "learning_rate": 9.989841626426162e-05,
+      "loss": 0.1715,
+      "step": 564
+    },
+    {
+      "epoch": 0.11859781696053737,
+      "grad_norm": 0.1743435561656952,
+      "learning_rate": 9.989606839895208e-05,
+      "loss": 0.1725,
+      "step": 565
+    },
+    {
+      "epoch": 0.11880772460117549,
+      "grad_norm": 0.14333437383174896,
+      "learning_rate": 9.989369373831407e-05,
+      "loss": 0.1731,
+      "step": 566
+    },
+    {
+      "epoch": 0.1190176322418136,
+      "grad_norm": 0.16472546756267548,
+      "learning_rate": 9.989129228362284e-05,
+      "loss": 0.1641,
+      "step": 567
+    },
+    {
+      "epoch": 0.11922753988245172,
+      "grad_norm": 0.19595298171043396,
+      "learning_rate": 9.988886403616802e-05,
+      "loss": 0.1789,
+      "step": 568
+    },
+    {
+      "epoch": 0.11943744752308984,
+      "grad_norm": 0.15337203443050385,
+      "learning_rate": 9.988640899725361e-05,
+      "loss": 0.166,
+      "step": 569
+    },
+    {
+      "epoch": 0.11964735516372796,
+      "grad_norm": 0.13052469491958618,
+      "learning_rate": 9.988392716819806e-05,
+      "loss": 0.1762,
+      "step": 570
+    },
+    {
+      "epoch": 0.11985726280436608,
+      "grad_norm": 0.17160564661026,
+      "learning_rate": 9.988141855033415e-05,
+      "loss": 0.1742,
+      "step": 571
+    },
+    {
+      "epoch": 0.1200671704450042,
+      "grad_norm": 0.15600250661373138,
+      "learning_rate": 9.987888314500906e-05,
+      "loss": 0.1819,
+      "step": 572
+    },
+    {
+      "epoch": 0.12027707808564232,
+      "grad_norm": 0.17436926066875458,
+      "learning_rate": 9.987632095358437e-05,
+      "loss": 0.154,
+      "step": 573
+    },
+    {
+      "epoch": 0.12048698572628044,
+      "grad_norm": 0.1583249568939209,
+      "learning_rate": 9.987373197743603e-05,
+      "loss": 0.1812,
+      "step": 574
+    },
+    {
+      "epoch": 0.12069689336691855,
+      "grad_norm": 0.14877773821353912,
+      "learning_rate": 9.987111621795437e-05,
+      "loss": 0.1591,
+      "step": 575
+    },
+    {
+      "epoch": 0.12090680100755667,
+      "grad_norm": 0.13348308205604553,
+      "learning_rate": 9.986847367654414e-05,
+      "loss": 0.164,
+      "step": 576
+    },
+    {
+      "epoch": 0.12111670864819479,
+      "grad_norm": 0.16104738414287567,
+      "learning_rate": 9.986580435462443e-05,
+      "loss": 0.1568,
+      "step": 577
+    },
+    {
+      "epoch": 0.12132661628883291,
+      "grad_norm": 0.1393367350101471,
+      "learning_rate": 9.98631082536287e-05,
+      "loss": 0.1504,
+      "step": 578
+    },
+    {
+      "epoch": 0.12153652392947104,
+      "grad_norm": 0.17166094481945038,
+      "learning_rate": 9.986038537500488e-05,
+      "loss": 0.1746,
+      "step": 579
+    },
+    {
+      "epoch": 0.12174643157010916,
+      "grad_norm": 0.19518902897834778,
+      "learning_rate": 9.985763572021516e-05,
+      "loss": 0.1961,
+      "step": 580
+    },
+    {
+      "epoch": 0.12195633921074728,
+      "grad_norm": 0.17867891490459442,
+      "learning_rate": 9.985485929073619e-05,
+      "loss": 0.1836,
+      "step": 581
+    },
+    {
+      "epoch": 0.12216624685138538,
+      "grad_norm": 0.2270008772611618,
+      "learning_rate": 9.9852056088059e-05,
+      "loss": 0.1847,
+      "step": 582
+    },
+    {
+      "epoch": 0.1223761544920235,
+      "grad_norm": 0.16426697373390198,
+      "learning_rate": 9.984922611368892e-05,
+      "loss": 0.1587,
+      "step": 583
+    },
+    {
+      "epoch": 0.12258606213266163,
+      "grad_norm": 0.16188107430934906,
+      "learning_rate": 9.984636936914575e-05,
+      "loss": 0.1783,
+      "step": 584
+    },
+    {
+      "epoch": 0.12279596977329975,
+      "grad_norm": 0.16157647967338562,
+      "learning_rate": 9.984348585596361e-05,
+      "loss": 0.1612,
+      "step": 585
+    },
+    {
+      "epoch": 0.12300587741393787,
+      "grad_norm": 0.17590083181858063,
+      "learning_rate": 9.984057557569104e-05,
+      "loss": 0.1852,
+      "step": 586
+    },
+    {
+      "epoch": 0.12321578505457599,
+      "grad_norm": 0.13868169486522675,
+      "learning_rate": 9.983763852989088e-05,
+      "loss": 0.1915,
+      "step": 587
+    },
+    {
+      "epoch": 0.12342569269521411,
+      "grad_norm": 0.15027277171611786,
+      "learning_rate": 9.983467472014043e-05,
+      "loss": 0.1924,
+      "step": 588
+    },
+    {
+      "epoch": 0.12363560033585222,
+      "grad_norm": 0.16492144763469696,
+      "learning_rate": 9.983168414803132e-05,
+      "loss": 0.1768,
+      "step": 589
+    },
+    {
+      "epoch": 0.12384550797649034,
+      "grad_norm": 0.16169002652168274,
+      "learning_rate": 9.982866681516954e-05,
+      "loss": 0.1704,
+      "step": 590
+    },
+    {
+      "epoch": 0.12405541561712846,
+      "grad_norm": 0.16292813420295715,
+      "learning_rate": 9.982562272317546e-05,
+      "loss": 0.1687,
+      "step": 591
+    },
+    {
+      "epoch": 0.12426532325776658,
+      "grad_norm": 0.16432535648345947,
+      "learning_rate": 9.982255187368386e-05,
+      "loss": 0.1702,
+      "step": 592
+    },
+    {
+      "epoch": 0.1244752308984047,
+      "grad_norm": 0.13967812061309814,
+      "learning_rate": 9.981945426834382e-05,
+      "loss": 0.1858,
+      "step": 593
+    },
+    {
+      "epoch": 0.12468513853904283,
+      "grad_norm": 0.16663575172424316,
+      "learning_rate": 9.981632990881885e-05,
+      "loss": 0.158,
+      "step": 594
+    },
+    {
+      "epoch": 0.12489504617968095,
+      "grad_norm": 0.17746829986572266,
+      "learning_rate": 9.981317879678679e-05,
+      "loss": 0.1679,
+      "step": 595
+    },
+    {
+      "epoch": 0.12510495382031905,
+      "grad_norm": 0.15006081759929657,
+      "learning_rate": 9.981000093393986e-05,
+      "loss": 0.1659,
+      "step": 596
+    },
+    {
+      "epoch": 0.1253148614609572,
+      "grad_norm": 0.15521980822086334,
+      "learning_rate": 9.980679632198466e-05,
+      "loss": 0.1587,
+      "step": 597
+    },
+    {
+      "epoch": 0.1255247691015953,
+      "grad_norm": 0.17882917821407318,
+      "learning_rate": 9.980356496264212e-05,
+      "loss": 0.1571,
+      "step": 598
+    },
+    {
+      "epoch": 0.12573467674223343,
+      "grad_norm": 0.17295823991298676,
+      "learning_rate": 9.980030685764754e-05,
+      "loss": 0.1936,
+      "step": 599
+    },
+    {
+      "epoch": 0.12594458438287154,
+      "grad_norm": 0.1699322760105133,
+      "learning_rate": 9.979702200875065e-05,
+      "loss": 0.1555,
+      "step": 600
+    },
+    {
+      "epoch": 0.12615449202350965,
+      "grad_norm": 0.15549197793006897,
+      "learning_rate": 9.979371041771543e-05,
+      "loss": 0.1569,
+      "step": 601
+    },
+    {
+      "epoch": 0.12636439966414778,
+      "grad_norm": 0.16207100450992584,
+      "learning_rate": 9.979037208632034e-05,
+      "loss": 0.1609,
+      "step": 602
+    },
+    {
+      "epoch": 0.1265743073047859,
+      "grad_norm": 0.156686931848526,
+      "learning_rate": 9.978700701635807e-05,
+      "loss": 0.165,
+      "step": 603
+    },
+    {
+      "epoch": 0.12678421494542402,
+      "grad_norm": 0.16895289719104767,
+      "learning_rate": 9.97836152096358e-05,
+      "loss": 0.1623,
+      "step": 604
+    },
+    {
+      "epoch": 0.12699412258606213,
+      "grad_norm": 0.14077980816364288,
+      "learning_rate": 9.978019666797498e-05,
+      "loss": 0.1573,
+      "step": 605
+    },
+    {
+      "epoch": 0.12720403022670027,
+      "grad_norm": 0.14448733627796173,
+      "learning_rate": 9.977675139321146e-05,
+      "loss": 0.1671,
+      "step": 606
+    },
+    {
+      "epoch": 0.12741393786733837,
+      "grad_norm": 0.1879337728023529,
+      "learning_rate": 9.977327938719541e-05,
+      "loss": 0.1644,
+      "step": 607
+    },
+    {
+      "epoch": 0.12762384550797648,
+      "grad_norm": 0.22603319585323334,
+      "learning_rate": 9.976978065179138e-05,
+      "loss": 0.1817,
+      "step": 608
+    },
+    {
+      "epoch": 0.12783375314861462,
+      "grad_norm": 0.19939488172531128,
+      "learning_rate": 9.976625518887828e-05,
+      "loss": 0.1925,
+      "step": 609
+    },
+    {
+      "epoch": 0.12804366078925272,
+      "grad_norm": 0.17629611492156982,
+      "learning_rate": 9.976270300034936e-05,
+      "loss": 0.1603,
+      "step": 610
+    },
+    {
+      "epoch": 0.12825356842989086,
+      "grad_norm": 0.17778918147087097,
+      "learning_rate": 9.975912408811223e-05,
+      "loss": 0.1849,
+      "step": 611
+    },
+    {
+      "epoch": 0.12846347607052896,
+      "grad_norm": 0.2002590447664261,
+      "learning_rate": 9.975551845408886e-05,
+      "loss": 0.166,
+      "step": 612
+    },
+    {
+      "epoch": 0.1286733837111671,
+      "grad_norm": 0.20324808359146118,
+      "learning_rate": 9.975188610021553e-05,
+      "loss": 0.1673,
+      "step": 613
+    },
+    {
+      "epoch": 0.1288832913518052,
+      "grad_norm": 0.20010024309158325,
+      "learning_rate": 9.974822702844291e-05,
+      "loss": 0.1529,
+      "step": 614
+    },
+    {
+      "epoch": 0.12909319899244331,
+      "grad_norm": 0.18944194912910461,
+      "learning_rate": 9.974454124073603e-05,
+      "loss": 0.1652,
+      "step": 615
+    },
+    {
+      "epoch": 0.12930310663308145,
+      "grad_norm": 0.16956883668899536,
+      "learning_rate": 9.974082873907418e-05,
+      "loss": 0.1812,
+      "step": 616
+    },
+    {
+      "epoch": 0.12951301427371956,
+      "grad_norm": 0.21157757937908173,
+      "learning_rate": 9.973708952545111e-05,
+      "loss": 0.1862,
+      "step": 617
+    },
+    {
+      "epoch": 0.1297229219143577,
+      "grad_norm": 0.20554983615875244,
+      "learning_rate": 9.973332360187486e-05,
+      "loss": 0.1705,
+      "step": 618
+    },
+    {
