diff --git "a/checkpoint-2729/trainer_state.json" "b/checkpoint-2729/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2729/trainer_state.json"
@@ -0,0 +1,19168 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 683,
+  "global_step": 2729,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00036643459142543056,
+      "grad_norm": 0.46817973256111145,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.2732,
+      "step": 1
+    },
+    {
+      "epoch": 0.00036643459142543056,
+      "eval_loss": 1.0894134044647217,
+      "eval_runtime": 793.7236,
+      "eval_samples_per_second": 3.417,
+      "eval_steps_per_second": 0.427,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007328691828508611,
+      "grad_norm": 0.47947484254837036,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3138,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010993037742762918,
+      "grad_norm": 0.49031010270118713,
+      "learning_rate": 6e-06,
+      "loss": 1.3379,
+      "step": 3
+    },
+    {
+      "epoch": 0.0014657383657017222,
+      "grad_norm": 0.48090797662734985,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.2044,
+      "step": 4
+    },
+    {
+      "epoch": 0.001832172957127153,
+      "grad_norm": 0.4987533390522003,
+      "learning_rate": 1e-05,
+      "loss": 1.3141,
+      "step": 5
+    },
+    {
+      "epoch": 0.0021986075485525836,
+      "grad_norm": 0.500445544719696,
+      "learning_rate": 1.2e-05,
+      "loss": 1.3083,
+      "step": 6
+    },
+    {
+      "epoch": 0.002565042139978014,
+      "grad_norm": 0.44339627027511597,
+      "learning_rate": 1.4e-05,
+      "loss": 1.2854,
+      "step": 7
+    },
+    {
+      "epoch": 0.0029314767314034445,
+      "grad_norm": 0.48581796884536743,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.1958,
+      "step": 8
+    },
+    {
+      "epoch": 0.003297911322828875,
+      "grad_norm": 0.4820937514305115,
+      "learning_rate": 1.8e-05,
+      "loss": 1.3483,
+      "step": 9
+    },
+    {
+      "epoch": 0.003664345914254306,
+      "grad_norm": 0.44360584020614624,
+      "learning_rate": 2e-05,
+      "loss": 1.3793,
+      "step": 10
+    },
+    {
+      "epoch": 0.004030780505679736,
+      "grad_norm": 0.4145188331604004,
+      "learning_rate": 1.999999332499975e-05,
+      "loss": 0.8327,
+      "step": 11
+    },
+    {
+      "epoch": 0.004397215097105167,
+      "grad_norm": 0.3894279897212982,
+      "learning_rate": 1.9999973300007917e-05,
+      "loss": 1.2966,
+      "step": 12
+    },
+    {
+      "epoch": 0.004763649688530597,
+      "grad_norm": 0.38675037026405334,
+      "learning_rate": 1.999993992505123e-05,
+      "loss": 1.2477,
+      "step": 13
+    },
+    {
+      "epoch": 0.005130084279956028,
+      "grad_norm": 0.37546542286872864,
+      "learning_rate": 1.999989320017424e-05,
+      "loss": 1.2395,
+      "step": 14
+    },
+    {
+      "epoch": 0.005496518871381458,
+      "grad_norm": 0.3815326988697052,
+      "learning_rate": 1.9999833125439333e-05,
+      "loss": 1.2152,
+      "step": 15
+    },
+    {
+      "epoch": 0.005862953462806889,
+      "grad_norm": 0.4222187399864197,
+      "learning_rate": 1.9999759700926705e-05,
+      "loss": 1.3211,
+      "step": 16
+    },
+    {
+      "epoch": 0.00622938805423232,
+      "grad_norm": 0.3901718556880951,
+      "learning_rate": 1.9999672926734378e-05,
+      "loss": 1.3281,
+      "step": 17
+    },
+    {
+      "epoch": 0.00659582264565775,
+      "grad_norm": 0.4249936044216156,
+      "learning_rate": 1.99995728029782e-05,
+      "loss": 1.1635,
+      "step": 18
+    },
+    {
+      "epoch": 0.006962257237083181,
+      "grad_norm": 0.3950119912624359,
+      "learning_rate": 1.999945932979183e-05,
+      "loss": 1.2472,
+      "step": 19
+    },
+    {
+      "epoch": 0.007328691828508612,
+      "grad_norm": 0.47602978348731995,
+      "learning_rate": 1.9999332507326753e-05,
+      "loss": 1.2014,
+      "step": 20
+    },
+    {
+      "epoch": 0.007695126419934042,
+      "grad_norm": 0.43517524003982544,
+      "learning_rate": 1.9999192335752284e-05,
+      "loss": 1.2921,
+      "step": 21
+    },
+    {
+      "epoch": 0.008061561011359472,
+      "grad_norm": 0.41316503286361694,
+      "learning_rate": 1.999903881525555e-05,
+      "loss": 1.2258,
+      "step": 22
+    },
+    {
+      "epoch": 0.008427995602784902,
+      "grad_norm": 0.4038918912410736,
+      "learning_rate": 1.9998871946041497e-05,
+      "loss": 1.2234,
+      "step": 23
+    },
+    {
+      "epoch": 0.008794430194210334,
+      "grad_norm": 0.43499597907066345,
+      "learning_rate": 1.99986917283329e-05,
+      "loss": 1.2151,
+      "step": 24
+    },
+    {
+      "epoch": 0.009160864785635764,
+      "grad_norm": 0.4008733928203583,
+      "learning_rate": 1.9998498162370347e-05,
+      "loss": 1.1845,
+      "step": 25
+    },
+    {
+      "epoch": 0.009527299377061194,
+      "grad_norm": 0.42672181129455566,
+      "learning_rate": 1.9998291248412253e-05,
+      "loss": 1.111,
+      "step": 26
+    },
+    {
+      "epoch": 0.009893733968486624,
+      "grad_norm": 0.39955198764801025,
+      "learning_rate": 1.999807098673484e-05,
+      "loss": 1.2242,
+      "step": 27
+    },
+    {
+      "epoch": 0.010260168559912056,
+      "grad_norm": 0.4173682928085327,
+      "learning_rate": 1.999783737763216e-05,
+      "loss": 1.1515,
+      "step": 28
+    },
+    {
+      "epoch": 0.010626603151337486,
+      "grad_norm": 0.4032392203807831,
+      "learning_rate": 1.999759042141609e-05,
+      "loss": 1.2194,
+      "step": 29
+    },
+    {
+      "epoch": 0.010993037742762916,
+      "grad_norm": 0.391091525554657,
+      "learning_rate": 1.9997330118416305e-05,
+      "loss": 1.204,
+      "step": 30
+    },
+    {
+      "epoch": 0.011359472334188348,
+      "grad_norm": 0.3903694152832031,
+      "learning_rate": 1.9997056468980315e-05,
+      "loss": 1.2606,
+      "step": 31
+    },
+    {
+      "epoch": 0.011725906925613778,
+      "grad_norm": 0.3887855112552643,
+      "learning_rate": 1.999676947347344e-05,
+      "loss": 1.2385,
+      "step": 32
+    },
+    {
+      "epoch": 0.012092341517039208,
+      "grad_norm": 0.4681757390499115,
+      "learning_rate": 1.999646913227882e-05,
+      "loss": 1.1089,
+      "step": 33
+    },
+    {
+      "epoch": 0.01245877610846464,
+      "grad_norm": 0.42056483030319214,
+      "learning_rate": 1.9996155445797415e-05,
+      "loss": 1.1911,
+      "step": 34
+    },
+    {
+      "epoch": 0.01282521069989007,
+      "grad_norm": 0.4387955665588379,
+      "learning_rate": 1.999582841444799e-05,
+      "loss": 1.2162,
+      "step": 35
+    },
+    {
+      "epoch": 0.0131916452913155,
+      "grad_norm": 0.4079417884349823,
+      "learning_rate": 1.999548803866713e-05,
+      "loss": 1.1769,
+      "step": 36
+    },
+    {
+      "epoch": 0.013558079882740931,
+      "grad_norm": 0.48628127574920654,
+      "learning_rate": 1.9995134318909247e-05,
+      "loss": 1.0853,
+      "step": 37
+    },
+    {
+      "epoch": 0.013924514474166361,
+      "grad_norm": 0.5449658632278442,
+      "learning_rate": 1.9994767255646548e-05,
+      "loss": 1.0893,
+      "step": 38
+    },
+    {
+      "epoch": 0.014290949065591791,
+      "grad_norm": 0.4337720274925232,
+      "learning_rate": 1.9994386849369062e-05,
+      "loss": 1.2324,
+      "step": 39
+    },
+    {
+      "epoch": 0.014657383657017223,
+      "grad_norm": 0.4274596571922302,
+      "learning_rate": 1.999399310058464e-05,
+      "loss": 1.0204,
+      "step": 40
+    },
+    {
+      "epoch": 0.015023818248442653,
+      "grad_norm": 0.420960009098053,
+      "learning_rate": 1.9993586009818927e-05,
+      "loss": 1.1547,
+      "step": 41
+    },
+    {
+      "epoch": 0.015390252839868083,
+      "grad_norm": 0.42221662402153015,
+      "learning_rate": 1.9993165577615393e-05,
+      "loss": 1.1781,
+      "step": 42
+    },
+    {
+      "epoch": 0.015756687431293513,
+      "grad_norm": 0.4561188220977783,
+      "learning_rate": 1.999273180453532e-05,
+      "loss": 1.2284,
+      "step": 43
+    },
+    {
+      "epoch": 0.016123122022718945,
+      "grad_norm": 0.43053948879241943,
+      "learning_rate": 1.9992284691157785e-05,
+      "loss": 1.1378,
+      "step": 44
+    },
+    {
+      "epoch": 0.016489556614144377,
+      "grad_norm": 0.480121374130249,
+      "learning_rate": 1.999182423807969e-05,
+      "loss": 1.2033,
+      "step": 45
+    },
+    {
+      "epoch": 0.016855991205569805,
+      "grad_norm": 0.4473513960838318,
+      "learning_rate": 1.9991350445915745e-05,
+      "loss": 1.1538,
+      "step": 46
+    },
+    {
+      "epoch": 0.017222425796995237,
+      "grad_norm": 0.5381205081939697,
+      "learning_rate": 1.999086331529845e-05,
+      "loss": 1.1332,
+      "step": 47
+    },
+    {
+      "epoch": 0.01758886038842067,
+      "grad_norm": 0.4490923583507538,
+      "learning_rate": 1.999036284687814e-05,
+      "loss": 1.1733,
+      "step": 48
+    },
+    {
+      "epoch": 0.017955294979846097,
+      "grad_norm": 0.44338756799697876,
+      "learning_rate": 1.998984904132293e-05,
+      "loss": 1.1781,
+      "step": 49
+    },
+    {
+      "epoch": 0.01832172957127153,
+      "grad_norm": 0.5198048949241638,
+      "learning_rate": 1.998932189931875e-05,
+      "loss": 1.1102,
+      "step": 50
+    },
+    {
+      "epoch": 0.01868816416269696,
+      "grad_norm": 0.6394063234329224,
+      "learning_rate": 1.9988781421569333e-05,
+      "loss": 1.0399,
+      "step": 51
+    },
+    {
+      "epoch": 0.01905459875412239,
+      "grad_norm": 0.5123482942581177,
+      "learning_rate": 1.9988227608796227e-05,
+      "loss": 1.145,
+      "step": 52
+    },
+    {
+      "epoch": 0.01942103334554782,
+      "grad_norm": 0.5353620052337646,
+      "learning_rate": 1.9987660461738763e-05,
+      "loss": 1.0634,
+      "step": 53
+    },
+    {
+      "epoch": 0.01978746793697325,
+      "grad_norm": 0.49093300104141235,
+      "learning_rate": 1.998707998115409e-05,
+      "loss": 1.1711,
+      "step": 54
+    },
+    {
+      "epoch": 0.02015390252839868,
+      "grad_norm": 0.4911767542362213,
+      "learning_rate": 1.9986486167817136e-05,
+      "loss": 1.1343,
+      "step": 55
+    },
+    {
+      "epoch": 0.020520337119824112,
+      "grad_norm": 0.5678600072860718,
+      "learning_rate": 1.998587902252066e-05,
+      "loss": 1.1299,
+      "step": 56
+    },
+    {
+      "epoch": 0.02088677171124954,
+      "grad_norm": 0.5117340683937073,
+      "learning_rate": 1.9985258546075187e-05,
+      "loss": 1.1147,
+      "step": 57
+    },
+    {
+      "epoch": 0.021253206302674972,
+      "grad_norm": 0.5707929730415344,
+      "learning_rate": 1.9984624739309056e-05,
+      "loss": 1.1359,
+      "step": 58
+    },
+    {
+      "epoch": 0.021619640894100404,
+      "grad_norm": 0.539131224155426,
+      "learning_rate": 1.9983977603068404e-05,
+      "loss": 1.1615,
+      "step": 59
+    },
+    {
+      "epoch": 0.021986075485525832,
+      "grad_norm": 0.5231873393058777,
+      "learning_rate": 1.9983317138217155e-05,
+      "loss": 1.1158,
+      "step": 60
+    },
+    {
+      "epoch": 0.022352510076951264,
+      "grad_norm": 0.5066128373146057,
+      "learning_rate": 1.9982643345637027e-05,
+      "loss": 1.1435,
+      "step": 61
+    },
+    {
+      "epoch": 0.022718944668376696,
+      "grad_norm": 0.6198028922080994,
+      "learning_rate": 1.9981956226227537e-05,
+      "loss": 1.056,
+      "step": 62
+    },
+    {
+      "epoch": 0.023085379259802124,
+      "grad_norm": 0.5626301765441895,
+      "learning_rate": 1.9981255780905988e-05,
+      "loss": 1.1502,
+      "step": 63
+    },
+    {
+      "epoch": 0.023451813851227556,
+      "grad_norm": 0.6188763380050659,
+      "learning_rate": 1.9980542010607476e-05,
+      "loss": 1.0754,
+      "step": 64
+    },
+    {
+      "epoch": 0.023818248442652987,
+      "grad_norm": 0.6580602526664734,
+      "learning_rate": 1.997981491628488e-05,
+      "loss": 1.1088,
+      "step": 65
+    },
+    {
+      "epoch": 0.024184683034078416,
+      "grad_norm": 0.7106578350067139,
+      "learning_rate": 1.9979074498908876e-05,
+      "loss": 1.0783,
+      "step": 66
+    },
+    {
+      "epoch": 0.024551117625503847,
+      "grad_norm": 0.5700773596763611,
+      "learning_rate": 1.997832075946792e-05,
+      "loss": 1.0747,
+      "step": 67
+    },
+    {
+      "epoch": 0.02491755221692928,
+      "grad_norm": 0.6465053558349609,
+      "learning_rate": 1.9977553698968252e-05,
+      "loss": 1.1031,
+      "step": 68
+    },
+    {
+      "epoch": 0.025283986808354707,
+      "grad_norm": 0.5877029299736023,
+      "learning_rate": 1.9976773318433896e-05,
+      "loss": 1.1117,
+      "step": 69
+    },
+    {
+      "epoch": 0.02565042139978014,
+      "grad_norm": 0.6212673187255859,
+      "learning_rate": 1.9975979618906665e-05,
+      "loss": 1.0253,
+      "step": 70
+    },
+    {
+      "epoch": 0.02601685599120557,
+      "grad_norm": 0.6170660853385925,
+      "learning_rate": 1.9975172601446147e-05,
+      "loss": 0.9862,
+      "step": 71
+    },
+    {
+      "epoch": 0.026383290582631,
+      "grad_norm": 0.6177666783332825,
+      "learning_rate": 1.997435226712971e-05,
+      "loss": 1.0611,
+      "step": 72
+    },
+    {
+      "epoch": 0.02674972517405643,
+      "grad_norm": 0.6978894472122192,
+      "learning_rate": 1.99735186170525e-05,
+      "loss": 1.0911,
+      "step": 73
+    },
+    {
+      "epoch": 0.027116159765481863,
+      "grad_norm": 0.635191023349762,
+      "learning_rate": 1.997267165232744e-05,
+      "loss": 1.1159,
+      "step": 74
+    },
+    {
+      "epoch": 0.02748259435690729,
+      "grad_norm": 0.567520260810852,
+      "learning_rate": 1.9971811374085228e-05,
+      "loss": 1.113,
+      "step": 75
+    },
+    {
+      "epoch": 0.027849028948332723,
+      "grad_norm": 0.5939634442329407,
+      "learning_rate": 1.9970937783474333e-05,
+      "loss": 1.1586,
+      "step": 76
+    },
+    {
+      "epoch": 0.028215463539758154,
+      "grad_norm": 0.6401302218437195,
+      "learning_rate": 1.9970050881661e-05,
+      "loss": 1.0674,
+      "step": 77
+    },
+    {
+      "epoch": 0.028581898131183583,
+      "grad_norm": 0.6552327871322632,
+      "learning_rate": 1.9969150669829242e-05,
+      "loss": 1.088,
+      "step": 78
+    },
+    {
+      "epoch": 0.028948332722609015,
+      "grad_norm": 0.5755530595779419,
+      "learning_rate": 1.996823714918085e-05,
+      "loss": 1.0717,
+      "step": 79
+    },
+    {
+      "epoch": 0.029314767314034446,
+      "grad_norm": 0.6203620433807373,
+      "learning_rate": 1.996731032093536e-05,
+      "loss": 1.0951,
+      "step": 80
+    },
+    {
+      "epoch": 0.029681201905459875,
+      "grad_norm": 0.6365242004394531,
+      "learning_rate": 1.99663701863301e-05,
+      "loss": 1.1612,
+      "step": 81
+    },
+    {
+      "epoch": 0.030047636496885306,
+      "grad_norm": 0.7320545315742493,
+      "learning_rate": 1.9965416746620147e-05,
+      "loss": 1.0666,
+      "step": 82
+    },
+    {
+      "epoch": 0.030414071088310738,
+      "grad_norm": 0.6602171659469604,
+      "learning_rate": 1.9964450003078336e-05,
+      "loss": 1.0418,
+      "step": 83
+    },
+    {
+      "epoch": 0.030780505679736166,
+      "grad_norm": 0.6662451028823853,
+      "learning_rate": 1.9963469956995278e-05,
+      "loss": 1.0383,
+      "step": 84
+    },
+    {
+      "epoch": 0.031146940271161598,
+      "grad_norm": 0.7394860982894897,
+      "learning_rate": 1.9962476609679328e-05,
+      "loss": 1.0124,
+      "step": 85
+    },
+    {
+      "epoch": 0.031513374862587026,
+      "grad_norm": 0.6913645267486572,
+      "learning_rate": 1.9961469962456613e-05,
+      "loss": 1.1247,
+      "step": 86
+    },
+    {
+      "epoch": 0.03187980945401246,
+      "grad_norm": 0.8122132420539856,
+      "learning_rate": 1.9960450016670997e-05,
+      "loss": 0.9204,
+      "step": 87
+    },
+    {
+      "epoch": 0.03224624404543789,
+      "grad_norm": 0.6972271800041199,
+      "learning_rate": 1.9959416773684113e-05,
+      "loss": 1.0335,
+      "step": 88
+    },
+    {
+      "epoch": 0.03261267863686332,
+      "grad_norm": 0.7278566956520081,
+      "learning_rate": 1.9958370234875345e-05,
+      "loss": 1.022,
+      "step": 89
+    },
+    {
+      "epoch": 0.03297911322828875,
+      "grad_norm": 0.6476576328277588,
+      "learning_rate": 1.9957310401641817e-05,
+      "loss": 1.0114,
+      "step": 90
+    },
+    {
+      "epoch": 0.03334554781971418,
+      "grad_norm": 0.7813729643821716,
+      "learning_rate": 1.9956237275398402e-05,
+      "loss": 0.9962,
+      "step": 91
+    },
+    {
+      "epoch": 0.03371198241113961,
+      "grad_norm": 0.7484748959541321,
+      "learning_rate": 1.9955150857577733e-05,
+      "loss": 1.1148,
+      "step": 92
+    },
+    {
+      "epoch": 0.034078417002565045,
+      "grad_norm": 0.6738268136978149,
+      "learning_rate": 1.995405114963017e-05,
+      "loss": 1.1177,
+      "step": 93
+    },
+    {
+      "epoch": 0.03444485159399047,
+      "grad_norm": 0.8074272274971008,
+      "learning_rate": 1.9952938153023828e-05,
+      "loss": 1.0971,
+      "step": 94
+    },
+    {
+      "epoch": 0.0348112861854159,
+      "grad_norm": 0.6509901881217957,
+      "learning_rate": 1.995181186924456e-05,
+      "loss": 1.1152,
+      "step": 95
+    },
+    {
+      "epoch": 0.03517772077684134,
+      "grad_norm": 0.7371155619621277,
+      "learning_rate": 1.9950672299795947e-05,
+      "loss": 1.0771,
+      "step": 96
+    },
+    {
+      "epoch": 0.035544155368266765,
+      "grad_norm": 0.8736119270324707,
+      "learning_rate": 1.994951944619932e-05,
+      "loss": 1.0211,
+      "step": 97
+    },
+    {
+      "epoch": 0.035910589959692193,
+      "grad_norm": 0.8305104970932007,
+      "learning_rate": 1.9948353309993735e-05,
+      "loss": 0.9812,
+      "step": 98
+    },
+    {
+      "epoch": 0.03627702455111763,
+      "grad_norm": 0.7285597920417786,
+      "learning_rate": 1.994717389273599e-05,
+      "loss": 1.0571,
+      "step": 99
+    },
+    {
+      "epoch": 0.03664345914254306,
+      "grad_norm": 0.6941920518875122,
+      "learning_rate": 1.9945981196000598e-05,
+      "loss": 1.0796,
+      "step": 100
+    },
+    {
+      "epoch": 0.037009893733968485,
+      "grad_norm": 0.7027592062950134,
+      "learning_rate": 1.9944775221379817e-05,
+      "loss": 1.0951,
+      "step": 101
+    },
+    {
+      "epoch": 0.03737632832539392,
+      "grad_norm": 0.7003068327903748,
+      "learning_rate": 1.9943555970483615e-05,
+      "loss": 1.0421,
+      "step": 102
+    },
+    {
+      "epoch": 0.03774276291681935,
+      "grad_norm": 0.7429030537605286,
+      "learning_rate": 1.9942323444939706e-05,
+      "loss": 1.1032,
+      "step": 103
+    },
+    {
+      "epoch": 0.03810919750824478,
+      "grad_norm": 0.7443472743034363,
+      "learning_rate": 1.99410776463935e-05,
+      "loss": 1.0299,
+      "step": 104
+    },
+    {
+      "epoch": 0.03847563209967021,
+      "grad_norm": 0.7034825086593628,
+      "learning_rate": 1.9939818576508138e-05,
+      "loss": 1.0014,
+      "step": 105
+    },
+    {
+      "epoch": 0.03884206669109564,
+      "grad_norm": 0.6462461948394775,
+      "learning_rate": 1.9938546236964482e-05,
+      "loss": 1.034,
+      "step": 106
+    },
+    {
+      "epoch": 0.03920850128252107,
+      "grad_norm": 0.981590747833252,
+      "learning_rate": 1.9937260629461107e-05,
+      "loss": 1.0843,
+      "step": 107
+    },
+    {
+      "epoch": 0.0395749358739465,
+      "grad_norm": 0.746054470539093,
+      "learning_rate": 1.9935961755714298e-05,
+      "loss": 1.0168,
+      "step": 108
+    },
+    {
+      "epoch": 0.03994137046537193,
+      "grad_norm": 0.6923157572746277,
+      "learning_rate": 1.9934649617458044e-05,
+      "loss": 1.0926,
+      "step": 109
+    },
+    {
+      "epoch": 0.04030780505679736,
+      "grad_norm": 0.8634763360023499,
+      "learning_rate": 1.9933324216444064e-05,
+      "loss": 1.093,
+      "step": 110
+    },
+    {
+      "epoch": 0.04067423964822279,
+      "grad_norm": 0.773468554019928,
+      "learning_rate": 1.9931985554441754e-05,
+      "loss": 1.0596,
+      "step": 111
+    },
+    {
+      "epoch": 0.041040674239648224,
+      "grad_norm": 0.6959455609321594,
+      "learning_rate": 1.993063363323824e-05,
+      "loss": 1.085,
+      "step": 112
+    },
+    {
+      "epoch": 0.04140710883107365,
+      "grad_norm": 0.7528244256973267,
+      "learning_rate": 1.9929268454638326e-05,
+      "loss": 0.9813,
+      "step": 113
+    },
+    {
+      "epoch": 0.04177354342249908,
+      "grad_norm": 0.6986820697784424,
+      "learning_rate": 1.9927890020464535e-05,
+      "loss": 1.0718,
+      "step": 114
+    },
+    {
+      "epoch": 0.042139978013924516,
+      "grad_norm": 0.790373682975769,
+      "learning_rate": 1.992649833255707e-05,
+      "loss": 0.9903,
+      "step": 115
+    },
+    {
+      "epoch": 0.042506412605349944,
+      "grad_norm": 0.7352178692817688,
+      "learning_rate": 1.9925093392773837e-05,
+      "loss": 1.0697,
+      "step": 116
+    },
+    {
+      "epoch": 0.04287284719677537,
+      "grad_norm": 0.8071436882019043,
+      "learning_rate": 1.9923675202990428e-05,
+      "loss": 0.927,
+      "step": 117
+    },
+    {
+      "epoch": 0.04323928178820081,
+      "grad_norm": 0.912890613079071,
+      "learning_rate": 1.9922243765100133e-05,
+      "loss": 0.9728,
+      "step": 118
+    },
+    {
+      "epoch": 0.043605716379626236,
+      "grad_norm": 0.9719120860099792,
+      "learning_rate": 1.992079908101392e-05,
+      "loss": 0.989,
+      "step": 119
+    },
+    {
+      "epoch": 0.043972150971051664,
+      "grad_norm": 0.7970651984214783,
+      "learning_rate": 1.9919341152660433e-05,
+      "loss": 1.0516,
+      "step": 120
+    },
+    {
+      "epoch": 0.0443385855624771,
+      "grad_norm": 0.7409046292304993,
+      "learning_rate": 1.9917869981986014e-05,
+      "loss": 1.0232,
+      "step": 121
+    },
+    {
+      "epoch": 0.04470502015390253,
+      "grad_norm": 0.7631481885910034,
+      "learning_rate": 1.991638557095468e-05,
+      "loss": 1.0237,
+      "step": 122
+    },
+    {
+      "epoch": 0.045071454745327956,
+      "grad_norm": 0.7730304002761841,
+      "learning_rate": 1.9914887921548112e-05,
+      "loss": 0.9949,
+      "step": 123
+    },
+    {
+      "epoch": 0.04543788933675339,
+      "grad_norm": 0.8115593791007996,
+      "learning_rate": 1.9913377035765677e-05,
+      "loss": 1.0673,
+      "step": 124
+    },
+    {
+      "epoch": 0.04580432392817882,
+      "grad_norm": 0.8105559349060059,
+      "learning_rate": 1.9911852915624405e-05,
+      "loss": 1.0031,
+      "step": 125
+    },
+    {
+      "epoch": 0.04617075851960425,
+      "grad_norm": 0.807749330997467,
+      "learning_rate": 1.9910315563159e-05,
+      "loss": 1.0177,
+      "step": 126
+    },
+    {
+      "epoch": 0.04653719311102968,
+      "grad_norm": 0.8281621336936951,
+      "learning_rate": 1.990876498042182e-05,
+      "loss": 1.0317,
+      "step": 127
+    },
+    {
+      "epoch": 0.04690362770245511,
+      "grad_norm": 0.7854720950126648,
+      "learning_rate": 1.9907201169482894e-05,
+      "loss": 1.0084,
+      "step": 128
+    },
+    {
+      "epoch": 0.04727006229388054,
+      "grad_norm": 0.7980220913887024,
+      "learning_rate": 1.9905624132429922e-05,
+      "loss": 1.005,
+      "step": 129
+    },
+    {
+      "epoch": 0.047636496885305975,
+      "grad_norm": 0.7824934720993042,
+      "learning_rate": 1.9904033871368232e-05,
+      "loss": 1.0137,
+      "step": 130
+    },
+    {
+      "epoch": 0.0480029314767314,
+      "grad_norm": 0.8496504426002502,
+      "learning_rate": 1.990243038842084e-05,
+      "loss": 1.069,
+      "step": 131
+    },
+    {
+      "epoch": 0.04836936606815683,
+      "grad_norm": 0.8010658025741577,
+      "learning_rate": 1.990081368572838e-05,
+      "loss": 0.9737,
+      "step": 132
+    },
+    {
+      "epoch": 0.04873580065958227,
+      "grad_norm": 0.8724714517593384,
+      "learning_rate": 1.9899183765449157e-05,
+      "loss": 0.9907,
+      "step": 133
+    },
+    {
+      "epoch": 0.049102235251007695,
+      "grad_norm": 0.8537774085998535,
+      "learning_rate": 1.9897540629759116e-05,
+      "loss": 1.0092,
+      "step": 134
+    },
+    {
+      "epoch": 0.04946866984243312,
+      "grad_norm": 0.7882579565048218,
+      "learning_rate": 1.9895884280851843e-05,
+      "loss": 1.0597,
+      "step": 135
+    },
+    {
+      "epoch": 0.04983510443385856,
+      "grad_norm": 0.9498627781867981,
+      "learning_rate": 1.989421472093856e-05,
+      "loss": 0.8423,
+      "step": 136
+    },
+    {
+      "epoch": 0.05020153902528399,
+      "grad_norm": 0.774014413356781,
+      "learning_rate": 1.9892531952248132e-05,
+      "loss": 1.0462,
+      "step": 137
+    },
+    {
+      "epoch": 0.050567973616709415,
+      "grad_norm": 0.7702706456184387,
+      "learning_rate": 1.989083597702705e-05,
+      "loss": 1.0541,
+      "step": 138
+    },
+    {
+      "epoch": 0.05093440820813485,
+      "grad_norm": 0.8901228308677673,
+      "learning_rate": 1.9889126797539454e-05,
+      "loss": 0.9953,
+      "step": 139
+    },
+    {
+      "epoch": 0.05130084279956028,
+      "grad_norm": 0.9667432308197021,
+      "learning_rate": 1.988740441606709e-05,
+      "loss": 0.8933,
+      "step": 140
+    },
+    {
+      "epoch": 0.05166727739098571,
+      "grad_norm": 0.8098576068878174,
+      "learning_rate": 1.9885668834909335e-05,
+      "loss": 0.9936,
+      "step": 141
+    },
+    {
+      "epoch": 0.05203371198241114,
+      "grad_norm": 0.8478096723556519,
+      "learning_rate": 1.9883920056383195e-05,
+      "loss": 1.0868,
+      "step": 142
+    },
+    {
+      "epoch": 0.05240014657383657,
+      "grad_norm": 1.008238434791565,
+      "learning_rate": 1.9882158082823287e-05,
+      "loss": 0.9707,
+      "step": 143
+    },
+    {
+      "epoch": 0.052766581165262,
+      "grad_norm": 0.9024046063423157,
+      "learning_rate": 1.9880382916581848e-05,
+      "loss": 0.9777,
+      "step": 144
+    },
+    {
+      "epoch": 0.053133015756687434,
+      "grad_norm": 0.8705045580863953,
+      "learning_rate": 1.9878594560028725e-05,
+      "loss": 1.0243,
+      "step": 145
+    },
+    {
+      "epoch": 0.05349945034811286,
+      "grad_norm": 0.7713609337806702,
+      "learning_rate": 1.987679301555137e-05,
+      "loss": 1.011,
+      "step": 146
+    },
+    {
+      "epoch": 0.05386588493953829,
+      "grad_norm": 0.9589905142784119,
+      "learning_rate": 1.9874978285554846e-05,
+      "loss": 0.9648,
+      "step": 147
+    },
+    {
+      "epoch": 0.054232319530963725,
+      "grad_norm": 0.8619270920753479,
+      "learning_rate": 1.987315037246182e-05,
+      "loss": 0.9608,
+      "step": 148
+    },
+    {
+      "epoch": 0.054598754122389154,
+      "grad_norm": 0.789116382598877,
+      "learning_rate": 1.9871309278712558e-05,
+      "loss": 1.0212,
+      "step": 149
+    },
+    {
+      "epoch": 0.05496518871381458,
+      "grad_norm": 0.8430724740028381,
+      "learning_rate": 1.986945500676491e-05,
+      "loss": 1.0186,
+      "step": 150
+    },
+    {
+      "epoch": 0.05533162330524002,
+      "grad_norm": 0.9624394774436951,
+      "learning_rate": 1.9867587559094343e-05,
+      "loss": 0.9897,
+      "step": 151
+    },
+    {
+      "epoch": 0.055698057896665445,
+      "grad_norm": 0.9204763770103455,
+      "learning_rate": 1.986570693819389e-05,
+      "loss": 0.9457,
+      "step": 152
+    },
+    {
+      "epoch": 0.056064492488090874,
+      "grad_norm": 1.0764364004135132,
+      "learning_rate": 1.9863813146574186e-05,
+      "loss": 0.9249,
+      "step": 153
+    },
+    {
+      "epoch": 0.05643092707951631,
+      "grad_norm": 0.8509868383407593,
+      "learning_rate": 1.9861906186763443e-05,
+      "loss": 0.9667,
+      "step": 154
+    },
+    {
+      "epoch": 0.05679736167094174,
+      "grad_norm": 0.7725467681884766,
+      "learning_rate": 1.9859986061307443e-05,
+      "loss": 0.9902,
+      "step": 155
+    },
+    {
+      "epoch": 0.057163796262367166,
+      "grad_norm": 0.933456301689148,
+      "learning_rate": 1.985805277276956e-05,
+      "loss": 0.9678,
+      "step": 156
+    },
+    {
+      "epoch": 0.0575302308537926,
+      "grad_norm": 0.8996209502220154,
+      "learning_rate": 1.9856106323730742e-05,
+      "loss": 0.9796,
+      "step": 157
+    },
+    {
+      "epoch": 0.05789666544521803,
+      "grad_norm": 0.8602309823036194,
+      "learning_rate": 1.9854146716789488e-05,
+      "loss": 1.0265,
+      "step": 158
+    },
+    {
+      "epoch": 0.05826310003664346,
+      "grad_norm": 0.9406113624572754,
+      "learning_rate": 1.9852173954561875e-05,
+      "loss": 0.9452,
+      "step": 159
+    },
+    {
+      "epoch": 0.05862953462806889,
+      "grad_norm": 0.7956333160400391,
+      "learning_rate": 1.985018803968154e-05,
+      "loss": 1.061,
+      "step": 160
+    },
+    {
+      "epoch": 0.05899596921949432,
+      "grad_norm": 0.7723748087882996,
+      "learning_rate": 1.9848188974799686e-05,
+      "loss": 1.0782,
+      "step": 161
+    },
+    {
+      "epoch": 0.05936240381091975,
+      "grad_norm": 1.1234358549118042,
+      "learning_rate": 1.9846176762585058e-05,
+      "loss": 0.8828,
+      "step": 162
+    },
+    {