+      "epoch": 0.1299328295549958,
+      "grad_norm": 0.21964187920093536,
+      "learning_rate": 9.97295309703678e-05,
+      "loss": 0.1807,
+      "step": 619
+    },
+    {
+      "epoch": 0.13014273719563393,
+      "grad_norm": 0.15438182651996613,
+      "learning_rate": 9.972571163296666e-05,
+      "loss": 0.1951,
+      "step": 620
+    },
+    {
+      "epoch": 0.13035264483627204,
+      "grad_norm": 0.17805755138397217,
+      "learning_rate": 9.972186559172253e-05,
+      "loss": 0.1543,
+      "step": 621
+    },
+    {
+      "epoch": 0.13056255247691015,
+      "grad_norm": 0.1914507895708084,
+      "learning_rate": 9.97179928487008e-05,
+      "loss": 0.1815,
+      "step": 622
+    },
+    {
+      "epoch": 0.13077246011754828,
+      "grad_norm": 0.18952740728855133,
+      "learning_rate": 9.971409340598123e-05,
+      "loss": 0.1601,
+      "step": 623
+    },
+    {
+      "epoch": 0.1309823677581864,
+      "grad_norm": 0.1550108641386032,
+      "learning_rate": 9.971016726565791e-05,
+      "loss": 0.1726,
+      "step": 624
+    },
+    {
+      "epoch": 0.13119227539882453,
+      "grad_norm": 0.15205375850200653,
+      "learning_rate": 9.970621442983929e-05,
+      "loss": 0.1536,
+      "step": 625
+    },
+    {
+      "epoch": 0.13140218303946263,
+      "grad_norm": 0.15574775636196136,
+      "learning_rate": 9.970223490064809e-05,
+      "loss": 0.167,
+      "step": 626
+    },
+    {
+      "epoch": 0.13161209068010077,
+      "grad_norm": 0.18050767481327057,
+      "learning_rate": 9.969822868022143e-05,
+      "loss": 0.1752,
+      "step": 627
+    },
+    {
+      "epoch": 0.13182199832073888,
+      "grad_norm": 0.1507614701986313,
+      "learning_rate": 9.969419577071076e-05,
+      "loss": 0.1712,
+      "step": 628
+    },
+    {
+      "epoch": 0.13203190596137698,
+      "grad_norm": 0.20433206856250763,
+      "learning_rate": 9.96901361742818e-05,
+      "loss": 0.1782,
+      "step": 629
+    },
+    {
+      "epoch": 0.13224181360201512,
+      "grad_norm": 0.14705337584018707,
+      "learning_rate": 9.968604989311467e-05,
+      "loss": 0.1554,
+      "step": 630
+    },
+    {
+      "epoch": 0.13245172124265323,
+      "grad_norm": 0.1573822945356369,
+      "learning_rate": 9.968193692940382e-05,
+      "loss": 0.1762,
+      "step": 631
+    },
+    {
+      "epoch": 0.13266162888329136,
+      "grad_norm": 0.17816203832626343,
+      "learning_rate": 9.967779728535797e-05,
+      "loss": 0.1783,
+      "step": 632
+    },
+    {
+      "epoch": 0.13287153652392947,
+      "grad_norm": 0.1705409288406372,
+      "learning_rate": 9.967363096320022e-05,
+      "loss": 0.1734,
+      "step": 633
+    },
+    {
+      "epoch": 0.1330814441645676,
+      "grad_norm": 0.14574241638183594,
+      "learning_rate": 9.966943796516798e-05,
+      "loss": 0.1599,
+      "step": 634
+    },
+    {
+      "epoch": 0.1332913518052057,
+      "grad_norm": 0.1341760754585266,
+      "learning_rate": 9.966521829351297e-05,
+      "loss": 0.1833,
+      "step": 635
+    },
+    {
+      "epoch": 0.13350125944584382,
+      "grad_norm": 0.1680106520652771,
+      "learning_rate": 9.966097195050128e-05,
+      "loss": 0.1701,
+      "step": 636
+    },
+    {
+      "epoch": 0.13371116708648195,
+      "grad_norm": 0.1687798947095871,
+      "learning_rate": 9.965669893841326e-05,
+      "loss": 0.1669,
+      "step": 637
+    },
+    {
+      "epoch": 0.13392107472712006,
+      "grad_norm": 0.1792893260717392,
+      "learning_rate": 9.965239925954364e-05,
+      "loss": 0.1705,
+      "step": 638
+    },
+    {
+      "epoch": 0.1341309823677582,
+      "grad_norm": 0.15732963383197784,
+      "learning_rate": 9.964807291620144e-05,
+      "loss": 0.176,
+      "step": 639
+    },
+    {
+      "epoch": 0.1343408900083963,
+      "grad_norm": 0.14176061749458313,
+      "learning_rate": 9.964371991070999e-05,
+      "loss": 0.1755,
+      "step": 640
+    },
+    {
+      "epoch": 0.13455079764903444,
+      "grad_norm": 0.1526860147714615,
+      "learning_rate": 9.963934024540698e-05,
+      "loss": 0.1589,
+      "step": 641
+    },
+    {
+      "epoch": 0.13476070528967254,
+      "grad_norm": 0.15608391165733337,
+      "learning_rate": 9.963493392264435e-05,
+      "loss": 0.1652,
+      "step": 642
+    },
+    {
+      "epoch": 0.13497061293031065,
+      "grad_norm": 0.14473848044872284,
+      "learning_rate": 9.963050094478845e-05,
+      "loss": 0.1742,
+      "step": 643
+    },
+    {
+      "epoch": 0.1351805205709488,
+      "grad_norm": 0.1382717341184616,
+      "learning_rate": 9.962604131421984e-05,
+      "loss": 0.1713,
+      "step": 644
+    },
+    {
+      "epoch": 0.1353904282115869,
+      "grad_norm": 0.14345118403434753,
+      "learning_rate": 9.962155503333348e-05,
+      "loss": 0.1648,
+      "step": 645
+    },
+    {
+      "epoch": 0.13560033585222503,
+      "grad_norm": 0.1398243010044098,
+      "learning_rate": 9.96170421045386e-05,
+      "loss": 0.1889,
+      "step": 646
+    },
+    {
+      "epoch": 0.13581024349286314,
+      "grad_norm": 0.14956693351268768,
+      "learning_rate": 9.96125025302587e-05,
+      "loss": 0.1744,
+      "step": 647
+    },
+    {
+      "epoch": 0.13602015113350127,
+      "grad_norm": 0.15064063668251038,
+      "learning_rate": 9.96079363129317e-05,
+      "loss": 0.1625,
+      "step": 648
+    },
+    {
+      "epoch": 0.13623005877413938,
+      "grad_norm": 0.15413698554039001,
+      "learning_rate": 9.960334345500974e-05,
+      "loss": 0.1745,
+      "step": 649
+    },
+    {
+      "epoch": 0.13643996641477749,
+      "grad_norm": 0.18636789917945862,
+      "learning_rate": 9.959872395895929e-05,
+      "loss": 0.1639,
+      "step": 650
+    },
+    {
+      "epoch": 0.13664987405541562,
+      "grad_norm": 0.2444021999835968,
+      "learning_rate": 9.959407782726108e-05,
+      "loss": 0.1826,
+      "step": 651
+    },
+    {
+      "epoch": 0.13685978169605373,
+      "grad_norm": 0.16845685243606567,
+      "learning_rate": 9.958940506241026e-05,
+      "loss": 0.1793,
+      "step": 652
+    },
+    {
+      "epoch": 0.13706968933669186,
+      "grad_norm": 0.16233941912651062,
+      "learning_rate": 9.958470566691618e-05,
+      "loss": 0.18,
+      "step": 653
+    },
+    {
+      "epoch": 0.13727959697732997,
+      "grad_norm": 0.18665330111980438,
+      "learning_rate": 9.95799796433025e-05,
+      "loss": 0.1662,
+      "step": 654
+    },
+    {
+      "epoch": 0.1374895046179681,
+      "grad_norm": 0.18732531368732452,
+      "learning_rate": 9.957522699410723e-05,
+      "loss": 0.1646,
+      "step": 655
+    },
+    {
+      "epoch": 0.1376994122586062,
+      "grad_norm": 0.12580807507038116,
+      "learning_rate": 9.957044772188266e-05,
+      "loss": 0.1616,
+      "step": 656
+    },
+    {
+      "epoch": 0.13790931989924432,
+      "grad_norm": 0.17410576343536377,
+      "learning_rate": 9.956564182919535e-05,
+      "loss": 0.1628,
+      "step": 657
+    },
+    {
+      "epoch": 0.13811922753988246,
+      "grad_norm": 0.1753510683774948,
+      "learning_rate": 9.95608093186262e-05,
+      "loss": 0.1867,
+      "step": 658
+    },
+    {
+      "epoch": 0.13832913518052056,
+      "grad_norm": 0.17664988338947296,
+      "learning_rate": 9.955595019277032e-05,
+      "loss": 0.1819,
+      "step": 659
+    },
+    {
+      "epoch": 0.1385390428211587,
+      "grad_norm": 0.1787986010313034,
+      "learning_rate": 9.955106445423722e-05,
+      "loss": 0.1668,
+      "step": 660
+    },
+    {
+      "epoch": 0.1387489504617968,
+      "grad_norm": 0.14035004377365112,
+      "learning_rate": 9.954615210565065e-05,
+      "loss": 0.1651,
+      "step": 661
+    },
+    {
+      "epoch": 0.13895885810243494,
+      "grad_norm": 0.20008955895900726,
+      "learning_rate": 9.954121314964864e-05,
+      "loss": 0.179,
+      "step": 662
+    },
+    {
+      "epoch": 0.13916876574307305,
+      "grad_norm": 0.22378800809383392,
+      "learning_rate": 9.953624758888352e-05,
+      "loss": 0.1729,
+      "step": 663
+    },
+    {
+      "epoch": 0.13937867338371115,
+      "grad_norm": 0.18687045574188232,
+      "learning_rate": 9.953125542602193e-05,
+      "loss": 0.1756,
+      "step": 664
+    },
+    {
+      "epoch": 0.1395885810243493,
+      "grad_norm": 0.1506877988576889,
+      "learning_rate": 9.952623666374475e-05,
+      "loss": 0.1616,
+      "step": 665
+    },
+    {
+      "epoch": 0.1397984886649874,
+      "grad_norm": 0.20918136835098267,
+      "learning_rate": 9.95211913047472e-05,
+      "loss": 0.1888,
+      "step": 666
+    },
+    {
+      "epoch": 0.14000839630562553,
+      "grad_norm": 0.17521612346172333,
+      "learning_rate": 9.951611935173872e-05,
+      "loss": 0.1886,
+      "step": 667
+    },
+    {
+      "epoch": 0.14021830394626364,
+      "grad_norm": 0.14788690209388733,
+      "learning_rate": 9.951102080744308e-05,
+      "loss": 0.162,
+      "step": 668
+    },
+    {
+      "epoch": 0.14042821158690177,
+      "grad_norm": 0.18560314178466797,
+      "learning_rate": 9.950589567459832e-05,
+      "loss": 0.1573,
+      "step": 669
+    },
+    {
+      "epoch": 0.14063811922753988,
+      "grad_norm": 0.16913674771785736,
+      "learning_rate": 9.950074395595675e-05,
+      "loss": 0.1713,
+      "step": 670
+    },
+    {
+      "epoch": 0.140848026868178,
+      "grad_norm": 0.13768184185028076,
+      "learning_rate": 9.949556565428496e-05,
+      "loss": 0.1733,
+      "step": 671
+    },
+    {
+      "epoch": 0.14105793450881612,
+      "grad_norm": 0.19472239911556244,
+      "learning_rate": 9.949036077236382e-05,
+      "loss": 0.1638,
+      "step": 672
+    },
+    {
+      "epoch": 0.14126784214945423,
+      "grad_norm": 0.17684867978096008,
+      "learning_rate": 9.948512931298846e-05,
+      "loss": 0.1686,
+      "step": 673
+    },
+    {
+      "epoch": 0.14147774979009237,
+      "grad_norm": 0.20061515271663666,
+      "learning_rate": 9.94798712789683e-05,
+      "loss": 0.1921,
+      "step": 674
+    },
+    {
+      "epoch": 0.14168765743073047,
+      "grad_norm": 0.17213481664657593,
+      "learning_rate": 9.9474586673127e-05,
+      "loss": 0.1636,
+      "step": 675
+    },
+    {
+      "epoch": 0.1418975650713686,
+      "grad_norm": 0.16144217550754547,
+      "learning_rate": 9.946927549830258e-05,
+      "loss": 0.1594,
+      "step": 676
+    },
+    {
+      "epoch": 0.14210747271200672,
+      "grad_norm": 0.16045495867729187,
+      "learning_rate": 9.946393775734719e-05,
+      "loss": 0.1585,
+      "step": 677
+    },
+    {
+      "epoch": 0.14231738035264482,
+      "grad_norm": 0.168419748544693,
+      "learning_rate": 9.945857345312735e-05,
+      "loss": 0.1618,
+      "step": 678
+    },
+    {
+      "epoch": 0.14252728799328296,
+      "grad_norm": 0.16631141304969788,
+      "learning_rate": 9.945318258852383e-05,
+      "loss": 0.1648,
+      "step": 679
+    },
+    {
+      "epoch": 0.14273719563392107,
+      "grad_norm": 0.17133933305740356,
+      "learning_rate": 9.944776516643161e-05,
+      "loss": 0.1902,
+      "step": 680
+    },
+    {
+      "epoch": 0.1429471032745592,
+      "grad_norm": 0.144994854927063,
+      "learning_rate": 9.944232118976e-05,
+      "loss": 0.1645,
+      "step": 681
+    },
+    {
+      "epoch": 0.1431570109151973,
+      "grad_norm": 0.13521502912044525,
+      "learning_rate": 9.943685066143252e-05,
+      "loss": 0.1679,
+      "step": 682
+    },
+    {
+      "epoch": 0.14336691855583544,
+      "grad_norm": 0.1505574733018875,
+      "learning_rate": 9.943135358438698e-05,
+      "loss": 0.1497,
+      "step": 683
+    },
+    {
+      "epoch": 0.14357682619647355,
+      "grad_norm": 0.1701841652393341,
+      "learning_rate": 9.942582996157544e-05,
+      "loss": 0.141,
+      "step": 684
+    },
+    {
+      "epoch": 0.14378673383711166,
+      "grad_norm": 0.16892337799072266,
+      "learning_rate": 9.94202797959642e-05,
+      "loss": 0.1845,
+      "step": 685
+    },
+    {
+      "epoch": 0.1439966414777498,
+      "grad_norm": 0.1322741061449051,
+      "learning_rate": 9.941470309053384e-05,
+      "loss": 0.1635,
+      "step": 686
+    },
+    {
+      "epoch": 0.1442065491183879,
+      "grad_norm": 0.18180270493030548,
+      "learning_rate": 9.940909984827915e-05,
+      "loss": 0.1521,
+      "step": 687
+    },
+    {
+      "epoch": 0.14441645675902604,
+      "grad_norm": 0.17136745154857635,
+      "learning_rate": 9.940347007220924e-05,
+      "loss": 0.1694,
+      "step": 688
+    },
+    {
+      "epoch": 0.14462636439966414,
+      "grad_norm": 0.16122983396053314,
+      "learning_rate": 9.93978137653474e-05,
+      "loss": 0.1726,
+      "step": 689
+    },
+    {
+      "epoch": 0.14483627204030228,
+      "grad_norm": 0.15024663507938385,
+      "learning_rate": 9.939213093073118e-05,
+      "loss": 0.1703,
+      "step": 690
+    },
+    {
+      "epoch": 0.14504617968094038,
+      "grad_norm": 0.14193399250507355,
+      "learning_rate": 9.938642157141245e-05,
+      "loss": 0.1837,
+      "step": 691
+    },
+    {
+      "epoch": 0.1452560873215785,
+      "grad_norm": 0.15650463104248047,
+      "learning_rate": 9.938068569045721e-05,
+      "loss": 0.1665,
+      "step": 692
+    },
+    {
+      "epoch": 0.14546599496221663,
+      "grad_norm": 0.14254000782966614,
+      "learning_rate": 9.937492329094577e-05,
+      "loss": 0.175,
+      "step": 693
+    },
+    {
+      "epoch": 0.14567590260285473,
+      "grad_norm": 0.17051447927951813,
+      "learning_rate": 9.93691343759727e-05,
+      "loss": 0.1664,
+      "step": 694
+    },
+    {
+      "epoch": 0.14588581024349287,
+      "grad_norm": 0.16692955791950226,
+      "learning_rate": 9.936331894864677e-05,
+      "loss": 0.1682,
+      "step": 695
+    },
+    {
+      "epoch": 0.14609571788413098,
+      "grad_norm": 0.20158237218856812,
+      "learning_rate": 9.935747701209096e-05,
+      "loss": 0.1574,
+      "step": 696
+    },
+    {
+      "epoch": 0.1463056255247691,
+      "grad_norm": 0.1351911425590515,
+      "learning_rate": 9.935160856944257e-05,
+      "loss": 0.1618,
+      "step": 697
+    },
+    {
+      "epoch": 0.14651553316540722,
+      "grad_norm": 0.1611570417881012,
+      "learning_rate": 9.934571362385305e-05,
+      "loss": 0.164,
+      "step": 698
+    },
+    {
+      "epoch": 0.14672544080604533,
+      "grad_norm": 0.22555968165397644,
+      "learning_rate": 9.933979217848815e-05,
+      "loss": 0.2044,
+      "step": 699
+    },
+    {
+      "epoch": 0.14693534844668346,
+      "grad_norm": 0.17471735179424286,
+      "learning_rate": 9.93338442365278e-05,
+      "loss": 0.1809,
+      "step": 700
+    },
+    {
+      "epoch": 0.14714525608732157,
+      "grad_norm": 0.12285126745700836,
+      "learning_rate": 9.93278698011662e-05,
+      "loss": 0.1644,
+      "step": 701
+    },
+    {
+      "epoch": 0.1473551637279597,
+      "grad_norm": 0.1364145576953888,
+      "learning_rate": 9.932186887561175e-05,
+      "loss": 0.1655,
+      "step": 702
+    },
+    {
+      "epoch": 0.1475650713685978,
+      "grad_norm": 0.14189580082893372,
+      "learning_rate": 9.931584146308708e-05,
+      "loss": 0.1729,
+      "step": 703
+    },
+    {
+      "epoch": 0.14777497900923595,
+      "grad_norm": 0.2254076600074768,
+      "learning_rate": 9.930978756682905e-05,
+      "loss": 0.1731,
+      "step": 704
+    },
+    {
+      "epoch": 0.14798488664987405,
+      "grad_norm": 0.15398503839969635,
+      "learning_rate": 9.930370719008875e-05,
+      "loss": 0.1813,
+      "step": 705
+    },
+    {
+      "epoch": 0.1481947942905122,
+      "grad_norm": 0.15267014503479004,
+      "learning_rate": 9.929760033613146e-05,
+      "loss": 0.1681,
+      "step": 706
+    },
+    {
+      "epoch": 0.1484047019311503,
+      "grad_norm": 0.1611442267894745,
+      "learning_rate": 9.929146700823671e-05,
+      "loss": 0.1925,
+      "step": 707
+    },
+    {
+      "epoch": 0.1486146095717884,
+      "grad_norm": 0.15641498565673828,
+      "learning_rate": 9.928530720969827e-05,
+      "loss": 0.1733,
+      "step": 708
+    },
+    {
+      "epoch": 0.14882451721242654,
+      "grad_norm": 0.12724818289279938,
+      "learning_rate": 9.927912094382403e-05,
+      "loss": 0.1735,
+      "step": 709
+    },
+    {
+      "epoch": 0.14903442485306465,
+      "grad_norm": 0.17570020258426666,
+      "learning_rate": 9.92729082139362e-05,
+      "loss": 0.1732,
+      "step": 710
+    },
+    {
+      "epoch": 0.14924433249370278,
+      "grad_norm": 0.19477427005767822,
+      "learning_rate": 9.926666902337115e-05,
+      "loss": 0.1551,
+      "step": 711
+    },
+    {
+      "epoch": 0.1494542401343409,
+      "grad_norm": 0.16798420250415802,
+      "learning_rate": 9.926040337547946e-05,
+      "loss": 0.1906,
+      "step": 712
+    },
+    {
+      "epoch": 0.14966414777497902,
+      "grad_norm": 0.17551501095294952,
+      "learning_rate": 9.925411127362594e-05,
+      "loss": 0.1747,
+      "step": 713
+    },
+    {
+      "epoch": 0.14987405541561713,
+      "grad_norm": 0.1842852234840393,
+      "learning_rate": 9.924779272118957e-05,
+      "loss": 0.1563,
+      "step": 714
+    },
+    {
+      "epoch": 0.15008396305625524,
+      "grad_norm": 0.1805478185415268,
+      "learning_rate": 9.924144772156358e-05,
+      "loss": 0.165,
+      "step": 715
+    },
+    {
+      "epoch": 0.15029387069689337,
+      "grad_norm": 0.21542084217071533,
+      "learning_rate": 9.923507627815536e-05,
+      "loss": 0.1469,
+      "step": 716
+    },
+    {
+      "epoch": 0.15050377833753148,
+      "grad_norm": 0.15649262070655823,
+      "learning_rate": 9.922867839438654e-05,
+      "loss": 0.1706,
+      "step": 717
+    },
+    {
+      "epoch": 0.15071368597816961,
+      "grad_norm": 0.14313052594661713,
+      "learning_rate": 9.92222540736929e-05,
+      "loss": 0.1737,
+      "step": 718
+    },
+    {
+      "epoch": 0.15092359361880772,
+      "grad_norm": 0.13190749287605286,
+      "learning_rate": 9.92158033195245e-05,
+      "loss": 0.1742,
+      "step": 719
+    },
+    {
+      "epoch": 0.15113350125944586,
+      "grad_norm": 0.18187177181243896,
+      "learning_rate": 9.920932613534549e-05,
+      "loss": 0.1748,
+      "step": 720
+    },
+    {
+      "epoch": 0.15134340890008396,
+      "grad_norm": 0.14302362501621246,
+      "learning_rate": 9.920282252463429e-05,
+      "loss": 0.1711,
+      "step": 721
+    },
+    {
+      "epoch": 0.15155331654072207,
+      "grad_norm": 0.16898708045482635,
+      "learning_rate": 9.919629249088347e-05,
+      "loss": 0.1616,
+      "step": 722
+    },
+    {
+      "epoch": 0.1517632241813602,
+      "grad_norm": 0.15528154373168945,
+      "learning_rate": 9.918973603759984e-05,
+      "loss": 0.1652,
+      "step": 723
+    },
+    {
+      "epoch": 0.15197313182199831,
+      "grad_norm": 0.1718195527791977,
+      "learning_rate": 9.918315316830434e-05,
+      "loss": 0.1758,
+      "step": 724
+    },
+    {
+      "epoch": 0.15218303946263645,
+      "grad_norm": 0.15404529869556427,
+      "learning_rate": 9.917654388653211e-05,
+      "loss": 0.162,
+      "step": 725
+    },
+    {
+      "epoch": 0.15239294710327456,
+      "grad_norm": 0.19148094952106476,
+      "learning_rate": 9.916990819583252e-05,
+      "loss": 0.1887,
+      "step": 726
+    },
+    {
+      "epoch": 0.1526028547439127,
+      "grad_norm": 0.21298371255397797,
+      "learning_rate": 9.916324609976906e-05,
+      "loss": 0.1712,
+      "step": 727
+    },
+    {
+      "epoch": 0.1528127623845508,
+      "grad_norm": 0.2041487991809845,