+      "epoch": 0.059728838402345184,
+      "grad_norm": 0.8539266586303711,
+      "learning_rate": 1.9844151405723962e-05,
+      "loss": 0.9929,
+      "step": 163
+    },
+    {
+      "epoch": 0.06009527299377061,
+      "grad_norm": 0.8535122871398926,
+      "learning_rate": 1.9842112906920247e-05,
+      "loss": 0.9884,
+      "step": 164
+    },
+    {
+      "epoch": 0.06046170758519604,
+      "grad_norm": 0.7588261365890503,
+      "learning_rate": 1.9840061268895315e-05,
+      "loss": 1.0081,
+      "step": 165
+    },
+    {
+      "epoch": 0.060828142176621476,
+      "grad_norm": 0.7827401757240295,
+      "learning_rate": 1.9837996494388098e-05,
+      "loss": 0.9491,
+      "step": 166
+    },
+    {
+      "epoch": 0.061194576768046904,
+      "grad_norm": 0.7755241990089417,
+      "learning_rate": 1.983591858615507e-05,
+      "loss": 1.056,
+      "step": 167
+    },
+    {
+      "epoch": 0.06156101135947233,
+      "grad_norm": 0.8388570547103882,
+      "learning_rate": 1.983382754697024e-05,
+      "loss": 0.9922,
+      "step": 168
+    },
+    {
+      "epoch": 0.06192744595089777,
+      "grad_norm": 0.9278474450111389,
+      "learning_rate": 1.983172337962515e-05,
+      "loss": 0.912,
+      "step": 169
+    },
+    {
+      "epoch": 0.062293880542323196,
+      "grad_norm": 0.847387969493866,
+      "learning_rate": 1.982960608692885e-05,
+      "loss": 1.0021,
+      "step": 170
+    },
+    {
+      "epoch": 0.06266031513374863,
+      "grad_norm": 0.8056350350379944,
+      "learning_rate": 1.982747567170794e-05,
+      "loss": 1.0151,
+      "step": 171
+    },
+    {
+      "epoch": 0.06302674972517405,
+      "grad_norm": 0.6489182114601135,
+      "learning_rate": 1.9825332136806516e-05,
+      "loss": 0.6902,
+      "step": 172
+    },
+    {
+      "epoch": 0.06339318431659949,
+      "grad_norm": 0.8523520827293396,
+      "learning_rate": 1.9823175485086203e-05,
+      "loss": 1.0431,
+      "step": 173
+    },
+    {
+      "epoch": 0.06375961890802492,
+      "grad_norm": 0.8439933061599731,
+      "learning_rate": 1.982100571942613e-05,
+      "loss": 1.0139,
+      "step": 174
+    },
+    {
+      "epoch": 0.06412605349945034,
+      "grad_norm": 0.7784318327903748,
+      "learning_rate": 1.9818822842722926e-05,
+      "loss": 0.9976,
+      "step": 175
+    },
+    {
+      "epoch": 0.06449248809087578,
+      "grad_norm": 0.9022171497344971,
+      "learning_rate": 1.981662685789074e-05,
+      "loss": 0.9754,
+      "step": 176
+    },
+    {
+      "epoch": 0.06485892268230121,
+      "grad_norm": 0.8426647186279297,
+      "learning_rate": 1.9814417767861216e-05,
+      "loss": 0.9734,
+      "step": 177
+    },
+    {
+      "epoch": 0.06522535727372664,
+      "grad_norm": 0.8823671340942383,
+      "learning_rate": 1.9812195575583478e-05,
+      "loss": 1.0413,
+      "step": 178
+    },
+    {
+      "epoch": 0.06559179186515207,
+      "grad_norm": 0.8977214694023132,
+      "learning_rate": 1.980996028402416e-05,
+      "loss": 0.959,
+      "step": 179
+    },
+    {
+      "epoch": 0.0659582264565775,
+      "grad_norm": 0.8814698457717896,
+      "learning_rate": 1.980771189616737e-05,
+      "loss": 0.9576,
+      "step": 180
+    },
+    {
+      "epoch": 0.06632466104800293,
+      "grad_norm": 1.0457162857055664,
+      "learning_rate": 1.9805450415014714e-05,
+      "loss": 0.9335,
+      "step": 181
+    },
+    {
+      "epoch": 0.06669109563942836,
+      "grad_norm": 0.9762332439422607,
+      "learning_rate": 1.9803175843585268e-05,
+      "loss": 0.944,
+      "step": 182
+    },
+    {
+      "epoch": 0.0670575302308538,
+      "grad_norm": 0.8236628770828247,
+      "learning_rate": 1.9800888184915582e-05,
+      "loss": 1.0402,
+      "step": 183
+    },
+    {
+      "epoch": 0.06742396482227922,
+      "grad_norm": 1.0684088468551636,
+      "learning_rate": 1.9798587442059676e-05,
+      "loss": 1.011,
+      "step": 184
+    },
+    {
+      "epoch": 0.06779039941370466,
+      "grad_norm": 0.9889763593673706,
+      "learning_rate": 1.979627361808905e-05,
+      "loss": 0.9743,
+      "step": 185
+    },
+    {
+      "epoch": 0.06815683400513009,
+      "grad_norm": 0.8603727221488953,
+      "learning_rate": 1.9793946716092658e-05,
+      "loss": 1.0072,
+      "step": 186
+    },
+    {
+      "epoch": 0.06852326859655551,
+      "grad_norm": 0.998291552066803,
+      "learning_rate": 1.979160673917691e-05,
+      "loss": 0.8558,
+      "step": 187
+    },
+    {
+      "epoch": 0.06888970318798095,
+      "grad_norm": 0.8995135426521301,
+      "learning_rate": 1.978925369046568e-05,
+      "loss": 1.0158,
+      "step": 188
+    },
+    {
+      "epoch": 0.06925613777940638,
+      "grad_norm": 1.0453314781188965,
+      "learning_rate": 1.9786887573100283e-05,
+      "loss": 0.9511,
+      "step": 189
+    },
+    {
+      "epoch": 0.0696225723708318,
+      "grad_norm": 0.883664608001709,
+      "learning_rate": 1.9784508390239492e-05,
+      "loss": 0.9613,
+      "step": 190
+    },
+    {
+      "epoch": 0.06998900696225724,
+      "grad_norm": 0.8757372498512268,
+      "learning_rate": 1.978211614505951e-05,
+      "loss": 1.0289,
+      "step": 191
+    },
+    {
+      "epoch": 0.07035544155368267,
+      "grad_norm": 0.9564153552055359,
+      "learning_rate": 1.977971084075399e-05,
+      "loss": 0.9337,
+      "step": 192
+    },
+    {
+      "epoch": 0.0707218761451081,
+      "grad_norm": 0.9205679893493652,
+      "learning_rate": 1.9777292480534008e-05,
+      "loss": 0.9985,
+      "step": 193
+    },
+    {
+      "epoch": 0.07108831073653353,
+      "grad_norm": 0.8132325410842896,
+      "learning_rate": 1.977486106762808e-05,
+      "loss": 0.9598,
+      "step": 194
+    },
+    {
+      "epoch": 0.07145474532795897,
+      "grad_norm": 0.9424161911010742,
+      "learning_rate": 1.9772416605282135e-05,
+      "loss": 0.9823,
+      "step": 195
+    },
+    {
+      "epoch": 0.07182117991938439,
+      "grad_norm": 0.9367775917053223,
+      "learning_rate": 1.9769959096759538e-05,
+      "loss": 1.0072,
+      "step": 196
+    },
+    {
+      "epoch": 0.07218761451080982,
+      "grad_norm": 0.8410287499427795,
+      "learning_rate": 1.9767488545341062e-05,
+      "loss": 1.0155,
+      "step": 197
+    },
+    {
+      "epoch": 0.07255404910223526,
+      "grad_norm": 1.0502501726150513,
+      "learning_rate": 1.9765004954324892e-05,
+      "loss": 0.9272,
+      "step": 198
+    },
+    {
+      "epoch": 0.07292048369366068,
+      "grad_norm": 0.9136397838592529,
+      "learning_rate": 1.976250832702662e-05,
+      "loss": 0.9435,
+      "step": 199
+    },
+    {
+      "epoch": 0.07328691828508611,
+      "grad_norm": 1.0021872520446777,
+      "learning_rate": 1.9759998666779245e-05,
+      "loss": 0.9796,
+      "step": 200
+    },
+    {
+      "epoch": 0.07365335287651155,
+      "grad_norm": 0.9359725713729858,
+      "learning_rate": 1.9757475976933165e-05,
+      "loss": 0.987,
+      "step": 201
+    },
+    {
+      "epoch": 0.07401978746793697,
+      "grad_norm": 1.0036094188690186,
+      "learning_rate": 1.9754940260856172e-05,
+      "loss": 0.8347,
+      "step": 202
+    },
+    {
+      "epoch": 0.0743862220593624,
+      "grad_norm": 0.9520359039306641,
+      "learning_rate": 1.975239152193344e-05,
+      "loss": 0.9691,
+      "step": 203
+    },
+    {
+      "epoch": 0.07475265665078784,
+      "grad_norm": 0.9743679761886597,
+      "learning_rate": 1.9749829763567545e-05,
+      "loss": 0.9516,
+      "step": 204
+    },
+    {
+      "epoch": 0.07511909124221326,
+      "grad_norm": 0.9027923941612244,
+      "learning_rate": 1.974725498917843e-05,
+      "loss": 0.9631,
+      "step": 205
+    },
+    {
+      "epoch": 0.0754855258336387,
+      "grad_norm": 0.9035953283309937,
+      "learning_rate": 1.9744667202203418e-05,
+      "loss": 1.0298,
+      "step": 206
+    },
+    {
+      "epoch": 0.07585196042506413,
+      "grad_norm": 0.9111297726631165,
+      "learning_rate": 1.9742066406097206e-05,
+      "loss": 0.9723,
+      "step": 207
+    },
+    {
+      "epoch": 0.07621839501648955,
+      "grad_norm": 0.8764698505401611,
+      "learning_rate": 1.9739452604331853e-05,
+      "loss": 1.0585,
+      "step": 208
+    },
+    {
+      "epoch": 0.07658482960791499,
+      "grad_norm": 1.0158146619796753,
+      "learning_rate": 1.9736825800396796e-05,
+      "loss": 1.0051,
+      "step": 209
+    },
+    {
+      "epoch": 0.07695126419934042,
+      "grad_norm": 0.9445521831512451,
+      "learning_rate": 1.9734185997798805e-05,
+      "loss": 1.0046,
+      "step": 210
+    },
+    {
+      "epoch": 0.07731769879076585,
+      "grad_norm": 1.1613709926605225,
+      "learning_rate": 1.9731533200062026e-05,
+      "loss": 0.8739,
+      "step": 211
+    },
+    {
+      "epoch": 0.07768413338219128,
+      "grad_norm": 0.8824067115783691,
+      "learning_rate": 1.9728867410727936e-05,
+      "loss": 1.0177,
+      "step": 212
+    },
+    {
+      "epoch": 0.07805056797361672,
+      "grad_norm": 0.8832942843437195,
+      "learning_rate": 1.9726188633355373e-05,
+      "loss": 1.1069,
+      "step": 213
+    },
+    {
+      "epoch": 0.07841700256504214,
+      "grad_norm": 0.895889163017273,
+      "learning_rate": 1.9723496871520494e-05,
+      "loss": 0.9416,
+      "step": 214
+    },
+    {
+      "epoch": 0.07878343715646757,
+      "grad_norm": 0.9018993377685547,
+      "learning_rate": 1.9720792128816812e-05,
+      "loss": 0.9647,
+      "step": 215
+    },
+    {
+      "epoch": 0.079149871747893,
+      "grad_norm": 0.9014832973480225,
+      "learning_rate": 1.9718074408855156e-05,
+      "loss": 1.0068,
+      "step": 216
+    },
+    {
+      "epoch": 0.07951630633931843,
+      "grad_norm": 0.9585338234901428,
+      "learning_rate": 1.9715343715263676e-05,
+      "loss": 1.0059,
+      "step": 217
+    },
+    {
+      "epoch": 0.07988274093074386,
+      "grad_norm": 0.8547597527503967,
+      "learning_rate": 1.9712600051687853e-05,
+      "loss": 1.0381,
+      "step": 218
+    },
+    {
+      "epoch": 0.08024917552216929,
+      "grad_norm": 0.9184234142303467,
+      "learning_rate": 1.9709843421790478e-05,
+      "loss": 0.9581,
+      "step": 219
+    },
+    {
+      "epoch": 0.08061561011359472,
+      "grad_norm": 0.9227030873298645,
+      "learning_rate": 1.970707382925165e-05,
+      "loss": 0.9675,
+      "step": 220
+    },
+    {
+      "epoch": 0.08098204470502016,
+      "grad_norm": 0.8544616103172302,
+      "learning_rate": 1.9704291277768777e-05,
+      "loss": 0.9977,
+      "step": 221
+    },
+    {
+      "epoch": 0.08134847929644558,
+      "grad_norm": 0.9003406167030334,
+      "learning_rate": 1.970149577105657e-05,
+      "loss": 0.9491,
+      "step": 222
+    },
+    {
+      "epoch": 0.08171491388787101,
+      "grad_norm": 0.9470295310020447,
+      "learning_rate": 1.969868731284702e-05,
+      "loss": 0.9453,
+      "step": 223
+    },
+    {
+      "epoch": 0.08208134847929645,
+      "grad_norm": 0.9368393421173096,
+      "learning_rate": 1.9695865906889428e-05,
+      "loss": 0.9945,
+      "step": 224
+    },
+    {
+      "epoch": 0.08244778307072187,
+      "grad_norm": 0.8223913908004761,
+      "learning_rate": 1.9693031556950367e-05,
+      "loss": 0.9732,
+      "step": 225
+    },
+    {
+      "epoch": 0.0828142176621473,
+      "grad_norm": 0.8167693018913269,
+      "learning_rate": 1.969018426681369e-05,
+      "loss": 0.9945,
+      "step": 226
+    },
+    {
+      "epoch": 0.08318065225357274,
+      "grad_norm": 0.9292685389518738,
+      "learning_rate": 1.968732404028054e-05,
+      "loss": 0.9716,
+      "step": 227
+    },
+    {
+      "epoch": 0.08354708684499816,
+      "grad_norm": 0.864932119846344,
+      "learning_rate": 1.968445088116931e-05,
+      "loss": 0.9755,
+      "step": 228
+    },
+    {
+      "epoch": 0.0839135214364236,
+      "grad_norm": 1.0593254566192627,
+      "learning_rate": 1.9681564793315676e-05,
+      "loss": 0.9336,
+      "step": 229
+    },
+    {
+      "epoch": 0.08427995602784903,
+      "grad_norm": 0.8241183161735535,
+      "learning_rate": 1.967866578057256e-05,
+      "loss": 0.9613,
+      "step": 230
+    },
+    {
+      "epoch": 0.08464639061927445,
+      "grad_norm": 0.9200891852378845,
+      "learning_rate": 1.967575384681014e-05,
+      "loss": 0.9725,
+      "step": 231
+    },
+    {
+      "epoch": 0.08501282521069989,
+      "grad_norm": 0.9276204109191895,
+      "learning_rate": 1.967282899591586e-05,
+      "loss": 0.9295,
+      "step": 232
+    },
+    {
+      "epoch": 0.08537925980212532,
+      "grad_norm": 1.0053272247314453,
+      "learning_rate": 1.9669891231794383e-05,
+      "loss": 0.9801,
+      "step": 233
+    },
+    {
+      "epoch": 0.08574569439355074,
+      "grad_norm": 0.9977024793624878,
+      "learning_rate": 1.9666940558367635e-05,
+      "loss": 0.9326,
+      "step": 234
+    },
+    {
+      "epoch": 0.08611212898497618,
+      "grad_norm": 0.9312626123428345,
+      "learning_rate": 1.9663976979574753e-05,
+      "loss": 0.895,
+      "step": 235
+    },
+    {
+      "epoch": 0.08647856357640162,
+      "grad_norm": 0.917569100856781,
+      "learning_rate": 1.9661000499372127e-05,
+      "loss": 0.9973,
+      "step": 236
+    },
+    {
+      "epoch": 0.08684499816782704,
+      "grad_norm": 1.0833723545074463,
+      "learning_rate": 1.9658011121733353e-05,
+      "loss": 0.9263,
+      "step": 237
+    },
+    {
+      "epoch": 0.08721143275925247,
+      "grad_norm": 0.8841248750686646,
+      "learning_rate": 1.9655008850649252e-05,
+      "loss": 0.9975,
+      "step": 238
+    },
+    {
+      "epoch": 0.08757786735067791,
+      "grad_norm": 0.9899241328239441,
+      "learning_rate": 1.965199369012785e-05,
+      "loss": 0.9774,
+      "step": 239
+    },
+    {
+      "epoch": 0.08794430194210333,
+      "grad_norm": 0.9991698265075684,
+      "learning_rate": 1.9648965644194396e-05,
+      "loss": 0.9299,
+      "step": 240
+    },
+    {
+      "epoch": 0.08831073653352876,
+      "grad_norm": 0.8291311264038086,
+      "learning_rate": 1.9645924716891326e-05,
+      "loss": 1.0274,
+      "step": 241
+    },
+    {
+      "epoch": 0.0886771711249542,
+      "grad_norm": 0.9432514309883118,
+      "learning_rate": 1.9642870912278275e-05,
+      "loss": 0.98,
+      "step": 242
+    },
+    {
+      "epoch": 0.08904360571637962,
+      "grad_norm": 0.9141779541969299,
+      "learning_rate": 1.963980423443208e-05,
+      "loss": 1.0134,
+      "step": 243
+    },
+    {
+      "epoch": 0.08941004030780506,
+      "grad_norm": 1.1006799936294556,
+      "learning_rate": 1.963672468744675e-05,
+      "loss": 0.9813,
+      "step": 244
+    },
+    {
+      "epoch": 0.08977647489923049,
+      "grad_norm": 1.022298812866211,
+      "learning_rate": 1.963363227543348e-05,
+      "loss": 0.9569,
+      "step": 245
+    },
+    {
+      "epoch": 0.09014290949065591,
+      "grad_norm": 0.8487532734870911,
+      "learning_rate": 1.9630527002520646e-05,
+      "loss": 1.0137,
+      "step": 246
+    },
+    {
+      "epoch": 0.09050934408208135,
+      "grad_norm": 1.1966487169265747,
+      "learning_rate": 1.9627408872853788e-05,
+      "loss": 0.8798,
+      "step": 247
+    },
+    {
+      "epoch": 0.09087577867350678,
+      "grad_norm": 0.9319120049476624,
+      "learning_rate": 1.96242778905956e-05,
+      "loss": 0.9981,
+      "step": 248
+    },
+    {
+      "epoch": 0.0912422132649322,
+      "grad_norm": 0.9164144992828369,
+      "learning_rate": 1.962113405992595e-05,
+      "loss": 0.9609,
+      "step": 249
+    },
+    {
+      "epoch": 0.09160864785635764,
+      "grad_norm": 1.06501042842865,
+      "learning_rate": 1.961797738504185e-05,
+      "loss": 0.9288,
+      "step": 250
+    },
+    {
+      "epoch": 0.09197508244778307,
+      "grad_norm": 0.9936481714248657,
+      "learning_rate": 1.961480787015746e-05,
+      "loss": 0.9782,
+      "step": 251
+    },
+    {
+      "epoch": 0.0923415170392085,
+      "grad_norm": 0.989457368850708,
+      "learning_rate": 1.961162551950409e-05,
+      "loss": 0.969,
+      "step": 252
+    },
+    {
+      "epoch": 0.09270795163063393,
+      "grad_norm": 1.0204919576644897,
+      "learning_rate": 1.9608430337330172e-05,
+      "loss": 1.0042,
+      "step": 253
+    },
+    {
+      "epoch": 0.09307438622205937,
+      "grad_norm": 0.9626120924949646,
+      "learning_rate": 1.9605222327901274e-05,
+      "loss": 0.9647,
+      "step": 254
+    },
+    {
+      "epoch": 0.09344082081348479,
+      "grad_norm": 0.9840561151504517,
+      "learning_rate": 1.9602001495500087e-05,
+      "loss": 0.9892,
+      "step": 255
+    },
+    {
+      "epoch": 0.09380725540491022,
+      "grad_norm": 0.9287786483764648,
+      "learning_rate": 1.959876784442643e-05,
+      "loss": 0.9548,
+      "step": 256
+    },
+    {
+      "epoch": 0.09417368999633566,
+      "grad_norm": 0.9564248323440552,
+      "learning_rate": 1.959552137899722e-05,
+      "loss": 0.9827,
+      "step": 257
+    },
+    {
+      "epoch": 0.09454012458776108,
+      "grad_norm": 0.9497889280319214,
+      "learning_rate": 1.959226210354649e-05,
+      "loss": 0.9986,
+      "step": 258
+    },
+    {
+      "epoch": 0.09490655917918651,
+      "grad_norm": 1.0784432888031006,
+      "learning_rate": 1.9588990022425378e-05,
+      "loss": 0.9232,
+      "step": 259
+    },
+    {
+      "epoch": 0.09527299377061195,
+      "grad_norm": 1.0418896675109863,
+      "learning_rate": 1.9585705140002107e-05,
+      "loss": 0.8728,
+      "step": 260
+    },
+    {
+      "epoch": 0.09563942836203737,
+      "grad_norm": 0.9301762580871582,
+      "learning_rate": 1.9582407460661995e-05,
+      "loss": 0.958,
+      "step": 261
+    },
+    {
+      "epoch": 0.0960058629534628,
+      "grad_norm": 1.0677168369293213,
+      "learning_rate": 1.9579096988807448e-05,
+      "loss": 0.8912,
+      "step": 262
+    },
+    {
+      "epoch": 0.09637229754488824,
+      "grad_norm": 0.9866990447044373,
+      "learning_rate": 1.9575773728857946e-05,
+      "loss": 0.955,
+      "step": 263
+    },
+    {
+      "epoch": 0.09673873213631366,
+      "grad_norm": 1.1164122819900513,
+      "learning_rate": 1.9572437685250038e-05,
+      "loss": 1.0106,
+      "step": 264
+    },
+    {
+      "epoch": 0.0971051667277391,
+      "grad_norm": 1.0720348358154297,
+      "learning_rate": 1.956908886243734e-05,
+      "loss": 0.9836,
+      "step": 265
+    },
+    {
+      "epoch": 0.09747160131916453,
+      "grad_norm": 1.0078245401382446,
+      "learning_rate": 1.956572726489054e-05,
+      "loss": 0.882,
+      "step": 266
+    },
+    {
+      "epoch": 0.09783803591058995,
+      "grad_norm": 0.9684966802597046,
+      "learning_rate": 1.9562352897097358e-05,
+      "loss": 1.0168,
+      "step": 267
+    },
+    {
+      "epoch": 0.09820447050201539,
+      "grad_norm": 1.0070780515670776,
+      "learning_rate": 1.9558965763562588e-05,
+      "loss": 0.9401,
+      "step": 268
+    },
+    {
+      "epoch": 0.09857090509344082,
+      "grad_norm": 1.065958023071289,
+      "learning_rate": 1.9555565868808047e-05,
+      "loss": 0.9685,
+      "step": 269
+    },
+    {
+      "epoch": 0.09893733968486625,
+      "grad_norm": 1.1749391555786133,
+      "learning_rate": 1.9552153217372592e-05,
+      "loss": 0.8862,
+      "step": 270
+    },
+    {
+      "epoch": 0.09930377427629168,
+      "grad_norm": 1.0479028224945068,
+      "learning_rate": 1.9548727813812117e-05,
+      "loss": 0.9078,
+      "step": 271
+    },
+    {
+      "epoch": 0.09967020886771712,
+      "grad_norm": 1.0470627546310425,
+      "learning_rate": 1.9545289662699537e-05,
+      "loss": 0.9752,
+      "step": 272
+    },
+    {
+      "epoch": 0.10003664345914254,
+      "grad_norm": 0.9723828434944153,
+      "learning_rate": 1.9541838768624785e-05,
+      "loss": 0.9602,
+      "step": 273
+    },
+    {
+      "epoch": 0.10040307805056797,
+      "grad_norm": 0.97696852684021,
+      "learning_rate": 1.9538375136194794e-05,
+      "loss": 0.9589,
+      "step": 274
+    },
+    {
+      "epoch": 0.10076951264199341,
+      "grad_norm": 0.9613191485404968,
+      "learning_rate": 1.9534898770033525e-05,
+      "loss": 0.9577,
+      "step": 275
+    },
+    {
+      "epoch": 0.10113594723341883,
+      "grad_norm": 0.932762086391449,
+      "learning_rate": 1.953140967478193e-05,
+      "loss": 0.9514,
+      "step": 276
+    },
+    {
+      "epoch": 0.10150238182484426,
+      "grad_norm": 0.9490351676940918,
+      "learning_rate": 1.9527907855097933e-05,
+      "loss": 1.0122,
+      "step": 277
+    },
+    {
+      "epoch": 0.1018688164162697,
+      "grad_norm": 0.9454355835914612,
+      "learning_rate": 1.9524393315656484e-05,
+      "loss": 0.9801,
+      "step": 278
+    },
+    {
+      "epoch": 0.10223525100769512,
+      "grad_norm": 0.9905188679695129,
+      "learning_rate": 1.952086606114948e-05,
+      "loss": 0.9325,
+      "step": 279
+    },
+    {
+      "epoch": 0.10260168559912056,
+      "grad_norm": 1.023087978363037,
+      "learning_rate": 1.9517326096285808e-05,
+      "loss": 0.9594,
+      "step": 280
+    },
+    {
+      "epoch": 0.10296812019054599,
+      "grad_norm": 0.9482645988464355,
+      "learning_rate": 1.951377342579133e-05,
+      "loss": 0.9682,
+      "step": 281
+    },
+    {
+      "epoch": 0.10333455478197141,
+      "grad_norm": 1.154199481010437,
+      "learning_rate": 1.951020805440885e-05,
+      "loss": 0.8358,
+      "step": 282
+    },
+    {
+      "epoch": 0.10370098937339685,
+      "grad_norm": 1.1318973302841187,
+      "learning_rate": 1.9506629986898144e-05,
+      "loss": 1.0204,
+      "step": 283
+    },
+    {
+      "epoch": 0.10406742396482228,
+      "grad_norm": 1.0413382053375244,
+      "learning_rate": 1.9503039228035936e-05,
+      "loss": 0.9936,
+      "step": 284
+    },
+    {
+      "epoch": 0.1044338585562477,
+      "grad_norm": 1.1161582469940186,
+      "learning_rate": 1.9499435782615882e-05,
+      "loss": 0.8052,
+      "step": 285
+    },
+    {
+      "epoch": 0.10480029314767314,
+      "grad_norm": 1.00542151927948,
+      "learning_rate": 1.9495819655448588e-05,
+      "loss": 0.9014,
+      "step": 286
+    },
+    {
+      "epoch": 0.10516672773909858,
+      "grad_norm": 1.0385977029800415,
+      "learning_rate": 1.9492190851361576e-05,
+      "loss": 0.995,
+      "step": 287
+    },
+    {
+      "epoch": 0.105533162330524,
+      "grad_norm": 1.0507909059524536,
+      "learning_rate": 1.948854937519931e-05,
+      "loss": 0.974,
+      "step": 288
+    },
+    {
+      "epoch": 0.10589959692194943,
+      "grad_norm": 0.9018959403038025,
+      "learning_rate": 1.9484895231823153e-05,
+      "loss": 0.9413,
+      "step": 289
+    },
+    {
+      "epoch": 0.10626603151337487,
+      "grad_norm": 0.9558141827583313,
+      "learning_rate": 1.9481228426111392e-05,
+      "loss": 0.978,
+      "step": 290
+    },
+    {
+      "epoch": 0.10663246610480029,
+      "grad_norm": 0.975355863571167,
+      "learning_rate": 1.9477548962959207e-05,
+      "loss": 1.0181,
+      "step": 291
+    },
+    {
+      "epoch": 0.10699890069622572,
+      "grad_norm": 0.9691599607467651,
+      "learning_rate": 1.9473856847278688e-05,
+      "loss": 0.9769,
+      "step": 292
+    },
+    {
+      "epoch": 0.10736533528765116,
+      "grad_norm": 1.0449618101119995,
+      "learning_rate": 1.94701520839988e-05,
+      "loss": 0.9867,
+      "step": 293
+    },
+    {
+      "epoch": 0.10773176987907658,
+      "grad_norm": 0.9220532178878784,
+      "learning_rate": 1.9466434678065416e-05,
+      "loss": 0.9805,
+      "step": 294
+    },
+    {
+      "epoch": 0.10809820447050202,
+      "grad_norm": 1.054179310798645,
+      "learning_rate": 1.946270463444126e-05,
+      "loss": 0.9521,
+      "step": 295
+    },
+    {
+      "epoch": 0.10846463906192745,
+      "grad_norm": 1.110876202583313,
+      "learning_rate": 1.9458961958105948e-05,
+      "loss": 0.9413,
+      "step": 296
+    },
+    {
+      "epoch": 0.10883107365335287,
+      "grad_norm": 0.9981120824813843,
+      "learning_rate": 1.945520665405595e-05,
+      "loss": 0.9817,
+      "step": 297
+    },
+    {
+      "epoch": 0.10919750824477831,
+      "grad_norm": 1.1343566179275513,
+      "learning_rate": 1.94514387273046e-05,
+      "loss": 0.9903,
+      "step": 298
+    },
+    {
+      "epoch": 0.10956394283620374,
+      "grad_norm": 1.007285237312317,
+      "learning_rate": 1.944765818288208e-05,
+      "loss": 1.0138,
+      "step": 299
+    },
+    {
+      "epoch": 0.10993037742762916,
+      "grad_norm": 1.0697396993637085,
+      "learning_rate": 1.944386502583541e-05,
+      "loss": 0.9171,
+      "step": 300
+    },
+    {
+      "epoch": 0.1102968120190546,
+      "grad_norm": 0.9989778995513916,
+      "learning_rate": 1.9440059261228463e-05,
+      "loss": 1.0735,
+      "step": 301
+    },
+    {
+      "epoch": 0.11066324661048003,
+      "grad_norm": 1.0370742082595825,
+      "learning_rate": 1.9436240894141933e-05,
+      "loss": 0.9576,
+      "step": 302
+    },
+    {
+      "epoch": 0.11102968120190546,
+      "grad_norm": 1.0512537956237793,
+      "learning_rate": 1.9432409929673336e-05,
+      "loss": 0.9519,
+      "step": 303
+    },
+    {
+      "epoch": 0.11139611579333089,
+      "grad_norm": 1.1380470991134644,
+      "learning_rate": 1.942856637293702e-05,
+      "loss": 0.93,
+      "step": 304
+    },
+    {
+      "epoch": 0.11176255038475633,
+      "grad_norm": 1.0147130489349365,
+      "learning_rate": 1.9424710229064117e-05,
+      "loss": 0.9551,
+      "step": 305
+    },
+    {
+      "epoch": 0.11212898497618175,
+      "grad_norm": 0.6644235849380493,
+      "learning_rate": 1.94208415032026e-05,
+      "loss": 0.6299,
+      "step": 306
+    },
+    {
+      "epoch": 0.11249541956760718,
+      "grad_norm": 1.1452592611312866,
+      "learning_rate": 1.94169602005172e-05,
+      "loss": 0.9289,
+      "step": 307
+    },
+    {
+      "epoch": 0.11286185415903262,
+      "grad_norm": 0.7639943361282349,
+      "learning_rate": 1.9413066326189463e-05,
+      "loss": 0.6014,
+      "step": 308
+    },
+    {
+      "epoch": 0.11322828875045804,
+      "grad_norm": 0.9849272966384888,
+      "learning_rate": 1.9409159885417717e-05,
+      "loss": 0.9686,
+      "step": 309
+    },
+    {
+      "epoch": 0.11359472334188347,
+      "grad_norm": 0.9223613142967224,
+      "learning_rate": 1.9405240883417052e-05,
+      "loss": 0.8893,
+      "step": 310
+    },
+    {
+      "epoch": 0.11396115793330891,
+      "grad_norm": 1.0126756429672241,
+      "learning_rate": 1.940130932541934e-05,
+      "loss": 0.9326,
+      "step": 311
+    },
+    {
+      "epoch": 0.11432759252473433,
+      "grad_norm": 1.0185225009918213,
+      "learning_rate": 1.939736521667321e-05,
+      "loss": 0.9744,
+      "step": 312
+    },
+    {
+      "epoch": 0.11469402711615977,
+      "grad_norm": 1.1406480073928833,
+      "learning_rate": 1.9393408562444048e-05,
+      "loss": 1.003,
+      "step": 313
+    },
+    {
+      "epoch": 0.1150604617075852,
+      "grad_norm": 1.078837513923645,
+      "learning_rate": 1.9389439368013987e-05,
+      "loss": 0.9552,
+      "step": 314
+    },
+    {
+      "epoch": 0.11542689629901062,
+      "grad_norm": 1.042822003364563,
+      "learning_rate": 1.9385457638681902e-05,
+      "loss": 0.9658,
+      "step": 315
+    },
+    {
+      "epoch": 0.11579333089043606,
+      "grad_norm": 0.9646812677383423,
+      "learning_rate": 1.9381463379763403e-05,
+      "loss": 0.9181,
+      "step": 316
+    },
+    {
+      "epoch": 0.1161597654818615,
+      "grad_norm": 1.1890491247177124,
+      "learning_rate": 1.9377456596590826e-05,
+      "loss": 0.885,
+      "step": 317
+    },
+    {
+      "epoch": 0.11652620007328691,
+      "grad_norm": 1.0086746215820312,
+      "learning_rate": 1.9373437294513225e-05,
+      "loss": 1.0218,
+      "step": 318
+    },
+    {
+      "epoch": 0.11689263466471235,
+      "grad_norm": 0.7628529071807861,
+      "learning_rate": 1.9369405478896367e-05,
+      "loss": 0.5956,
+      "step": 319
+    },
+    {
+      "epoch": 0.11725906925613779,
+      "grad_norm": 0.9907658100128174,
+      "learning_rate": 1.936536115512273e-05,
+      "loss": 0.9748,
+      "step": 320
+    },
+    {
+      "epoch": 0.1176255038475632,
+      "grad_norm": 0.9829033613204956,
+      "learning_rate": 1.9361304328591485e-05,
+      "loss": 0.9372,
+      "step": 321
+    },
+    {
+      "epoch": 0.11799193843898864,
+      "grad_norm": 0.9079287052154541,
+      "learning_rate": 1.9357235004718493e-05,
+      "loss": 0.9166,
+      "step": 322
+    },
+    {
+      "epoch": 0.11835837303041408,
+      "grad_norm": 0.9061397314071655,
+      "learning_rate": 1.9353153188936304e-05,
+      "loss": 0.913,
+      "step": 323
+    },
+    {
+      "epoch": 0.1187248076218395,
+      "grad_norm": 1.019878625869751,
+      "learning_rate": 1.9349058886694146e-05,
+      "loss": 0.9803,
+      "step": 324
+    },
+    {
+      "epoch": 0.11909124221326493,
+      "grad_norm": 1.2982869148254395,
+      "learning_rate": 1.9344952103457906e-05,
+      "loss": 0.9391,
+      "step": 325
+    },
+    {
+      "epoch": 0.11945767680469037,
+      "grad_norm": 1.0518271923065186,
+      "learning_rate": 1.9340832844710143e-05,
+      "loss": 0.9108,
+      "step": 326
+    },
+    {
+      "epoch": 0.11982411139611579,
+      "grad_norm": 1.0201774835586548,