+      "learning_rate": 9.915655760191944e-05,
+      "loss": 0.167,
+      "step": 728
+    },
+    {
+      "epoch": 0.1530226700251889,
+      "grad_norm": 0.17242367565631866,
+      "learning_rate": 9.914984270587552e-05,
+      "loss": 0.163,
+      "step": 729
+    },
+    {
+      "epoch": 0.15323257766582704,
+      "grad_norm": 0.1646365076303482,
+      "learning_rate": 9.914310141524339e-05,
+      "loss": 0.1638,
+      "step": 730
+    },
+    {
+      "epoch": 0.15344248530646515,
+      "grad_norm": 0.1947726458311081,
+      "learning_rate": 9.913633373364324e-05,
+      "loss": 0.1828,
+      "step": 731
+    },
+    {
+      "epoch": 0.15365239294710328,
+      "grad_norm": 0.17962804436683655,
+      "learning_rate": 9.912953966470948e-05,
+      "loss": 0.1699,
+      "step": 732
+    },
+    {
+      "epoch": 0.1538623005877414,
+      "grad_norm": 0.17121249437332153,
+      "learning_rate": 9.912271921209068e-05,
+      "loss": 0.1605,
+      "step": 733
+    },
+    {
+      "epoch": 0.15407220822837953,
+      "grad_norm": 0.17192313075065613,
+      "learning_rate": 9.911587237944959e-05,
+      "loss": 0.1746,
+      "step": 734
+    },
+    {
+      "epoch": 0.15428211586901763,
+      "grad_norm": 0.1262722760438919,
+      "learning_rate": 9.910899917046311e-05,
+      "loss": 0.1677,
+      "step": 735
+    },
+    {
+      "epoch": 0.15449202350965574,
+      "grad_norm": 0.15364903211593628,
+      "learning_rate": 9.910209958882231e-05,
+      "loss": 0.1792,
+      "step": 736
+    },
+    {
+      "epoch": 0.15470193115029388,
+      "grad_norm": 0.17038469016551971,
+      "learning_rate": 9.909517363823241e-05,
+      "loss": 0.1588,
+      "step": 737
+    },
+    {
+      "epoch": 0.15491183879093198,
+      "grad_norm": 0.14676110446453094,
+      "learning_rate": 9.908822132241281e-05,
+      "loss": 0.1639,
+      "step": 738
+    },
+    {
+      "epoch": 0.15512174643157012,
+      "grad_norm": 0.15300148725509644,
+      "learning_rate": 9.908124264509707e-05,
+      "loss": 0.1752,
+      "step": 739
+    },
+    {
+      "epoch": 0.15533165407220823,
+      "grad_norm": 0.1372271180152893,
+      "learning_rate": 9.90742376100329e-05,
+      "loss": 0.1728,
+      "step": 740
+    },
+    {
+      "epoch": 0.15554156171284636,
+      "grad_norm": 0.13885709643363953,
+      "learning_rate": 9.906720622098215e-05,
+      "loss": 0.1625,
+      "step": 741
+    },
+    {
+      "epoch": 0.15575146935348447,
+      "grad_norm": 0.12582066655158997,
+      "learning_rate": 9.906014848172086e-05,
+      "loss": 0.17,
+      "step": 742
+    },
+    {
+      "epoch": 0.15596137699412257,
+      "grad_norm": 0.13245654106140137,
+      "learning_rate": 9.905306439603918e-05,
+      "loss": 0.1711,
+      "step": 743
+    },
+    {
+      "epoch": 0.1561712846347607,
+      "grad_norm": 0.12488420307636261,
+      "learning_rate": 9.904595396774142e-05,
+      "loss": 0.1725,
+      "step": 744
+    },
+    {
+      "epoch": 0.15638119227539882,
+      "grad_norm": 0.1577821522951126,
+      "learning_rate": 9.903881720064606e-05,
+      "loss": 0.1707,
+      "step": 745
+    },
+    {
+      "epoch": 0.15659109991603695,
+      "grad_norm": 0.16098785400390625,
+      "learning_rate": 9.903165409858567e-05,
+      "loss": 0.1608,
+      "step": 746
+    },
+    {
+      "epoch": 0.15680100755667506,
+      "grad_norm": 0.14498507976531982,
+      "learning_rate": 9.902446466540707e-05,
+      "loss": 0.1776,
+      "step": 747
+    },
+    {
+      "epoch": 0.1570109151973132,
+      "grad_norm": 0.15581083297729492,
+      "learning_rate": 9.901724890497109e-05,
+      "loss": 0.1746,
+      "step": 748
+    },
+    {
+      "epoch": 0.1572208228379513,
+      "grad_norm": 0.16081508994102478,
+      "learning_rate": 9.901000682115276e-05,
+      "loss": 0.1869,
+      "step": 749
+    },
+    {
+      "epoch": 0.1574307304785894,
+      "grad_norm": 0.16249604523181915,
+      "learning_rate": 9.900273841784126e-05,
+      "loss": 0.1629,
+      "step": 750
+    },
+    {
+      "epoch": 0.15764063811922754,
+      "grad_norm": 0.1613830178976059,
+      "learning_rate": 9.899544369893992e-05,
+      "loss": 0.1588,
+      "step": 751
+    },
+    {
+      "epoch": 0.15785054575986565,
+      "grad_norm": 0.15741245448589325,
+      "learning_rate": 9.898812266836613e-05,
+      "loss": 0.1708,
+      "step": 752
+    },
+    {
+      "epoch": 0.1580604534005038,
+      "grad_norm": 0.12306373566389084,
+      "learning_rate": 9.898077533005144e-05,
+      "loss": 0.1751,
+      "step": 753
+    },
+    {
+      "epoch": 0.1582703610411419,
+      "grad_norm": 0.15071968734264374,
+      "learning_rate": 9.897340168794155e-05,
+      "loss": 0.1686,
+      "step": 754
+    },
+    {
+      "epoch": 0.15848026868178003,
+      "grad_norm": 0.14102348685264587,
+      "learning_rate": 9.896600174599632e-05,
+      "loss": 0.1701,
+      "step": 755
+    },
+    {
+      "epoch": 0.15869017632241814,
+      "grad_norm": 0.16288180649280548,
+      "learning_rate": 9.895857550818963e-05,
+      "loss": 0.1652,
+      "step": 756
+    },
+    {
+      "epoch": 0.15890008396305624,
+      "grad_norm": 0.17342409491539001,
+      "learning_rate": 9.895112297850956e-05,
+      "loss": 0.1702,
+      "step": 757
+    },
+    {
+      "epoch": 0.15910999160369438,
+      "grad_norm": 0.13485445082187653,
+      "learning_rate": 9.894364416095829e-05,
+      "loss": 0.1705,
+      "step": 758
+    },
+    {
+      "epoch": 0.15931989924433249,
+      "grad_norm": 0.17127734422683716,
+      "learning_rate": 9.893613905955211e-05,
+      "loss": 0.1831,
+      "step": 759
+    },
+    {
+      "epoch": 0.15952980688497062,
+      "grad_norm": 0.13848379254341125,
+      "learning_rate": 9.892860767832144e-05,
+      "loss": 0.161,
+      "step": 760
+    },
+    {
+      "epoch": 0.15973971452560873,
+      "grad_norm": 0.12154096364974976,
+      "learning_rate": 9.892105002131081e-05,
+      "loss": 0.1637,
+      "step": 761
+    },
+    {
+      "epoch": 0.15994962216624686,
+      "grad_norm": 0.15433917939662933,
+      "learning_rate": 9.891346609257882e-05,
+      "loss": 0.1765,
+      "step": 762
+    },
+    {
+      "epoch": 0.16015952980688497,
+      "grad_norm": 0.16946150362491608,
+      "learning_rate": 9.890585589619825e-05,
+      "loss": 0.1766,
+      "step": 763
+    },
+    {
+      "epoch": 0.16036943744752308,
+      "grad_norm": 0.14503143727779388,
+      "learning_rate": 9.889821943625594e-05,
+      "loss": 0.1458,
+      "step": 764
+    },
+    {
+      "epoch": 0.1605793450881612,
+      "grad_norm": 0.16491486132144928,
+      "learning_rate": 9.889055671685283e-05,
+      "loss": 0.1861,
+      "step": 765
+    },
+    {
+      "epoch": 0.16078925272879932,
+      "grad_norm": 0.15087951719760895,
+      "learning_rate": 9.888286774210398e-05,
+      "loss": 0.1745,
+      "step": 766
+    },
+    {
+      "epoch": 0.16099916036943746,
+      "grad_norm": 0.19266510009765625,
+      "learning_rate": 9.887515251613857e-05,
+      "loss": 0.1711,
+      "step": 767
+    },
+    {
+      "epoch": 0.16120906801007556,
+      "grad_norm": 0.19419905543327332,
+      "learning_rate": 9.886741104309981e-05,
+      "loss": 0.1705,
+      "step": 768
+    },
+    {
+      "epoch": 0.1614189756507137,
+      "grad_norm": 0.13408306241035461,
+      "learning_rate": 9.885964332714508e-05,
+      "loss": 0.1629,
+      "step": 769
+    },
+    {
+      "epoch": 0.1616288832913518,
+      "grad_norm": 0.17757900059223175,
+      "learning_rate": 9.885184937244581e-05,
+      "loss": 0.1643,
+      "step": 770
+    },
+    {
+      "epoch": 0.1618387909319899,
+      "grad_norm": 0.16376204788684845,
+      "learning_rate": 9.884402918318754e-05,
+      "loss": 0.1869,
+      "step": 771
+    },
+    {
+      "epoch": 0.16204869857262805,
+      "grad_norm": 0.1362319141626358,
+      "learning_rate": 9.883618276356988e-05,
+      "loss": 0.1736,
+      "step": 772
+    },
+    {
+      "epoch": 0.16225860621326615,
+      "grad_norm": 0.15015824139118195,
+      "learning_rate": 9.882831011780653e-05,
+      "loss": 0.1598,
+      "step": 773
+    },
+    {
+      "epoch": 0.1624685138539043,
+      "grad_norm": 0.1815597116947174,
+      "learning_rate": 9.882041125012528e-05,
+      "loss": 0.1597,
+      "step": 774
+    },
+    {
+      "epoch": 0.1626784214945424,
+      "grad_norm": 0.16493958234786987,
+      "learning_rate": 9.881248616476803e-05,
+      "loss": 0.181,
+      "step": 775
+    },
+    {
+      "epoch": 0.16288832913518053,
+      "grad_norm": 0.1213487982749939,
+      "learning_rate": 9.880453486599072e-05,
+      "loss": 0.1644,
+      "step": 776
+    },
+    {
+      "epoch": 0.16309823677581864,
+      "grad_norm": 0.1440252661705017,
+      "learning_rate": 9.879655735806337e-05,
+      "loss": 0.1657,
+      "step": 777
+    },
+    {
+      "epoch": 0.16330814441645675,
+      "grad_norm": 0.1406947672367096,
+      "learning_rate": 9.878855364527007e-05,
+      "loss": 0.172,
+      "step": 778
+    },
+    {
+      "epoch": 0.16351805205709488,
+      "grad_norm": 0.17073537409305573,
+      "learning_rate": 9.878052373190902e-05,
+      "loss": 0.1852,
+      "step": 779
+    },
+    {
+      "epoch": 0.163727959697733,
+      "grad_norm": 0.15924111008644104,
+      "learning_rate": 9.877246762229247e-05,
+      "loss": 0.18,
+      "step": 780
+    },
+    {
+      "epoch": 0.16393786733837112,
+      "grad_norm": 0.14216449856758118,
+      "learning_rate": 9.876438532074672e-05,
+      "loss": 0.1575,
+      "step": 781
+    },
+    {
+      "epoch": 0.16414777497900923,
+      "grad_norm": 0.12731285393238068,
+      "learning_rate": 9.875627683161217e-05,
+      "loss": 0.161,
+      "step": 782
+    },
+    {