+      "learning_rate": 1.9336701115950066e-05,
+      "loss": 0.9951,
+      "step": 327
+    },
+    {
+      "epoch": 0.12019054598754123,
+      "grad_norm": 1.065725564956665,
+      "learning_rate": 1.9332556922693537e-05,
+      "loss": 0.9327,
+      "step": 328
+    },
+    {
+      "epoch": 0.12055698057896666,
+      "grad_norm": 0.9934291839599609,
+      "learning_rate": 1.9328400270473046e-05,
+      "loss": 0.883,
+      "step": 329
+    },
+    {
+      "epoch": 0.12092341517039208,
+      "grad_norm": 1.004043459892273,
+      "learning_rate": 1.9324231164837733e-05,
+      "loss": 0.9671,
+      "step": 330
+    },
+    {
+      "epoch": 0.12128984976181752,
+      "grad_norm": 1.190549612045288,
+      "learning_rate": 1.9320049611353357e-05,
+      "loss": 0.8884,
+      "step": 331
+    },
+    {
+      "epoch": 0.12165628435324295,
+      "grad_norm": 1.0553498268127441,
+      "learning_rate": 1.9315855615602278e-05,
+      "loss": 1.0202,
+      "step": 332
+    },
+    {
+      "epoch": 0.12202271894466837,
+      "grad_norm": 1.1617465019226074,
+      "learning_rate": 1.931164918318349e-05,
+      "loss": 0.9008,
+      "step": 333
+    },
+    {
+      "epoch": 0.12238915353609381,
+      "grad_norm": 0.9961878657341003,
+      "learning_rate": 1.930743031971258e-05,
+      "loss": 1.0226,
+      "step": 334
+    },
+    {
+      "epoch": 0.12275558812751924,
+      "grad_norm": 1.0462695360183716,
+      "learning_rate": 1.9303199030821725e-05,
+      "loss": 0.9614,
+      "step": 335
+    },
+    {
+      "epoch": 0.12312202271894467,
+      "grad_norm": 1.0623806715011597,
+      "learning_rate": 1.9298955322159708e-05,
+      "loss": 0.9325,
+      "step": 336
+    },
+    {
+      "epoch": 0.1234884573103701,
+      "grad_norm": 1.0319890975952148,
+      "learning_rate": 1.9294699199391868e-05,
+      "loss": 1.0047,
+      "step": 337
+    },
+    {
+      "epoch": 0.12385489190179554,
+      "grad_norm": 1.0860097408294678,
+      "learning_rate": 1.9290430668200134e-05,
+      "loss": 0.977,
+      "step": 338
+    },
+    {
+      "epoch": 0.12422132649322096,
+      "grad_norm": 1.0328339338302612,
+      "learning_rate": 1.9286149734282998e-05,
+      "loss": 0.9675,
+      "step": 339
+    },
+    {
+      "epoch": 0.12458776108464639,
+      "grad_norm": 0.9981843829154968,
+      "learning_rate": 1.9281856403355503e-05,
+      "loss": 1.0054,
+      "step": 340
+    },
+    {
+      "epoch": 0.12495419567607183,
+      "grad_norm": 1.0083777904510498,
+      "learning_rate": 1.9277550681149246e-05,
+      "loss": 0.9567,
+      "step": 341
+    },
+    {
+      "epoch": 0.12532063026749726,
+      "grad_norm": 0.9963028430938721,
+      "learning_rate": 1.9273232573412372e-05,
+      "loss": 0.9788,
+      "step": 342
+    },
+    {
+      "epoch": 0.12568706485892267,
+      "grad_norm": 1.0055075883865356,
+      "learning_rate": 1.9268902085909547e-05,
+      "loss": 0.9923,
+      "step": 343
+    },
+    {
+      "epoch": 0.1260534994503481,
+      "grad_norm": 1.0248360633850098,
+      "learning_rate": 1.9264559224421975e-05,
+      "loss": 0.9404,
+      "step": 344
+    },
+    {
+      "epoch": 0.12641993404177354,
+      "grad_norm": 1.0669928789138794,
+      "learning_rate": 1.926020399474738e-05,
+      "loss": 0.9099,
+      "step": 345
+    },
+    {
+      "epoch": 0.12678636863319898,
+      "grad_norm": 1.0234222412109375,
+      "learning_rate": 1.925583640269999e-05,
+      "loss": 0.993,
+      "step": 346
+    },
+    {
+      "epoch": 0.1271528032246244,
+      "grad_norm": 1.0699959993362427,
+      "learning_rate": 1.925145645411054e-05,
+      "loss": 0.8935,
+      "step": 347
+    },
+    {
+      "epoch": 0.12751923781604985,
+      "grad_norm": 1.0531938076019287,
+      "learning_rate": 1.9247064154826268e-05,
+      "loss": 0.996,
+      "step": 348
+    },
+    {
+      "epoch": 0.12788567240747525,
+      "grad_norm": 1.0274754762649536,
+      "learning_rate": 1.9242659510710885e-05,
+      "loss": 0.9521,
+      "step": 349
+    },
+    {
+      "epoch": 0.1282521069989007,
+      "grad_norm": 0.9951926469802856,
+      "learning_rate": 1.9238242527644595e-05,
+      "loss": 0.9261,
+      "step": 350
+    },
+    {
+      "epoch": 0.12861854159032612,
+      "grad_norm": 1.2358548641204834,
+      "learning_rate": 1.923381321152407e-05,
+      "loss": 0.8526,
+      "step": 351
+    },
+    {
+      "epoch": 0.12898497618175156,
+      "grad_norm": 0.9344589114189148,
+      "learning_rate": 1.9229371568262447e-05,
+      "loss": 0.9233,
+      "step": 352
+    },
+    {
+      "epoch": 0.129351410773177,
+      "grad_norm": 1.0159389972686768,
+      "learning_rate": 1.9224917603789323e-05,
+      "loss": 0.9833,
+      "step": 353
+    },
+    {
+      "epoch": 0.12971784536460243,
+      "grad_norm": 1.0737959146499634,
+      "learning_rate": 1.922045132405074e-05,
+      "loss": 0.9128,
+      "step": 354
+    },
+    {
+      "epoch": 0.13008427995602784,
+      "grad_norm": 1.0360455513000488,
+      "learning_rate": 1.9215972735009178e-05,
+      "loss": 0.9552,
+      "step": 355
+    },
+    {
+      "epoch": 0.13045071454745327,
+      "grad_norm": 1.0206668376922607,
+      "learning_rate": 1.9211481842643557e-05,
+      "loss": 0.936,
+      "step": 356
+    },
+    {
+      "epoch": 0.1308171491388787,
+      "grad_norm": 1.0223139524459839,
+      "learning_rate": 1.920697865294922e-05,
+      "loss": 0.9847,
+      "step": 357
+    },
+    {
+      "epoch": 0.13118358373030414,
+      "grad_norm": 1.066007375717163,
+      "learning_rate": 1.9202463171937917e-05,
+      "loss": 0.9398,
+      "step": 358
+    },
+    {
+      "epoch": 0.13155001832172958,
+      "grad_norm": 1.1262496709823608,
+      "learning_rate": 1.9197935405637824e-05,
+      "loss": 0.951,
+      "step": 359
+    },
+    {
+      "epoch": 0.131916452913155,
+      "grad_norm": 0.9768517017364502,
+      "learning_rate": 1.9193395360093507e-05,
+      "loss": 0.915,
+      "step": 360
+    },
+    {
+      "epoch": 0.13228288750458042,
+      "grad_norm": 0.9572597146034241,
+      "learning_rate": 1.9188843041365927e-05,
+      "loss": 0.9051,
+      "step": 361
+    },
+    {
+      "epoch": 0.13264932209600586,
+      "grad_norm": 0.963023841381073,
+      "learning_rate": 1.9184278455532427e-05,
+      "loss": 0.9233,
+      "step": 362
+    },
+    {
+      "epoch": 0.1330157566874313,
+      "grad_norm": 1.2647531032562256,
+      "learning_rate": 1.9179701608686734e-05,
+      "loss": 0.895,
+      "step": 363
+    },
+    {
+      "epoch": 0.13338219127885673,
+      "grad_norm": 1.116768717765808,
+      "learning_rate": 1.9175112506938934e-05,
+      "loss": 0.8941,
+      "step": 364
+    },
+    {
+      "epoch": 0.13374862587028216,
+      "grad_norm": 1.072059154510498,
+      "learning_rate": 1.9170511156415485e-05,
+      "loss": 0.9278,
+      "step": 365
+    },
+    {
+      "epoch": 0.1341150604617076,
+      "grad_norm": 1.1003344058990479,
+      "learning_rate": 1.916589756325918e-05,
+      "loss": 0.8847,
+      "step": 366
+    },
+    {
+      "epoch": 0.134481495053133,
+      "grad_norm": 1.0306068658828735,
+      "learning_rate": 1.9161271733629174e-05,
+      "loss": 0.938,
+      "step": 367
+    },
+    {
+      "epoch": 0.13484792964455844,
+      "grad_norm": 1.0388495922088623,
+      "learning_rate": 1.915663367370095e-05,
+      "loss": 1.0279,
+      "step": 368
+    },
+    {
+      "epoch": 0.13521436423598387,
+      "grad_norm": 1.0442951917648315,
+      "learning_rate": 1.9151983389666312e-05,
+      "loss": 0.8832,
+      "step": 369
+    },
+    {
+      "epoch": 0.1355807988274093,
+      "grad_norm": 1.3007270097732544,
+      "learning_rate": 1.9147320887733392e-05,
+      "loss": 0.8581,
+      "step": 370
+    },
+    {
+      "epoch": 0.13594723341883475,
+      "grad_norm": 1.0682867765426636,
+      "learning_rate": 1.9142646174126634e-05,
+      "loss": 0.9914,
+      "step": 371
+    },
+    {
+      "epoch": 0.13631366801026018,
+      "grad_norm": 1.3072115182876587,
+      "learning_rate": 1.913795925508678e-05,
+      "loss": 1.019,
+      "step": 372
+    },
+    {
+      "epoch": 0.1366801026016856,
+      "grad_norm": 1.0528311729431152,
+      "learning_rate": 1.9133260136870864e-05,
+      "loss": 0.9199,
+      "step": 373
+    },
+    {
+      "epoch": 0.13704653719311102,
+      "grad_norm": 0.9990109205245972,
+      "learning_rate": 1.912854882575221e-05,
+      "loss": 0.9312,
+      "step": 374
+    },
+    {
+      "epoch": 0.13741297178453646,
+      "grad_norm": 1.0572776794433594,
+      "learning_rate": 1.9123825328020417e-05,
+      "loss": 0.8832,
+      "step": 375
+    },
+    {
+      "epoch": 0.1377794063759619,
+      "grad_norm": 1.1701194047927856,
+      "learning_rate": 1.9119089649981362e-05,
+      "loss": 0.8376,
+      "step": 376
+    },
+    {
+      "epoch": 0.13814584096738733,
+      "grad_norm": 1.025697946548462,
+      "learning_rate": 1.9114341797957166e-05,
+      "loss": 1.0191,
+      "step": 377
+    },
+    {
+      "epoch": 0.13851227555881276,
+      "grad_norm": 1.1118934154510498,
+      "learning_rate": 1.9109581778286217e-05,
+      "loss": 0.9342,
+      "step": 378
+    },
+    {
+      "epoch": 0.13887871015023817,
+      "grad_norm": 1.0030187368392944,
+      "learning_rate": 1.910480959732314e-05,
+      "loss": 0.9565,
+      "step": 379
+    },
+    {
+      "epoch": 0.1392451447416636,
+      "grad_norm": 0.994335949420929,
+      "learning_rate": 1.9100025261438798e-05,
+      "loss": 0.9429,
+      "step": 380
+    },
+    {
+      "epoch": 0.13961157933308904,
+      "grad_norm": 1.1278151273727417,
+      "learning_rate": 1.909522877702028e-05,
+      "loss": 0.9287,
+      "step": 381
+    },
+    {
+      "epoch": 0.13997801392451448,
+      "grad_norm": 1.0508943796157837,
+      "learning_rate": 1.909042015047089e-05,
+      "loss": 0.9015,
+      "step": 382
+    },
+    {
+      "epoch": 0.1403444485159399,
+      "grad_norm": 1.0135952234268188,
+      "learning_rate": 1.9085599388210148e-05,
+      "loss": 1.0046,
+      "step": 383
+    },
+    {
+      "epoch": 0.14071088310736535,
+      "grad_norm": 1.016960859298706,
+      "learning_rate": 1.908076649667377e-05,
+      "loss": 1.0349,
+      "step": 384
+    },
+    {
+      "epoch": 0.14107731769879076,
+      "grad_norm": 1.0783015489578247,
+      "learning_rate": 1.9075921482313668e-05,
+      "loss": 0.9434,
+      "step": 385
+    },
+    {
+      "epoch": 0.1414437522902162,
+      "grad_norm": 1.0530004501342773,
+      "learning_rate": 1.907106435159793e-05,
+      "loss": 0.8487,
+      "step": 386
+    },
+    {
+      "epoch": 0.14181018688164163,
+      "grad_norm": 1.109596848487854,
+      "learning_rate": 1.9066195111010834e-05,
+      "loss": 0.9367,
+      "step": 387
+    },
+    {
+      "epoch": 0.14217662147306706,
+      "grad_norm": 1.0377064943313599,
+      "learning_rate": 1.9061313767052815e-05,
+      "loss": 1.0746,
+      "step": 388
+    },
+    {
+      "epoch": 0.1425430560644925,
+      "grad_norm": 1.1181172132492065,
+      "learning_rate": 1.9056420326240466e-05,
+      "loss": 0.9222,
+      "step": 389
+    },
+    {
+      "epoch": 0.14290949065591793,
+      "grad_norm": 1.0245875120162964,
+      "learning_rate": 1.9051514795106528e-05,
+      "loss": 0.9345,
+      "step": 390
+    },
+    {
+      "epoch": 0.14327592524734334,
+      "grad_norm": 1.1641144752502441,
+      "learning_rate": 1.904659718019989e-05,
+      "loss": 0.9257,
+      "step": 391
+    },
+    {
+      "epoch": 0.14364235983876877,
+      "grad_norm": 1.039544701576233,
+      "learning_rate": 1.9041667488085566e-05,
+      "loss": 0.9521,
+      "step": 392
+    },
+    {
+      "epoch": 0.1440087944301942,
+      "grad_norm": 1.0379778146743774,
+      "learning_rate": 1.9036725725344693e-05,
+      "loss": 0.9192,
+      "step": 393
+    },
+    {
+      "epoch": 0.14437522902161964,
+      "grad_norm": 1.2251654863357544,
+      "learning_rate": 1.903177189857453e-05,
+      "loss": 0.8815,
+      "step": 394
+    },
+    {
+      "epoch": 0.14474166361304508,
+      "grad_norm": 1.0674530267715454,
+      "learning_rate": 1.902680601438843e-05,
+      "loss": 0.9159,
+      "step": 395
+    },
+    {
+      "epoch": 0.14510809820447051,
+      "grad_norm": 1.0083818435668945,
+      "learning_rate": 1.9021828079415852e-05,
+      "loss": 0.8784,
+      "step": 396
+    },
+    {
+      "epoch": 0.14547453279589592,
+      "grad_norm": 0.8914971947669983,
+      "learning_rate": 1.9016838100302336e-05,
+      "loss": 0.5819,
+      "step": 397
+    },
+    {
+      "epoch": 0.14584096738732136,
+      "grad_norm": 1.086569905281067,
+      "learning_rate": 1.9011836083709513e-05,
+      "loss": 0.9501,
+      "step": 398
+    },
+    {
+      "epoch": 0.1462074019787468,
+      "grad_norm": 1.1220741271972656,
+      "learning_rate": 1.9006822036315066e-05,
+      "loss": 0.8849,
+      "step": 399
+    },
+    {
+      "epoch": 0.14657383657017223,
+      "grad_norm": 1.1456749439239502,
+      "learning_rate": 1.900179596481275e-05,
+      "loss": 0.9831,
+      "step": 400
+    },
+    {
+      "epoch": 0.14694027116159766,
+      "grad_norm": 1.1781152486801147,
+      "learning_rate": 1.899675787591238e-05,
+      "loss": 0.9056,
+      "step": 401
+    },
+    {
+      "epoch": 0.1473067057530231,
+      "grad_norm": 1.0729511976242065,
+      "learning_rate": 1.899170777633979e-05,
+      "loss": 0.9269,
+      "step": 402
+    },
+    {
+      "epoch": 0.1476731403444485,
+      "grad_norm": 1.3148356676101685,
+      "learning_rate": 1.8986645672836873e-05,
+      "loss": 0.8364,
+      "step": 403
+    },
+    {
+      "epoch": 0.14803957493587394,
+      "grad_norm": 1.1030545234680176,
+      "learning_rate": 1.898157157216154e-05,
+      "loss": 0.9535,
+      "step": 404
+    },
+    {
+      "epoch": 0.14840600952729938,
+      "grad_norm": 1.0783032178878784,
+      "learning_rate": 1.8976485481087707e-05,
+      "loss": 0.9164,
+      "step": 405
+    },
+    {
+      "epoch": 0.1487724441187248,
+      "grad_norm": 1.2315967082977295,
+      "learning_rate": 1.8971387406405307e-05,
+      "loss": 0.864,
+      "step": 406
+    },
+    {
+      "epoch": 0.14913887871015025,
+      "grad_norm": 1.0555812120437622,
+      "learning_rate": 1.8966277354920278e-05,
+      "loss": 0.9607,
+      "step": 407
+    },
+    {
+      "epoch": 0.14950531330157568,
+      "grad_norm": 1.1461411714553833,
+      "learning_rate": 1.8961155333454534e-05,
+      "loss": 0.9076,
+      "step": 408
+    },
+    {
+      "epoch": 0.1498717478930011,
+      "grad_norm": 1.1050230264663696,
+      "learning_rate": 1.895602134884597e-05,
+      "loss": 1.0034,
+      "step": 409
+    },
+    {
+      "epoch": 0.15023818248442652,
+      "grad_norm": 1.09053373336792,
+      "learning_rate": 1.8950875407948462e-05,
+      "loss": 0.9454,
+      "step": 410
+    },
+    {
+      "epoch": 0.15060461707585196,
+      "grad_norm": 1.027106761932373,
+      "learning_rate": 1.8945717517631833e-05,
+      "loss": 0.8855,
+      "step": 411
+    },
+    {
+      "epoch": 0.1509710516672774,
+      "grad_norm": 0.9812170267105103,
+      "learning_rate": 1.894054768478188e-05,
+      "loss": 0.885,
+      "step": 412
+    },
+    {
+      "epoch": 0.15133748625870283,
+      "grad_norm": 0.9762282967567444,
+      "learning_rate": 1.893536591630032e-05,
+      "loss": 0.9308,
+      "step": 413
+    },
+    {
+      "epoch": 0.15170392085012827,
+      "grad_norm": 1.1730345487594604,
+      "learning_rate": 1.8930172219104815e-05,
+      "loss": 0.8527,
+      "step": 414
+    },
+    {
+      "epoch": 0.15207035544155367,
+      "grad_norm": 1.147405982017517,
+      "learning_rate": 1.8924966600128955e-05,
+      "loss": 0.9975,
+      "step": 415
+    },
+    {
+      "epoch": 0.1524367900329791,
+      "grad_norm": 1.1086004972457886,
+      "learning_rate": 1.8919749066322238e-05,
+      "loss": 0.9845,
+      "step": 416
+    },
+    {
+      "epoch": 0.15280322462440454,
+      "grad_norm": 1.0477776527404785,
+      "learning_rate": 1.8914519624650073e-05,
+      "loss": 0.9967,
+      "step": 417
+    },
+    {
+      "epoch": 0.15316965921582998,
+      "grad_norm": 1.0343297719955444,
+      "learning_rate": 1.8909278282093768e-05,
+      "loss": 1.0273,
+      "step": 418
+    },
+    {
+      "epoch": 0.1535360938072554,
+      "grad_norm": 1.105607271194458,
+      "learning_rate": 1.890402504565051e-05,
+      "loss": 0.9371,
+      "step": 419
+    },
+    {
+      "epoch": 0.15390252839868085,
+      "grad_norm": 1.175635814666748,
+      "learning_rate": 1.8898759922333374e-05,
+      "loss": 0.986,
+      "step": 420
+    },
+    {
+      "epoch": 0.15426896299010626,
+      "grad_norm": 1.0726004838943481,
+      "learning_rate": 1.88934829191713e-05,
+      "loss": 0.9664,
+      "step": 421
+    },
+    {
+      "epoch": 0.1546353975815317,
+      "grad_norm": 1.0392824411392212,
+      "learning_rate": 1.8888194043209082e-05,
+      "loss": 0.9196,
+      "step": 422
+    },
+    {
+      "epoch": 0.15500183217295713,
+      "grad_norm": 1.1240150928497314,
+      "learning_rate": 1.888289330150738e-05,
+      "loss": 0.92,
+      "step": 423
+    },
+    {
+      "epoch": 0.15536826676438256,
+      "grad_norm": 1.0844699144363403,
+      "learning_rate": 1.887758070114267e-05,
+      "loss": 0.9914,
+      "step": 424
+    },
+    {
+      "epoch": 0.155734701355808,
+      "grad_norm": 1.0138442516326904,
+      "learning_rate": 1.887225624920729e-05,
+      "loss": 0.9292,
+      "step": 425
+    },
+    {
+      "epoch": 0.15610113594723343,
+      "grad_norm": 1.0808322429656982,
+      "learning_rate": 1.8866919952809374e-05,
+      "loss": 0.8664,
+      "step": 426
+    },
+    {
+      "epoch": 0.15646757053865884,
+      "grad_norm": 0.9856568574905396,
+      "learning_rate": 1.8861571819072874e-05,
+      "loss": 0.9507,
+      "step": 427
+    },
+    {
+      "epoch": 0.15683400513008428,
+      "grad_norm": 1.2238532304763794,
+      "learning_rate": 1.885621185513756e-05,
+      "loss": 0.8829,
+      "step": 428
+    },
+    {
+      "epoch": 0.1572004397215097,
+      "grad_norm": 1.194239854812622,
+      "learning_rate": 1.8850840068158975e-05,
+      "loss": 1.0353,
+      "step": 429
+    },
+    {
+      "epoch": 0.15756687431293515,
+      "grad_norm": 1.12733793258667,
+      "learning_rate": 1.884545646530846e-05,
+      "loss": 0.903,
+      "step": 430
+    },
+    {
+      "epoch": 0.15793330890436058,
+      "grad_norm": 1.1467331647872925,
+      "learning_rate": 1.8840061053773124e-05,
+      "loss": 0.9051,
+      "step": 431
+    },
+    {
+      "epoch": 0.158299743495786,
+      "grad_norm": 1.0669034719467163,
+      "learning_rate": 1.883465384075584e-05,
+      "loss": 0.8962,
+      "step": 432
+    },
+    {
+      "epoch": 0.15866617808721142,
+      "grad_norm": 1.2676420211791992,
+      "learning_rate": 1.882923483347524e-05,
+      "loss": 0.965,
+      "step": 433
+    },
+    {
+      "epoch": 0.15903261267863686,
+      "grad_norm": 1.1339911222457886,
+      "learning_rate": 1.8823804039165697e-05,
+      "loss": 0.9037,
+      "step": 434
+    },
+    {
+      "epoch": 0.1593990472700623,
+      "grad_norm": 1.0925798416137695,
+      "learning_rate": 1.881836146507732e-05,
+      "loss": 0.9377,
+      "step": 435
+    },
+    {
+      "epoch": 0.15976548186148773,
+      "grad_norm": 1.2156996726989746,
+      "learning_rate": 1.8812907118475952e-05,
+      "loss": 0.8098,
+      "step": 436
+    },
+    {
+      "epoch": 0.16013191645291316,
+      "grad_norm": 1.1337549686431885,
+      "learning_rate": 1.8807441006643138e-05,
+      "loss": 0.9229,
+      "step": 437
+    },
+    {
+      "epoch": 0.16049835104433857,
+      "grad_norm": 1.2260011434555054,
+      "learning_rate": 1.8801963136876146e-05,
+      "loss": 0.9081,
+      "step": 438
+    },
+    {
+      "epoch": 0.160864785635764,
+      "grad_norm": 1.1918498277664185,
+      "learning_rate": 1.8796473516487923e-05,
+      "loss": 0.9247,
+      "step": 439
+    },
+    {
+      "epoch": 0.16123122022718944,
+      "grad_norm": 1.2885518074035645,
+      "learning_rate": 1.8790972152807122e-05,
+      "loss": 0.9089,
+      "step": 440
+    },
+    {
+      "epoch": 0.16159765481861488,
+      "grad_norm": 1.2859281301498413,
+      "learning_rate": 1.8785459053178056e-05,
+      "loss": 0.9167,
+      "step": 441
+    },
+    {
+      "epoch": 0.1619640894100403,
+      "grad_norm": 1.0627514123916626,
+      "learning_rate": 1.877993422496072e-05,
+      "loss": 0.8814,
+      "step": 442
+    },
+    {
+      "epoch": 0.16233052400146575,
+      "grad_norm": 1.1065446138381958,
+      "learning_rate": 1.8774397675530754e-05,
+      "loss": 0.9363,
+      "step": 443
+    },
+    {
+      "epoch": 0.16269695859289116,
+      "grad_norm": 1.2442569732666016,
+      "learning_rate": 1.8768849412279456e-05,
+      "loss": 0.9993,
+      "step": 444
+    },
+    {
+      "epoch": 0.1630633931843166,
+      "grad_norm": 0.9614419341087341,
+      "learning_rate": 1.8763289442613753e-05,
+      "loss": 0.9399,
+      "step": 445
+    },
+    {
+      "epoch": 0.16342982777574203,
+      "grad_norm": 1.0689506530761719,
+      "learning_rate": 1.875771777395621e-05,
+      "loss": 0.9789,
+      "step": 446
+    },
+    {
+      "epoch": 0.16379626236716746,
+      "grad_norm": 1.1324421167373657,
+      "learning_rate": 1.8752134413745003e-05,
+      "loss": 0.9286,
+      "step": 447
+    },
+    {
+      "epoch": 0.1641626969585929,
+      "grad_norm": 1.3980520963668823,
+      "learning_rate": 1.874653936943392e-05,
+      "loss": 0.8706,
+      "step": 448
+    },
+    {
+      "epoch": 0.16452913155001833,
+      "grad_norm": 1.0804977416992188,
+      "learning_rate": 1.8740932648492337e-05,
+      "loss": 0.9385,
+      "step": 449
+    },
+    {
+      "epoch": 0.16489556614144374,
+      "grad_norm": 1.0773653984069824,
+      "learning_rate": 1.8735314258405237e-05,
+      "loss": 0.9025,
+      "step": 450
+    },
+    {
+      "epoch": 0.16526200073286917,
+      "grad_norm": 1.056978464126587,
+      "learning_rate": 1.8729684206673167e-05,
+      "loss": 0.9076,
+      "step": 451
+    },
+    {
+      "epoch": 0.1656284353242946,
+      "grad_norm": 1.0701828002929688,
+      "learning_rate": 1.872404250081225e-05,
+      "loss": 0.9008,
+      "step": 452
+    },
+    {
+      "epoch": 0.16599486991572004,
+      "grad_norm": 1.309175729751587,
+      "learning_rate": 1.871838914835415e-05,
+      "loss": 0.9097,
+      "step": 453
+    },
+    {
+      "epoch": 0.16636130450714548,
+      "grad_norm": 0.9885631203651428,
+      "learning_rate": 1.871272415684611e-05,
+      "loss": 0.9232,
+      "step": 454
+    },
+    {
+      "epoch": 0.16672773909857092,
+      "grad_norm": 1.313955307006836,
+      "learning_rate": 1.8707047533850885e-05,
+      "loss": 0.7769,
+      "step": 455
+    },
+    {
+      "epoch": 0.16709417368999632,
+      "grad_norm": 1.122620701789856,
+      "learning_rate": 1.870135928694677e-05,
+      "loss": 0.893,
+      "step": 456
+    },
+    {
+      "epoch": 0.16746060828142176,
+      "grad_norm": 1.1263620853424072,
+      "learning_rate": 1.8695659423727574e-05,
+      "loss": 0.9634,
+      "step": 457
+    },
+    {
+      "epoch": 0.1678270428728472,
+      "grad_norm": 1.491286277770996,
+      "learning_rate": 1.8689947951802616e-05,
+      "loss": 0.7905,
+      "step": 458
+    },
+    {
+      "epoch": 0.16819347746427263,
+      "grad_norm": 1.1821624040603638,
+      "learning_rate": 1.868422487879671e-05,
+      "loss": 0.9576,
+      "step": 459
+    },
+    {
+      "epoch": 0.16855991205569806,
+      "grad_norm": 1.2670471668243408,
+      "learning_rate": 1.8678490212350154e-05,
+      "loss": 0.9215,
+      "step": 460
+    },
+    {
+      "epoch": 0.1689263466471235,
+      "grad_norm": 1.2155308723449707,
+      "learning_rate": 1.8672743960118733e-05,
+      "loss": 0.9208,
+      "step": 461
+    },
+    {
+      "epoch": 0.1692927812385489,
+      "grad_norm": 1.062874436378479,
+      "learning_rate": 1.8666986129773695e-05,
+      "loss": 0.94,
+      "step": 462
+    },
+    {
+      "epoch": 0.16965921582997434,
+      "grad_norm": 0.9979032874107361,
+      "learning_rate": 1.8661216729001743e-05,
+      "loss": 0.9396,
+      "step": 463
+    },
+    {
+      "epoch": 0.17002565042139978,
+      "grad_norm": 0.9946810007095337,
+      "learning_rate": 1.8655435765505023e-05,
+      "loss": 0.9787,
+      "step": 464
+    },
+    {
+      "epoch": 0.1703920850128252,
+      "grad_norm": 1.2013760805130005,
+      "learning_rate": 1.864964324700113e-05,
+      "loss": 0.9518,
+      "step": 465
+    },
+    {
+      "epoch": 0.17075851960425065,
+      "grad_norm": 1.1290384531021118,
+      "learning_rate": 1.8643839181223068e-05,
+      "loss": 0.9795,
+      "step": 466
+    },
+    {
+      "epoch": 0.17112495419567608,
+      "grad_norm": 1.229293704032898,
+      "learning_rate": 1.863802357591927e-05,
+      "loss": 0.9091,
+      "step": 467
+    },
+    {
+      "epoch": 0.1714913887871015,
+      "grad_norm": 1.096228003501892,
+      "learning_rate": 1.8632196438853567e-05,
+      "loss": 0.952,
+      "step": 468
+    },
+    {
+      "epoch": 0.17185782337852692,
+      "grad_norm": 1.2736438512802124,
+      "learning_rate": 1.862635777780519e-05,
+      "loss": 0.941,
+      "step": 469
+    },
+    {
+      "epoch": 0.17222425796995236,
+      "grad_norm": 1.2913713455200195,
+      "learning_rate": 1.862050760056875e-05,
+      "loss": 0.8991,
+      "step": 470
+    },
+    {
+      "epoch": 0.1725906925613778,
+      "grad_norm": 1.2449238300323486,
+      "learning_rate": 1.861464591495423e-05,
+      "loss": 0.9098,
+      "step": 471
+    },
+    {
+      "epoch": 0.17295712715280323,
+      "grad_norm": 1.1578012704849243,
+      "learning_rate": 1.860877272878699e-05,
+      "loss": 0.8868,
+      "step": 472
+    },
+    {
+      "epoch": 0.17332356174422867,
+      "grad_norm": 1.0383613109588623,
+      "learning_rate": 1.8602888049907725e-05,
+      "loss": 0.9155,
+      "step": 473
+    },
+    {
+      "epoch": 0.17368999633565407,
+      "grad_norm": 1.028679370880127,
+      "learning_rate": 1.859699188617249e-05,
+      "loss": 0.9295,
+      "step": 474
+    },
+    {
+      "epoch": 0.1740564309270795,
+      "grad_norm": 1.1471338272094727,
+      "learning_rate": 1.8591084245452654e-05,
+      "loss": 0.8889,
+      "step": 475
+    },
+    {
+      "epoch": 0.17442286551850494,
+      "grad_norm": 1.1970279216766357,
+      "learning_rate": 1.8585165135634926e-05,
+      "loss": 0.8554,
+      "step": 476
+    },
+    {
+      "epoch": 0.17478930010993038,
+      "grad_norm": 1.0609773397445679,
+      "learning_rate": 1.857923456462131e-05,
+      "loss": 0.8865,
+      "step": 477
+    },
+    {
+      "epoch": 0.17515573470135581,
+      "grad_norm": 1.135913610458374,
+      "learning_rate": 1.8573292540329126e-05,
+      "loss": 0.9691,
+      "step": 478
+    },
+    {
+      "epoch": 0.17552216929278125,
+      "grad_norm": 1.1295160055160522,
+      "learning_rate": 1.8567339070690972e-05,
+      "loss": 0.9278,
+      "step": 479
+    },
+    {
+      "epoch": 0.17588860388420666,
+      "grad_norm": 1.269704818725586,
+      "learning_rate": 1.856137416365473e-05,
+      "loss": 0.8976,
+      "step": 480
+    },
+    {
+      "epoch": 0.1762550384756321,
+      "grad_norm": 1.2077471017837524,
+      "learning_rate": 1.8555397827183557e-05,
+      "loss": 0.8924,
+      "step": 481
+    },
+    {
+      "epoch": 0.17662147306705753,
+      "grad_norm": 1.1305952072143555,
+      "learning_rate": 1.8549410069255863e-05,
+      "loss": 0.9248,
+      "step": 482
+    },
+    {
+      "epoch": 0.17698790765848296,
+      "grad_norm": 1.317244529724121,
+      "learning_rate": 1.8543410897865295e-05,
+      "loss": 0.8066,
+      "step": 483
+    },
+    {
+      "epoch": 0.1773543422499084,
+      "grad_norm": 1.138925552368164,
+      "learning_rate": 1.8537400321020757e-05,
+      "loss": 0.9526,
+      "step": 484
+    },
+    {
+      "epoch": 0.17772077684133383,
+      "grad_norm": 1.3495008945465088,
+      "learning_rate": 1.853137834674636e-05,
+      "loss": 0.8081,
+      "step": 485
+    },
+    {
+      "epoch": 0.17808721143275924,
+      "grad_norm": 1.4174842834472656,
+      "learning_rate": 1.852534498308145e-05,
+      "loss": 0.7942,
+      "step": 486
+    },
+    {
+      "epoch": 0.17845364602418468,
+      "grad_norm": 1.2117899656295776,
+      "learning_rate": 1.851930023808056e-05,
+      "loss": 0.9396,
+      "step": 487
+    },
+    {
+      "epoch": 0.1788200806156101,
+      "grad_norm": 1.273874282836914,
+      "learning_rate": 1.8513244119813428e-05,
+      "loss": 0.9593,
+      "step": 488
+    },
+    {
+      "epoch": 0.17918651520703555,
+      "grad_norm": 1.2929598093032837,
+      "learning_rate": 1.8507176636364973e-05,
+      "loss": 0.9401,
+      "step": 489
+    },
+    {
+      "epoch": 0.17955294979846098,
+      "grad_norm": 1.0998018980026245,
+      "learning_rate": 1.8501097795835284e-05,
+      "loss": 0.9874,
+      "step": 490
+    },
+    {
+      "epoch": 0.17991938438988642,
+      "grad_norm": 1.18282949924469,
+      "learning_rate": 1.8495007606339615e-05,
+      "loss": 0.9393,
+      "step": 491
+    },
+    {
+      "epoch": 0.18028581898131182,
+      "grad_norm": 1.1368458271026611,
+      "learning_rate": 1.8488906076008368e-05,
+      "loss": 0.9364,
+      "step": 492
+    },
+    {
+      "epoch": 0.18065225357273726,
+      "grad_norm": 1.1537795066833496,
+      "learning_rate": 1.8482793212987088e-05,
+      "loss": 0.9196,
+      "step": 493
+    },
+    {
+      "epoch": 0.1810186881641627,
+      "grad_norm": 1.2160780429840088,
+      "learning_rate": 1.8476669025436448e-05,
+      "loss": 0.8206,
+      "step": 494
+    },
+    {
+      "epoch": 0.18138512275558813,
+      "grad_norm": 1.1482170820236206,
+      "learning_rate": 1.8470533521532235e-05,
+      "loss": 0.9044,
+      "step": 495
+    },
+    {
+      "epoch": 0.18175155734701356,
+      "grad_norm": 1.2820298671722412,
+      "learning_rate": 1.846438670946535e-05,
+      "loss": 0.9536,
+      "step": 496
+    },
+    {
+      "epoch": 0.182117991938439,
+      "grad_norm": 1.358842134475708,
+      "learning_rate": 1.8458228597441783e-05,
+      "loss": 0.7656,
+      "step": 497
+    },
+    {
+      "epoch": 0.1824844265298644,
+      "grad_norm": 1.266168236732483,
+      "learning_rate": 1.8452059193682622e-05,
+      "loss": 0.9818,
+      "step": 498
+    },
+    {
+      "epoch": 0.18285086112128984,
+      "grad_norm": 1.104447841644287,
+      "learning_rate": 1.8445878506424013e-05,
+      "loss": 0.9381,
+      "step": 499
+    },
+    {
+      "epoch": 0.18321729571271528,
+      "grad_norm": 1.1257696151733398,
+      "learning_rate": 1.843968654391718e-05,
+      "loss": 0.8861,
+      "step": 500
+    },
+    {
+      "epoch": 0.1835837303041407,
+      "grad_norm": 1.0890069007873535,
+      "learning_rate": 1.843348331442839e-05,
+      "loss": 0.9278,
+      "step": 501
+    },
+    {
+      "epoch": 0.18395016489556615,
+      "grad_norm": 1.0580757856369019,
+      "learning_rate": 1.8427268826238954e-05,
+      "loss": 0.9762,
+      "step": 502
+    },
+    {
+      "epoch": 0.18431659948699158,
+      "grad_norm": 1.441726565361023,
+      "learning_rate": 1.8421043087645216e-05,
+      "loss": 0.8981,
+      "step": 503
+    },
+    {
+      "epoch": 0.184683034078417,
+      "grad_norm": 1.272265076637268,
+      "learning_rate": 1.8414806106958535e-05,
+      "loss": 0.9042,
+      "step": 504
+    },
+    {
+      "epoch": 0.18504946866984243,
+      "grad_norm": 1.1588517427444458,
+      "learning_rate": 1.8408557892505285e-05,
+      "loss": 0.9226,
+      "step": 505
+    },
+    {
+      "epoch": 0.18541590326126786,
+      "grad_norm": 1.3338249921798706,
+      "learning_rate": 1.840229845262683e-05,
+      "loss": 0.9488,
+      "step": 506
+    },
+    {
+      "epoch": 0.1857823378526933,
+      "grad_norm": 1.1822744607925415,
+      "learning_rate": 1.839602779567952e-05,
+      "loss": 0.8742,
+      "step": 507
+    },
+    {
+      "epoch": 0.18614877244411873,
+      "grad_norm": 1.1773830652236938,
+      "learning_rate": 1.8389745930034686e-05,
+      "loss": 0.9124,
+      "step": 508
+    },
+    {
+      "epoch": 0.18651520703554417,
+      "grad_norm": 1.115365982055664,
+      "learning_rate": 1.8383452864078618e-05,
+      "loss": 0.8988,
+      "step": 509
+    },
+    {
+      "epoch": 0.18688164162696957,
+      "grad_norm": 1.1186784505844116,
+      "learning_rate": 1.8377148606212558e-05,
+      "loss": 0.903,
+      "step": 510
+    },
+    {
+      "epoch": 0.187248076218395,
+      "grad_norm": 1.2818164825439453,
+      "learning_rate": 1.8370833164852694e-05,
+      "loss": 0.8521,
+      "step": 511
+    },
+    {
+      "epoch": 0.18761451080982045,
+      "grad_norm": 1.0865288972854614,
+      "learning_rate": 1.8364506548430133e-05,
+      "loss": 0.905,
+      "step": 512
+    },
+    {
+      "epoch": 0.18798094540124588,
+      "grad_norm": 1.1884108781814575,
+      "learning_rate": 1.835816876539092e-05,
+      "loss": 0.9714,
+      "step": 513
+    },
+    {
+      "epoch": 0.18834737999267132,
+      "grad_norm": 1.1109466552734375,
+      "learning_rate": 1.8351819824195988e-05,
+      "loss": 0.9337,
+      "step": 514
+    },
+    {
+      "epoch": 0.18871381458409675,
+      "grad_norm": 1.3398234844207764,
+      "learning_rate": 1.834545973332117e-05,
+      "loss": 1.011,
+      "step": 515
+    },
+    {
+      "epoch": 0.18908024917552216,
+      "grad_norm": 1.2325025796890259,
+      "learning_rate": 1.8339088501257194e-05,
+      "loss": 0.9101,
+      "step": 516
+    },
+    {
+      "epoch": 0.1894466837669476,
+      "grad_norm": 1.158435344696045,
+      "learning_rate": 1.8332706136509654e-05,
+      "loss": 0.9724,
+      "step": 517
+    },
+    {
+      "epoch": 0.18981311835837303,
+      "grad_norm": 1.3573687076568604,
+      "learning_rate": 1.832631264759901e-05,
+      "loss": 0.9551,
+      "step": 518
+    },
+    {
+      "epoch": 0.19017955294979846,
+      "grad_norm": 1.2095463275909424,
+      "learning_rate": 1.8319908043060562e-05,
+      "loss": 0.8611,
+      "step": 519
+    },
+    {
+      "epoch": 0.1905459875412239,
+      "grad_norm": 1.2500462532043457,
+      "learning_rate": 1.8313492331444468e-05,
+      "loss": 0.8936,
+      "step": 520
+    },
+    {
+      "epoch": 0.19091242213264933,
+      "grad_norm": 1.3058297634124756,
+      "learning_rate": 1.830706552131569e-05,
+      "loss": 0.9843,
+      "step": 521
+    },
+    {
+      "epoch": 0.19127885672407474,
+      "grad_norm": 1.47245454788208,
+      "learning_rate": 1.8300627621254038e-05,
+      "loss": 0.7822,
+      "step": 522
+    },
+    {
+      "epoch": 0.19164529131550018,
+      "grad_norm": 1.1628376245498657,
+      "learning_rate": 1.829417863985409e-05,
+      "loss": 0.9191,
+      "step": 523
+    },
+    {
+      "epoch": 0.1920117259069256,
+      "grad_norm": 1.1671644449234009,
+      "learning_rate": 1.828771858572525e-05,
+      "loss": 0.8991,
+      "step": 524
+    },
+    {
+      "epoch": 0.19237816049835105,
+      "grad_norm": 1.2752845287322998,
+      "learning_rate": 1.828124746749168e-05,
+      "loss": 0.844,
+      "step": 525
+    },
+    {
+      "epoch": 0.19274459508977648,
+      "grad_norm": 1.2671666145324707,
+      "learning_rate": 1.8274765293792328e-05,
+      "loss": 0.7955,
+      "step": 526
+    },
+    {
+      "epoch": 0.19311102968120192,
+      "grad_norm": 1.3220008611679077,
+      "learning_rate": 1.8268272073280902e-05,
+      "loss": 0.8403,
+      "step": 527
+    },
+    {
+      "epoch": 0.19347746427262733,
+      "grad_norm": 1.256608009338379,
+      "learning_rate": 1.8261767814625844e-05,
+      "loss": 0.9375,
+      "step": 528
+    },
+    {
+      "epoch": 0.19384389886405276,
+      "grad_norm": 1.2269988059997559,
+      "learning_rate": 1.8255252526510344e-05,
+      "loss": 0.9036,
+      "step": 529
+    },
+    {
+      "epoch": 0.1942103334554782,
+      "grad_norm": 1.3886499404907227,
+      "learning_rate": 1.824872621763231e-05,
+      "loss": 0.9434,
+      "step": 530
+    },
+    {
+      "epoch": 0.19457676804690363,
+      "grad_norm": 1.1359144449234009,
+      "learning_rate": 1.8242188896704362e-05,
+      "loss": 0.9783,
+      "step": 531
+    },
+    {
+      "epoch": 0.19494320263832907,
+      "grad_norm": 1.2621866464614868,
+      "learning_rate": 1.823564057245383e-05,
+      "loss": 0.9,
+      "step": 532
+    },
+    {
+      "epoch": 0.1953096372297545,
+      "grad_norm": 1.1670093536376953,
+      "learning_rate": 1.8229081253622718e-05,
+      "loss": 0.9509,
+      "step": 533
+    },
+    {
+      "epoch": 0.1956760718211799,
+      "grad_norm": 1.209097146987915,
+      "learning_rate": 1.8222510948967726e-05,
+      "loss": 0.9837,
+      "step": 534
+    },
+    {
+      "epoch": 0.19604250641260534,
+      "grad_norm": 1.249265193939209,
+      "learning_rate": 1.8215929667260208e-05,
+      "loss": 0.8725,
+      "step": 535
+    },
+    {
+      "epoch": 0.19640894100403078,
+      "grad_norm": 1.1079298257827759,
+      "learning_rate": 1.8209337417286173e-05,
+      "loss": 0.9533,
+      "step": 536
+    },
+    {
+      "epoch": 0.19677537559545621,
+      "grad_norm": 1.3172224760055542,
+      "learning_rate": 1.8202734207846284e-05,
+      "loss": 0.8488,
+      "step": 537
+    },
+    {
+      "epoch": 0.19714181018688165,
+      "grad_norm": 1.2152198553085327,
+      "learning_rate": 1.819612004775581e-05,
+      "loss": 0.9447,
+      "step": 538
+    },
+    {
+      "epoch": 0.19750824477830708,
+      "grad_norm": 1.367629885673523,
+      "learning_rate": 1.8189494945844673e-05,
+      "loss": 0.8947,
+      "step": 539
+    },
+    {
+      "epoch": 0.1978746793697325,
+      "grad_norm": 1.0554075241088867,
+      "learning_rate": 1.818285891095737e-05,
+      "loss": 0.9616,
+      "step": 540
+    },
+    {
+      "epoch": 0.19824111396115793,
+      "grad_norm": 1.3592339754104614,
+      "learning_rate": 1.8176211951953015e-05,
+      "loss": 0.7161,
+      "step": 541
+    },
+    {
+      "epoch": 0.19860754855258336,
+      "grad_norm": 1.1595011949539185,
+      "learning_rate": 1.8169554077705295e-05,
+      "loss": 0.9479,
+      "step": 542
+    },
+    {
+      "epoch": 0.1989739831440088,
+      "grad_norm": 1.3406450748443604,
+      "learning_rate": 1.8162885297102475e-05,
+      "loss": 0.8909,
+      "step": 543
+    },
+    {
+      "epoch": 0.19934041773543423,
+      "grad_norm": 1.242037296295166,
+      "learning_rate": 1.815620561904738e-05,
+      "loss": 0.9435,
+      "step": 544
+    },
+    {
+      "epoch": 0.19970685232685967,
+      "grad_norm": 1.229488492012024,
+      "learning_rate": 1.8149515052457375e-05,
+      "loss": 0.8382,
+      "step": 545
+    },
+    {
+      "epoch": 0.20007328691828508,
+      "grad_norm": 1.371599793434143,
+      "learning_rate": 1.814281360626437e-05,
+      "loss": 0.8936,
+      "step": 546
+    },
+    {
+      "epoch": 0.2004397215097105,
+      "grad_norm": 1.2121071815490723,
+      "learning_rate": 1.8136101289414797e-05,
+      "loss": 0.8484,
+      "step": 547
+    },
+    {
+      "epoch": 0.20080615610113595,
+      "grad_norm": 1.1973918676376343,
+      "learning_rate": 1.8129378110869592e-05,
+      "loss": 0.9279,
+      "step": 548
+    },
+    {
+      "epoch": 0.20117259069256138,
+      "grad_norm": 1.1278659105300903,
+      "learning_rate": 1.812264407960421e-05,
+      "loss": 0.8555,
+      "step": 549
+    },
+    {
+      "epoch": 0.20153902528398682,
+      "grad_norm": 1.1396849155426025,
+      "learning_rate": 1.8115899204608575e-05,
+      "loss": 0.8995,
+      "step": 550
+    },
+    {
+      "epoch": 0.20190545987541225,
+      "grad_norm": 1.102533221244812,
+      "learning_rate": 1.8109143494887097e-05,
+      "loss": 0.901,
+      "step": 551
+    },
+    {
+      "epoch": 0.20227189446683766,
+      "grad_norm": 1.2453352212905884,
+      "learning_rate": 1.810237695945865e-05,
+      "loss": 0.8631,
+      "step": 552
+    },
+    {
+      "epoch": 0.2026383290582631,
+      "grad_norm": 1.165541172027588,
+      "learning_rate": 1.8095599607356556e-05,
+      "loss": 0.9209,
+      "step": 553
+    },
+    {
+      "epoch": 0.20300476364968853,
+      "grad_norm": 1.2267589569091797,
+      "learning_rate": 1.808881144762859e-05,
+      "loss": 0.857,
+      "step": 554
+    },
+    {
+      "epoch": 0.20337119824111397,
+      "grad_norm": 1.1667150259017944,
+      "learning_rate": 1.8082012489336934e-05,
+      "loss": 0.9138,
+      "step": 555
+    },
+    {
+      "epoch": 0.2037376328325394,
+      "grad_norm": 1.3724040985107422,
+      "learning_rate": 1.80752027415582e-05,
+      "loss": 0.8979,
+      "step": 556
+    },
+    {
+      "epoch": 0.20410406742396484,
+      "grad_norm": 1.4671728610992432,
+      "learning_rate": 1.806838221338341e-05,
+      "loss": 0.8206,
+      "step": 557
+    },
+    {
+      "epoch": 0.20447050201539024,
+      "grad_norm": 1.176984429359436,
+      "learning_rate": 1.806155091391796e-05,
+      "loss": 0.9374,
+      "step": 558
+    },
+    {
+      "epoch": 0.20483693660681568,
+      "grad_norm": 1.1816294193267822,
+      "learning_rate": 1.8054708852281638e-05,
+      "loss": 0.9485,
+      "step": 559
+    },
+    {
+      "epoch": 0.2052033711982411,
+      "grad_norm": 1.2052361965179443,
+      "learning_rate": 1.80478560376086e-05,
+      "loss": 0.9411,
+      "step": 560
+    },
+    {
+      "epoch": 0.20556980578966655,
+      "grad_norm": 1.2356535196304321,
+      "learning_rate": 1.804099247904735e-05,
+      "loss": 0.9236,
+      "step": 561
+    },
+    {
+      "epoch": 0.20593624038109198,
+      "grad_norm": 1.0646589994430542,
+      "learning_rate": 1.803411818576074e-05,
+      "loss": 0.931,
+      "step": 562
+    },
+    {
+      "epoch": 0.20630267497251742,
+      "grad_norm": 1.207443118095398,
+      "learning_rate": 1.802723316692595e-05,
+      "loss": 0.9101,
+      "step": 563
+    },
+    {
+      "epoch": 0.20666910956394283,
+      "grad_norm": 1.25215744972229,
+      "learning_rate": 1.8020337431734484e-05,
+      "loss": 0.9273,
+      "step": 564
+    },
+    {
+      "epoch": 0.20703554415536826,
+      "grad_norm": 1.1377441883087158,
+      "learning_rate": 1.8013430989392146e-05,
+      "loss": 0.857,
+      "step": 565
+    },
+    {
+      "epoch": 0.2074019787467937,
+      "grad_norm": 1.185183048248291,
+      "learning_rate": 1.800651384911904e-05,
+      "loss": 0.9233,
+      "step": 566
+    },
+    {
+      "epoch": 0.20776841333821913,
+      "grad_norm": 1.4221135377883911,
+      "learning_rate": 1.7999586020149546e-05,
+      "loss": 0.8213,
+      "step": 567
+    },
+    {
+      "epoch": 0.20813484792964457,
+      "grad_norm": 1.2080090045928955,
+      "learning_rate": 1.799264751173232e-05,
+      "loss": 0.8868,
+      "step": 568
+    },
+    {
+      "epoch": 0.20850128252107,
+      "grad_norm": 1.3872716426849365,
+      "learning_rate": 1.7985698333130263e-05,
+      "loss": 0.8625,
+      "step": 569
+    },
+    {
+      "epoch": 0.2088677171124954,
+      "grad_norm": 1.171733021736145,
+      "learning_rate": 1.7978738493620534e-05,
+      "loss": 0.9601,
+      "step": 570
+    },
+    {
+      "epoch": 0.20923415170392085,
+      "grad_norm": 1.206907033920288,
+      "learning_rate": 1.797176800249452e-05,
+      "loss": 0.9184,
+      "step": 571
+    },
+    {
+      "epoch": 0.20960058629534628,
+      "grad_norm": 1.104447603225708,
+      "learning_rate": 1.7964786869057826e-05,
+      "loss": 0.9092,
+      "step": 572
+    },
+    {
+      "epoch": 0.20996702088677172,
+      "grad_norm": 1.2680176496505737,
+      "learning_rate": 1.7957795102630265e-05,
+      "loss": 0.9294,
+      "step": 573
+    },
+    {
+      "epoch": 0.21033345547819715,
+      "grad_norm": 1.1666079759597778,
+      "learning_rate": 1.7950792712545846e-05,
+      "loss": 0.9258,
+      "step": 574
+    },
+    {
+      "epoch": 0.21069989006962256,
+      "grad_norm": 1.1464416980743408,
+      "learning_rate": 1.7943779708152763e-05,
+      "loss": 0.9132,
+      "step": 575
+    },
+    {
+      "epoch": 0.211066324661048,
+      "grad_norm": 1.243110179901123,
+      "learning_rate": 1.793675609881337e-05,
+      "loss": 0.8997,
+      "step": 576
+    },
+    {
+      "epoch": 0.21143275925247343,
+      "grad_norm": 1.3148272037506104,
+      "learning_rate": 1.792972189390419e-05,
+      "loss": 0.8708,
+      "step": 577
+    },
+    {
+      "epoch": 0.21179919384389886,
+      "grad_norm": 1.1587334871292114,
+      "learning_rate": 1.7922677102815886e-05,
+      "loss": 0.9239,
+      "step": 578
+    },
+    {
+      "epoch": 0.2121656284353243,
+      "grad_norm": 1.2452987432479858,
+      "learning_rate": 1.7915621734953255e-05,
+      "loss": 0.9119,
+      "step": 579
+    },
+    {
+      "epoch": 0.21253206302674973,
+      "grad_norm": 1.1972370147705078,
+      "learning_rate": 1.7908555799735217e-05,
+      "loss": 0.9042,
+      "step": 580
+    },
+    {
+      "epoch": 0.21289849761817514,
+      "grad_norm": 1.3831322193145752,
+      "learning_rate": 1.7901479306594794e-05,
+      "loss": 0.9228,
+      "step": 581
+    },
+    {
+      "epoch": 0.21326493220960058,
+      "grad_norm": 1.3178349733352661,
+      "learning_rate": 1.78943922649791e-05,
+      "loss": 0.8248,
+      "step": 582
+    },
+    {
+      "epoch": 0.213631366801026,
+      "grad_norm": 1.0427035093307495,
+      "learning_rate": 1.788729468434934e-05,
+      "loss": 0.8347,
+      "step": 583
+    },
+    {
+      "epoch": 0.21399780139245145,
+      "grad_norm": 1.55295991897583,
+      "learning_rate": 1.7880186574180782e-05,
+      "loss": 0.9322,
+      "step": 584
+    },
+    {
+      "epoch": 0.21436423598387688,
+      "grad_norm": 1.244181513786316,
+      "learning_rate": 1.7873067943962756e-05,
+      "loss": 0.8777,
+      "step": 585
+    },
+    {
+      "epoch": 0.21473067057530232,
+      "grad_norm": 1.3763676881790161,
+      "learning_rate": 1.7865938803198635e-05,
+      "loss": 0.9281,
+      "step": 586
+    },
+    {
+      "epoch": 0.21509710516672773,
+      "grad_norm": 1.5503166913986206,
+      "learning_rate": 1.7858799161405818e-05,
+      "loss": 0.9044,
+      "step": 587
+    },
+    {
+      "epoch": 0.21546353975815316,
+      "grad_norm": 1.391393780708313,
+      "learning_rate": 1.7851649028115728e-05,
+      "loss": 0.8613,
+      "step": 588
+    },
+    {
+      "epoch": 0.2158299743495786,
+      "grad_norm": 1.203378677368164,
+      "learning_rate": 1.7844488412873798e-05,
+      "loss": 0.84,
+      "step": 589
+    },
+    {
+      "epoch": 0.21619640894100403,
+      "grad_norm": 1.3878751993179321,
+      "learning_rate": 1.783731732523944e-05,
+      "loss": 0.9575,
+      "step": 590
+    },
+    {
+      "epoch": 0.21656284353242947,
+      "grad_norm": 1.523260235786438,
+      "learning_rate": 1.783013577478607e-05,
+      "loss": 0.88,
+      "step": 591
+    },
+    {
+      "epoch": 0.2169292781238549,
+      "grad_norm": 1.2326833009719849,
+      "learning_rate": 1.7822943771101047e-05,
+      "loss": 0.8595,
+      "step": 592
+    },
+    {
+      "epoch": 0.2172957127152803,
+      "grad_norm": 1.4009530544281006,
+      "learning_rate": 1.78157413237857e-05,
+      "loss": 0.8396,
+      "step": 593
+    },
+    {
+      "epoch": 0.21766214730670574,
+      "grad_norm": 1.1550780534744263,
+      "learning_rate": 1.78085284424553e-05,
+      "loss": 0.903,
+      "step": 594
+    },
+    {
+      "epoch": 0.21802858189813118,
+      "grad_norm": 1.2234489917755127,
+      "learning_rate": 1.7801305136739038e-05,
+      "loss": 0.8886,
+      "step": 595
+    },
+    {
+      "epoch": 0.21839501648955661,
+      "grad_norm": 1.4011855125427246,
+      "learning_rate": 1.779407141628003e-05,
+      "loss": 0.8986,
+      "step": 596
+    },
+    {
+      "epoch": 0.21876145108098205,
+      "grad_norm": 1.4119914770126343,
+      "learning_rate": 1.7786827290735295e-05,
+      "loss": 0.8956,
+      "step": 597
+    },
+    {
+      "epoch": 0.21912788567240749,
+      "grad_norm": 1.356140375137329,
+      "learning_rate": 1.7779572769775738e-05,
+      "loss": 0.915,
+      "step": 598
+    },
+    {
+      "epoch": 0.2194943202638329,
+      "grad_norm": 1.4082356691360474,
+      "learning_rate": 1.777230786308614e-05,
+      "loss": 0.8769,
+      "step": 599
+    },
+    {
+      "epoch": 0.21986075485525833,
+      "grad_norm": 1.1782652139663696,
+      "learning_rate": 1.7765032580365167e-05,
+      "loss": 0.9401,
+      "step": 600
+    },
+    {
+      "epoch": 0.22022718944668376,
+      "grad_norm": 1.3177205324172974,
+      "learning_rate": 1.775774693132531e-05,
+      "loss": 0.8825,
+      "step": 601
+    },
+    {
+      "epoch": 0.2205936240381092,
+      "grad_norm": 1.1520065069198608,
+      "learning_rate": 1.775045092569291e-05,
+      "loss": 0.8783,
+      "step": 602
+    },
+    {
+      "epoch": 0.22096005862953463,
+      "grad_norm": 1.222068428993225,
+      "learning_rate": 1.774314457320814e-05,
+      "loss": 0.9228,
+      "step": 603
+    },
+    {
+      "epoch": 0.22132649322096007,
+      "grad_norm": 1.2122035026550293,
+      "learning_rate": 1.773582788362498e-05,
+      "loss": 0.9379,
+      "step": 604
+    },
+    {
+      "epoch": 0.22169292781238548,
+      "grad_norm": 1.2832188606262207,
+      "learning_rate": 1.772850086671121e-05,
+      "loss": 0.9291,
+      "step": 605
+    },
+    {
+      "epoch": 0.2220593624038109,
+      "grad_norm": 1.3288195133209229,
+      "learning_rate": 1.7721163532248397e-05,
+      "loss": 0.9414,
+      "step": 606
+    },
+    {
+      "epoch": 0.22242579699523635,
+      "grad_norm": 1.1746708154678345,
+      "learning_rate": 1.7713815890031888e-05,
+      "loss": 0.8746,
+      "step": 607
+    },
+    {
+      "epoch": 0.22279223158666178,
+      "grad_norm": 0.9558817744255066,
+      "learning_rate": 1.7706457949870777e-05,
+      "loss": 0.5374,
+      "step": 608
+    },
+    {
+      "epoch": 0.22315866617808722,
+      "grad_norm": 1.473559856414795,
+      "learning_rate": 1.7699089721587925e-05,
+      "loss": 0.7729,
+      "step": 609
+    },
+    {
+      "epoch": 0.22352510076951265,
+      "grad_norm": 1.2237671613693237,
+      "learning_rate": 1.7691711215019905e-05,
+      "loss": 0.9161,
+      "step": 610
+    },
+    {
+      "epoch": 0.22389153536093806,
+      "grad_norm": 1.43886399269104,
+      "learning_rate": 1.768432244001703e-05,
+      "loss": 0.88,
+      "step": 611
+    },
+    {
+      "epoch": 0.2242579699523635,
+      "grad_norm": 1.2342700958251953,
+      "learning_rate": 1.7676923406443318e-05,
+      "loss": 0.9185,
+      "step": 612
+    },
+    {
+      "epoch": 0.22462440454378893,
+      "grad_norm": 1.2138959169387817,
+      "learning_rate": 1.7669514124176476e-05,
+      "loss": 0.9151,
+      "step": 613
+    },
+    {
+      "epoch": 0.22499083913521437,
+      "grad_norm": 1.2613773345947266,
+      "learning_rate": 1.7662094603107894e-05,
+      "loss": 0.9497,
+      "step": 614
+    },
+    {
+      "epoch": 0.2253572737266398,
+      "grad_norm": 1.2515242099761963,
+      "learning_rate": 1.765466485314264e-05,
+      "loss": 0.9384,
+      "step": 615
+    },
+    {
+      "epoch": 0.22572370831806524,
+      "grad_norm": 1.3340543508529663,
+      "learning_rate": 1.7647224884199423e-05,
+      "loss": 0.9223,
+      "step": 616
+    },
+    {
+      "epoch": 0.22609014290949064,
+      "grad_norm": 1.1793503761291504,
+      "learning_rate": 1.7639774706210605e-05,
+      "loss": 0.9514,
+      "step": 617
+    },
+    {
+      "epoch": 0.22645657750091608,
+      "grad_norm": 1.1925681829452515,
+      "learning_rate": 1.7632314329122172e-05,
+      "loss": 0.9048,
+      "step": 618
+    },
+    {
+      "epoch": 0.2268230120923415,
+      "grad_norm": 1.3044099807739258,
+      "learning_rate": 1.7624843762893734e-05,
+      "loss": 0.858,
+      "step": 619
+    },
+    {
+      "epoch": 0.22718944668376695,
+      "grad_norm": 1.2997506856918335,
+      "learning_rate": 1.761736301749849e-05,
+      "loss": 0.9041,
+      "step": 620
+    },
+    {
+      "epoch": 0.22755588127519238,
+      "grad_norm": 1.2851029634475708,
+      "learning_rate": 1.7609872102923237e-05,
+      "loss": 0.9003,
+      "step": 621
+    },
+    {
+      "epoch": 0.22792231586661782,
+      "grad_norm": 1.3808014392852783,
+      "learning_rate": 1.760237102916835e-05,
+      "loss": 0.8519,
+      "step": 622
+    },
+    {
+      "epoch": 0.22828875045804323,
+      "grad_norm": 1.4921283721923828,
+      "learning_rate": 1.759485980624776e-05,
+      "loss": 0.8629,
+      "step": 623
+    },
+    {
+      "epoch": 0.22865518504946866,
+      "grad_norm": 1.1690940856933594,
+      "learning_rate": 1.758733844418895e-05,
+      "loss": 0.9156,
+      "step": 624
+    },
+    {
+      "epoch": 0.2290216196408941,
+      "grad_norm": 1.3847805261611938,
+      "learning_rate": 1.757980695303294e-05,
+      "loss": 0.8851,
+      "step": 625
+    },
+    {
+      "epoch": 0.22938805423231953,
+      "grad_norm": 1.3306790590286255,
+      "learning_rate": 1.757226534283427e-05,
+      "loss": 0.9196,
+      "step": 626
+    },
+    {
+      "epoch": 0.22975448882374497,
+      "grad_norm": 1.3025264739990234,
+      "learning_rate": 1.7564713623660988e-05,
+      "loss": 0.8605,
+      "step": 627
+    },
+    {
+      "epoch": 0.2301209234151704,
+      "grad_norm": 1.1992818117141724,
+      "learning_rate": 1.7557151805594647e-05,
+      "loss": 0.9457,
+      "step": 628
+    },
+    {
+      "epoch": 0.2304873580065958,
+      "grad_norm": 1.2719687223434448,
+      "learning_rate": 1.7549579898730266e-05,
+      "loss": 0.928,
+      "step": 629
+    },
+    {
+      "epoch": 0.23085379259802125,
+      "grad_norm": 1.1798957586288452,
+      "learning_rate": 1.7541997913176345e-05,
+      "loss": 0.9542,
+      "step": 630
+    },
+    {
+      "epoch": 0.23122022718944668,
+      "grad_norm": 1.288134217262268,
+      "learning_rate": 1.753440585905483e-05,
+      "loss": 0.9595,
+      "step": 631
+    },
+    {
+      "epoch": 0.23158666178087212,
+      "grad_norm": 1.3597036600112915,
+      "learning_rate": 1.7526803746501125e-05,
+      "loss": 0.8699,
+      "step": 632
+    },
+    {
+      "epoch": 0.23195309637229755,
+      "grad_norm": 1.2226777076721191,
+      "learning_rate": 1.751919158566404e-05,
+      "loss": 0.9487,
+      "step": 633
+    },
+    {
+      "epoch": 0.232319530963723,
+      "grad_norm": 1.3789327144622803,
+      "learning_rate": 1.7511569386705816e-05,
+      "loss": 0.8582,
+      "step": 634
+    },
+    {
+      "epoch": 0.2326859655551484,
+      "grad_norm": 1.6425267457962036,
+      "learning_rate": 1.7503937159802084e-05,
+      "loss": 0.7309,
+      "step": 635
+    },
+    {
+      "epoch": 0.23305240014657383,
+      "grad_norm": 1.3353691101074219,
+      "learning_rate": 1.7496294915141875e-05,
+      "loss": 0.9254,
+      "step": 636
+    },
+    {
+      "epoch": 0.23341883473799926,
+      "grad_norm": 1.4697070121765137,
+      "learning_rate": 1.7488642662927574e-05,
+      "loss": 0.955,
+      "step": 637
+    },
+    {
+      "epoch": 0.2337852693294247,
+      "grad_norm": 1.4180868864059448,
+      "learning_rate": 1.748098041337495e-05,
+      "loss": 0.9065,
+      "step": 638
+    },
+    {
+      "epoch": 0.23415170392085013,
+      "grad_norm": 1.406212568283081,
+      "learning_rate": 1.74733081767131e-05,
+      "loss": 0.9334,
+      "step": 639
+    },
+    {
+      "epoch": 0.23451813851227557,
+      "grad_norm": 1.3697319030761719,
+      "learning_rate": 1.7465625963184464e-05,
+      "loss": 0.8559,
+      "step": 640
+    },
+    {
+      "epoch": 0.23488457310370098,
+      "grad_norm": 1.55168616771698,
+      "learning_rate": 1.745793378304479e-05,
+      "loss": 0.8157,
+      "step": 641
+    },
+    {
+      "epoch": 0.2352510076951264,
+      "grad_norm": 1.291481375694275,
+      "learning_rate": 1.7450231646563148e-05,
+      "loss": 0.9607,
+      "step": 642
+    },
+    {
+      "epoch": 0.23561744228655185,
+      "grad_norm": 1.2459636926651,
+      "learning_rate": 1.744251956402189e-05,
+      "loss": 0.9208,
+      "step": 643
+    },
+    {
+      "epoch": 0.23598387687797728,
+      "grad_norm": 1.4352821111679077,
+      "learning_rate": 1.7434797545716636e-05,
+      "loss": 0.905,
+      "step": 644
+    },
+    {
+      "epoch": 0.23635031146940272,
+      "grad_norm": 1.3828191757202148,
+      "learning_rate": 1.7427065601956292e-05,
+      "loss": 0.8894,
+      "step": 645
+    },
+    {
+      "epoch": 0.23671674606082815,
+      "grad_norm": 1.2361446619033813,
+      "learning_rate": 1.7419323743063e-05,
+      "loss": 0.9492,
+      "step": 646
+    },
+    {
+      "epoch": 0.23708318065225356,
+      "grad_norm": 1.5865951776504517,
+      "learning_rate": 1.7411571979372143e-05,
+      "loss": 0.869,
+      "step": 647
+    },
+    {
+      "epoch": 0.237449615243679,
+      "grad_norm": 1.345902681350708,
+      "learning_rate": 1.7403810321232324e-05,
+      "loss": 0.953,
+      "step": 648
+    },
+    {
+      "epoch": 0.23781604983510443,
+      "grad_norm": 1.2080811262130737,
+      "learning_rate": 1.7396038779005358e-05,
+      "loss": 0.9318,
+      "step": 649
+    },
+    {
+      "epoch": 0.23818248442652987,
+      "grad_norm": 1.2221083641052246,
+      "learning_rate": 1.738825736306625e-05,
+      "loss": 0.9838,
+      "step": 650
+    },
+    {
+      "epoch": 0.2385489190179553,
+      "grad_norm": 1.365902066230774,
+      "learning_rate": 1.7380466083803198e-05,
+      "loss": 0.8987,
+      "step": 651
+    },
+    {
+      "epoch": 0.23891535360938074,
+      "grad_norm": 1.3669956922531128,
+      "learning_rate": 1.7372664951617557e-05,
+      "loss": 0.9071,
+      "step": 652
+    },
+    {
+      "epoch": 0.23928178820080614,
+      "grad_norm": 1.4014561176300049,
+      "learning_rate": 1.7364853976923835e-05,
+      "loss": 0.8403,
+      "step": 653
+    },
+    {
+      "epoch": 0.23964822279223158,
+      "grad_norm": 1.188332200050354,
+      "learning_rate": 1.7357033170149692e-05,
+      "loss": 0.9235,
+      "step": 654
+    },
+    {
+      "epoch": 0.24001465738365702,
+      "grad_norm": 1.2374694347381592,
+      "learning_rate": 1.7349202541735895e-05,
+      "loss": 0.9465,
+      "step": 655
+    },
+    {
+      "epoch": 0.24038109197508245,
+      "grad_norm": 1.3026034832000732,