+      "epoch": 0.16435768261964737,
+      "grad_norm": 0.16211947798728943,
+      "learning_rate": 9.874814215924324e-05,
+      "loss": 0.1592,
+      "step": 783
+    },
+    {
+      "epoch": 0.16456759026028547,
+      "grad_norm": 0.1463523507118225,
+      "learning_rate": 9.873998130800844e-05,
+      "loss": 0.1773,
+      "step": 784
+    },
+    {
+      "epoch": 0.16477749790092358,
+      "grad_norm": 0.14601489901542664,
+      "learning_rate": 9.873179428229033e-05,
+      "loss": 0.1685,
+      "step": 785
+    },
+    {
+      "epoch": 0.16498740554156172,
+      "grad_norm": 0.1590069830417633,
+      "learning_rate": 9.872358108648557e-05,
+      "loss": 0.1709,
+      "step": 786
+    },
+    {
+      "epoch": 0.16519731318219982,
+      "grad_norm": 0.14027854800224304,
+      "learning_rate": 9.871534172500479e-05,
+      "loss": 0.1842,
+      "step": 787
+    },
+    {
+      "epoch": 0.16540722082283796,
+      "grad_norm": 0.12902334332466125,
+      "learning_rate": 9.870707620227271e-05,
+      "loss": 0.1484,
+      "step": 788
+    },
+    {
+      "epoch": 0.16561712846347607,
+      "grad_norm": 0.13427670300006866,
+      "learning_rate": 9.869878452272812e-05,
+      "loss": 0.1599,
+      "step": 789
+    },
+    {
+      "epoch": 0.1658270361041142,
+      "grad_norm": 0.17016825079917908,
+      "learning_rate": 9.869046669082386e-05,
+      "loss": 0.1506,
+      "step": 790
+    },
+    {
+      "epoch": 0.1660369437447523,
+      "grad_norm": 0.13757233321666718,
+      "learning_rate": 9.868212271102678e-05,
+      "loss": 0.1574,
+      "step": 791
+    },
+    {
+      "epoch": 0.16624685138539042,
+      "grad_norm": 0.19485127925872803,
+      "learning_rate": 9.867375258781778e-05,
+      "loss": 0.1686,
+      "step": 792
+    },
+    {
+      "epoch": 0.16645675902602855,
+      "grad_norm": 0.16131381690502167,
+      "learning_rate": 9.866535632569182e-05,
+      "loss": 0.1693,
+      "step": 793
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.16035085916519165,
+      "learning_rate": 9.865693392915787e-05,
+      "loss": 0.1589,
+      "step": 794
+    },
+    {
+      "epoch": 0.1668765743073048,
+      "grad_norm": 0.16287830471992493,
+      "learning_rate": 9.864848540273897e-05,
+      "loss": 0.1749,
+      "step": 795
+    },
+    {
+      "epoch": 0.1670864819479429,
+      "grad_norm": 0.16353754699230194,
+      "learning_rate": 9.864001075097214e-05,
+      "loss": 0.1846,
+      "step": 796
+    },
+    {
+      "epoch": 0.16729638958858103,
+      "grad_norm": 0.19826947152614594,
+      "learning_rate": 9.863150997840849e-05,
+      "loss": 0.1624,
+      "step": 797
+    },
+    {
+      "epoch": 0.16750629722921914,
+      "grad_norm": 0.14296779036521912,
+      "learning_rate": 9.862298308961313e-05,
+      "loss": 0.1823,
+      "step": 798
+    },
+    {
+      "epoch": 0.16771620486985725,
+      "grad_norm": 0.1371513456106186,
+      "learning_rate": 9.861443008916517e-05,
+      "loss": 0.1742,
+      "step": 799
+    },
+    {
+      "epoch": 0.16792611251049538,
+      "grad_norm": 0.17043526470661163,
+      "learning_rate": 9.86058509816578e-05,
+      "loss": 0.1661,
+      "step": 800
+    },
+    {
+      "epoch": 0.1681360201511335,
+      "grad_norm": 0.17045053839683533,
+      "learning_rate": 9.859724577169815e-05,
+      "loss": 0.1773,
+      "step": 801
+    },
+    {
+      "epoch": 0.16834592779177163,
+      "grad_norm": 0.1564718633890152,
+      "learning_rate": 9.858861446390748e-05,
+      "loss": 0.1636,
+      "step": 802
+    },
+    {
+      "epoch": 0.16855583543240973,
+      "grad_norm": 0.13335298001766205,
+      "learning_rate": 9.857995706292092e-05,
+      "loss": 0.18,
+      "step": 803
+    },
+    {
+      "epoch": 0.16876574307304787,
+      "grad_norm": 0.14552432298660278,
+      "learning_rate": 9.857127357338775e-05,
+      "loss": 0.1599,
+      "step": 804
+    },
+    {
+      "epoch": 0.16897565071368598,
+      "grad_norm": 0.1437654346227646,
+      "learning_rate": 9.856256399997119e-05,
+      "loss": 0.178,
+      "step": 805
+    },
+    {
+      "epoch": 0.16918555835432408,
+      "grad_norm": 0.143344447016716,
+      "learning_rate": 9.855382834734848e-05,
+      "loss": 0.159,
+      "step": 806
+    },
+    {
+      "epoch": 0.16939546599496222,
+      "grad_norm": 0.15504872798919678,
+      "learning_rate": 9.854506662021085e-05,
+      "loss": 0.1944,
+      "step": 807
+    },
+    {
+      "epoch": 0.16960537363560033,
+      "grad_norm": 0.12894919514656067,
+      "learning_rate": 9.853627882326357e-05,
+      "loss": 0.1677,
+      "step": 808
+    },
+    {
+      "epoch": 0.16981528127623846,
+      "grad_norm": 0.1339186728000641,
+      "learning_rate": 9.852746496122587e-05,
+      "loss": 0.1666,
+      "step": 809
+    },
+    {
+      "epoch": 0.17002518891687657,
+      "grad_norm": 0.11661262065172195,
+      "learning_rate": 9.8518625038831e-05,
+      "loss": 0.1611,
+      "step": 810
+    },
+    {
+      "epoch": 0.1702350965575147,
+      "grad_norm": 0.14091713726520538,
+      "learning_rate": 9.85097590608262e-05,
+      "loss": 0.1483,
+      "step": 811
+    },
+    {
+      "epoch": 0.1704450041981528,
+      "grad_norm": 0.13658110797405243,
+      "learning_rate": 9.85008670319727e-05,
+      "loss": 0.1667,
+      "step": 812
+    },
+    {
+      "epoch": 0.17065491183879095,
+      "grad_norm": 0.13288968801498413,
+      "learning_rate": 9.849194895704575e-05,
+      "loss": 0.1685,
+      "step": 813
+    },
+    {
+      "epoch": 0.17086481947942905,
+      "grad_norm": 0.1870601624250412,
+      "learning_rate": 9.84830048408345e-05,
+      "loss": 0.1794,
+      "step": 814
+    },
+    {
+      "epoch": 0.17107472712006716,
+      "grad_norm": 0.15742088854312897,
+      "learning_rate": 9.84740346881422e-05,
+      "loss": 0.1546,
+      "step": 815
+    },
+    {
+      "epoch": 0.1712846347607053,
+      "grad_norm": 0.1625792235136032,
+      "learning_rate": 9.846503850378602e-05,
+      "loss": 0.1506,
+      "step": 816
+    },
+    {
+      "epoch": 0.1714945424013434,
+      "grad_norm": 0.1315995752811432,
+      "learning_rate": 9.845601629259708e-05,
+      "loss": 0.1611,
+      "step": 817
+    },
+    {
+      "epoch": 0.17170445004198154,
+      "grad_norm": 0.16079425811767578,
+      "learning_rate": 9.844696805942055e-05,
+      "loss": 0.1588,
+      "step": 818
+    },
+    {
+      "epoch": 0.17191435768261965,
+      "grad_norm": 0.16755861043930054,
+      "learning_rate": 9.843789380911554e-05,
+      "loss": 0.1629,
+      "step": 819
+    },
+    {
+      "epoch": 0.17212426532325778,
+      "grad_norm": 0.12923209369182587,
+      "learning_rate": 9.84287935465551e-05,
+      "loss": 0.1675,
+      "step": 820
+    },
+    {
+      "epoch": 0.1723341729638959,
+      "grad_norm": 0.14494512975215912,
+      "learning_rate": 9.84196672766263e-05,
+      "loss": 0.1674,
+      "step": 821
+    },
+    {
+      "epoch": 0.172544080604534,
+      "grad_norm": 0.13111597299575806,
+      "learning_rate": 9.841051500423014e-05,
+      "loss": 0.1835,
+      "step": 822
+    },
+    {
+      "epoch": 0.17275398824517213,
+      "grad_norm": 0.1468263864517212,
+      "learning_rate": 9.840133673428162e-05,
+      "loss": 0.1635,
+      "step": 823
+    },
+    {
+      "epoch": 0.17296389588581024,
+      "grad_norm": 0.14243565499782562,
+      "learning_rate": 9.839213247170967e-05,
+      "loss": 0.1617,
+      "step": 824
+    },
+    {
+      "epoch": 0.17317380352644837,
+      "grad_norm": 0.14563241600990295,
+      "learning_rate": 9.838290222145718e-05,
+      "loss": 0.1614,
+      "step": 825
+    },
+    {
+      "epoch": 0.17338371116708648,
+      "grad_norm": 0.15182636678218842,
+      "learning_rate": 9.837364598848102e-05,
+      "loss": 0.1712,
+      "step": 826
+    },
+    {
+      "epoch": 0.17359361880772461,
+      "grad_norm": 0.15270951390266418,
+      "learning_rate": 9.8364363777752e-05,
+      "loss": 0.1721,
+      "step": 827
+    },
+    {
+      "epoch": 0.17380352644836272,
+      "grad_norm": 0.14734511077404022,
+      "learning_rate": 9.835505559425487e-05,
+      "loss": 0.1751,
+      "step": 828
+    },
+    {
+      "epoch": 0.17401343408900083,
+      "grad_norm": 0.13699984550476074,
+      "learning_rate": 9.834572144298834e-05,
+      "loss": 0.1895,
+      "step": 829
+    },
+    {
+      "epoch": 0.17422334172963896,
+      "grad_norm": 0.1422858089208603,
+      "learning_rate": 9.833636132896505e-05,
+      "loss": 0.1757,
+      "step": 830
+    },
+    {
+      "epoch": 0.17443324937027707,
+      "grad_norm": 0.1300913542509079,
+      "learning_rate": 9.832697525721161e-05,
+      "loss": 0.1774,
+      "step": 831
+    },
+    {
+      "epoch": 0.1746431570109152,
+      "grad_norm": 0.16247855126857758,
+      "learning_rate": 9.831756323276856e-05,
+      "loss": 0.1499,
+      "step": 832
+    },
+    {
+      "epoch": 0.1748530646515533,
+      "grad_norm": 0.13687850534915924,
+      "learning_rate": 9.830812526069036e-05,
+      "loss": 0.1739,
+      "step": 833
+    },
+    {
+      "epoch": 0.17506297229219145,
+      "grad_norm": 0.16242316365242004,
+      "learning_rate": 9.829866134604543e-05,
+      "loss": 0.1703,
+      "step": 834
+    },
+    {
+      "epoch": 0.17527287993282956,
+      "grad_norm": 0.1983039677143097,
+      "learning_rate": 9.82891714939161e-05,
+      "loss": 0.1726,
+      "step": 835
+    },
+    {
+      "epoch": 0.17548278757346766,
+      "grad_norm": 0.17525269091129303,
+      "learning_rate": 9.827965570939861e-05,
+      "loss": 0.1839,
+      "step": 836
+    },
+    {
+      "epoch": 0.1756926952141058,
+      "grad_norm": 0.1517772525548935,
+      "learning_rate": 9.827011399760319e-05,
+      "loss": 0.17,