+      "learning_rate": 1.734136210213634e-05,
+      "loss": 0.9048,
+      "step": 656
+    },
+    {
+      "epoch": 0.24074752656650789,
+      "grad_norm": 1.3879164457321167,
+      "learning_rate": 1.733351186181802e-05,
+      "loss": 0.8876,
+      "step": 657
+    },
+    {
+      "epoch": 0.24111396115793332,
+      "grad_norm": 1.3328919410705566,
+      "learning_rate": 1.732565183126099e-05,
+      "loss": 0.8552,
+      "step": 658
+    },
+    {
+      "epoch": 0.24148039574935873,
+      "grad_norm": 1.3270750045776367,
+      "learning_rate": 1.7317782020958403e-05,
+      "loss": 0.8985,
+      "step": 659
+    },
+    {
+      "epoch": 0.24184683034078416,
+      "grad_norm": 1.2228972911834717,
+      "learning_rate": 1.730990244141646e-05,
+      "loss": 0.8599,
+      "step": 660
+    },
+    {
+      "epoch": 0.2422132649322096,
+      "grad_norm": 1.3068444728851318,
+      "learning_rate": 1.7302013103154386e-05,
+      "loss": 0.9188,
+      "step": 661
+    },
+    {
+      "epoch": 0.24257969952363503,
+      "grad_norm": 1.3953183889389038,
+      "learning_rate": 1.7294114016704458e-05,
+      "loss": 0.8665,
+      "step": 662
+    },
+    {
+      "epoch": 0.24294613411506047,
+      "grad_norm": 1.2629730701446533,
+      "learning_rate": 1.728620519261196e-05,
+      "loss": 0.9255,
+      "step": 663
+    },
+    {
+      "epoch": 0.2433125687064859,
+      "grad_norm": 1.3211989402770996,
+      "learning_rate": 1.727828664143516e-05,
+      "loss": 0.9707,
+      "step": 664
+    },
+    {
+      "epoch": 0.2436790032979113,
+      "grad_norm": 1.4088813066482544,
+      "learning_rate": 1.7270358373745334e-05,
+      "loss": 0.9134,
+      "step": 665
+    },
+    {
+      "epoch": 0.24404543788933675,
+      "grad_norm": 1.3637491464614868,
+      "learning_rate": 1.7262420400126717e-05,
+      "loss": 0.8627,
+      "step": 666
+    },
+    {
+      "epoch": 0.24441187248076218,
+      "grad_norm": 1.5863977670669556,
+      "learning_rate": 1.72544727311765e-05,
+      "loss": 0.8357,
+      "step": 667
+    },
+    {
+      "epoch": 0.24477830707218762,
+      "grad_norm": 1.2949635982513428,
+      "learning_rate": 1.724651537750483e-05,
+      "loss": 0.9373,
+      "step": 668
+    },
+    {
+      "epoch": 0.24514474166361305,
+      "grad_norm": 1.5079573392868042,
+      "learning_rate": 1.7238548349734773e-05,
+      "loss": 0.869,
+      "step": 669
+    },
+    {
+      "epoch": 0.2455111762550385,
+      "grad_norm": 1.920493483543396,
+      "learning_rate": 1.7230571658502303e-05,
+      "loss": 0.9121,
+      "step": 670
+    },
+    {
+      "epoch": 0.2458776108464639,
+      "grad_norm": 1.531829833984375,
+      "learning_rate": 1.7222585314456312e-05,
+      "loss": 0.8534,
+      "step": 671
+    },
+    {
+      "epoch": 0.24624404543788933,
+      "grad_norm": 1.6908535957336426,
+      "learning_rate": 1.7214589328258565e-05,
+      "loss": 0.8029,
+      "step": 672
+    },
+    {
+      "epoch": 0.24661048002931477,
+      "grad_norm": 1.3261759281158447,
+      "learning_rate": 1.7206583710583707e-05,
+      "loss": 0.9446,
+      "step": 673
+    },
+    {
+      "epoch": 0.2469769146207402,
+      "grad_norm": 1.347502589225769,
+      "learning_rate": 1.719856847211924e-05,
+      "loss": 0.9782,
+      "step": 674
+    },
+    {
+      "epoch": 0.24734334921216564,
+      "grad_norm": 1.199223518371582,
+      "learning_rate": 1.7190543623565498e-05,
+      "loss": 0.9558,
+      "step": 675
+    },
+    {
+      "epoch": 0.24770978380359107,
+      "grad_norm": 1.1717863082885742,
+      "learning_rate": 1.718250917563566e-05,
+      "loss": 0.9372,
+      "step": 676
+    },
+    {
+      "epoch": 0.24807621839501648,
+      "grad_norm": 1.303219199180603,
+      "learning_rate": 1.717446513905572e-05,
+      "loss": 0.923,
+      "step": 677
+    },
+    {
+      "epoch": 0.24844265298644191,
+      "grad_norm": 1.276764988899231,
+      "learning_rate": 1.716641152456446e-05,
+      "loss": 0.8635,
+      "step": 678
+    },
+    {
+      "epoch": 0.24880908757786735,
+      "grad_norm": 1.1870089769363403,
+      "learning_rate": 1.7158348342913458e-05,
+      "loss": 0.8853,
+      "step": 679
+    },
+    {
+      "epoch": 0.24917552216929278,
+      "grad_norm": 1.3337750434875488,
+      "learning_rate": 1.715027560486706e-05,
+      "loss": 0.8399,
+      "step": 680
+    },
+    {
+      "epoch": 0.24954195676071822,
+      "grad_norm": 1.6048245429992676,
+      "learning_rate": 1.7142193321202375e-05,
+      "loss": 0.8518,
+      "step": 681
+    },
+    {
+      "epoch": 0.24990839135214366,
+      "grad_norm": 1.397851586341858,
+      "learning_rate": 1.7134101502709252e-05,
+      "loss": 0.8988,
+      "step": 682
+    },
+    {
+      "epoch": 0.2502748259435691,
+      "grad_norm": 1.3171697854995728,
+      "learning_rate": 1.712600016019027e-05,
+      "loss": 0.8739,
+      "step": 683
+    },
+    {
+      "epoch": 0.2502748259435691,
+      "eval_loss": 0.8078320622444153,
+      "eval_runtime": 796.026,
+      "eval_samples_per_second": 3.407,
+      "eval_steps_per_second": 0.426,
+      "step": 683
+    },
+    {
+      "epoch": 0.2506412605349945,
+      "grad_norm": 1.196366548538208,
+      "learning_rate": 1.711788930446071e-05,
+      "loss": 0.9273,
+      "step": 684
+    },
+    {
+      "epoch": 0.25100769512641996,
+      "grad_norm": 1.5654913187026978,
+      "learning_rate": 1.7109768946348583e-05,
+      "loss": 0.7878,
+      "step": 685
+    },
+    {
+      "epoch": 0.25137412971784534,
+      "grad_norm": 1.4333325624465942,
+      "learning_rate": 1.7101639096694556e-05,
+      "loss": 0.9278,
+      "step": 686
+    },
+    {
+      "epoch": 0.2517405643092708,
+      "grad_norm": 1.56187903881073,
+      "learning_rate": 1.709349976635198e-05,
+      "loss": 0.8064,
+      "step": 687
+    },
+    {
+      "epoch": 0.2521069989006962,
+      "grad_norm": 1.242186188697815,
+      "learning_rate": 1.7085350966186865e-05,
+      "loss": 0.8858,
+      "step": 688
+    },
+    {
+      "epoch": 0.25247343349212165,
+      "grad_norm": 1.4079562425613403,
+      "learning_rate": 1.7077192707077854e-05,
+      "loss": 0.9868,
+      "step": 689
+    },
+    {
+      "epoch": 0.2528398680835471,
+      "grad_norm": 1.2565964460372925,
+      "learning_rate": 1.706902499991623e-05,
+      "loss": 0.9144,
+      "step": 690
+    },
+    {
+      "epoch": 0.2532063026749725,
+      "grad_norm": 1.2119742631912231,
+      "learning_rate": 1.7060847855605876e-05,
+      "loss": 0.9579,
+      "step": 691
+    },
+    {
+      "epoch": 0.25357273726639795,
+      "grad_norm": 1.210569143295288,
+      "learning_rate": 1.7052661285063287e-05,
+      "loss": 0.8889,
+      "step": 692
+    },
+    {
+      "epoch": 0.2539391718578234,
+      "grad_norm": 1.507573127746582,
+      "learning_rate": 1.7044465299217526e-05,
+      "loss": 0.8981,
+      "step": 693
+    },
+    {
+      "epoch": 0.2543056064492488,
+      "grad_norm": 1.2725515365600586,
+      "learning_rate": 1.7036259909010246e-05,
+      "loss": 0.9347,
+      "step": 694
+    },
+    {
+      "epoch": 0.25467204104067426,
+      "grad_norm": 1.3163098096847534,
+      "learning_rate": 1.7028045125395635e-05,
+      "loss": 0.8776,
+      "step": 695
+    },
+    {
+      "epoch": 0.2550384756320997,
+      "grad_norm": 1.2225093841552734,
+      "learning_rate": 1.7019820959340426e-05,
+      "loss": 0.8471,
+      "step": 696
+    },
+    {
+      "epoch": 0.25540491022352513,
+      "grad_norm": 1.6468865871429443,
+      "learning_rate": 1.7011587421823894e-05,
+      "loss": 0.8818,
+      "step": 697
+    },
+    {
+      "epoch": 0.2557713448149505,
+      "grad_norm": 1.2895342111587524,
+      "learning_rate": 1.7003344523837803e-05,
+      "loss": 0.917,
+      "step": 698
+    },
+    {
+      "epoch": 0.25613777940637594,
+      "grad_norm": 1.3217147588729858,
+      "learning_rate": 1.699509227638642e-05,
+      "loss": 0.9249,
+      "step": 699
+    },
+    {
+      "epoch": 0.2565042139978014,
+      "grad_norm": 1.495193600654602,
+      "learning_rate": 1.6986830690486507e-05,
+      "loss": 0.8172,
+      "step": 700
+    },
+    {
+      "epoch": 0.2568706485892268,
+      "grad_norm": 1.4029464721679688,
+      "learning_rate": 1.697855977716727e-05,
+      "loss": 0.8741,
+      "step": 701
+    },
+    {
+      "epoch": 0.25723708318065225,
+      "grad_norm": 1.5721744298934937,
+      "learning_rate": 1.6970279547470382e-05,
+      "loss": 0.9183,
+      "step": 702
+    },
+    {
+      "epoch": 0.2576035177720777,
+      "grad_norm": 1.5230120420455933,
+      "learning_rate": 1.6961990012449955e-05,
+      "loss": 0.8991,
+      "step": 703
+    },
+    {
+      "epoch": 0.2579699523635031,
+      "grad_norm": 1.426965355873108,
+      "learning_rate": 1.695369118317251e-05,
+      "loss": 0.8526,
+      "step": 704
+    },
+    {
+      "epoch": 0.25833638695492855,
+      "grad_norm": 1.231719970703125,
+      "learning_rate": 1.694538307071699e-05,
+      "loss": 0.8871,
+      "step": 705
+    },
+    {
+      "epoch": 0.258702821546354,
+      "grad_norm": 1.3230870962142944,
+      "learning_rate": 1.6937065686174728e-05,
+      "loss": 0.9368,
+      "step": 706
+    },
+    {
+      "epoch": 0.2590692561377794,
+      "grad_norm": 1.194519281387329,
+      "learning_rate": 1.692873904064943e-05,
+      "loss": 0.9452,
+      "step": 707
+    },
+    {
+      "epoch": 0.25943569072920486,
+      "grad_norm": 1.2940521240234375,
+      "learning_rate": 1.692040314525716e-05,
+      "loss": 0.9218,
+      "step": 708
+    },
+    {
+      "epoch": 0.2598021253206303,
+      "grad_norm": 1.2810473442077637,
+      "learning_rate": 1.6912058011126356e-05,
+      "loss": 0.9535,
+      "step": 709
+    },
+    {
+      "epoch": 0.2601685599120557,
+      "grad_norm": 1.210940957069397,
+      "learning_rate": 1.690370364939776e-05,
+      "loss": 0.9243,
+      "step": 710
+    },
+    {
+      "epoch": 0.2605349945034811,
+      "grad_norm": 1.413370132446289,
+      "learning_rate": 1.6895340071224446e-05,
+      "loss": 0.9291,
+      "step": 711
+    },
+    {
+      "epoch": 0.26090142909490655,
+      "grad_norm": 1.3101171255111694,
+      "learning_rate": 1.68869672877718e-05,
+      "loss": 0.8924,
+      "step": 712
+    },
+    {
+      "epoch": 0.261267863686332,
+      "grad_norm": 1.2237186431884766,
+      "learning_rate": 1.6878585310217472e-05,
+      "loss": 0.9246,
+      "step": 713
+    },
+    {
+      "epoch": 0.2616342982777574,
+      "grad_norm": 1.2307909727096558,
+      "learning_rate": 1.6870194149751417e-05,
+      "loss": 0.9654,
+      "step": 714
+    },
+    {
+      "epoch": 0.26200073286918285,
+      "grad_norm": 1.3479371070861816,
+      "learning_rate": 1.686179381757583e-05,
+      "loss": 0.8926,
+      "step": 715
+    },
+    {
+      "epoch": 0.2623671674606083,
+      "grad_norm": 1.4196346998214722,
+      "learning_rate": 1.685338432490516e-05,
+      "loss": 0.8783,
+      "step": 716
+    },
+    {
+      "epoch": 0.2627336020520337,
+      "grad_norm": 1.2949929237365723,
+      "learning_rate": 1.6844965682966067e-05,
+      "loss": 0.9335,
+      "step": 717
+    },
+    {
+      "epoch": 0.26310003664345916,
+      "grad_norm": 1.2742618322372437,
+      "learning_rate": 1.683653790299745e-05,
+      "loss": 0.9229,
+      "step": 718
+    },
+    {
+      "epoch": 0.2634664712348846,
+      "grad_norm": 1.468691110610962,
+      "learning_rate": 1.682810099625039e-05,
+      "loss": 0.7729,
+      "step": 719
+    },
+    {
+      "epoch": 0.26383290582631,
+      "grad_norm": 1.296600580215454,
+      "learning_rate": 1.6819654973988165e-05,
+      "loss": 0.9306,
+      "step": 720
+    },
+    {
+      "epoch": 0.2641993404177354,
+      "grad_norm": 1.3135846853256226,
+      "learning_rate": 1.6811199847486205e-05,
+      "loss": 0.8773,
+      "step": 721
+    },
+    {
+      "epoch": 0.26456577500916084,
+      "grad_norm": 1.3318557739257812,
+      "learning_rate": 1.680273562803211e-05,
+      "loss": 0.8922,
+      "step": 722
+    },
+    {
+      "epoch": 0.2649322096005863,
+      "grad_norm": 1.3065025806427002,
+      "learning_rate": 1.6794262326925616e-05,
+      "loss": 0.9231,
+      "step": 723
+    },
+    {
+      "epoch": 0.2652986441920117,
+      "grad_norm": 1.2115901708602905,
+      "learning_rate": 1.6785779955478572e-05,
+      "loss": 0.8406,
+      "step": 724
+    },
+    {
+      "epoch": 0.26566507878343715,
+      "grad_norm": 1.3239742517471313,
+      "learning_rate": 1.6777288525014958e-05,
+      "loss": 0.9726,
+      "step": 725
+    },
+    {
+      "epoch": 0.2660315133748626,
+      "grad_norm": 1.3156651258468628,
+      "learning_rate": 1.6768788046870817e-05,
+      "loss": 0.9525,
+      "step": 726
+    },
+    {
+      "epoch": 0.266397947966288,
+      "grad_norm": 1.624005675315857,
+      "learning_rate": 1.67602785323943e-05,
+      "loss": 0.8564,
+      "step": 727
+    },
+    {
+      "epoch": 0.26676438255771345,
+      "grad_norm": 1.3210386037826538,
+      "learning_rate": 1.675175999294561e-05,
+      "loss": 0.858,
+      "step": 728
+    },
+    {
+      "epoch": 0.2671308171491389,
+      "grad_norm": 1.3814237117767334,
+      "learning_rate": 1.674323243989699e-05,
+      "loss": 0.8559,
+      "step": 729
+    },
+    {
+      "epoch": 0.2674972517405643,
+      "grad_norm": 1.353384256362915,
+      "learning_rate": 1.6734695884632723e-05,
+      "loss": 0.8697,
+      "step": 730
+    },
+    {
+      "epoch": 0.26786368633198976,
+      "grad_norm": 1.5542417764663696,
+      "learning_rate": 1.6726150338549117e-05,
+      "loss": 0.9452,
+      "step": 731
+    },
+    {
+      "epoch": 0.2682301209234152,
+      "grad_norm": 1.4018888473510742,
+      "learning_rate": 1.671759581305448e-05,
+      "loss": 0.9139,
+      "step": 732
+    },
+    {
+      "epoch": 0.2685965555148406,
+      "grad_norm": 1.525744080543518,
+      "learning_rate": 1.6709032319569088e-05,
+      "loss": 0.9375,
+      "step": 733
+    },
+    {
+      "epoch": 0.268962990106266,
+      "grad_norm": 1.343700647354126,
+      "learning_rate": 1.670045986952522e-05,
+      "loss": 0.9302,
+      "step": 734
+    },
+    {
+      "epoch": 0.26932942469769144,
+      "grad_norm": 1.2460815906524658,
+      "learning_rate": 1.66918784743671e-05,
+      "loss": 0.8872,
+      "step": 735
+    },
+    {
+      "epoch": 0.2696958592891169,
+      "grad_norm": 1.4924426078796387,
+      "learning_rate": 1.6683288145550873e-05,
+      "loss": 0.8286,
+      "step": 736
+    },
+    {
+      "epoch": 0.2700622938805423,
+      "grad_norm": 1.3389033079147339,
+      "learning_rate": 1.667468889454464e-05,
+      "loss": 0.8757,
+      "step": 737
+    },
+    {
+      "epoch": 0.27042872847196775,
+      "grad_norm": 1.362754464149475,
+      "learning_rate": 1.6666080732828405e-05,
+      "loss": 0.9041,
+      "step": 738
+    },
+    {
+      "epoch": 0.2707951630633932,
+      "grad_norm": 1.2567392587661743,
+      "learning_rate": 1.6657463671894055e-05,
+      "loss": 0.9453,
+      "step": 739
+    },
+    {
+      "epoch": 0.2711615976548186,
+      "grad_norm": 1.3294352293014526,
+      "learning_rate": 1.6648837723245374e-05,
+      "loss": 0.8718,
+      "step": 740
+    },
+    {
+      "epoch": 0.27152803224624406,
+      "grad_norm": 1.354974627494812,
+      "learning_rate": 1.6640202898398007e-05,
+      "loss": 0.9177,
+      "step": 741
+    },
+    {
+      "epoch": 0.2718944668376695,
+      "grad_norm": 1.6166919469833374,
+      "learning_rate": 1.6631559208879432e-05,
+      "loss": 0.7697,
+      "step": 742
+    },
+    {
+      "epoch": 0.2722609014290949,
+      "grad_norm": 1.4290088415145874,
+      "learning_rate": 1.6622906666228983e-05,
+      "loss": 0.9476,
+      "step": 743
+    },
+    {
+      "epoch": 0.27262733602052036,
+      "grad_norm": 1.226045846939087,
+      "learning_rate": 1.661424528199781e-05,
+      "loss": 0.9194,
+      "step": 744
+    },
+    {
+      "epoch": 0.27299377061194574,
+      "grad_norm": 1.4980007410049438,
+      "learning_rate": 1.660557506774885e-05,
+      "loss": 0.9057,
+      "step": 745
+    },
+    {
+      "epoch": 0.2733602052033712,
+      "grad_norm": 1.4683295488357544,
+      "learning_rate": 1.6596896035056853e-05,
+      "loss": 0.8626,
+      "step": 746
+    },
+    {
+      "epoch": 0.2737266397947966,
+      "grad_norm": 1.223743200302124,
+      "learning_rate": 1.6588208195508316e-05,
+      "loss": 0.9407,
+      "step": 747
+    },
+    {
+      "epoch": 0.27409307438622205,
+      "grad_norm": 1.3194293975830078,
+      "learning_rate": 1.6579511560701513e-05,
+      "loss": 0.9266,
+      "step": 748
+    },
+    {
+      "epoch": 0.2744595089776475,
+      "grad_norm": 1.3832615613937378,
+      "learning_rate": 1.6570806142246448e-05,
+      "loss": 0.91,
+      "step": 749
+    },
+    {
+      "epoch": 0.2748259435690729,
+      "grad_norm": 1.4815707206726074,
+      "learning_rate": 1.656209195176485e-05,
+      "loss": 0.8604,
+      "step": 750
+    },
+    {
+      "epoch": 0.27519237816049835,
+      "grad_norm": 1.290116786956787,
+      "learning_rate": 1.655336900089018e-05,
+      "loss": 0.9047,
+      "step": 751
+    },
+    {
+      "epoch": 0.2755588127519238,
+      "grad_norm": 1.3829827308654785,
+      "learning_rate": 1.6544637301267558e-05,
+      "loss": 0.8946,
+      "step": 752
+    },
+    {
+      "epoch": 0.2759252473433492,
+      "grad_norm": 1.3662350177764893,
+      "learning_rate": 1.6535896864553818e-05,
+      "loss": 0.9196,
+      "step": 753
+    },
+    {
+      "epoch": 0.27629168193477466,
+      "grad_norm": 1.3391906023025513,
+      "learning_rate": 1.652714770241744e-05,
+      "loss": 0.9328,
+      "step": 754
+    },
+    {
+      "epoch": 0.2766581165262001,
+      "grad_norm": 1.3064299821853638,
+      "learning_rate": 1.651838982653855e-05,
+      "loss": 0.8811,
+      "step": 755
+    },
+    {
+      "epoch": 0.27702455111762553,
+      "grad_norm": 1.387911081314087,
+      "learning_rate": 1.650962324860892e-05,
+      "loss": 0.973,
+      "step": 756
+    },
+    {
+      "epoch": 0.2773909857090509,
+      "grad_norm": 1.292227029800415,
+      "learning_rate": 1.6500847980331923e-05,
+      "loss": 0.9254,
+      "step": 757
+    },
+    {
+      "epoch": 0.27775742030047634,
+      "grad_norm": 1.4158412218093872,
+      "learning_rate": 1.6492064033422554e-05,
+      "loss": 0.9298,
+      "step": 758
+    },
+    {
+      "epoch": 0.2781238548919018,
+      "grad_norm": 1.4395830631256104,
+      "learning_rate": 1.648327141960738e-05,
+      "loss": 0.9033,
+      "step": 759
+    },
+    {
+      "epoch": 0.2784902894833272,
+      "grad_norm": 1.3105932474136353,
+      "learning_rate": 1.6474470150624533e-05,
+      "loss": 0.8909,
+      "step": 760
+    },
+    {
+      "epoch": 0.27885672407475265,
+      "grad_norm": 1.6311067342758179,
+      "learning_rate": 1.646566023822371e-05,
+      "loss": 0.929,
+      "step": 761
+    },
+    {
+      "epoch": 0.2792231586661781,
+      "grad_norm": 1.4089274406433105,
+      "learning_rate": 1.6456841694166154e-05,
+      "loss": 0.8728,
+      "step": 762
+    },
+    {
+      "epoch": 0.2795895932576035,
+      "grad_norm": 1.506455659866333,
+      "learning_rate": 1.644801453022461e-05,
+      "loss": 0.8659,
+      "step": 763
+    },
+    {
+      "epoch": 0.27995602784902895,
+      "grad_norm": 1.3019893169403076,
+      "learning_rate": 1.643917875818335e-05,
+      "loss": 0.8047,
+      "step": 764
+    },
+    {
+      "epoch": 0.2803224624404544,
+      "grad_norm": 1.3945775032043457,
+      "learning_rate": 1.6430334389838126e-05,
+      "loss": 0.9586,
+      "step": 765
+    },
+    {
+      "epoch": 0.2806888970318798,
+      "grad_norm": 1.3473881483078003,
+      "learning_rate": 1.6421481436996173e-05,
+      "loss": 0.8552,
+      "step": 766
+    },
+    {
+      "epoch": 0.28105533162330526,
+      "grad_norm": 1.5067903995513916,
+      "learning_rate": 1.641261991147618e-05,
+      "loss": 0.9138,
+      "step": 767
+    },
+    {
+      "epoch": 0.2814217662147307,
+      "grad_norm": 1.4067862033843994,
+      "learning_rate": 1.6403749825108286e-05,
+      "loss": 0.8866,
+      "step": 768
+    },
+    {
+      "epoch": 0.2817882008061561,
+      "grad_norm": 1.6273012161254883,
+      "learning_rate": 1.6394871189734057e-05,
+      "loss": 0.8841,
+      "step": 769
+    },
+    {
+      "epoch": 0.2821546353975815,
+      "grad_norm": 1.539231538772583,
+      "learning_rate": 1.638598401720647e-05,
+      "loss": 0.8788,
+      "step": 770
+    },
+    {
+      "epoch": 0.28252106998900695,
+      "grad_norm": 1.7438411712646484,
+      "learning_rate": 1.6377088319389904e-05,
+      "loss": 0.8567,
+      "step": 771
+    },
+    {
+      "epoch": 0.2828875045804324,
+      "grad_norm": 1.4629979133605957,
+      "learning_rate": 1.6368184108160114e-05,
+      "loss": 0.8732,
+      "step": 772
+    },
+    {
+      "epoch": 0.2832539391718578,
+      "grad_norm": 1.3000826835632324,
+      "learning_rate": 1.635927139540422e-05,
+      "loss": 0.8814,
+      "step": 773
+    },
+    {
+      "epoch": 0.28362037376328325,
+      "grad_norm": 1.4041448831558228,
+      "learning_rate": 1.63503501930207e-05,
+      "loss": 0.9357,
+      "step": 774
+    },
+    {
+      "epoch": 0.2839868083547087,
+      "grad_norm": 1.3833987712860107,
+      "learning_rate": 1.6341420512919355e-05,
+      "loss": 0.8805,
+      "step": 775
+    },
+    {
+      "epoch": 0.2843532429461341,
+      "grad_norm": 1.5424768924713135,
+      "learning_rate": 1.6332482367021308e-05,
+      "loss": 0.8681,
+      "step": 776
+    },
+    {
+      "epoch": 0.28471967753755956,
+      "grad_norm": 1.3150067329406738,
+      "learning_rate": 1.6323535767258987e-05,
+      "loss": 0.904,
+      "step": 777
+    },
+    {
+      "epoch": 0.285086112128985,
+      "grad_norm": 1.4194296598434448,
+      "learning_rate": 1.63145807255761e-05,
+      "loss": 0.9059,
+      "step": 778
+    },
+    {
+      "epoch": 0.2854525467204104,
+      "grad_norm": 1.305395245552063,
+      "learning_rate": 1.6305617253927635e-05,
+      "loss": 0.9258,
+      "step": 779
+    },
+    {
+      "epoch": 0.28581898131183586,
+      "grad_norm": 1.4417041540145874,
+      "learning_rate": 1.6296645364279816e-05,
+      "loss": 0.886,
+      "step": 780
+    },
+    {
+      "epoch": 0.28618541590326124,
+      "grad_norm": 1.3115651607513428,
+      "learning_rate": 1.6287665068610127e-05,
+      "loss": 0.9332,
+      "step": 781
+    },
+    {
+      "epoch": 0.2865518504946867,
+      "grad_norm": 1.4982367753982544,
+      "learning_rate": 1.6278676378907258e-05,
+      "loss": 0.8913,
+      "step": 782
+    },
+    {
+      "epoch": 0.2869182850861121,
+      "grad_norm": 1.3922163248062134,
+      "learning_rate": 1.6269679307171108e-05,
+      "loss": 0.9012,
+      "step": 783
+    },
+    {
+      "epoch": 0.28728471967753755,
+      "grad_norm": 1.4132483005523682,
+      "learning_rate": 1.626067386541277e-05,
+      "loss": 0.8996,
+      "step": 784
+    },
+    {
+      "epoch": 0.287651154268963,
+      "grad_norm": 1.4408533573150635,
+      "learning_rate": 1.6251660065654513e-05,
+      "loss": 0.9164,
+      "step": 785
+    },
+    {
+      "epoch": 0.2880175888603884,
+      "grad_norm": 1.3485194444656372,
+      "learning_rate": 1.6242637919929752e-05,
+      "loss": 0.9189,
+      "step": 786
+    },
+    {
+      "epoch": 0.28838402345181385,
+      "grad_norm": 1.4074796438217163,
+      "learning_rate": 1.6233607440283066e-05,
+      "loss": 0.8408,
+      "step": 787
+    },
+    {
+      "epoch": 0.2887504580432393,
+      "grad_norm": 1.3223270177841187,
+      "learning_rate": 1.6224568638770127e-05,
+      "loss": 0.9216,
+      "step": 788
+    },
+    {
+      "epoch": 0.2891168926346647,
+      "grad_norm": 1.3968936204910278,
+      "learning_rate": 1.6215521527457752e-05,
+      "loss": 0.8787,
+      "step": 789
+    },
+    {
+      "epoch": 0.28948332722609016,
+      "grad_norm": 1.5885320901870728,
+      "learning_rate": 1.6206466118423824e-05,
+      "loss": 0.8786,
+      "step": 790
+    },
+    {
+      "epoch": 0.2898497618175156,
+      "grad_norm": 1.3874950408935547,
+      "learning_rate": 1.6197402423757323e-05,
+      "loss": 0.902,
+      "step": 791
+    },
+    {
+      "epoch": 0.29021619640894103,
+      "grad_norm": 1.279403805732727,
+      "learning_rate": 1.6188330455558277e-05,
+      "loss": 0.9003,
+      "step": 792
+    },
+    {
+      "epoch": 0.2905826310003664,
+      "grad_norm": 1.3086025714874268,
+      "learning_rate": 1.6179250225937764e-05,
+      "loss": 0.8931,
+      "step": 793
+    },
+    {
+      "epoch": 0.29094906559179184,
+      "grad_norm": 1.386427402496338,
+      "learning_rate": 1.617016174701789e-05,
+      "loss": 0.9941,
+      "step": 794
+    },
+    {
+      "epoch": 0.2913155001832173,
+      "grad_norm": 1.4497308731079102,
+      "learning_rate": 1.6161065030931777e-05,
+      "loss": 0.8774,
+      "step": 795
+    },
+    {
+      "epoch": 0.2916819347746427,
+      "grad_norm": 1.4114580154418945,
+      "learning_rate": 1.6151960089823542e-05,
+      "loss": 0.8589,
+      "step": 796
+    },
+    {
+      "epoch": 0.29204836936606815,
+      "grad_norm": 1.627975344657898,
+      "learning_rate": 1.614284693584828e-05,
+      "loss": 0.9294,
+      "step": 797
+    },
+    {
+      "epoch": 0.2924148039574936,
+      "grad_norm": 1.55379056930542,
+      "learning_rate": 1.6133725581172056e-05,
+      "loss": 0.7997,
+      "step": 798
+    },
+    {
+      "epoch": 0.292781238548919,
+      "grad_norm": 1.4049032926559448,
+      "learning_rate": 1.6124596037971867e-05,
+      "loss": 0.8726,
+      "step": 799
+    },
+    {
+      "epoch": 0.29314767314034446,
+      "grad_norm": 1.6755776405334473,
+      "learning_rate": 1.611545831843567e-05,
+      "loss": 0.8759,
+      "step": 800
+    },
+    {
+      "epoch": 0.2935141077317699,
+      "grad_norm": 1.4013546705245972,
+      "learning_rate": 1.6106312434762315e-05,
+      "loss": 0.9327,
+      "step": 801
+    },
+    {
+      "epoch": 0.2938805423231953,
+      "grad_norm": 1.3094841241836548,
+      "learning_rate": 1.609715839916155e-05,
+      "loss": 0.9164,
+      "step": 802
+    },
+    {
+      "epoch": 0.29424697691462076,
+      "grad_norm": 1.4455472230911255,
+      "learning_rate": 1.608799622385402e-05,
+      "loss": 0.8573,
+      "step": 803
+    },
+    {
+      "epoch": 0.2946134115060462,
+      "grad_norm": 1.357675552368164,
+      "learning_rate": 1.607882592107123e-05,
+      "loss": 0.8871,
+      "step": 804
+    },
+    {
+      "epoch": 0.2949798460974716,
+      "grad_norm": 1.3851970434188843,
+      "learning_rate": 1.6069647503055532e-05,
+      "loss": 0.8517,
+      "step": 805
+    },
+    {
+      "epoch": 0.295346280688897,
+      "grad_norm": 1.5386179685592651,
+      "learning_rate": 1.6060460982060116e-05,
+      "loss": 0.854,
+      "step": 806
+    },
+    {
+      "epoch": 0.29571271528032245,
+      "grad_norm": 1.4447675943374634,
+      "learning_rate": 1.6051266370348985e-05,
+      "loss": 0.845,
+      "step": 807
+    },
+    {
+      "epoch": 0.2960791498717479,
+      "grad_norm": 1.5048695802688599,
+      "learning_rate": 1.604206368019695e-05,
+      "loss": 0.852,
+      "step": 808
+    },
+    {
+      "epoch": 0.2964455844631733,
+      "grad_norm": 1.702054500579834,
+      "learning_rate": 1.60328529238896e-05,
+      "loss": 0.8974,
+      "step": 809
+    },
+    {
+      "epoch": 0.29681201905459875,
+      "grad_norm": 1.6284360885620117,
+      "learning_rate": 1.6023634113723297e-05,
+      "loss": 0.9014,
+      "step": 810
+    },
+    {
+      "epoch": 0.2971784536460242,
+      "grad_norm": 1.4068750143051147,
+      "learning_rate": 1.6014407262005152e-05,
+      "loss": 0.8157,
+      "step": 811
+    },
+    {
+      "epoch": 0.2975448882374496,
+      "grad_norm": 1.3355695009231567,
+      "learning_rate": 1.6005172381053012e-05,
+      "loss": 0.8513,
+      "step": 812
+    },
+    {
+      "epoch": 0.29791132282887506,
+      "grad_norm": 1.4846398830413818,
+      "learning_rate": 1.5995929483195443e-05,
+      "loss": 0.8568,
+      "step": 813
+    },
+    {
+      "epoch": 0.2982777574203005,
+      "grad_norm": 1.259242296218872,
+      "learning_rate": 1.598667858077172e-05,
+      "loss": 0.9381,
+      "step": 814
+    },
+    {
+      "epoch": 0.29864419201172593,
+      "grad_norm": 1.4263806343078613,
+      "learning_rate": 1.597741968613179e-05,
+      "loss": 0.8763,
+      "step": 815
+    },
+    {
+      "epoch": 0.29901062660315136,
+      "grad_norm": 1.3765134811401367,
+      "learning_rate": 1.5968152811636283e-05,
+      "loss": 0.8603,
+      "step": 816
+    },
+    {
+      "epoch": 0.29937706119457674,
+      "grad_norm": 1.2649085521697998,
+      "learning_rate": 1.595887796965647e-05,
+      "loss": 0.9276,
+      "step": 817
+    },
+    {
+      "epoch": 0.2997434957860022,
+      "grad_norm": 1.4955521821975708,
+      "learning_rate": 1.594959517257428e-05,
+      "loss": 0.844,
+      "step": 818
+    },
+    {
+      "epoch": 0.3001099303774276,
+      "grad_norm": 1.2310223579406738,
+      "learning_rate": 1.5940304432782233e-05,
+      "loss": 0.8783,
+      "step": 819
+    },
+    {
+      "epoch": 0.30047636496885305,
+      "grad_norm": 1.319631576538086,
+      "learning_rate": 1.5931005762683473e-05,
+      "loss": 0.9017,
+      "step": 820
+    },
+    {
+      "epoch": 0.3008427995602785,
+      "grad_norm": 1.3764492273330688,
+      "learning_rate": 1.5921699174691724e-05,
+      "loss": 0.832,
+      "step": 821
+    },
+    {
+      "epoch": 0.3012092341517039,
+      "grad_norm": 1.4033286571502686,
+      "learning_rate": 1.5912384681231282e-05,
+      "loss": 0.8962,
+      "step": 822
+    },
+    {
+      "epoch": 0.30157566874312935,
+      "grad_norm": 1.5620622634887695,
+      "learning_rate": 1.5903062294737e-05,
+      "loss": 0.9079,
+      "step": 823
+    },
+    {
+      "epoch": 0.3019421033345548,
+      "grad_norm": 1.4058263301849365,
+      "learning_rate": 1.5893732027654258e-05,
+      "loss": 0.9625,
+      "step": 824
+    },
+    {
+      "epoch": 0.3023085379259802,
+      "grad_norm": 1.4301252365112305,
+      "learning_rate": 1.5884393892438962e-05,
+      "loss": 0.8528,
+      "step": 825
+    },
+    {
+      "epoch": 0.30267497251740566,
+      "grad_norm": 1.5510419607162476,
+      "learning_rate": 1.587504790155753e-05,
+      "loss": 0.858,
+      "step": 826
+    },
+    {
+      "epoch": 0.3030414071088311,
+      "grad_norm": 1.5547479391098022,
+      "learning_rate": 1.586569406748686e-05,
+      "loss": 0.9054,
+      "step": 827
+    },
+    {
+      "epoch": 0.30340784170025653,
+      "grad_norm": 1.5077707767486572,
+      "learning_rate": 1.585633240271431e-05,
+      "loss": 0.8879,
+      "step": 828
+    },
+    {
+      "epoch": 0.3037742762916819,
+      "grad_norm": 1.394148588180542,
+      "learning_rate": 1.5846962919737717e-05,
+      "loss": 0.8883,
+      "step": 829
+    },
+    {
+      "epoch": 0.30414071088310735,
+      "grad_norm": 1.3555182218551636,
+      "learning_rate": 1.5837585631065334e-05,
+      "loss": 0.904,
+      "step": 830
+    },
+    {
+      "epoch": 0.3045071454745328,
+      "grad_norm": 1.3965303897857666,
+      "learning_rate": 1.5828200549215843e-05,
+      "loss": 0.8471,
+      "step": 831
+    },
+    {
+      "epoch": 0.3048735800659582,
+      "grad_norm": 1.4720205068588257,
+      "learning_rate": 1.581880768671833e-05,
+      "loss": 0.8241,
+      "step": 832
+    },
+    {
+      "epoch": 0.30524001465738365,
+      "grad_norm": 1.3568192720413208,
+      "learning_rate": 1.580940705611226e-05,
+      "loss": 0.8405,
+      "step": 833
+    },
+    {
+      "epoch": 0.3056064492488091,
+      "grad_norm": 1.3553125858306885,
+      "learning_rate": 1.5799998669947485e-05,
+      "loss": 0.8948,
+      "step": 834
+    },
+    {
+      "epoch": 0.3059728838402345,
+      "grad_norm": 1.5014175176620483,
+      "learning_rate": 1.5790582540784196e-05,
+      "loss": 0.8631,
+      "step": 835
+    },
+    {
+      "epoch": 0.30633931843165996,
+      "grad_norm": 1.5766485929489136,
+      "learning_rate": 1.5781158681192928e-05,
+      "loss": 0.8352,
+      "step": 836
+    },
+    {
+      "epoch": 0.3067057530230854,
+      "grad_norm": 1.4493907690048218,
+      "learning_rate": 1.577172710375453e-05,
+      "loss": 0.8646,
+      "step": 837
+    },
+    {
+      "epoch": 0.3070721876145108,
+      "grad_norm": 1.5850780010223389,
+      "learning_rate": 1.5762287821060164e-05,
+      "loss": 0.8462,
+      "step": 838
+    },
+    {
+      "epoch": 0.30743862220593626,
+      "grad_norm": 1.5644487142562866,
+      "learning_rate": 1.575284084571127e-05,
+      "loss": 0.9149,
+      "step": 839
+    },
+    {
+      "epoch": 0.3078050567973617,
+      "grad_norm": 1.4901156425476074,
+      "learning_rate": 1.5743386190319556e-05,
+      "loss": 0.8686,
+      "step": 840
+    },
+    {
+      "epoch": 0.3081714913887871,
+      "grad_norm": 1.4039537906646729,
+      "learning_rate": 1.5733923867506995e-05,
+      "loss": 0.8827,
+      "step": 841
+    },
+    {
+      "epoch": 0.3085379259802125,
+      "grad_norm": 1.3826498985290527,
+      "learning_rate": 1.5724453889905787e-05,
+      "loss": 0.9294,
+      "step": 842
+    },
+    {
+      "epoch": 0.30890436057163795,
+      "grad_norm": 1.324507474899292,
+      "learning_rate": 1.571497627015835e-05,
+      "loss": 0.9455,
+      "step": 843
+    },
+    {
+      "epoch": 0.3092707951630634,
+      "grad_norm": 1.5789650678634644,
+      "learning_rate": 1.570549102091731e-05,
+      "loss": 0.8778,
+      "step": 844
+    },
+    {
+      "epoch": 0.3096372297544888,
+      "grad_norm": 1.2591837644577026,
+      "learning_rate": 1.5695998154845468e-05,
+      "loss": 0.9207,
+      "step": 845
+    },
+    {
+      "epoch": 0.31000366434591425,
+      "grad_norm": 1.3110920190811157,
+      "learning_rate": 1.5686497684615804e-05,
+      "loss": 0.9022,
+      "step": 846
+    },
+    {
+      "epoch": 0.3103700989373397,
+      "grad_norm": 1.377200722694397,
+      "learning_rate": 1.5676989622911457e-05,
+      "loss": 0.8624,
+      "step": 847
+    },
+    {
+      "epoch": 0.3107365335287651,
+      "grad_norm": 1.3229527473449707,
+      "learning_rate": 1.5667473982425673e-05,
+      "loss": 0.8857,
+      "step": 848
+    },
+    {
+      "epoch": 0.31110296812019056,
+      "grad_norm": 1.7923849821090698,
+      "learning_rate": 1.5657950775861842e-05,
+      "loss": 0.8445,
+      "step": 849
+    },
+    {
+      "epoch": 0.311469402711616,
+      "grad_norm": 1.4395514726638794,
+      "learning_rate": 1.5648420015933444e-05,
+      "loss": 0.9183,
+      "step": 850
+    },
+    {
+      "epoch": 0.31183583730304143,
+      "grad_norm": 1.5209360122680664,
+      "learning_rate": 1.5638881715364042e-05,
+      "loss": 0.879,
+      "step": 851
+    },
+    {
+      "epoch": 0.31220227189446687,
+      "grad_norm": 1.3047930002212524,
+      "learning_rate": 1.5629335886887272e-05,
+      "loss": 0.9366,
+      "step": 852
+    },
+    {
+      "epoch": 0.31256870648589224,
+      "grad_norm": 1.5573649406433105,
+      "learning_rate": 1.5619782543246813e-05,
+      "loss": 0.8785,
+      "step": 853
+    },
+    {
+      "epoch": 0.3129351410773177,
+      "grad_norm": 1.3706642389297485,
+      "learning_rate": 1.561022169719638e-05,
+      "loss": 0.9162,
+      "step": 854
+    },
+    {
+      "epoch": 0.3133015756687431,
+      "grad_norm": 1.303852915763855,
+      "learning_rate": 1.56006533614997e-05,
+      "loss": 0.8853,
+      "step": 855
+    },
+    {
+      "epoch": 0.31366801026016855,
+      "grad_norm": 1.4067341089248657,
+      "learning_rate": 1.55910775489305e-05,
+      "loss": 0.9383,
+      "step": 856
+    },
+    {
+      "epoch": 0.314034444851594,
+      "grad_norm": 1.5358949899673462,
+      "learning_rate": 1.55814942722725e-05,
+      "loss": 0.8933,
+      "step": 857
+    },
+    {
+      "epoch": 0.3144008794430194,
+      "grad_norm": 1.4367671012878418,
+      "learning_rate": 1.5571903544319365e-05,
+      "loss": 0.8596,
+      "step": 858
+    },
+    {
+      "epoch": 0.31476731403444486,
+      "grad_norm": 1.280152440071106,
+      "learning_rate": 1.556230537787472e-05,
+      "loss": 0.8472,
+      "step": 859
+    },
+    {
+      "epoch": 0.3151337486258703,
+      "grad_norm": 1.437868595123291,
+      "learning_rate": 1.555269978575212e-05,
+      "loss": 0.8706,
+      "step": 860
+    },
+    {
+      "epoch": 0.3155001832172957,
+      "grad_norm": 1.5736864805221558,
+      "learning_rate": 1.554308678077503e-05,
+      "loss": 0.8014,
+      "step": 861
+    },
+    {
+      "epoch": 0.31586661780872116,
+      "grad_norm": 1.3937286138534546,
+      "learning_rate": 1.5533466375776812e-05,
+      "loss": 0.9138,
+      "step": 862
+    },
+    {
+      "epoch": 0.3162330524001466,
+      "grad_norm": 1.6589691638946533,
+      "learning_rate": 1.5523838583600706e-05,
+      "loss": 0.8216,
+      "step": 863
+    },
+    {
+      "epoch": 0.316599486991572,
+      "grad_norm": 1.3993836641311646,
+      "learning_rate": 1.5514203417099816e-05,
+      "loss": 0.8878,
+      "step": 864
+    },
+    {
+      "epoch": 0.3169659215829974,
+      "grad_norm": 1.3711109161376953,
+      "learning_rate": 1.550456088913709e-05,
+      "loss": 0.919,
+      "step": 865
+    },
+    {
+      "epoch": 0.31733235617442285,
+      "grad_norm": 1.5643397569656372,
+      "learning_rate": 1.5494911012585303e-05,
+      "loss": 0.9188,
+      "step": 866
+    },
+    {
+      "epoch": 0.3176987907658483,
+      "grad_norm": 1.5048398971557617,
+      "learning_rate": 1.5485253800327043e-05,
+      "loss": 0.9647,
+      "step": 867
+    },
+    {
+      "epoch": 0.3180652253572737,
+      "grad_norm": 1.6283950805664062,
+      "learning_rate": 1.547558926525468e-05,
+      "loss": 0.813,
+      "step": 868
+    },
+    {
+      "epoch": 0.31843165994869915,
+      "grad_norm": 1.706950068473816,
+      "learning_rate": 1.546591742027038e-05,
+      "loss": 0.7791,
+      "step": 869
+    },
+    {
+      "epoch": 0.3187980945401246,
+      "grad_norm": 1.3241078853607178,
+      "learning_rate": 1.545623827828605e-05,
+      "loss": 0.922,
+      "step": 870
+    },
+    {
+      "epoch": 0.31916452913155,
+      "grad_norm": 1.6119778156280518,
+      "learning_rate": 1.5446551852223346e-05,
+      "loss": 0.8562,
+      "step": 871
+    },
+    {
+      "epoch": 0.31953096372297546,
+      "grad_norm": 1.4082729816436768,
+      "learning_rate": 1.5436858155013653e-05,
+      "loss": 0.9317,
+      "step": 872
+    },
+    {
+      "epoch": 0.3198973983144009,
+      "grad_norm": 1.4613032341003418,
+      "learning_rate": 1.5427157199598048e-05,
+      "loss": 0.8964,
+      "step": 873
+    },
+    {
+      "epoch": 0.32026383290582633,
+      "grad_norm": 1.325289249420166,
+      "learning_rate": 1.541744899892731e-05,
+      "loss": 0.9091,
+      "step": 874
+    },
+    {
+      "epoch": 0.32063026749725176,
+      "grad_norm": 1.4540269374847412,
+      "learning_rate": 1.5407733565961894e-05,
+      "loss": 0.8688,
+      "step": 875
+    },
+    {
+      "epoch": 0.32099670208867714,
+      "grad_norm": 1.56600821018219,
+      "learning_rate": 1.5398010913671894e-05,
+      "loss": 0.9103,
+      "step": 876
+    },
+    {
+      "epoch": 0.3213631366801026,
+      "grad_norm": 1.4235271215438843,
+      "learning_rate": 1.5388281055037054e-05,
+      "loss": 0.9022,
+      "step": 877
+    },
+    {
+      "epoch": 0.321729571271528,
+      "grad_norm": 1.42256498336792,
+      "learning_rate": 1.5378544003046743e-05,
+      "loss": 0.8512,
+      "step": 878
+    },
+    {
+      "epoch": 0.32209600586295345,
+      "grad_norm": 1.3547056913375854,
+      "learning_rate": 1.5368799770699915e-05,
+      "loss": 0.9118,
+      "step": 879
+    },
+    {
+      "epoch": 0.3224624404543789,
+      "grad_norm": 1.5990328788757324,
+      "learning_rate": 1.5359048371005128e-05,
+      "loss": 0.8412,
+      "step": 880
+    },
+    {
+      "epoch": 0.3228288750458043,
+      "grad_norm": 1.3500826358795166,
+      "learning_rate": 1.5349289816980498e-05,
+      "loss": 0.8604,
+      "step": 881
+    },
+    {
+      "epoch": 0.32319530963722976,
+      "grad_norm": 1.513796091079712,
+      "learning_rate": 1.5339524121653698e-05,
+      "loss": 0.8849,
+      "step": 882
+    },
+    {
+      "epoch": 0.3235617442286552,
+      "grad_norm": 1.3060262203216553,
+      "learning_rate": 1.532975129806193e-05,
+      "loss": 0.9494,
+      "step": 883
+    },
+    {
+      "epoch": 0.3239281788200806,
+      "grad_norm": 1.295326828956604,
+      "learning_rate": 1.531997135925191e-05,
+      "loss": 0.8887,
+      "step": 884
+    },
+    {
+      "epoch": 0.32429461341150606,
+      "grad_norm": 1.467307686805725,
+      "learning_rate": 1.5310184318279864e-05,
+      "loss": 0.8879,
+      "step": 885
+    },
+    {
+      "epoch": 0.3246610480029315,
+      "grad_norm": 1.5006343126296997,
+      "learning_rate": 1.530039018821149e-05,
+      "loss": 0.8955,
+      "step": 886
+    },
+    {
+      "epoch": 0.32502748259435693,
+      "grad_norm": 1.5569864511489868,
+      "learning_rate": 1.5290588982121945e-05,
+      "loss": 0.8606,
+      "step": 887
+    },
+    {
+      "epoch": 0.3253939171857823,
+      "grad_norm": 1.6372853517532349,
+      "learning_rate": 1.5280780713095852e-05,
+      "loss": 0.9186,
+      "step": 888
+    },
+    {
+      "epoch": 0.32576035177720775,
+      "grad_norm": 1.3921681642532349,
+      "learning_rate": 1.5270965394227237e-05,
+      "loss": 0.8669,
+      "step": 889
+    },
+    {
+      "epoch": 0.3261267863686332,
+      "grad_norm": 1.4064171314239502,
+      "learning_rate": 1.5261143038619562e-05,
+      "loss": 0.8973,
+      "step": 890
+    },
+    {
+      "epoch": 0.3264932209600586,
+      "grad_norm": 1.840980052947998,
+      "learning_rate": 1.525131365938567e-05,
+      "loss": 0.8078,
+      "step": 891
+    },
+    {
+      "epoch": 0.32685965555148405,
+      "grad_norm": 1.7794231176376343,
+      "learning_rate": 1.524147726964778e-05,
+      "loss": 0.7925,
+      "step": 892
+    },
+    {
+      "epoch": 0.3272260901429095,
+      "grad_norm": 1.3070979118347168,
+      "learning_rate": 1.5231633882537474e-05,
+      "loss": 0.8818,
+      "step": 893
+    },
+    {
+      "epoch": 0.3275925247343349,
+      "grad_norm": 1.4535861015319824,
+      "learning_rate": 1.5221783511195674e-05,
+      "loss": 0.8376,
+      "step": 894
+    },
+    {
+      "epoch": 0.32795895932576036,
+      "grad_norm": 1.4390013217926025,
+      "learning_rate": 1.5211926168772626e-05,
+      "loss": 0.8658,
+      "step": 895
+    },
+    {
+      "epoch": 0.3283253939171858,
+      "grad_norm": 1.4747962951660156,
+      "learning_rate": 1.5202061868427884e-05,
+      "loss": 0.9033,
+      "step": 896
+    },
+    {
+      "epoch": 0.32869182850861123,
+      "grad_norm": 1.7829028367996216,
+      "learning_rate": 1.519219062333029e-05,
+      "loss": 0.7904,
+      "step": 897
+    },
+    {
+      "epoch": 0.32905826310003666,
+      "grad_norm": 1.575325608253479,
+      "learning_rate": 1.5182312446657954e-05,
+      "loss": 0.8527,
+      "step": 898
+    },
+    {
+      "epoch": 0.3294246976914621,
+      "grad_norm": 1.4097166061401367,
+      "learning_rate": 1.5172427351598243e-05,
+      "loss": 0.8692,
+      "step": 899
+    },
+    {
+      "epoch": 0.3297911322828875,
+      "grad_norm": 1.6933417320251465,
+      "learning_rate": 1.516253535134776e-05,
+      "loss": 0.8665,
+      "step": 900
+    },
+    {
+      "epoch": 0.3301575668743129,
+      "grad_norm": 1.4317610263824463,
+      "learning_rate": 1.5152636459112327e-05,
+      "loss": 0.9007,
+      "step": 901
+    },
+    {
+      "epoch": 0.33052400146573835,
+      "grad_norm": 1.781380295753479,
+      "learning_rate": 1.5142730688106966e-05,
+      "loss": 0.835,
+      "step": 902
+    },
+    {
+      "epoch": 0.3308904360571638,
+      "grad_norm": 1.495296835899353,
+      "learning_rate": 1.5132818051555878e-05,
+      "loss": 0.9437,
+      "step": 903
+    },
+    {
+      "epoch": 0.3312568706485892,
+      "grad_norm": 1.6618905067443848,
+      "learning_rate": 1.5122898562692437e-05,
+      "loss": 0.8886,
+      "step": 904
+    },
+    {
+      "epoch": 0.33162330524001465,
+      "grad_norm": 1.5304425954818726,
+      "learning_rate": 1.5112972234759155e-05,
+      "loss": 0.889,
+      "step": 905
+    },
+    {
+      "epoch": 0.3319897398314401,
+      "grad_norm": 1.4052519798278809,
+      "learning_rate": 1.5103039081007691e-05,
+      "loss": 0.8726,
+      "step": 906
+    },
+    {
+      "epoch": 0.3323561744228655,
+      "grad_norm": 1.5464938879013062,
+      "learning_rate": 1.5093099114698797e-05,
+      "loss": 0.878,
+      "step": 907
+    },
+    {
+      "epoch": 0.33272260901429096,
+      "grad_norm": 1.5896848440170288,
+      "learning_rate": 1.5083152349102332e-05,
+      "loss": 0.9362,
+      "step": 908
+    },
+    {
+      "epoch": 0.3330890436057164,
+      "grad_norm": 1.5317317247390747,
+      "learning_rate": 1.5073198797497228e-05,
+      "loss": 0.8346,
+      "step": 909
+    },
+    {
+      "epoch": 0.33345547819714183,
+      "grad_norm": 1.634902834892273,
+      "learning_rate": 1.5063238473171475e-05,
+      "loss": 0.9317,
+      "step": 910
+    },
+    {
+      "epoch": 0.33382191278856727,
+      "grad_norm": 1.6703317165374756,
+      "learning_rate": 1.5053271389422111e-05,
+      "loss": 0.8432,
+      "step": 911
+    },
+    {
+      "epoch": 0.33418834737999265,
+      "grad_norm": 1.6900526285171509,
+      "learning_rate": 1.5043297559555188e-05,
+      "loss": 0.7934,
+      "step": 912
+    },
+    {
+      "epoch": 0.3345547819714181,
+      "grad_norm": 1.5616822242736816,
+      "learning_rate": 1.5033316996885772e-05,
+      "loss": 0.8728,
+      "step": 913
+    },
+    {
+      "epoch": 0.3349212165628435,
+      "grad_norm": 1.455959677696228,
+      "learning_rate": 1.5023329714737917e-05,
+      "loss": 0.8429,
+      "step": 914
+    },
+    {
+      "epoch": 0.33528765115426895,
+      "grad_norm": 1.4425702095031738,
+      "learning_rate": 1.5013335726444642e-05,
+      "loss": 0.9135,
+      "step": 915
+    },
+    {
+      "epoch": 0.3356540857456944,
+      "grad_norm": 1.6457642316818237,
+      "learning_rate": 1.5003335045347921e-05,
+      "loss": 0.8378,
+      "step": 916
+    },
+    {
+      "epoch": 0.3360205203371198,
+      "grad_norm": 1.4372854232788086,
+      "learning_rate": 1.4993327684798667e-05,
+      "loss": 0.9147,
+      "step": 917
+    },
+    {
+      "epoch": 0.33638695492854526,
+      "grad_norm": 1.5803145170211792,
+      "learning_rate": 1.4983313658156704e-05,
+      "loss": 0.8997,
+      "step": 918
+    },
+    {
+      "epoch": 0.3367533895199707,
+      "grad_norm": 1.4921263456344604,
+      "learning_rate": 1.497329297879076e-05,
+      "loss": 0.9059,
+      "step": 919
+    },
+    {
+      "epoch": 0.3371198241113961,
+      "grad_norm": 1.6669347286224365,
+      "learning_rate": 1.4963265660078441e-05,
+      "loss": 0.87,
+      "step": 920
+    },
+    {
+      "epoch": 0.33748625870282156,
+      "grad_norm": 1.3645561933517456,
+      "learning_rate": 1.4953231715406219e-05,
+      "loss": 0.896,
+      "step": 921
+    },
+    {
+      "epoch": 0.337852693294247,
+      "grad_norm": 1.6434755325317383,
+      "learning_rate": 1.494319115816941e-05,
+      "loss": 0.7876,
+      "step": 922
+    },
+    {
+      "epoch": 0.33821912788567243,
+      "grad_norm": 1.6944279670715332,
+      "learning_rate": 1.493314400177216e-05,
+      "loss": 0.8852,
+      "step": 923
+    },
+    {
+      "epoch": 0.3385855624770978,
+      "grad_norm": 1.7454200983047485,
+      "learning_rate": 1.4923090259627422e-05,
+      "loss": 0.7508,
+      "step": 924
+    },
+    {
+      "epoch": 0.33895199706852325,
+      "grad_norm": 1.5646865367889404,
+      "learning_rate": 1.491302994515694e-05,
+      "loss": 0.9465,
+      "step": 925
+    },
+    {
+      "epoch": 0.3393184316599487,
+      "grad_norm": 1.416644811630249,
+      "learning_rate": 1.490296307179124e-05,
+      "loss": 0.8758,
+      "step": 926
+    },
+    {
+      "epoch": 0.3396848662513741,
+      "grad_norm": 1.4960588216781616,
+      "learning_rate": 1.489288965296959e-05,
+      "loss": 0.9026,
+      "step": 927
+    },
+    {
+      "epoch": 0.34005130084279955,
+      "grad_norm": 1.7496960163116455,
+      "learning_rate": 1.4882809702140014e-05,
+      "loss": 0.8591,
+      "step": 928
+    },
+    {
+      "epoch": 0.340417735434225,
+      "grad_norm": 1.5206427574157715,
+      "learning_rate": 1.4872723232759243e-05,
+      "loss": 0.8428,
+      "step": 929
+    },
+    {
+      "epoch": 0.3407841700256504,
+      "grad_norm": 1.3755855560302734,
+      "learning_rate": 1.486263025829271e-05,
+      "loss": 0.9219,
+      "step": 930
+    },
+    {
+      "epoch": 0.34115060461707586,
+      "grad_norm": 1.4779146909713745,
+      "learning_rate": 1.4852530792214541e-05,
+      "loss": 0.8751,
+      "step": 931
+    },
+    {
+      "epoch": 0.3415170392085013,
+      "grad_norm": 1.670133352279663,
+      "learning_rate": 1.4842424848007526e-05,
+      "loss": 0.8551,
+      "step": 932
+    },
+    {
+      "epoch": 0.34188347379992673,
+      "grad_norm": 1.8766719102859497,
+      "learning_rate": 1.4832312439163095e-05,
+      "loss": 0.7814,
+      "step": 933
+    },
+    {
+      "epoch": 0.34224990839135216,
+      "grad_norm": 1.638991117477417,
+      "learning_rate": 1.4822193579181318e-05,
+      "loss": 0.8566,
+      "step": 934
+    },
+    {
+      "epoch": 0.3426163429827776,
+      "grad_norm": 1.517900824546814,
+      "learning_rate": 1.4812068281570871e-05,
+      "loss": 0.9234,
+      "step": 935
+    },
+    {
+      "epoch": 0.342982777574203,
+      "grad_norm": 1.4461380243301392,
+      "learning_rate": 1.4801936559849028e-05,
+      "loss": 0.9009,
+      "step": 936
+    },
+    {
+      "epoch": 0.3433492121656284,
+      "grad_norm": 1.518513560295105,
+      "learning_rate": 1.4791798427541644e-05,
+      "loss": 0.9789,
+      "step": 937
+    },
+    {
+      "epoch": 0.34371564675705385,
+      "grad_norm": 1.5770090818405151,
+      "learning_rate": 1.4781653898183116e-05,
+      "loss": 0.9328,
+      "step": 938
+    },
+    {
+      "epoch": 0.3440820813484793,
+      "grad_norm": 1.4077907800674438,
+      "learning_rate": 1.4771502985316395e-05,
+      "loss": 0.8984,
+      "step": 939
+    },
+    {
+      "epoch": 0.3444485159399047,
+      "grad_norm": 1.5162850618362427,
+      "learning_rate": 1.4761345702492955e-05,
+      "loss": 0.8382,
+      "step": 940
+    },
+    {
+      "epoch": 0.34481495053133016,
+      "grad_norm": 1.5433305501937866,
+      "learning_rate": 1.4751182063272763e-05,
+      "loss": 0.872,
+      "step": 941
+    },
+    {
+      "epoch": 0.3451813851227556,
+      "grad_norm": 1.393364429473877,
+      "learning_rate": 1.474101208122428e-05,
+      "loss": 0.9207,
+      "step": 942
+    },
+    {
+      "epoch": 0.345547819714181,
+      "grad_norm": 1.435567855834961,
+      "learning_rate": 1.4730835769924432e-05,
+      "loss": 0.8501,
+      "step": 943
+    },
+    {
+      "epoch": 0.34591425430560646,
+      "grad_norm": 1.5786954164505005,
+      "learning_rate": 1.4720653142958595e-05,
+      "loss": 0.8781,
+      "step": 944
+    },
+    {
+      "epoch": 0.3462806888970319,
+      "grad_norm": 1.5463298559188843,
+      "learning_rate": 1.4710464213920579e-05,
+      "loss": 0.858,
+      "step": 945
+    },
+    {
+      "epoch": 0.34664712348845733,
+      "grad_norm": 1.3941972255706787,
+      "learning_rate": 1.4700268996412601e-05,
+      "loss": 0.9108,
+      "step": 946
+    },
+    {
+      "epoch": 0.34701355807988277,
+      "grad_norm": 1.7360153198242188,
+      "learning_rate": 1.4690067504045282e-05,
+      "loss": 0.8832,
+      "step": 947
+    },
+    {
+      "epoch": 0.34737999267130815,
+      "grad_norm": 1.4550964832305908,
+      "learning_rate": 1.467985975043761e-05,
+      "loss": 0.9578,
+      "step": 948
+    },
+    {
+      "epoch": 0.3477464272627336,
+      "grad_norm": 1.9805035591125488,
+      "learning_rate": 1.4669645749216936e-05,
+      "loss": 0.8997,
+      "step": 949
+    },
+    {
+      "epoch": 0.348112861854159,
+      "grad_norm": 1.642874836921692,
+      "learning_rate": 1.4659425514018956e-05,
+      "loss": 0.8798,
+      "step": 950
+    },
+    {
+      "epoch": 0.34847929644558445,
+      "grad_norm": 1.7270559072494507,
+      "learning_rate": 1.4649199058487682e-05,
+      "loss": 0.8678,
+      "step": 951
+    },
+    {
+      "epoch": 0.3488457310370099,
+      "grad_norm": 1.448574185371399,
+      "learning_rate": 1.4638966396275434e-05,
+      "loss": 0.8721,
+      "step": 952
+    },
+    {
+      "epoch": 0.3492121656284353,
+      "grad_norm": 1.5349843502044678,
+      "learning_rate": 1.4628727541042815e-05,
+      "loss": 0.9147,
+      "step": 953
+    },
+    {
+      "epoch": 0.34957860021986076,
+      "grad_norm": 1.7318191528320312,
+      "learning_rate": 1.4618482506458702e-05,
+      "loss": 0.9098,
+      "step": 954
+    },
+    {
+      "epoch": 0.3499450348112862,
+      "grad_norm": 1.4772664308547974,
+      "learning_rate": 1.4608231306200211e-05,
+      "loss": 0.915,
+      "step": 955
+    },
+    {
+      "epoch": 0.35031146940271163,
+      "grad_norm": 1.7366101741790771,
+      "learning_rate": 1.4597973953952696e-05,
+      "loss": 0.855,
+      "step": 956
+    },
+    {
+      "epoch": 0.35067790399413706,
+      "grad_norm": 1.4249106645584106,
+      "learning_rate": 1.4587710463409721e-05,
+      "loss": 0.9456,
+      "step": 957
+    },
+    {
+      "epoch": 0.3510443385855625,
+      "grad_norm": 1.5543416738510132,
+      "learning_rate": 1.4577440848273052e-05,
+      "loss": 0.9116,
+      "step": 958
+    },
+    {
+      "epoch": 0.35141077317698793,
+      "grad_norm": 1.4858545064926147,
+      "learning_rate": 1.456716512225262e-05,
+      "loss": 0.8623,
+      "step": 959
+    },
+    {
+      "epoch": 0.3517772077684133,
+      "grad_norm": 1.6535437107086182,
+      "learning_rate": 1.4556883299066529e-05,
+      "loss": 0.8603,
+      "step": 960
+    },
+    {
+      "epoch": 0.35214364235983875,
+      "grad_norm": 1.3971456289291382,
+      "learning_rate": 1.4546595392441e-05,
+      "loss": 0.8705,
+      "step": 961
+    },
+    {
+      "epoch": 0.3525100769512642,
+      "grad_norm": 1.7072619199752808,
+      "learning_rate": 1.4536301416110398e-05,
+      "loss": 0.7356,
+      "step": 962
+    },
+    {
+      "epoch": 0.3528765115426896,
+      "grad_norm": 1.5533279180526733,
+      "learning_rate": 1.4526001383817182e-05,
+      "loss": 0.8684,
+      "step": 963
+    },
+    {
+      "epoch": 0.35324294613411505,
+      "grad_norm": 1.5234684944152832,
+      "learning_rate": 1.4515695309311887e-05,
+      "loss": 0.8985,
+      "step": 964
+    },
+    {
+      "epoch": 0.3536093807255405,
+      "grad_norm": 1.4784553050994873,
+      "learning_rate": 1.4505383206353133e-05,
+      "loss": 0.8635,
+      "step": 965
+    },
+    {
+      "epoch": 0.3539758153169659,
+      "grad_norm": 1.4752111434936523,
+      "learning_rate": 1.4495065088707576e-05,
+      "loss": 0.9312,
+      "step": 966
+    },
+    {
+      "epoch": 0.35434224990839136,
+      "grad_norm": 1.7725974321365356,
+      "learning_rate": 1.44847409701499e-05,
+      "loss": 0.9453,
+      "step": 967
+    },
+    {
+      "epoch": 0.3547086844998168,
+      "grad_norm": 1.5872156620025635,
+      "learning_rate": 1.4474410864462807e-05,
+      "loss": 0.8472,
+      "step": 968
+    },
+    {
+      "epoch": 0.35507511909124223,
+      "grad_norm": 1.4837459325790405,
+      "learning_rate": 1.4464074785436986e-05,
+      "loss": 0.8928,
+      "step": 969
+    },
+    {
+      "epoch": 0.35544155368266767,
+      "grad_norm": 1.5927033424377441,
+      "learning_rate": 1.44537327468711e-05,
+      "loss": 0.8754,
+      "step": 970
+    },
+    {
+      "epoch": 0.3558079882740931,
+      "grad_norm": 1.8709561824798584,
+      "learning_rate": 1.4443384762571781e-05,
+      "loss": 0.8707,
+      "step": 971
+    },
+    {
+      "epoch": 0.3561744228655185,
+      "grad_norm": 1.2936451435089111,
+      "learning_rate": 1.4433030846353582e-05,
+      "loss": 0.8956,
+      "step": 972
+    },
+    {
+      "epoch": 0.3565408574569439,
+      "grad_norm": 1.4189300537109375,
+      "learning_rate": 1.4422671012038982e-05,
+      "loss": 0.9132,
+      "step": 973
+    },
+    {
+      "epoch": 0.35690729204836935,
+      "grad_norm": 1.8624743223190308,
+      "learning_rate": 1.4412305273458361e-05,
+      "loss": 0.7831,
+      "step": 974
+    },
+    {
+      "epoch": 0.3572737266397948,
+      "grad_norm": 1.4298512935638428,
+      "learning_rate": 1.4401933644449977e-05,
+      "loss": 0.8054,
+      "step": 975
+    },
+    {
+      "epoch": 0.3576401612312202,
+      "grad_norm": 1.5367136001586914,
+      "learning_rate": 1.4391556138859962e-05,
+      "loss": 0.8131,
+      "step": 976
+    },
+    {
+      "epoch": 0.35800659582264566,
+      "grad_norm": 1.4892581701278687,
+      "learning_rate": 1.4381172770542283e-05,
+      "loss": 0.8886,
+      "step": 977
+    },
+    {
+      "epoch": 0.3583730304140711,
+      "grad_norm": 1.5779441595077515,
+      "learning_rate": 1.4370783553358735e-05,
+      "loss": 0.8548,
+      "step": 978
+    },
+    {
+      "epoch": 0.3587394650054965,
+      "grad_norm": 1.9342527389526367,
+      "learning_rate": 1.4360388501178925e-05,
+      "loss": 0.7853,
+      "step": 979
+    },
+    {
+      "epoch": 0.35910589959692196,
+      "grad_norm": 1.5269722938537598,
+      "learning_rate": 1.4349987627880251e-05,
+      "loss": 0.8765,
+      "step": 980
+    },
+    {
+      "epoch": 0.3594723341883474,
+      "grad_norm": 1.4872721433639526,
+      "learning_rate": 1.4339580947347877e-05,
+      "loss": 0.8588,
+      "step": 981
+    },
+    {
+      "epoch": 0.35983876877977283,
+      "grad_norm": 1.6207835674285889,
+      "learning_rate": 1.4329168473474722e-05,
+      "loss": 0.8713,
+      "step": 982
+    },
+    {
+      "epoch": 0.36020520337119827,
+      "grad_norm": 1.7222713232040405,
+      "learning_rate": 1.431875022016144e-05,
+      "loss": 0.9375,
+      "step": 983
+    },
+    {
+      "epoch": 0.36057163796262365,
+      "grad_norm": 1.682226300239563,
+      "learning_rate": 1.4308326201316399e-05,