+      "step": 837
+    },
+    {
+      "epoch": 0.1759026028547439,
+      "grad_norm": 0.17186671495437622,
+      "learning_rate": 9.826054636365396e-05,
+      "loss": 0.1526,
+      "step": 838
+    },
+    {
+      "epoch": 0.17611251049538204,
+      "grad_norm": 0.15066854655742645,
+      "learning_rate": 9.825095281268894e-05,
+      "loss": 0.1442,
+      "step": 839
+    },
+    {
+      "epoch": 0.17632241813602015,
+      "grad_norm": 0.14937272667884827,
+      "learning_rate": 9.82413333498601e-05,
+      "loss": 0.1641,
+      "step": 840
+    },
+    {
+      "epoch": 0.17653232577665828,
+      "grad_norm": 0.1843566745519638,
+      "learning_rate": 9.823168798033328e-05,
+      "loss": 0.1674,
+      "step": 841
+    },
+    {
+      "epoch": 0.1767422334172964,
+      "grad_norm": 0.15555419027805328,
+      "learning_rate": 9.82220167092883e-05,
+      "loss": 0.1551,
+      "step": 842
+    },
+    {
+      "epoch": 0.1769521410579345,
+      "grad_norm": 0.1585703343153,
+      "learning_rate": 9.821231954191885e-05,
+      "loss": 0.1578,
+      "step": 843
+    },
+    {
+      "epoch": 0.17716204869857263,
+      "grad_norm": 0.1662077009677887,
+      "learning_rate": 9.82025964834325e-05,
+      "loss": 0.1778,
+      "step": 844
+    },
+    {
+      "epoch": 0.17737195633921074,
+      "grad_norm": 0.1385970562696457,
+      "learning_rate": 9.819284753905078e-05,
+      "loss": 0.1653,
+      "step": 845
+    },
+    {
+      "epoch": 0.17758186397984888,
+      "grad_norm": 0.14429986476898193,
+      "learning_rate": 9.81830727140091e-05,
+      "loss": 0.1801,
+      "step": 846
+    },
+    {
+      "epoch": 0.17779177162048698,
+      "grad_norm": 0.16140583157539368,
+      "learning_rate": 9.817327201355675e-05,
+      "loss": 0.1808,
+      "step": 847
+    },
+    {
+      "epoch": 0.17800167926112512,
+      "grad_norm": 0.13548845052719116,
+      "learning_rate": 9.816344544295692e-05,
+      "loss": 0.1567,
+      "step": 848
+    },
+    {
+      "epoch": 0.17821158690176322,
+      "grad_norm": 0.14740432798862457,
+      "learning_rate": 9.815359300748674e-05,
+      "loss": 0.1737,
+      "step": 849
+    },
+    {
+      "epoch": 0.17842149454240133,
+      "grad_norm": 0.15687352418899536,
+      "learning_rate": 9.814371471243715e-05,
+      "loss": 0.1775,
+      "step": 850
+    },
+    {
+      "epoch": 0.17863140218303947,
+      "grad_norm": 0.14518170058727264,
+      "learning_rate": 9.813381056311307e-05,
+      "loss": 0.1536,
+      "step": 851
+    },
+    {
+      "epoch": 0.17884130982367757,
+      "grad_norm": 0.1569397896528244,
+      "learning_rate": 9.812388056483319e-05,
+      "loss": 0.1667,
+      "step": 852
+    },
+    {
+      "epoch": 0.1790512174643157,
+      "grad_norm": 0.13371489942073822,
+      "learning_rate": 9.81139247229302e-05,
+      "loss": 0.175,
+      "step": 853
+    },
+    {
+      "epoch": 0.17926112510495382,
+      "grad_norm": 0.15128138661384583,
+      "learning_rate": 9.810394304275058e-05,
+      "loss": 0.1601,
+      "step": 854
+    },
+    {
+      "epoch": 0.17947103274559195,
+      "grad_norm": 0.12941020727157593,
+      "learning_rate": 9.809393552965476e-05,
+      "loss": 0.1579,
+      "step": 855
+    },
+    {
+      "epoch": 0.17968094038623006,
+      "grad_norm": 0.1644691675901413,
+      "learning_rate": 9.808390218901696e-05,
+      "loss": 0.1737,
+      "step": 856
+    },
+    {
+      "epoch": 0.17989084802686817,
+      "grad_norm": 0.16271840035915375,
+      "learning_rate": 9.807384302622533e-05,
+      "loss": 0.1846,
+      "step": 857
+    },
+    {
+      "epoch": 0.1801007556675063,
+      "grad_norm": 0.15787339210510254,
+      "learning_rate": 9.806375804668189e-05,
+      "loss": 0.1733,
+      "step": 858
+    },
+    {
+      "epoch": 0.1803106633081444,
+      "grad_norm": 0.16740775108337402,
+      "learning_rate": 9.805364725580248e-05,
+      "loss": 0.1625,
+      "step": 859
+    },
+    {
+      "epoch": 0.18052057094878254,
+      "grad_norm": 0.11970685422420502,
+      "learning_rate": 9.804351065901682e-05,
+      "loss": 0.1554,
+      "step": 860
+    },
+    {
+      "epoch": 0.18073047858942065,
+      "grad_norm": 0.21359467506408691,
+      "learning_rate": 9.803334826176852e-05,
+      "loss": 0.1658,
+      "step": 861
+    },
+    {
+      "epoch": 0.1809403862300588,
+      "grad_norm": 0.1699121594429016,
+      "learning_rate": 9.8023160069515e-05,
+      "loss": 0.1608,
+      "step": 862
+    },
+    {
+      "epoch": 0.1811502938706969,
+      "grad_norm": 0.15073060989379883,
+      "learning_rate": 9.801294608772755e-05,
+      "loss": 0.156,
+      "step": 863
+    },
+    {
+      "epoch": 0.181360201511335,
+      "grad_norm": 0.17342859506607056,
+      "learning_rate": 9.800270632189133e-05,
+      "loss": 0.1829,
+      "step": 864
+    },
+    {
+      "epoch": 0.18157010915197314,
+      "grad_norm": 0.20463357865810394,
+      "learning_rate": 9.799244077750531e-05,
+      "loss": 0.1839,
+      "step": 865
+    },
+    {
+      "epoch": 0.18178001679261124,
+      "grad_norm": 0.17843082547187805,
+      "learning_rate": 9.798214946008234e-05,
+      "loss": 0.1707,
+      "step": 866
+    },
+    {
+      "epoch": 0.18198992443324938,
+      "grad_norm": 0.14360542595386505,
+      "learning_rate": 9.797183237514907e-05,
+      "loss": 0.1687,
+      "step": 867
+    },
+    {
+      "epoch": 0.18219983207388749,
+      "grad_norm": 0.15701772272586823,
+      "learning_rate": 9.796148952824603e-05,
+      "loss": 0.1584,
+      "step": 868
+    },
+    {
+      "epoch": 0.18240973971452562,
+      "grad_norm": 0.12531954050064087,
+      "learning_rate": 9.795112092492755e-05,
+      "loss": 0.1467,
+      "step": 869
+    },
+    {
+      "epoch": 0.18261964735516373,
+      "grad_norm": 0.16143764555454254,
+      "learning_rate": 9.794072657076182e-05,
+      "loss": 0.1687,
+      "step": 870
+    },
+    {
+      "epoch": 0.18282955499580184,
+      "grad_norm": 0.13665997982025146,
+      "learning_rate": 9.793030647133084e-05,
+      "loss": 0.1644,
+      "step": 871
+    },
+    {
+      "epoch": 0.18303946263643997,
+      "grad_norm": 0.15797339379787445,
+      "learning_rate": 9.791986063223045e-05,
+      "loss": 0.1737,
+      "step": 872
+    },
+    {
+      "epoch": 0.18324937027707808,
+      "grad_norm": 0.13726350665092468,
+      "learning_rate": 9.790938905907027e-05,
+      "loss": 0.1842,
+      "step": 873
+    },
+    {
+      "epoch": 0.1834592779177162,
+      "grad_norm": 0.13706736266613007,
+      "learning_rate": 9.78988917574738e-05,
+      "loss": 0.1699,
+      "step": 874
+    },
+    {
+      "epoch": 0.18366918555835432,
+      "grad_norm": 0.13394543528556824,
+      "learning_rate": 9.788836873307835e-05,
+      "loss": 0.1757,
+      "step": 875
+    },
+    {
+      "epoch": 0.18387909319899245,
+      "grad_norm": 0.13909773528575897,
+      "learning_rate": 9.7877819991535e-05,
+      "loss": 0.1761,
+      "step": 876
+    },
+    {
+      "epoch": 0.18408900083963056,
+      "grad_norm": 0.1416359543800354,
+      "learning_rate": 9.786724553850865e-05,
+      "loss": 0.1657,
+      "step": 877
+    },
+    {
+      "epoch": 0.18429890848026867,
+      "grad_norm": 0.13720481097698212,
+      "learning_rate": 9.785664537967806e-05,
+      "loss": 0.173,
+      "step": 878
+    },
+    {
+      "epoch": 0.1845088161209068,
+      "grad_norm": 0.14136558771133423,
+      "learning_rate": 9.784601952073573e-05,
+      "loss": 0.1625,
+      "step": 879
+    },
+    {
+      "epoch": 0.1847187237615449,
+      "grad_norm": 0.1424713134765625,
+      "learning_rate": 9.783536796738802e-05,
+      "loss": 0.1676,
+      "step": 880
+    },
+    {
+      "epoch": 0.18492863140218305,
+      "grad_norm": 0.163283571600914,
+      "learning_rate": 9.782469072535502e-05,
+      "loss": 0.1574,
+      "step": 881
+    },
+    {
+      "epoch": 0.18513853904282115,
+      "grad_norm": 0.16645042598247528,
+      "learning_rate": 9.781398780037067e-05,
+      "loss": 0.1576,
+      "step": 882
+    },
+    {
+      "epoch": 0.1853484466834593,
+      "grad_norm": 0.17722882330417633,
+      "learning_rate": 9.780325919818268e-05,
+      "loss": 0.1815,
+      "step": 883
+    },
+    {
+      "epoch": 0.1855583543240974,
+      "grad_norm": 0.1435573697090149,
+      "learning_rate": 9.779250492455257e-05,
+      "loss": 0.168,
+      "step": 884
+    },
+    {
+      "epoch": 0.1857682619647355,
+      "grad_norm": 0.13388660550117493,
+      "learning_rate": 9.778172498525559e-05,
+      "loss": 0.1493,
+      "step": 885
+    },
+    {
+      "epoch": 0.18597816960537364,
+      "grad_norm": 0.15198923647403717,
+      "learning_rate": 9.777091938608088e-05,
+      "loss": 0.1759,
+      "step": 886
+    },
+    {
+      "epoch": 0.18618807724601175,
+      "grad_norm": 0.14781691133975983,
+      "learning_rate": 9.776008813283125e-05,
+      "loss": 0.1521,
+      "step": 887
+    },
+    {
+      "epoch": 0.18639798488664988,
+      "grad_norm": 0.18348653614521027,
+      "learning_rate": 9.774923123132332e-05,
+      "loss": 0.1522,
+      "step": 888
+    },
+    {
+      "epoch": 0.186607892527288,
+      "grad_norm": 0.15346892178058624,
+      "learning_rate": 9.773834868738752e-05,
+      "loss": 0.1593,
+      "step": 889
+    },
+    {
+      "epoch": 0.18681780016792612,
+      "grad_norm": 0.1563442200422287,
+      "learning_rate": 9.7727440506868e-05,
+      "loss": 0.1569,
+      "step": 890
+    },
+    {
+      "epoch": 0.18702770780856423,
+      "grad_norm": 0.12280000746250153,
+      "learning_rate": 9.771650669562274e-05,
+      "loss": 0.1593,
+      "step": 891
+    },
+    {
+      "epoch": 0.18723761544920234,
+      "grad_norm": 0.1622755527496338,
+      "learning_rate": 9.770554725952341e-05,
+      "loss": 0.1736,
+      "step": 892
+    },
+    {
+      "epoch": 0.18744752308984047,
+      "grad_norm": 0.19185318052768707,
+      "learning_rate": 9.769456220445549e-05,
+      "loss": 0.165,
+      "step": 893
+    },
+    {
+      "epoch": 0.18765743073047858,
+      "grad_norm": 0.16040024161338806,
+      "learning_rate": 9.768355153631822e-05,
+      "loss": 0.1837,
+      "step": 894
+    },
+    {
+      "epoch": 0.18786733837111672,
+      "grad_norm": 0.11211330443620682,
+      "learning_rate": 9.767251526102456e-05,
+      "loss": 0.152,
+      "step": 895
+    },
+    {
+      "epoch": 0.18807724601175482,
+      "grad_norm": 0.16628898680210114,
+      "learning_rate": 9.766145338450125e-05,
+      "loss": 0.1729,
+      "step": 896
+    },
+    {
+      "epoch": 0.18828715365239296,
+      "grad_norm": 0.14260315895080566,
+      "learning_rate": 9.765036591268877e-05,
+      "loss": 0.165,
+      "step": 897
+    },
+    {
+      "epoch": 0.18849706129303107,
+      "grad_norm": 0.12734055519104004,
+      "learning_rate": 9.763925285154135e-05,
+      "loss": 0.1714,
+      "step": 898
+    },
+    {
+      "epoch": 0.18870696893366917,
+      "grad_norm": 0.1465056985616684,
+      "learning_rate": 9.762811420702693e-05,
+      "loss": 0.1805,
+      "step": 899
+    },
+    {
+      "epoch": 0.1889168765743073,
+      "grad_norm": 0.12369433790445328,
+      "learning_rate": 9.761694998512727e-05,
+      "loss": 0.1737,
+      "step": 900
+    },
+    {
+      "epoch": 0.18912678421494541,
+      "grad_norm": 0.13893358409404755,
+      "learning_rate": 9.760576019183775e-05,
+      "loss": 0.1502,
+      "step": 901
+    },
+    {
+      "epoch": 0.18933669185558355,
+      "grad_norm": 0.13517262041568756,
+      "learning_rate": 9.759454483316761e-05,
+      "loss": 0.1648,
+      "step": 902
+    },
+    {
+      "epoch": 0.18954659949622166,
+      "grad_norm": 0.13170479238033295,
+      "learning_rate": 9.75833039151397e-05,
+      "loss": 0.159,
+      "step": 903
+    },
+    {
+      "epoch": 0.1897565071368598,
+      "grad_norm": 0.13293495774269104,
+      "learning_rate": 9.757203744379067e-05,
+      "loss": 0.1667,
+      "step": 904
+    },
+    {
+      "epoch": 0.1899664147774979,
+      "grad_norm": 0.12381456047296524,
+      "learning_rate": 9.756074542517088e-05,
+      "loss": 0.1684,
+      "step": 905
+    },
+    {
+      "epoch": 0.190176322418136,
+      "grad_norm": 0.12754730880260468,
+      "learning_rate": 9.75494278653444e-05,
+      "loss": 0.1559,
+      "step": 906
+    },
+    {
+      "epoch": 0.19038623005877414,
+      "grad_norm": 0.17824846506118774,
+      "learning_rate": 9.753808477038899e-05,
+      "loss": 0.1533,
+      "step": 907
+    },
+    {
+      "epoch": 0.19059613769941225,
+      "grad_norm": 0.13958828151226044,
+      "learning_rate": 9.752671614639619e-05,
+      "loss": 0.1494,
+      "step": 908
+    },
+    {
+      "epoch": 0.19080604534005038,
+      "grad_norm": 0.14593012630939484,
+      "learning_rate": 9.75153219994712e-05,
+      "loss": 0.1598,
+      "step": 909
+    },
+    {
+      "epoch": 0.1910159529806885,
+      "grad_norm": 0.19262051582336426,
+      "learning_rate": 9.750390233573293e-05,
+      "loss": 0.1638,
+      "step": 910
+    },
+    {
+      "epoch": 0.19122586062132663,
+      "grad_norm": 0.15087257325649261,
+      "learning_rate": 9.7492457161314e-05,
+      "loss": 0.1644,
+      "step": 911
+    },
+    {
+      "epoch": 0.19143576826196473,
+      "grad_norm": 0.17646470665931702,
+      "learning_rate": 9.748098648236072e-05,
+      "loss": 0.1644,
+      "step": 912
+    },
+    {
+      "epoch": 0.19164567590260284,
+      "grad_norm": 0.15336105227470398,
+      "learning_rate": 9.746949030503312e-05,
+      "loss": 0.1717,
+      "step": 913
+    },
+    {
+      "epoch": 0.19185558354324098,
+      "grad_norm": 0.15364870429039001,
+      "learning_rate": 9.745796863550492e-05,
+      "loss": 0.1721,
+      "step": 914
+    },
+    {
+      "epoch": 0.19206549118387908,
+      "grad_norm": 0.1769437938928604,
+      "learning_rate": 9.74464214799635e-05,
+      "loss": 0.1485,
+      "step": 915
+    },
+    {
+      "epoch": 0.19227539882451722,
+      "grad_norm": 0.18178406357765198,
+      "learning_rate": 9.743484884460993e-05,
+      "loss": 0.1523,
+      "step": 916
+    },
+    {
+      "epoch": 0.19248530646515533,
+      "grad_norm": 0.16841888427734375,
+      "learning_rate": 9.742325073565905e-05,
+      "loss": 0.148,
+      "step": 917
+    },
+    {
+      "epoch": 0.19269521410579346,
+      "grad_norm": 0.13603579998016357,
+      "learning_rate": 9.741162715933924e-05,
+      "loss": 0.161,
+      "step": 918
+    },
+    {
+      "epoch": 0.19290512174643157,
+      "grad_norm": 0.1348285973072052,
+      "learning_rate": 9.739997812189265e-05,
+      "loss": 0.145,
+      "step": 919
+    },
+    {
+      "epoch": 0.19311502938706968,
+      "grad_norm": 0.13922441005706787,
+      "learning_rate": 9.738830362957508e-05,
+      "loss": 0.1607,
+      "step": 920
+    },
+    {
+      "epoch": 0.1933249370277078,
+      "grad_norm": 0.12017328292131424,
+      "learning_rate": 9.7376603688656e-05,
+      "loss": 0.152,
+      "step": 921
+    },
+    {
+      "epoch": 0.19353484466834592,
+      "grad_norm": 0.14594700932502747,
+      "learning_rate": 9.736487830541853e-05,
+      "loss": 0.1638,
+      "step": 922
+    },
+    {
+      "epoch": 0.19374475230898405,
+      "grad_norm": 0.14390654861927032,
+      "learning_rate": 9.73531274861595e-05,
+      "loss": 0.1608,
+      "step": 923
+    },
+    {
+      "epoch": 0.19395465994962216,
+      "grad_norm": 0.12394702434539795,
+      "learning_rate": 9.734135123718933e-05,
+      "loss": 0.1612,
+      "step": 924
+    },
+    {
+      "epoch": 0.1941645675902603,
+      "grad_norm": 0.16754676401615143,
+      "learning_rate": 9.732954956483218e-05,
+      "loss": 0.1791,
+      "step": 925
+    },
+    {
+      "epoch": 0.1943744752308984,
+      "grad_norm": 0.2816343605518341,
+      "learning_rate": 9.731772247542576e-05,
+      "loss": 0.1489,
+      "step": 926
+    },
+    {
+      "epoch": 0.19458438287153654,
+      "grad_norm": 0.15909412503242493,
+      "learning_rate": 9.730586997532155e-05,
+      "loss": 0.1531,
+      "step": 927
+    },
+    {
+      "epoch": 0.19479429051217464,
+      "grad_norm": 0.16626602411270142,
+      "learning_rate": 9.729399207088457e-05,
+      "loss": 0.1729,
+      "step": 928
+    },
+    {
+      "epoch": 0.19500419815281275,
+      "grad_norm": 0.14581038057804108,
+      "learning_rate": 9.728208876849354e-05,
+      "loss": 0.1616,
+      "step": 929
+    },
+    {
+      "epoch": 0.1952141057934509,
+      "grad_norm": 0.14985312521457672,
+      "learning_rate": 9.727016007454079e-05,
+      "loss": 0.1583,
+      "step": 930
+    },
+    {
+      "epoch": 0.195424013434089,
+      "grad_norm": 0.13958559930324554,
+      "learning_rate": 9.725820599543234e-05,
+      "loss": 0.1646,
+      "step": 931
+    },
+    {
+      "epoch": 0.19563392107472713,
+      "grad_norm": 0.1607862412929535,
+      "learning_rate": 9.724622653758777e-05,
+      "loss": 0.1549,
+      "step": 932
+    },
+    {
+      "epoch": 0.19584382871536524,
+      "grad_norm": 0.17007960379123688,
+      "learning_rate": 9.723422170744031e-05,
+      "loss": 0.1718,
+      "step": 933
+    },
+    {
+      "epoch": 0.19605373635600337,
+      "grad_norm": 0.1419927030801773,
+      "learning_rate": 9.722219151143688e-05,
+      "loss": 0.1689,
+      "step": 934
+    },
+    {
+      "epoch": 0.19626364399664148,
+      "grad_norm": 0.1631292998790741,
+      "learning_rate": 9.721013595603793e-05,
+      "loss": 0.1611,
+      "step": 935
+    },
+    {
+      "epoch": 0.1964735516372796,
+      "grad_norm": 0.19870012998580933,
+      "learning_rate": 9.719805504771758e-05,
+      "loss": 0.1836,
+      "step": 936
+    },
+    {
+      "epoch": 0.19668345927791772,
+      "grad_norm": 0.14150285720825195,
+      "learning_rate": 9.718594879296355e-05,
+      "loss": 0.1718,
+      "step": 937
+    },
+    {
+      "epoch": 0.19689336691855583,
+      "grad_norm": 0.1416793018579483,
+      "learning_rate": 9.717381719827716e-05,
+      "loss": 0.1511,
+      "step": 938
+    },
+    {
+      "epoch": 0.19710327455919396,
+      "grad_norm": 0.14615508913993835,
+      "learning_rate": 9.716166027017339e-05,
+      "loss": 0.1599,
+      "step": 939
+    },
+    {
+      "epoch": 0.19731318219983207,
+      "grad_norm": 0.13773533701896667,
+      "learning_rate": 9.714947801518076e-05,
+      "loss": 0.1765,
+      "step": 940
+    },
+    {
+      "epoch": 0.1975230898404702,
+      "grad_norm": 0.11306725442409515,
+      "learning_rate": 9.713727043984143e-05,
+      "loss": 0.1678,
+      "step": 941
+    },
+    {
+      "epoch": 0.1977329974811083,
+      "grad_norm": 0.14150364696979523,
+      "learning_rate": 9.712503755071115e-05,
+      "loss": 0.1535,
+      "step": 942
+    },
+    {
+      "epoch": 0.19794290512174642,
+      "grad_norm": 0.15026667714118958,
+      "learning_rate": 9.711277935435925e-05,
+      "loss": 0.155,
+      "step": 943
+    },
+    {
+      "epoch": 0.19815281276238456,
+      "grad_norm": 0.1540324091911316,
+      "learning_rate": 9.710049585736866e-05,
+      "loss": 0.1866,
+      "step": 944
+    },
+    {
+      "epoch": 0.19836272040302266,
+      "grad_norm": 0.12859384715557098,
+      "learning_rate": 9.708818706633591e-05,
+      "loss": 0.1512,
+      "step": 945
+    },
+    {
+      "epoch": 0.1985726280436608,
+      "grad_norm": 0.14280495047569275,
+      "learning_rate": 9.707585298787109e-05,
+      "loss": 0.1558,
+      "step": 946
+    },
+    {
+      "epoch": 0.1987825356842989,