+      "loss": 0.896,
+      "step": 984
+    },
+    {
+      "epoch": 0.3609380725540491,
+      "grad_norm": 1.8652957677841187,
+      "learning_rate": 1.4297896430855662e-05,
+      "loss": 0.826,
+      "step": 985
+    },
+    {
+      "epoch": 0.3613045071454745,
+      "grad_norm": 1.1483632326126099,
+      "learning_rate": 1.428746092270298e-05,
+      "loss": 0.5728,
+      "step": 986
+    },
+    {
+      "epoch": 0.36167094173689995,
+      "grad_norm": 1.5997368097305298,
+      "learning_rate": 1.4277019690789749e-05,
+      "loss": 0.9115,
+      "step": 987
+    },
+    {
+      "epoch": 0.3620373763283254,
+      "grad_norm": 1.8785408735275269,
+      "learning_rate": 1.4266572749055022e-05,
+      "loss": 0.8324,
+      "step": 988
+    },
+    {
+      "epoch": 0.3624038109197508,
+      "grad_norm": 1.4323786497116089,
+      "learning_rate": 1.425612011144546e-05,
+      "loss": 0.9204,
+      "step": 989
+    },
+    {
+      "epoch": 0.36277024551117626,
+      "grad_norm": 1.4402960538864136,
+      "learning_rate": 1.4245661791915336e-05,
+      "loss": 0.9272,
+      "step": 990
+    },
+    {
+      "epoch": 0.3631366801026017,
+      "grad_norm": 1.5204726457595825,
+      "learning_rate": 1.423519780442651e-05,
+      "loss": 0.8778,
+      "step": 991
+    },
+    {
+      "epoch": 0.36350311469402713,
+      "grad_norm": 1.6032034158706665,
+      "learning_rate": 1.4224728162948403e-05,
+      "loss": 0.8389,
+      "step": 992
+    },
+    {
+      "epoch": 0.36386954928545256,
+      "grad_norm": 2.3969008922576904,
+      "learning_rate": 1.4214252881457987e-05,
+      "loss": 0.7944,
+      "step": 993
+    },
+    {
+      "epoch": 0.364235983876878,
+      "grad_norm": 1.8171178102493286,
+      "learning_rate": 1.4203771973939769e-05,
+      "loss": 0.8405,
+      "step": 994
+    },
+    {
+      "epoch": 0.36460241846830344,
+      "grad_norm": 1.6910125017166138,
+      "learning_rate": 1.4193285454385755e-05,
+      "loss": 0.8838,
+      "step": 995
+    },
+    {
+      "epoch": 0.3649688530597288,
+      "grad_norm": 1.770917534828186,
+      "learning_rate": 1.4182793336795446e-05,
+      "loss": 0.7731,
+      "step": 996
+    },
+    {
+      "epoch": 0.36533528765115425,
+      "grad_norm": 1.5590789318084717,
+      "learning_rate": 1.417229563517583e-05,
+      "loss": 0.8358,
+      "step": 997
+    },
+    {
+      "epoch": 0.3657017222425797,
+      "grad_norm": 1.6439181566238403,
+      "learning_rate": 1.4161792363541329e-05,
+      "loss": 0.8601,
+      "step": 998
+    },
+    {
+      "epoch": 0.3660681568340051,
+      "grad_norm": 1.4935253858566284,
+      "learning_rate": 1.4151283535913819e-05,
+      "loss": 0.8239,
+      "step": 999
+    },
+    {
+      "epoch": 0.36643459142543056,
+      "grad_norm": 1.662537932395935,
+      "learning_rate": 1.414076916632258e-05,
+      "loss": 0.901,
+      "step": 1000
+    },
+    {
+      "epoch": 0.366801026016856,
+      "grad_norm": 1.5783257484436035,
+      "learning_rate": 1.4130249268804296e-05,
+      "loss": 0.7823,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3671674606082814,
+      "grad_norm": 1.5855540037155151,
+      "learning_rate": 1.4119723857403039e-05,
+      "loss": 0.9523,
+      "step": 1002
+    },
+    {
+      "epoch": 0.36753389519970686,
+      "grad_norm": 1.349881887435913,
+      "learning_rate": 1.410919294617022e-05,
+      "loss": 0.8282,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3679003297911323,
+      "grad_norm": 1.4568612575531006,
+      "learning_rate": 1.4098656549164615e-05,
+      "loss": 0.9527,
+      "step": 1004
+    },
+    {
+      "epoch": 0.36826676438255773,
+      "grad_norm": 1.461704969406128,
+      "learning_rate": 1.4088114680452314e-05,
+      "loss": 0.9008,
+      "step": 1005
+    },
+    {
+      "epoch": 0.36863319897398317,
+      "grad_norm": 1.5116150379180908,
+      "learning_rate": 1.4077567354106711e-05,
+      "loss": 0.8853,
+      "step": 1006
+    },
+    {
+      "epoch": 0.36899963356540855,
+      "grad_norm": 1.7470734119415283,
+      "learning_rate": 1.406701458420849e-05,
+      "loss": 0.8858,
+      "step": 1007
+    },
+    {
+      "epoch": 0.369366068156834,
+      "grad_norm": 1.6353769302368164,
+      "learning_rate": 1.4056456384845592e-05,
+      "loss": 0.8338,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3697325027482594,
+      "grad_norm": 1.7861933708190918,
+      "learning_rate": 1.4045892770113222e-05,
+      "loss": 0.8883,
+      "step": 1009
+    },
+    {
+      "epoch": 0.37009893733968485,
+      "grad_norm": 1.5234062671661377,
+      "learning_rate": 1.40353237541138e-05,
+      "loss": 0.9258,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3704653719311103,
+      "grad_norm": 1.6214821338653564,
+      "learning_rate": 1.4024749350956965e-05,
+      "loss": 0.8274,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3708318065225357,
+      "grad_norm": 1.4756487607955933,
+      "learning_rate": 1.4014169574759548e-05,
+      "loss": 0.8759,
+      "step": 1012
+    },
+    {
+      "epoch": 0.37119824111396116,
+      "grad_norm": 1.4953577518463135,
+      "learning_rate": 1.4003584439645545e-05,
+      "loss": 0.935,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3715646757053866,
+      "grad_norm": 1.6752123832702637,
+      "learning_rate": 1.399299395974612e-05,
+      "loss": 0.9121,
+      "step": 1014
+    },
+    {
+      "epoch": 0.37193111029681203,
+      "grad_norm": 1.8773505687713623,
+      "learning_rate": 1.398239814919956e-05,
+      "loss": 0.8101,
+      "step": 1015
+    },
+    {
+      "epoch": 0.37229754488823746,
+      "grad_norm": 1.6290862560272217,
+      "learning_rate": 1.3971797022151268e-05,
+      "loss": 0.9078,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3726639794796629,
+      "grad_norm": 1.563781976699829,
+      "learning_rate": 1.3961190592753757e-05,
+      "loss": 0.877,
+      "step": 1017
+    },
+    {
+      "epoch": 0.37303041407108833,
+      "grad_norm": 1.7269985675811768,
+      "learning_rate": 1.3950578875166608e-05,
+      "loss": 0.8254,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3733968486625137,
+      "grad_norm": 1.568712592124939,
+      "learning_rate": 1.3939961883556461e-05,
+      "loss": 0.8323,
+      "step": 1019
+    },
+    {
+      "epoch": 0.37376328325393915,
+      "grad_norm": 1.534967064857483,
+      "learning_rate": 1.3929339632097008e-05,
+      "loss": 0.8769,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3741297178453646,
+      "grad_norm": 1.5710378885269165,
+      "learning_rate": 1.3918712134968946e-05,
+      "loss": 0.8604,
+      "step": 1021
+    },
+    {
+      "epoch": 0.37449615243679,
+      "grad_norm": 1.882443904876709,
+      "learning_rate": 1.3908079406359991e-05,
+      "loss": 0.8033,
+      "step": 1022
+    },
+    {
+      "epoch": 0.37486258702821545,
+      "grad_norm": 1.520538568496704,
+      "learning_rate": 1.3897441460464834e-05,
+      "loss": 0.9032,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3752290216196409,
+      "grad_norm": 1.724283218383789,
+      "learning_rate": 1.3886798311485133e-05,
+      "loss": 0.8723,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3755954562110663,
+      "grad_norm": 1.5896741151809692,
+      "learning_rate": 1.3876149973629492e-05,
+      "loss": 0.7985,
+      "step": 1025
+    },
+    {
+      "epoch": 0.37596189080249176,
+      "grad_norm": 1.8196618556976318,
+      "learning_rate": 1.386549646111344e-05,
+      "loss": 0.701,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3763283253939172,
+      "grad_norm": 1.603785514831543,
+      "learning_rate": 1.3854837788159424e-05,
+      "loss": 0.8844,
+      "step": 1027
+    },
+    {
+      "epoch": 0.37669475998534263,
+      "grad_norm": 1.6494451761245728,
+      "learning_rate": 1.384417396899677e-05,
+      "loss": 0.8592,
+      "step": 1028
+    },
+    {
+      "epoch": 0.37706119457676807,
+      "grad_norm": 1.7041468620300293,
+      "learning_rate": 1.3833505017861674e-05,
+      "loss": 0.8691,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3774276291681935,
+      "grad_norm": 1.751847267150879,
+      "learning_rate": 1.382283094899719e-05,
+      "loss": 0.8117,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3777940637596189,
+      "grad_norm": 1.5712251663208008,
+      "learning_rate": 1.3812151776653195e-05,
+      "loss": 0.8617,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3781604983510443,
+      "grad_norm": 1.6731207370758057,
+      "learning_rate": 1.3801467515086391e-05,
+      "loss": 0.8124,
+      "step": 1032
+    },
+    {
+      "epoch": 0.37852693294246975,
+      "grad_norm": 1.4233975410461426,
+      "learning_rate": 1.3790778178560266e-05,
+      "loss": 0.8663,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3788933675338952,
+      "grad_norm": 1.6427613496780396,
+      "learning_rate": 1.3780083781345084e-05,
+      "loss": 0.9296,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3792598021253206,
+      "grad_norm": 1.4288830757141113,
+      "learning_rate": 1.3769384337717864e-05,
+      "loss": 0.8394,
+      "step": 1035
+    },
+    {
+      "epoch": 0.37962623671674606,
+      "grad_norm": 1.548366904258728,
+      "learning_rate": 1.3758679861962366e-05,
+      "loss": 0.8548,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3799926713081715,
+      "grad_norm": 1.6227176189422607,
+      "learning_rate": 1.3747970368369063e-05,
+      "loss": 0.8663,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3803591058995969,
+      "grad_norm": 1.6395370960235596,
+      "learning_rate": 1.3737255871235133e-05,
+      "loss": 0.7948,
+      "step": 1038
+    },
+    {
+      "epoch": 0.38072554049102236,
+      "grad_norm": 1.5784190893173218,
+      "learning_rate": 1.372653638486443e-05,
+      "loss": 0.8652,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3810919750824478,
+      "grad_norm": 1.3747224807739258,
+      "learning_rate": 1.3715811923567468e-05,
+      "loss": 0.8831,
+      "step": 1040
+    },
+    {
+      "epoch": 0.38145840967387323,
+      "grad_norm": 1.5469486713409424,
+      "learning_rate": 1.3705082501661402e-05,
+      "loss": 0.8871,
+      "step": 1041
+    },
+    {
+      "epoch": 0.38182484426529867,
+      "grad_norm": 1.5060057640075684,
+      "learning_rate": 1.369434813347001e-05,
+      "loss": 0.907,
+      "step": 1042
+    },
+    {
+      "epoch": 0.38219127885672405,
+      "grad_norm": 1.5088938474655151,
+      "learning_rate": 1.368360883332368e-05,
+      "loss": 0.8994,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3825577134481495,
+      "grad_norm": 1.589644432067871,
+      "learning_rate": 1.3672864615559372e-05,
+      "loss": 0.8767,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3829241480395749,
+      "grad_norm": 1.5223177671432495,
+      "learning_rate": 1.3662115494520618e-05,
+      "loss": 0.9506,
+      "step": 1045
+    },
+    {
+      "epoch": 0.38329058263100035,
+      "grad_norm": 1.5508954524993896,
+      "learning_rate": 1.36513614845575e-05,
+      "loss": 0.9139,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3836570172224258,
+      "grad_norm": 1.9155925512313843,
+      "learning_rate": 1.3640602600026613e-05,
+      "loss": 0.7311,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3840234518138512,
+      "grad_norm": 1.6248713731765747,
+      "learning_rate": 1.3629838855291078e-05,
+      "loss": 0.8691,
+      "step": 1048
+    },
+    {
+      "epoch": 0.38438988640527666,
+      "grad_norm": 1.4297411441802979,
+      "learning_rate": 1.3619070264720489e-05,
+      "loss": 0.8643,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3847563209967021,
+      "grad_norm": 1.5291481018066406,
+      "learning_rate": 1.3608296842690916e-05,
+      "loss": 0.8865,
+      "step": 1050
+    },
+    {
+      "epoch": 0.38512275558812753,
+      "grad_norm": 1.5512865781784058,
+      "learning_rate": 1.3597518603584878e-05,
+      "loss": 0.8837,
+      "step": 1051
+    },
+    {
+      "epoch": 0.38548919017955297,
+      "grad_norm": 1.4452577829360962,
+      "learning_rate": 1.3586735561791327e-05,
+      "loss": 0.9034,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3858556247709784,
+      "grad_norm": 1.3865025043487549,
+      "learning_rate": 1.3575947731705624e-05,
+      "loss": 0.8562,
+      "step": 1053
+    },
+    {
+      "epoch": 0.38622205936240384,
+      "grad_norm": 1.528044581413269,
+      "learning_rate": 1.3565155127729516e-05,
+      "loss": 0.9205,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3865884939538292,
+      "grad_norm": 1.539243221282959,
+      "learning_rate": 1.3554357764271138e-05,
+      "loss": 0.8596,
+      "step": 1055
+    },
+    {
+      "epoch": 0.38695492854525465,
+      "grad_norm": 1.7668794393539429,
+      "learning_rate": 1.3543555655744966e-05,
+      "loss": 0.8194,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3873213631366801,
+      "grad_norm": 1.5567411184310913,
+      "learning_rate": 1.3532748816571821e-05,
+      "loss": 0.9152,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3876877977281055,
+      "grad_norm": 1.575163722038269,
+      "learning_rate": 1.3521937261178826e-05,
+      "loss": 0.87,
+      "step": 1058
+    },
+    {
+      "epoch": 0.38805423231953096,
+      "grad_norm": 1.6608469486236572,
+      "learning_rate": 1.3511121003999414e-05,
+      "loss": 0.8094,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3884206669109564,
+      "grad_norm": 1.887389063835144,
+      "learning_rate": 1.3500300059473285e-05,
+      "loss": 0.7911,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3887871015023818,
+      "grad_norm": 1.8042851686477661,
+      "learning_rate": 1.3489474442046404e-05,
+      "loss": 0.8217,
+      "step": 1061
+    },
+    {
+      "epoch": 0.38915353609380726,
+      "grad_norm": 1.6864506006240845,
+      "learning_rate": 1.3478644166170968e-05,
+      "loss": 0.9135,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3895199706852327,
+      "grad_norm": 1.9867236614227295,
+      "learning_rate": 1.3467809246305398e-05,
+      "loss": 0.8669,
+      "step": 1063
+    },
+    {
+      "epoch": 0.38988640527665813,
+      "grad_norm": 1.8301372528076172,
+      "learning_rate": 1.3456969696914313e-05,
+      "loss": 0.813,
+      "step": 1064
+    },
+    {
+      "epoch": 0.39025283986808357,
+      "grad_norm": 1.6149121522903442,
+      "learning_rate": 1.3446125532468507e-05,
+      "loss": 0.8958,
+      "step": 1065
+    },
+    {
+      "epoch": 0.390619274459509,
+      "grad_norm": 1.5697036981582642,
+      "learning_rate": 1.3435276767444946e-05,
+      "loss": 0.921,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3909857090509344,
+      "grad_norm": 1.525789499282837,
+      "learning_rate": 1.3424423416326727e-05,
+      "loss": 0.8502,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3913521436423598,
+      "grad_norm": 1.7171850204467773,
+      "learning_rate": 1.341356549360308e-05,
+      "loss": 0.7838,
+      "step": 1068
+    },
+    {
+      "epoch": 0.39171857823378525,
+      "grad_norm": 1.521950364112854,
+      "learning_rate": 1.3402703013769326e-05,
+      "loss": 0.8726,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3920850128252107,
+      "grad_norm": 1.6256566047668457,
+      "learning_rate": 1.3391835991326882e-05,
+      "loss": 0.954,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3924514474166361,
+      "grad_norm": 1.6131467819213867,
+      "learning_rate": 1.338096444078322e-05,
+      "loss": 0.884,
+      "step": 1071
+    },
+    {
+      "epoch": 0.39281788200806156,
+      "grad_norm": 1.5478419065475464,
+      "learning_rate": 1.337008837665186e-05,
+      "loss": 0.8749,
+      "step": 1072
+    },
+    {
+      "epoch": 0.393184316599487,
+      "grad_norm": 1.5852874517440796,
+      "learning_rate": 1.3359207813452348e-05,
+      "loss": 0.8507,
+      "step": 1073
+    },
+    {
+      "epoch": 0.39355075119091243,
+      "grad_norm": 1.6789813041687012,
+      "learning_rate": 1.334832276571024e-05,
+      "loss": 0.8886,
+      "step": 1074
+    },
+    {
+      "epoch": 0.39391718578233786,
+      "grad_norm": 1.699114203453064,
+      "learning_rate": 1.3337433247957075e-05,
+      "loss": 0.8698,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3942836203737633,
+      "grad_norm": 1.515486240386963,
+      "learning_rate": 1.3326539274730354e-05,
+      "loss": 0.8625,
+      "step": 1076
+    },
+    {
+      "epoch": 0.39465005496518873,
+      "grad_norm": 1.5709253549575806,
+      "learning_rate": 1.3315640860573537e-05,
+      "loss": 0.8617,
+      "step": 1077
+    },
+    {
+      "epoch": 0.39501648955661417,
+      "grad_norm": 1.5914605855941772,
+      "learning_rate": 1.3304738020036006e-05,
+      "loss": 0.8855,
+      "step": 1078
+    },
+    {
+      "epoch": 0.39538292414803955,
+      "grad_norm": 1.4676405191421509,
+      "learning_rate": 1.3293830767673055e-05,
+      "loss": 0.8757,
+      "step": 1079
+    },
+    {
+      "epoch": 0.395749358739465,
+      "grad_norm": 1.7750840187072754,
+      "learning_rate": 1.3282919118045866e-05,
+      "loss": 0.8586,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3961157933308904,
+      "grad_norm": 1.5571708679199219,
+      "learning_rate": 1.3272003085721485e-05,
+      "loss": 0.9255,
+      "step": 1081
+    },
+    {
+      "epoch": 0.39648222792231586,
+      "grad_norm": 1.6908925771713257,
+      "learning_rate": 1.3261082685272827e-05,
+      "loss": 0.8494,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3968486625137413,
+      "grad_norm": 1.520818829536438,
+      "learning_rate": 1.325015793127862e-05,
+      "loss": 0.8573,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3972150971051667,
+      "grad_norm": 1.4602961540222168,
+      "learning_rate": 1.3239228838323414e-05,
+      "loss": 0.9259,
+      "step": 1084
+    },
+    {
+      "epoch": 0.39758153169659216,
+      "grad_norm": 1.6277791261672974,
+      "learning_rate": 1.3228295420997546e-05,
+      "loss": 0.8911,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3979479662880176,
+      "grad_norm": 1.7978880405426025,
+      "learning_rate": 1.3217357693897133e-05,
+      "loss": 0.9046,
+      "step": 1086
+    },
+    {
+      "epoch": 0.39831440087944303,
+      "grad_norm": 1.6153454780578613,
+      "learning_rate": 1.3206415671624036e-05,
+      "loss": 0.8602,
+      "step": 1087
+    },
+    {
+      "epoch": 0.39868083547086847,
+      "grad_norm": 1.5234719514846802,
+      "learning_rate": 1.3195469368785858e-05,
+      "loss": 0.9155,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3990472700622939,
+      "grad_norm": 1.5738810300827026,
+      "learning_rate": 1.3184518799995912e-05,
+      "loss": 0.853,
+      "step": 1089
+    },
+    {
+      "epoch": 0.39941370465371934,
+      "grad_norm": 1.6827136278152466,
+      "learning_rate": 1.3173563979873212e-05,
+      "loss": 0.9015,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3997801392451447,
+      "grad_norm": 1.7982146739959717,
+      "learning_rate": 1.316260492304244e-05,
+      "loss": 0.8608,
+      "step": 1091
+    },
+    {
+      "epoch": 0.40014657383657015,
+      "grad_norm": 1.711007833480835,
+      "learning_rate": 1.3151641644133942e-05,
+      "loss": 0.8817,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4005130084279956,
+      "grad_norm": 1.7627228498458862,
+      "learning_rate": 1.3140674157783686e-05,
+      "loss": 0.8585,
+      "step": 1093
+    },
+    {
+      "epoch": 0.400879443019421,
+      "grad_norm": 1.6841421127319336,
+      "learning_rate": 1.3129702478633275e-05,
+      "loss": 0.8468,
+      "step": 1094
+    },
+    {
+      "epoch": 0.40124587761084646,
+      "grad_norm": 1.6395200490951538,
+      "learning_rate": 1.31187266213299e-05,
+      "loss": 0.91,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4016123122022719,
+      "grad_norm": 1.6827080249786377,
+      "learning_rate": 1.310774660052633e-05,
+      "loss": 0.8466,
+      "step": 1096
+    },
+    {
+      "epoch": 0.40197874679369733,
+      "grad_norm": 1.5233503580093384,
+      "learning_rate": 1.3096762430880894e-05,
+      "loss": 0.9055,
+      "step": 1097
+    },
+    {
+      "epoch": 0.40234518138512276,
+      "grad_norm": 1.6009891033172607,
+      "learning_rate": 1.308577412705746e-05,
+      "loss": 0.8622,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4027116159765482,
+      "grad_norm": 1.6245487928390503,
+      "learning_rate": 1.307478170372541e-05,
+      "loss": 0.8656,
+      "step": 1099
+    },
+    {
+      "epoch": 0.40307805056797363,
+      "grad_norm": 1.6522853374481201,
+      "learning_rate": 1.3063785175559634e-05,
+      "loss": 0.9106,
+      "step": 1100
+    },
+    {
+      "epoch": 0.40344448515939907,
+      "grad_norm": 1.8431390523910522,
+      "learning_rate": 1.3052784557240495e-05,
+      "loss": 0.8102,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4038109197508245,
+      "grad_norm": 1.626137614250183,
+      "learning_rate": 1.3041779863453821e-05,
+      "loss": 0.8664,
+      "step": 1102
+    },
+    {
+      "epoch": 0.4041773543422499,
+      "grad_norm": 1.595874547958374,
+      "learning_rate": 1.3030771108890878e-05,
+      "loss": 0.8705,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4045437889336753,
+      "grad_norm": 1.4765444993972778,
+      "learning_rate": 1.3019758308248351e-05,
+      "loss": 0.8728,
+      "step": 1104
+    },
+    {
+      "epoch": 0.40491022352510075,
+      "grad_norm": 1.9476228952407837,
+      "learning_rate": 1.3008741476228335e-05,
+      "loss": 0.849,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4052766581165262,
+      "grad_norm": 1.7913713455200195,
+      "learning_rate": 1.2997720627538297e-05,
+      "loss": 0.8579,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4056430927079516,
+      "grad_norm": 1.8295342922210693,
+      "learning_rate": 1.2986695776891072e-05,
+      "loss": 0.8795,
+      "step": 1107
+    },
+    {
+      "epoch": 0.40600952729937706,
+      "grad_norm": 1.7225682735443115,
+      "learning_rate": 1.2975666939004839e-05,
+      "loss": 0.8328,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4063759618908025,
+      "grad_norm": 1.6390389204025269,
+      "learning_rate": 1.2964634128603092e-05,
+      "loss": 0.869,
+      "step": 1109
+    },
+    {
+      "epoch": 0.40674239648222793,
+      "grad_norm": 1.4864574670791626,
+      "learning_rate": 1.2953597360414636e-05,
+      "loss": 0.9074,
+      "step": 1110
+    },
+    {
+      "epoch": 0.40710883107365337,
+      "grad_norm": 1.5992265939712524,
+      "learning_rate": 1.2942556649173557e-05,
+      "loss": 0.8928,
+      "step": 1111
+    },
+    {
+      "epoch": 0.4074752656650788,
+      "grad_norm": 1.5651394128799438,
+      "learning_rate": 1.2931512009619202e-05,
+      "loss": 0.8772,
+      "step": 1112
+    },
+    {
+      "epoch": 0.40784170025650424,
+      "grad_norm": 1.669217586517334,
+      "learning_rate": 1.292046345649617e-05,
+      "loss": 0.9396,
+      "step": 1113
+    },
+    {
+      "epoch": 0.40820813484792967,
+      "grad_norm": 1.7099789381027222,
+      "learning_rate": 1.290941100455428e-05,
+      "loss": 0.8651,
+      "step": 1114
+    },
+    {
+      "epoch": 0.40857456943935505,
+      "grad_norm": 1.6064772605895996,
+      "learning_rate": 1.2898354668548554e-05,
+      "loss": 0.9055,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4089410040307805,
+      "grad_norm": 1.5557650327682495,
+      "learning_rate": 1.28872944632392e-05,
+      "loss": 0.8809,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4093074386222059,
+      "grad_norm": 1.7314295768737793,
+      "learning_rate": 1.2876230403391592e-05,
+      "loss": 0.8917,
+      "step": 1117
+    },
+    {
+      "epoch": 0.40967387321363136,
+      "grad_norm": 1.8550328016281128,
+      "learning_rate": 1.2865162503776257e-05,
+      "loss": 0.9015,
+      "step": 1118
+    },
+    {
+      "epoch": 0.4100403078050568,
+      "grad_norm": 1.5810977220535278,
+      "learning_rate": 1.285409077916883e-05,
+      "loss": 0.8637,
+      "step": 1119
+    },
+    {
+      "epoch": 0.4104067423964822,
+      "grad_norm": 1.6856402158737183,
+      "learning_rate": 1.2843015244350077e-05,
+      "loss": 0.8522,
+      "step": 1120
+    },
+    {
+      "epoch": 0.41077317698790766,
+      "grad_norm": 1.704975962638855,
+      "learning_rate": 1.2831935914105831e-05,
+      "loss": 0.8145,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4111396115793331,
+      "grad_norm": 1.7331234216690063,
+      "learning_rate": 1.2820852803226993e-05,
+      "loss": 0.9148,
+      "step": 1122
+    },
+    {
+      "epoch": 0.41150604617075853,
+      "grad_norm": 1.8245302438735962,
+      "learning_rate": 1.2809765926509527e-05,
+      "loss": 0.8974,
+      "step": 1123
+    },
+    {
+      "epoch": 0.41187248076218397,
+      "grad_norm": 1.6337167024612427,
+      "learning_rate": 1.2798675298754409e-05,
+      "loss": 0.8307,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4122389153536094,
+      "grad_norm": 1.7906094789505005,
+      "learning_rate": 1.2787580934767625e-05,
+      "loss": 0.7789,
+      "step": 1125
+    },
+    {
+      "epoch": 0.41260534994503484,
+      "grad_norm": 1.7973666191101074,
+      "learning_rate": 1.277648284936016e-05,
+      "loss": 0.8781,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4129717845364602,
+      "grad_norm": 1.5800189971923828,
+      "learning_rate": 1.2765381057347945e-05,
+      "loss": 0.8426,
+      "step": 1127
+    },
+    {
+      "epoch": 0.41333821912788565,
+      "grad_norm": 1.8464257717132568,
+      "learning_rate": 1.2754275573551886e-05,
+      "loss": 0.9326,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4137046537193111,
+      "grad_norm": 1.7335541248321533,
+      "learning_rate": 1.2743166412797799e-05,
+      "loss": 0.7967,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4140710883107365,
+      "grad_norm": 1.535427212715149,
+      "learning_rate": 1.273205358991641e-05,
+      "loss": 0.9106,
+      "step": 1130
+    },
+    {
+      "epoch": 0.41443752290216196,
+      "grad_norm": 1.7636157274246216,
+      "learning_rate": 1.2720937119743347e-05,
+      "loss": 0.8137,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4148039574935874,
+      "grad_norm": 1.6306973695755005,
+      "learning_rate": 1.2709817017119093e-05,
+      "loss": 0.8591,
+      "step": 1132
+    },
+    {
+      "epoch": 0.41517039208501283,
+      "grad_norm": 1.6848173141479492,
+      "learning_rate": 1.2698693296888983e-05,
+      "loss": 0.8442,
+      "step": 1133
+    },
+    {
+      "epoch": 0.41553682667643826,
+      "grad_norm": 1.9436910152435303,
+      "learning_rate": 1.2687565973903193e-05,
+      "loss": 0.8002,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4159032612678637,
+      "grad_norm": 1.861436128616333,
+      "learning_rate": 1.2676435063016688e-05,
+      "loss": 0.8841,
+      "step": 1135
+    },
+    {
+      "epoch": 0.41626969585928913,
+      "grad_norm": 1.7975494861602783,
+      "learning_rate": 1.2665300579089245e-05,
+      "loss": 0.8786,
+      "step": 1136
+    },
+    {
+      "epoch": 0.41663613045071457,
+      "grad_norm": 1.812900185585022,
+      "learning_rate": 1.2654162536985393e-05,
+      "loss": 0.8586,
+      "step": 1137
+    },
+    {
+      "epoch": 0.41700256504214,
+      "grad_norm": 1.8576949834823608,
+      "learning_rate": 1.2643020951574425e-05,
+      "loss": 0.8443,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4173689996335654,
+      "grad_norm": 1.931456208229065,
+      "learning_rate": 1.2631875837730355e-05,
+      "loss": 0.8383,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4177354342249908,
+      "grad_norm": 1.7419774532318115,
+      "learning_rate": 1.262072721033191e-05,
+      "loss": 0.8625,
+      "step": 1140
+    },
+    {
+      "epoch": 0.41810186881641626,
+      "grad_norm": 1.7812283039093018,
+      "learning_rate": 1.2609575084262506e-05,
+      "loss": 0.864,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4184683034078417,
+      "grad_norm": 1.7734988927841187,
+      "learning_rate": 1.2598419474410238e-05,
+      "loss": 0.8285,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4188347379992671,
+      "grad_norm": 1.7850379943847656,
+      "learning_rate": 1.2587260395667839e-05,
+      "loss": 0.8713,
+      "step": 1143
+    },
+    {
+      "epoch": 0.41920117259069256,
+      "grad_norm": 1.765945315361023,
+      "learning_rate": 1.2576097862932687e-05,
+      "loss": 0.8273,
+      "step": 1144
+    },
+    {
+      "epoch": 0.419567607182118,
+      "grad_norm": 1.604203701019287,
+      "learning_rate": 1.2564931891106755e-05,
+      "loss": 0.8301,
+      "step": 1145
+    },
+    {
+      "epoch": 0.41993404177354343,
+      "grad_norm": 1.6468043327331543,
+      "learning_rate": 1.2553762495096624e-05,
+      "loss": 0.9061,
+      "step": 1146
+    },
+    {
+      "epoch": 0.42030047636496887,
+      "grad_norm": 1.7401213645935059,
+      "learning_rate": 1.2542589689813433e-05,
+      "loss": 0.8487,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4206669109563943,
+      "grad_norm": 1.7834421396255493,
+      "learning_rate": 1.2531413490172882e-05,
+      "loss": 0.8003,