+      "grad_norm": 0.14329122006893158,
+      "learning_rate": 9.70634936285979e-05,
+      "loss": 0.153,
+      "step": 947
+    },
+    {
+      "epoch": 0.19899244332493704,
+      "grad_norm": 0.18129687011241913,
+      "learning_rate": 9.705110899515359e-05,
+      "loss": 0.1592,
+      "step": 948
+    },
+    {
+      "epoch": 0.19920235096557515,
+      "grad_norm": 0.13191339373588562,
+      "learning_rate": 9.7038699094189e-05,
+      "loss": 0.1565,
+      "step": 949
+    },
+    {
+      "epoch": 0.19941225860621326,
+      "grad_norm": 0.1381015032529831,
+      "learning_rate": 9.702626393236849e-05,
+      "loss": 0.1704,
+      "step": 950
+    },
+    {
+      "epoch": 0.1996221662468514,
+      "grad_norm": 0.12249460071325302,
+      "learning_rate": 9.701380351637007e-05,
+      "loss": 0.1534,
+      "step": 951
+    },
+    {
+      "epoch": 0.1998320738874895,
+      "grad_norm": 0.1820680946111679,
+      "learning_rate": 9.700131785288525e-05,
+      "loss": 0.1805,
+      "step": 952
+    },
+    {
+      "epoch": 0.20004198152812763,
+      "grad_norm": 0.13920508325099945,
+      "learning_rate": 9.698880694861913e-05,
+      "loss": 0.1652,
+      "step": 953
+    },
+    {
+      "epoch": 0.20025188916876574,
+      "grad_norm": 0.14306975901126862,
+      "learning_rate": 9.697627081029033e-05,
+      "loss": 0.1397,
+      "step": 954
+    },
+    {
+      "epoch": 0.20046179680940387,
+      "grad_norm": 0.13642224669456482,
+      "learning_rate": 9.696370944463104e-05,
+      "loss": 0.174,
+      "step": 955
+    },
+    {
+      "epoch": 0.20067170445004198,
+      "grad_norm": 0.12083397060632706,
+      "learning_rate": 9.695112285838704e-05,
+      "loss": 0.1807,
+      "step": 956
+    },
+    {
+      "epoch": 0.2008816120906801,
+      "grad_norm": 0.17191384732723236,
+      "learning_rate": 9.693851105831757e-05,
+      "loss": 0.1656,
+      "step": 957
+    },
+    {
+      "epoch": 0.20109151973131822,
+      "grad_norm": 0.1415136307477951,
+      "learning_rate": 9.692587405119549e-05,
+      "loss": 0.1726,
+      "step": 958
+    },
+    {
+      "epoch": 0.20130142737195633,
+      "grad_norm": 0.15210871398448944,
+      "learning_rate": 9.691321184380713e-05,
+      "loss": 0.1614,
+      "step": 959
+    },
+    {
+      "epoch": 0.20151133501259447,
+      "grad_norm": 0.14295780658721924,
+      "learning_rate": 9.690052444295239e-05,
+      "loss": 0.1884,
+      "step": 960
+    },
+    {
+      "epoch": 0.20172124265323257,
+      "grad_norm": 0.13754574954509735,
+      "learning_rate": 9.688781185544471e-05,
+      "loss": 0.1719,
+      "step": 961
+    },
+    {
+      "epoch": 0.2019311502938707,
+      "grad_norm": 0.1425933688879013,
+      "learning_rate": 9.687507408811104e-05,
+      "loss": 0.1564,
+      "step": 962
+    },
+    {
+      "epoch": 0.20214105793450882,
+      "grad_norm": 0.1202697604894638,
+      "learning_rate": 9.686231114779184e-05,
+      "loss": 0.1584,
+      "step": 963
+    },
+    {
+      "epoch": 0.20235096557514692,
+      "grad_norm": 0.12710309028625488,
+      "learning_rate": 9.684952304134111e-05,
+      "loss": 0.168,
+      "step": 964
+    },
+    {
+      "epoch": 0.20256087321578506,
+      "grad_norm": 0.14089150726795197,
+      "learning_rate": 9.683670977562633e-05,
+      "loss": 0.1852,
+      "step": 965
+    },
+    {
+      "epoch": 0.20277078085642317,
+      "grad_norm": 0.13602103292942047,
+      "learning_rate": 9.682387135752856e-05,
+      "loss": 0.1673,
+      "step": 966
+    },
+    {
+      "epoch": 0.2029806884970613,
+      "grad_norm": 0.13352209329605103,
+      "learning_rate": 9.68110077939423e-05,
+      "loss": 0.1571,
+      "step": 967
+    },
+    {
+      "epoch": 0.2031905961376994,
+      "grad_norm": 0.13534000515937805,
+      "learning_rate": 9.679811909177556e-05,
+      "loss": 0.1706,
+      "step": 968
+    },
+    {
+      "epoch": 0.20340050377833754,
+      "grad_norm": 0.12475798279047012,
+      "learning_rate": 9.67852052579499e-05,
+      "loss": 0.1532,
+      "step": 969
+    },
+    {
+      "epoch": 0.20361041141897565,
+      "grad_norm": 0.1589164137840271,
+      "learning_rate": 9.677226629940032e-05,
+      "loss": 0.1597,
+      "step": 970
+    },
+    {
+      "epoch": 0.20382031905961376,
+      "grad_norm": 0.14976951479911804,
+      "learning_rate": 9.675930222307537e-05,
+      "loss": 0.1731,
+      "step": 971
+    },
+    {
+      "epoch": 0.2040302267002519,
+      "grad_norm": 0.12664476037025452,
+      "learning_rate": 9.6746313035937e-05,
+      "loss": 0.1363,
+      "step": 972
+    },
+    {
+      "epoch": 0.20424013434089,
+      "grad_norm": 0.12275015562772751,
+      "learning_rate": 9.673329874496075e-05,
+      "loss": 0.1574,
+      "step": 973
+    },
+    {
+      "epoch": 0.20445004198152814,
+      "grad_norm": 0.13222196698188782,
+      "learning_rate": 9.672025935713556e-05,
+      "loss": 0.1535,
+      "step": 974
+    },
+    {
+      "epoch": 0.20465994962216624,
+      "grad_norm": 0.15173639357089996,
+      "learning_rate": 9.670719487946389e-05,
+      "loss": 0.1792,
+      "step": 975
+    },
+    {
+      "epoch": 0.20486985726280438,
+      "grad_norm": 0.17748403549194336,
+      "learning_rate": 9.669410531896167e-05,
+      "loss": 0.1891,
+      "step": 976
+    },
+    {
+      "epoch": 0.20507976490344249,
+      "grad_norm": 0.14893555641174316,
+      "learning_rate": 9.66809906826583e-05,
+      "loss": 0.1438,
+      "step": 977
+    },
+    {
+      "epoch": 0.2052896725440806,
+      "grad_norm": 0.16388140618801117,
+      "learning_rate": 9.66678509775966e-05,
+      "loss": 0.1641,
+      "step": 978
+    },
+    {
+      "epoch": 0.20549958018471873,
+      "grad_norm": 0.17024089395999908,
+      "learning_rate": 9.665468621083293e-05,
+      "loss": 0.1839,
+      "step": 979
+    },
+    {
+      "epoch": 0.20570948782535683,
+      "grad_norm": 0.14790277183055878,
+      "learning_rate": 9.664149638943707e-05,
+      "loss": 0.1701,
+      "step": 980
+    },
+    {
+      "epoch": 0.20591939546599497,
+      "grad_norm": 0.14114373922348022,
+      "learning_rate": 9.662828152049223e-05,
+      "loss": 0.1699,
+      "step": 981
+    },
+    {
+      "epoch": 0.20612930310663308,
+      "grad_norm": 0.12556609511375427,
+      "learning_rate": 9.661504161109513e-05,
+      "loss": 0.1607,
+      "step": 982
+    },
+    {
+      "epoch": 0.2063392107472712,
+      "grad_norm": 0.11868204176425934,
+      "learning_rate": 9.660177666835585e-05,
+      "loss": 0.1487,
+      "step": 983
+    },
+    {
+      "epoch": 0.20654911838790932,
+      "grad_norm": 0.12117776274681091,
+      "learning_rate": 9.658848669939805e-05,
+      "loss": 0.1668,
+      "step": 984
+    },
+    {
+      "epoch": 0.20675902602854743,
+      "grad_norm": 0.126139298081398,
+      "learning_rate": 9.657517171135866e-05,
+      "loss": 0.1535,
+      "step": 985
+    },
+    {
+      "epoch": 0.20696893366918556,
+      "grad_norm": 0.15042434632778168,
+      "learning_rate": 9.656183171138818e-05,
+      "loss": 0.168,
+      "step": 986
+    },
+    {
+      "epoch": 0.20717884130982367,
+      "grad_norm": 0.1438342183828354,
+      "learning_rate": 9.65484667066505e-05,
+      "loss": 0.1628,
+      "step": 987
+    },
+    {
+      "epoch": 0.2073887489504618,
+      "grad_norm": 0.1280035823583603,
+      "learning_rate": 9.65350767043229e-05,
+      "loss": 0.1655,
+      "step": 988
+    },
+    {
+      "epoch": 0.2075986565910999,
+      "grad_norm": 0.13997893035411835,
+      "learning_rate": 9.652166171159614e-05,
+      "loss": 0.1766,
+      "step": 989
+    },
+    {
+      "epoch": 0.20780856423173805,
+      "grad_norm": 0.1653011292219162,
+      "learning_rate": 9.650822173567438e-05,
+      "loss": 0.1744,
+      "step": 990
+    },
+    {
+      "epoch": 0.20801847187237615,
+      "grad_norm": 0.1532358080148697,
+      "learning_rate": 9.64947567837752e-05,
+      "loss": 0.1529,
+      "step": 991
+    },
+    {
+      "epoch": 0.20822837951301426,
+      "grad_norm": 0.13623711466789246,
+      "learning_rate": 9.648126686312955e-05,
+      "loss": 0.1632,
+      "step": 992
+    },
+    {
+      "epoch": 0.2084382871536524,
+      "grad_norm": 0.13220278918743134,
+      "learning_rate": 9.646775198098186e-05,
+      "loss": 0.1462,
+      "step": 993
+    },
+    {
+      "epoch": 0.2086481947942905,
+      "grad_norm": 0.16712597012519836,
+      "learning_rate": 9.645421214458992e-05,
+      "loss": 0.1658,
+      "step": 994
+    },
+    {
+      "epoch": 0.20885810243492864,
+      "grad_norm": 0.16033975780010223,
+      "learning_rate": 9.644064736122493e-05,
+      "loss": 0.1523,
+      "step": 995
+    },
+    {
+      "epoch": 0.20906801007556675,
+      "grad_norm": 0.13283276557922363,
+      "learning_rate": 9.64270576381715e-05,
+      "loss": 0.1738,
+      "step": 996
+    },
+    {
+      "epoch": 0.20927791771620488,
+      "grad_norm": 0.2072938233613968,
+      "learning_rate": 9.64134429827276e-05,
+      "loss": 0.1589,
+      "step": 997
+    },
+    {
+      "epoch": 0.209487825356843,
+      "grad_norm": 0.15322737395763397,
+      "learning_rate": 9.639980340220462e-05,
+      "loss": 0.1772,
+      "step": 998
+    },
+    {
+      "epoch": 0.2096977329974811,
+      "grad_norm": 0.21009685099124908,
+      "learning_rate": 9.638613890392734e-05,
+      "loss": 0.1619,
+      "step": 999
+    },
+    {
+      "epoch": 0.20990764063811923,
+      "grad_norm": 0.1521667093038559,
+      "learning_rate": 9.63724494952339e-05,
+      "loss": 0.1878,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4764,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.14184099585065e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}