+      "step": 1148
+    },
+    {
+      "epoch": 0.42103334554781974,
+      "grad_norm": 1.6583036184310913,
+      "learning_rate": 1.2520233911095194e-05,
+      "loss": 0.8486,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4213997801392451,
+      "grad_norm": 1.8208075761795044,
+      "learning_rate": 1.250905096750511e-05,
+      "loss": 0.8591,
+      "step": 1150
+    },
+    {
+      "epoch": 0.42176621473067055,
+      "grad_norm": 1.756757140159607,
+      "learning_rate": 1.2497864674331858e-05,
+      "loss": 0.9033,
+      "step": 1151
+    },
+    {
+      "epoch": 0.422132649322096,
+      "grad_norm": 1.716612458229065,
+      "learning_rate": 1.2486675046509143e-05,
+      "loss": 0.8855,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4224990839135214,
+      "grad_norm": 1.66440749168396,
+      "learning_rate": 1.2475482098975116e-05,
+      "loss": 0.8667,
+      "step": 1153
+    },
+    {
+      "epoch": 0.42286551850494686,
+      "grad_norm": 1.9110910892486572,
+      "learning_rate": 1.2464285846672365e-05,
+      "loss": 0.8465,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4232319530963723,
+      "grad_norm": 1.6943145990371704,
+      "learning_rate": 1.2453086304547885e-05,
+      "loss": 0.8959,
+      "step": 1155
+    },
+    {
+      "epoch": 0.42359838768779773,
+      "grad_norm": 1.7550324201583862,
+      "learning_rate": 1.2441883487553066e-05,
+      "loss": 0.8589,
+      "step": 1156
+    },
+    {
+      "epoch": 0.42396482227922316,
+      "grad_norm": 1.9641809463500977,
+      "learning_rate": 1.243067741064367e-05,
+      "loss": 0.8543,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4243312568706486,
+      "grad_norm": 1.8191912174224854,
+      "learning_rate": 1.241946808877981e-05,
+      "loss": 0.8622,
+      "step": 1158
+    },
+    {
+      "epoch": 0.42469769146207403,
+      "grad_norm": 1.6547656059265137,
+      "learning_rate": 1.240825553692593e-05,
+      "loss": 0.8519,
+      "step": 1159
+    },
+    {
+      "epoch": 0.42506412605349947,
+      "grad_norm": 1.656889796257019,
+      "learning_rate": 1.2397039770050787e-05,
+      "loss": 0.9292,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4254305606449249,
+      "grad_norm": 1.8495047092437744,
+      "learning_rate": 1.2385820803127436e-05,
+      "loss": 0.8511,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4257969952363503,
+      "grad_norm": 1.9155131578445435,
+      "learning_rate": 1.2374598651133191e-05,
+      "loss": 0.8477,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4261634298277757,
+      "grad_norm": 1.7109763622283936,
+      "learning_rate": 1.2363373329049629e-05,
+      "loss": 0.8628,
+      "step": 1163
+    },
+    {
+      "epoch": 0.42652986441920115,
+      "grad_norm": 1.6846760511398315,
+      "learning_rate": 1.2352144851862553e-05,
+      "loss": 0.8688,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4268962990106266,
+      "grad_norm": 1.8518428802490234,
+      "learning_rate": 1.2340913234561986e-05,
+      "loss": 0.8429,
+      "step": 1165
+    },
+    {
+      "epoch": 0.427262733602052,
+      "grad_norm": 1.9886471033096313,
+      "learning_rate": 1.2329678492142134e-05,
+      "loss": 0.7325,
+      "step": 1166
+    },
+    {
+      "epoch": 0.42762916819347746,
+      "grad_norm": 1.7306740283966064,
+      "learning_rate": 1.2318440639601378e-05,
+      "loss": 0.8877,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4279956027849029,
+      "grad_norm": 2.007984161376953,
+      "learning_rate": 1.2307199691942252e-05,
+      "loss": 0.7804,
+      "step": 1168
+    },
+    {
+      "epoch": 0.42836203737632833,
+      "grad_norm": 1.6732981204986572,
+      "learning_rate": 1.2295955664171421e-05,
+      "loss": 0.8736,
+      "step": 1169
+    },
+    {
+      "epoch": 0.42872847196775377,
+      "grad_norm": 1.597426414489746,
+      "learning_rate": 1.2284708571299669e-05,
+      "loss": 0.8723,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4290949065591792,
+      "grad_norm": 1.7663847208023071,
+      "learning_rate": 1.2273458428341856e-05,
+      "loss": 0.9053,
+      "step": 1171
+    },
+    {
+      "epoch": 0.42946134115060464,
+      "grad_norm": 1.6893157958984375,
+      "learning_rate": 1.226220525031693e-05,
+      "loss": 0.8663,
+      "step": 1172
+    },
+    {
+      "epoch": 0.42982777574203007,
+      "grad_norm": 1.6163278818130493,
+      "learning_rate": 1.2250949052247883e-05,
+      "loss": 0.9114,
+      "step": 1173
+    },
+    {
+      "epoch": 0.43019421033345545,
+      "grad_norm": 1.7489039897918701,
+      "learning_rate": 1.2239689849161738e-05,
+      "loss": 0.8597,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4305606449248809,
+      "grad_norm": 1.7178845405578613,
+      "learning_rate": 1.2228427656089532e-05,
+      "loss": 0.8663,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4309270795163063,
+      "grad_norm": 1.952564001083374,
+      "learning_rate": 1.2217162488066295e-05,
+      "loss": 0.8371,
+      "step": 1176
+    },
+    {
+      "epoch": 0.43129351410773176,
+      "grad_norm": 1.9330706596374512,
+      "learning_rate": 1.2205894360131026e-05,
+      "loss": 0.863,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4316599486991572,
+      "grad_norm": 1.6406487226486206,
+      "learning_rate": 1.2194623287326679e-05,
+      "loss": 0.9078,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4320263832905826,
+      "grad_norm": 1.7578526735305786,
+      "learning_rate": 1.218334928470013e-05,
+      "loss": 0.7632,
+      "step": 1179
+    },
+    {
+      "epoch": 0.43239281788200806,
+      "grad_norm": 1.9374595880508423,
+      "learning_rate": 1.2172072367302178e-05,
+      "loss": 0.7159,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4327592524734335,
+      "grad_norm": 1.8487389087677002,
+      "learning_rate": 1.2160792550187509e-05,
+      "loss": 0.8132,
+      "step": 1181
+    },
+    {
+      "epoch": 0.43312568706485893,
+      "grad_norm": 2.047722816467285,
+      "learning_rate": 1.2149509848414676e-05,
+      "loss": 0.7963,
+      "step": 1182
+    },
+    {
+      "epoch": 0.43349212165628437,
+      "grad_norm": 1.8400959968566895,
+      "learning_rate": 1.2138224277046089e-05,
+      "loss": 0.8553,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4338585562477098,
+      "grad_norm": 1.7545138597488403,
+      "learning_rate": 1.2126935851147987e-05,
+      "loss": 0.8333,
+      "step": 1184
+    },
+    {
+      "epoch": 0.43422499083913524,
+      "grad_norm": 1.6373624801635742,
+      "learning_rate": 1.2115644585790419e-05,
+      "loss": 0.9219,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4345914254305606,
+      "grad_norm": 1.748104453086853,
+      "learning_rate": 1.2104350496047221e-05,
+      "loss": 0.8288,
+      "step": 1186
+    },
+    {
+      "epoch": 0.43495786002198605,
+      "grad_norm": 1.9308727979660034,
+      "learning_rate": 1.2093053596996005e-05,
+      "loss": 0.8304,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4353242946134115,
+      "grad_norm": 1.7878676652908325,
+      "learning_rate": 1.2081753903718137e-05,
+      "loss": 0.8218,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4356907292048369,
+      "grad_norm": 1.71224045753479,
+      "learning_rate": 1.2070451431298702e-05,
+      "loss": 0.875,
+      "step": 1189
+    },
+    {
+      "epoch": 0.43605716379626236,
+      "grad_norm": 1.8005670309066772,
+      "learning_rate": 1.2059146194826503e-05,
+      "loss": 0.8343,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4364235983876878,
+      "grad_norm": 1.8443491458892822,
+      "learning_rate": 1.204783820939403e-05,
+      "loss": 0.7996,
+      "step": 1191
+    },
+    {
+      "epoch": 0.43679003297911323,
+      "grad_norm": 2.1138112545013428,
+      "learning_rate": 1.2036527490097445e-05,
+      "loss": 0.805,
+      "step": 1192
+    },
+    {
+      "epoch": 0.43715646757053866,
+      "grad_norm": 1.7158002853393555,
+      "learning_rate": 1.202521405203656e-05,
+      "loss": 0.8328,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4375229021619641,
+      "grad_norm": 1.6394296884536743,
+      "learning_rate": 1.2013897910314816e-05,
+      "loss": 0.8793,
+      "step": 1194
+    },
+    {
+      "epoch": 0.43788933675338954,
+      "grad_norm": 1.9240604639053345,
+      "learning_rate": 1.200257908003926e-05,
+      "loss": 0.8484,
+      "step": 1195
+    },
+    {
+      "epoch": 0.43825577134481497,
+      "grad_norm": 1.6958271265029907,
+      "learning_rate": 1.1991257576320537e-05,
+      "loss": 0.9014,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4386222059362404,
+      "grad_norm": 1.7841869592666626,
+      "learning_rate": 1.1979933414272844e-05,
+      "loss": 0.8969,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4389886405276658,
+      "grad_norm": 1.571500539779663,
+      "learning_rate": 1.1968606609013949e-05,
+      "loss": 0.8512,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4393550751190912,
+      "grad_norm": 1.7198463678359985,
+      "learning_rate": 1.195727717566513e-05,
+      "loss": 0.8946,
+      "step": 1199
+    },
+    {
+      "epoch": 0.43972150971051666,
+      "grad_norm": 1.7086559534072876,
+      "learning_rate": 1.1945945129351184e-05,
+      "loss": 0.8485,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4400879443019421,
+      "grad_norm": 1.6819955110549927,
+      "learning_rate": 1.1934610485200398e-05,
+      "loss": 0.8898,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4404543788933675,
+      "grad_norm": 1.8454056978225708,
+      "learning_rate": 1.1923273258344513e-05,
+      "loss": 0.8706,
+      "step": 1202
+    },
+    {
+      "epoch": 0.44082081348479296,
+      "grad_norm": 1.705040693283081,
+      "learning_rate": 1.1911933463918734e-05,
+      "loss": 0.8426,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4411872480762184,
+      "grad_norm": 1.9316009283065796,
+      "learning_rate": 1.1900591117061684e-05,
+      "loss": 0.7853,
+      "step": 1204
+    },
+    {
+      "epoch": 0.44155368266764383,
+      "grad_norm": 1.7013500928878784,
+      "learning_rate": 1.1889246232915399e-05,
+      "loss": 0.7973,
+      "step": 1205
+    },
+    {
+      "epoch": 0.44192011725906927,
+      "grad_norm": 1.675676941871643,
+      "learning_rate": 1.1877898826625298e-05,
+      "loss": 0.8407,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4422865518504947,
+      "grad_norm": 1.847190499305725,
+      "learning_rate": 1.1866548913340174e-05,
+      "loss": 0.8683,
+      "step": 1207
+    },
+    {
+      "epoch": 0.44265298644192014,
+      "grad_norm": 2.021604537963867,
+      "learning_rate": 1.1855196508212157e-05,
+      "loss": 0.8217,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4430194210333456,
+      "grad_norm": 1.7539337873458862,
+      "learning_rate": 1.1843841626396705e-05,
+      "loss": 0.8669,
+      "step": 1209
+    },
+    {
+      "epoch": 0.44338585562477095,
+      "grad_norm": 1.8103632926940918,
+      "learning_rate": 1.1832484283052592e-05,
+      "loss": 0.9176,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4437522902161964,
+      "grad_norm": 1.998170256614685,
+      "learning_rate": 1.1821124493341868e-05,
+      "loss": 0.9539,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4441187248076218,
+      "grad_norm": 1.938361644744873,
+      "learning_rate": 1.180976227242986e-05,
+      "loss": 0.7924,
+      "step": 1212
+    },
+    {
+      "epoch": 0.44448515939904726,
+      "grad_norm": 1.9992772340774536,
+      "learning_rate": 1.1798397635485124e-05,
+      "loss": 0.9023,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4448515939904727,
+      "grad_norm": 1.6736135482788086,
+      "learning_rate": 1.1787030597679456e-05,
+      "loss": 0.9319,
+      "step": 1214
+    },
+    {
+      "epoch": 0.44521802858189813,
+      "grad_norm": 1.6292542219161987,
+      "learning_rate": 1.1775661174187851e-05,
+      "loss": 0.8363,
+      "step": 1215
+    },
+    {
+      "epoch": 0.44558446317332356,
+      "grad_norm": 1.8140482902526855,
+      "learning_rate": 1.1764289380188492e-05,
+      "loss": 0.8774,
+      "step": 1216
+    },
+    {
+      "epoch": 0.445950897764749,
+      "grad_norm": 1.6420940160751343,
+      "learning_rate": 1.1752915230862718e-05,
+      "loss": 0.8684,
+      "step": 1217
+    },
+    {
+      "epoch": 0.44631733235617443,
+      "grad_norm": 1.891953945159912,
+      "learning_rate": 1.1741538741395028e-05,
+      "loss": 0.8682,
+      "step": 1218
+    },
+    {
+      "epoch": 0.44668376694759987,
+      "grad_norm": 1.9159663915634155,
+      "learning_rate": 1.173015992697303e-05,
+      "loss": 0.8118,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4470502015390253,
+      "grad_norm": 1.684457778930664,
+      "learning_rate": 1.1718778802787443e-05,
+      "loss": 0.8647,
+      "step": 1220
+    },
+    {
+      "epoch": 0.44741663613045074,
+      "grad_norm": 1.9485814571380615,
+      "learning_rate": 1.170739538403207e-05,
+      "loss": 0.8042,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4477830707218761,
+      "grad_norm": 1.6726328134536743,
+      "learning_rate": 1.1696009685903773e-05,
+      "loss": 0.8763,
+      "step": 1222
+    },
+    {
+      "epoch": 0.44814950531330155,
+      "grad_norm": 1.5836412906646729,
+      "learning_rate": 1.168462172360246e-05,
+      "loss": 0.882,
+      "step": 1223
+    },
+    {
+      "epoch": 0.448515939904727,
+      "grad_norm": 1.5830448865890503,
+      "learning_rate": 1.1673231512331065e-05,
+      "loss": 0.8298,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4488823744961524,
+      "grad_norm": 1.7107692956924438,
+      "learning_rate": 1.1661839067295515e-05,
+      "loss": 0.8391,
+      "step": 1225
+    },
+    {
+      "epoch": 0.44924880908757786,
+      "grad_norm": 1.7715849876403809,
+      "learning_rate": 1.1650444403704729e-05,
+      "loss": 0.8519,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4496152436790033,
+      "grad_norm": 2.0211827754974365,
+      "learning_rate": 1.1639047536770581e-05,
+      "loss": 0.7753,
+      "step": 1227
+    },
+    {
+      "epoch": 0.44998167827042873,
+      "grad_norm": 1.7676541805267334,
+      "learning_rate": 1.162764848170789e-05,
+      "loss": 0.8719,
+      "step": 1228
+    },
+    {
+      "epoch": 0.45034811286185417,
+      "grad_norm": 1.7997002601623535,
+      "learning_rate": 1.1616247253734394e-05,
+      "loss": 0.8708,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4507145474532796,
+      "grad_norm": 1.5801912546157837,
+      "learning_rate": 1.1604843868070737e-05,
+      "loss": 0.8994,
+      "step": 1230
+    },
+    {
+      "epoch": 0.45108098204470504,
+      "grad_norm": 1.8483695983886719,
+      "learning_rate": 1.1593438339940433e-05,
+      "loss": 0.8573,
+      "step": 1231
+    },
+    {
+      "epoch": 0.45144741663613047,
+      "grad_norm": 1.8525443077087402,
+      "learning_rate": 1.1582030684569867e-05,
+      "loss": 0.8523,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4518138512275559,
+      "grad_norm": 2.0187089443206787,
+      "learning_rate": 1.1570620917188258e-05,
+      "loss": 0.782,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4521802858189813,
+      "grad_norm": 1.7200089693069458,
+      "learning_rate": 1.1559209053027646e-05,
+      "loss": 0.8386,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4525467204104067,
+      "grad_norm": 1.782767653465271,
+      "learning_rate": 1.1547795107322872e-05,
+      "loss": 0.8477,
+      "step": 1235
+    },
+    {
+      "epoch": 0.45291315500183216,
+      "grad_norm": 1.6392277479171753,
+      "learning_rate": 1.1536379095311552e-05,
+      "loss": 0.8898,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4532795895932576,
+      "grad_norm": 1.6319552659988403,
+      "learning_rate": 1.1524961032234063e-05,
+      "loss": 0.8794,
+      "step": 1237
+    },
+    {
+      "epoch": 0.453646024184683,
+      "grad_norm": 1.7427445650100708,
+      "learning_rate": 1.1513540933333518e-05,
+      "loss": 0.8871,
+      "step": 1238
+    },
+    {
+      "epoch": 0.45401245877610846,
+      "grad_norm": 1.8319940567016602,
+      "learning_rate": 1.1502118813855756e-05,
+      "loss": 0.919,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4543788933675339,
+      "grad_norm": 2.074718713760376,
+      "learning_rate": 1.14906946890493e-05,
+      "loss": 0.8275,
+      "step": 1240
+    },
+    {
+      "epoch": 0.45474532795895933,
+      "grad_norm": 1.9303303956985474,
+      "learning_rate": 1.147926857416536e-05,
+      "loss": 0.8594,
+      "step": 1241
+    },
+    {
+      "epoch": 0.45511176255038477,
+      "grad_norm": 1.7829499244689941,
+      "learning_rate": 1.1467840484457802e-05,
+      "loss": 0.8238,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4554781971418102,
+      "grad_norm": 1.978939175605774,
+      "learning_rate": 1.1456410435183121e-05,
+      "loss": 0.8077,
+      "step": 1243
+    },
+    {
+      "epoch": 0.45584463173323564,
+      "grad_norm": 1.8652037382125854,
+      "learning_rate": 1.1444978441600437e-05,
+      "loss": 0.7969,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4562110663246611,
+      "grad_norm": 1.90079927444458,
+      "learning_rate": 1.1433544518971463e-05,
+      "loss": 0.8939,
+      "step": 1245
+    },
+    {
+      "epoch": 0.45657750091608645,
+      "grad_norm": 1.7367563247680664,
+      "learning_rate": 1.1422108682560484e-05,
+      "loss": 0.9262,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4569439355075119,
+      "grad_norm": 1.567033052444458,
+      "learning_rate": 1.1410670947634344e-05,
+      "loss": 0.8592,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4573103700989373,
+      "grad_norm": 1.6424545049667358,
+      "learning_rate": 1.1399231329462417e-05,
+      "loss": 0.8162,
+      "step": 1248
+    },
+    {
+      "epoch": 0.45767680469036276,
+      "grad_norm": 1.5910917520523071,
+      "learning_rate": 1.1387789843316595e-05,
+      "loss": 0.5761,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4580432392817882,
+      "grad_norm": 1.6359686851501465,
+      "learning_rate": 1.1376346504471265e-05,
+      "loss": 0.8552,
+      "step": 1250
+    },
+    {
+      "epoch": 0.45840967387321363,
+      "grad_norm": 1.6789294481277466,
+      "learning_rate": 1.1364901328203281e-05,
+      "loss": 0.8821,
+      "step": 1251
+    },
+    {
+      "epoch": 0.45877610846463907,
+      "grad_norm": 1.849265694618225,
+      "learning_rate": 1.135345432979196e-05,
+      "loss": 0.8076,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4591425430560645,
+      "grad_norm": 1.6932215690612793,
+      "learning_rate": 1.1342005524519038e-05,
+      "loss": 0.8663,
+      "step": 1253
+    },
+    {
+      "epoch": 0.45950897764748994,
+      "grad_norm": 1.7442898750305176,
+      "learning_rate": 1.1330554927668673e-05,
+      "loss": 0.8856,
+      "step": 1254
+    },
+    {
+      "epoch": 0.45987541223891537,
+      "grad_norm": 1.8422902822494507,
+      "learning_rate": 1.1319102554527417e-05,
+      "loss": 0.8571,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4602418468303408,
+      "grad_norm": 1.964569330215454,
+      "learning_rate": 1.1307648420384181e-05,
+      "loss": 0.8005,
+      "step": 1256
+    },
+    {
+      "epoch": 0.46060828142176624,
+      "grad_norm": 1.689928650856018,
+      "learning_rate": 1.129619254053024e-05,
+      "loss": 0.792,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4609747160131916,
+      "grad_norm": 1.8332197666168213,
+      "learning_rate": 1.1284734930259196e-05,
+      "loss": 0.812,
+      "step": 1258
+    },
+    {
+      "epoch": 0.46134115060461706,
+      "grad_norm": 1.7506823539733887,
+      "learning_rate": 1.1273275604866954e-05,
+      "loss": 0.8223,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4617075851960425,
+      "grad_norm": 2.092235803604126,
+      "learning_rate": 1.1261814579651715e-05,
+      "loss": 0.7873,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4620740197874679,
+      "grad_norm": 1.8410356044769287,
+      "learning_rate": 1.1250351869913949e-05,
+      "loss": 0.8787,
+      "step": 1261
+    },
+    {
+      "epoch": 0.46244045437889336,
+      "grad_norm": 1.8117483854293823,
+      "learning_rate": 1.1238887490956374e-05,
+      "loss": 0.8607,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4628068889703188,
+      "grad_norm": 2.216958522796631,
+      "learning_rate": 1.122742145808394e-05,
+      "loss": 0.7297,
+      "step": 1263
+    },
+    {
+      "epoch": 0.46317332356174423,
+      "grad_norm": 1.816941738128662,
+      "learning_rate": 1.1215953786603798e-05,
+      "loss": 0.8391,
+      "step": 1264
+    },
+    {
+      "epoch": 0.46353975815316967,
+      "grad_norm": 1.7820223569869995,
+      "learning_rate": 1.1204484491825289e-05,
+      "loss": 0.852,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4639061927445951,
+      "grad_norm": 1.9520539045333862,
+      "learning_rate": 1.1193013589059922e-05,
+      "loss": 0.8399,
+      "step": 1266
+    },
+    {
+      "epoch": 0.46427262733602054,
+      "grad_norm": 1.8202781677246094,
+      "learning_rate": 1.1181541093621356e-05,
+      "loss": 0.8408,
+      "step": 1267
+    },
+    {
+      "epoch": 0.464639061927446,
+      "grad_norm": 1.7640472650527954,
+      "learning_rate": 1.117006702082537e-05,
+      "loss": 0.8051,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4650054965188714,
+      "grad_norm": 1.6843233108520508,
+      "learning_rate": 1.1158591385989853e-05,
+      "loss": 0.8673,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4653719311102968,
+      "grad_norm": 1.7685338258743286,
+      "learning_rate": 1.1147114204434782e-05,
+      "loss": 0.8184,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4657383657017222,
+      "grad_norm": 2.0041980743408203,
+      "learning_rate": 1.1135635491482187e-05,
+      "loss": 0.7857,
+      "step": 1271
+    },
+    {
+      "epoch": 0.46610480029314766,
+      "grad_norm": 1.8624567985534668,
+      "learning_rate": 1.1124155262456153e-05,
+      "loss": 0.8569,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4664712348845731,
+      "grad_norm": 1.9397152662277222,
+      "learning_rate": 1.1112673532682792e-05,
+      "loss": 0.9033,
+      "step": 1273
+    },
+    {
+      "epoch": 0.46683766947599853,
+      "grad_norm": 1.7223361730575562,
+      "learning_rate": 1.1101190317490204e-05,
+      "loss": 0.8823,
+      "step": 1274
+    },
+    {
+      "epoch": 0.46720410406742396,
+      "grad_norm": 1.7353929281234741,
+      "learning_rate": 1.1089705632208492e-05,
+      "loss": 0.8342,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4675705386588494,
+      "grad_norm": 2.029233455657959,
+      "learning_rate": 1.1078219492169706e-05,
+      "loss": 0.8752,
+      "step": 1276
+    },
+    {
+      "epoch": 0.46793697325027483,
+      "grad_norm": 1.9964361190795898,
+      "learning_rate": 1.1066731912707842e-05,
+      "loss": 0.8093,
+      "step": 1277
+    },
+    {
+      "epoch": 0.46830340784170027,
+      "grad_norm": 1.8332338333129883,
+      "learning_rate": 1.1055242909158825e-05,
+      "loss": 0.8543,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4686698424331257,
+      "grad_norm": 2.361818790435791,
+      "learning_rate": 1.104375249686047e-05,
+      "loss": 0.8291,
+      "step": 1279
+    },
+    {
+      "epoch": 0.46903627702455114,
+      "grad_norm": 1.7925989627838135,
+      "learning_rate": 1.1032260691152476e-05,
+      "loss": 0.9158,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4694027116159766,
+      "grad_norm": 1.713956356048584,
+      "learning_rate": 1.1020767507376416e-05,
+      "loss": 0.8466,
+      "step": 1281
+    },
+    {
+      "epoch": 0.46976914620740196,
+      "grad_norm": 1.8565016984939575,
+      "learning_rate": 1.1009272960875676e-05,
+      "loss": 0.8416,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4701355807988274,
+      "grad_norm": 1.8900073766708374,
+      "learning_rate": 1.0997777066995485e-05,
+      "loss": 0.9124,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4705020153902528,
+      "grad_norm": 1.6251051425933838,
+      "learning_rate": 1.098627984108286e-05,
+      "loss": 0.8101,
+      "step": 1284
+    },
+    {
+      "epoch": 0.47086844998167826,
+      "grad_norm": 1.70287024974823,
+      "learning_rate": 1.0974781298486596e-05,
+      "loss": 0.8346,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4712348845731037,
+      "grad_norm": 1.5804128646850586,
+      "learning_rate": 1.0963281454557253e-05,
+      "loss": 0.8686,
+      "step": 1286
+    },
+    {
+      "epoch": 0.47160131916452913,
+      "grad_norm": 1.6563670635223389,
+      "learning_rate": 1.0951780324647121e-05,
+      "loss": 0.8963,
+      "step": 1287
+    },
+    {
+      "epoch": 0.47196775375595457,
+      "grad_norm": 1.984560489654541,
+      "learning_rate": 1.0940277924110206e-05,
+      "loss": 0.8542,
+      "step": 1288
+    },
+    {
+      "epoch": 0.47233418834738,
+      "grad_norm": 1.7603366374969482,
+      "learning_rate": 1.0928774268302216e-05,
+      "loss": 0.7671,
+      "step": 1289
+    },
+    {
+      "epoch": 0.47270062293880544,
+      "grad_norm": 1.6885719299316406,
+      "learning_rate": 1.091726937258053e-05,
+      "loss": 0.9038,
+      "step": 1290
+    },
+    {
+      "epoch": 0.47306705753023087,
+      "grad_norm": 1.794811487197876,
+      "learning_rate": 1.090576325230419e-05,
+      "loss": 0.8681,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4734334921216563,
+      "grad_norm": 1.6549447774887085,
+      "learning_rate": 1.0894255922833858e-05,
+      "loss": 0.8645,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4737999267130817,
+      "grad_norm": 1.6765258312225342,
+      "learning_rate": 1.0882747399531828e-05,
+      "loss": 0.8511,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4741663613045071,
+      "grad_norm": 1.7557064294815063,
+      "learning_rate": 1.0871237697761976e-05,
+      "loss": 0.8482,
+      "step": 1294
+    },
+    {
+      "epoch": 0.47453279589593256,
+      "grad_norm": 1.8007601499557495,
+      "learning_rate": 1.0859726832889752e-05,
+      "loss": 0.8592,
+      "step": 1295
+    },
+    {
+      "epoch": 0.474899230487358,
+      "grad_norm": 2.138437271118164,
+      "learning_rate": 1.0848214820282163e-05,
+      "loss": 0.8028,
+      "step": 1296
+    },
+    {
+      "epoch": 0.47526566507878343,
+      "grad_norm": 1.8807333707809448,
+      "learning_rate": 1.0836701675307748e-05,
+      "loss": 0.8282,
+      "step": 1297
+    },
+    {
+      "epoch": 0.47563209967020886,
+      "grad_norm": 1.9707701206207275,
+      "learning_rate": 1.0825187413336556e-05,
+      "loss": 0.8882,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4759985342616343,
+      "grad_norm": 2.054344892501831,
+      "learning_rate": 1.0813672049740126e-05,
+      "loss": 0.8822,
+      "step": 1299
+    },
+    {
+      "epoch": 0.47636496885305973,
+      "grad_norm": 2.028402328491211,
+      "learning_rate": 1.0802155599891466e-05,
+      "loss": 0.8012,
+      "step": 1300
+    },
+    {
+      "epoch": 0.47673140344448517,
+      "grad_norm": 1.7414155006408691,
+      "learning_rate": 1.0790638079165041e-05,
+      "loss": 0.9486,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4770978380359106,
+      "grad_norm": 1.7140246629714966,
+      "learning_rate": 1.0779119502936743e-05,
+      "loss": 0.7964,
+      "step": 1302
+    },
+    {
+      "epoch": 0.47746427262733604,
+      "grad_norm": 1.8553911447525024,
+      "learning_rate": 1.0767599886583866e-05,
+      "loss": 0.7817,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4778307072187615,
+      "grad_norm": 1.6952173709869385,
+      "learning_rate": 1.075607924548511e-05,
+      "loss": 0.7962,
+      "step": 1304
+    },
+    {
+      "epoch": 0.47819714181018685,
+      "grad_norm": 1.79563570022583,
+      "learning_rate": 1.0744557595020518e-05,
+      "loss": 0.8404,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4785635764016123,
+      "grad_norm": 1.712677240371704,
+      "learning_rate": 1.0733034950571497e-05,
+      "loss": 0.8422,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4789300109930377,
+      "grad_norm": 1.8360368013381958,
+      "learning_rate": 1.0721511327520784e-05,
+      "loss": 0.8429,
+      "step": 1307
+    },
+    {
+      "epoch": 0.47929644558446316,
+      "grad_norm": 1.819366216659546,
+      "learning_rate": 1.0709986741252412e-05,
+      "loss": 0.826,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4796628801758886,
+      "grad_norm": 1.797717809677124,
+      "learning_rate": 1.0698461207151707e-05,
+      "loss": 0.8412,
+      "step": 1309
+    },
+    {
+      "epoch": 0.48002931476731403,
+      "grad_norm": 1.9159932136535645,
+      "learning_rate": 1.0686934740605255e-05,
+      "loss": 0.9032,
+      "step": 1310
+    },
+    {
+      "epoch": 0.48039574935873947,
+      "grad_norm": 1.6352150440216064,
+      "learning_rate": 1.067540735700089e-05,
+      "loss": 0.8857,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4807621839501649,
+      "grad_norm": 1.794416904449463,
+      "learning_rate": 1.0663879071727667e-05,
+      "loss": 0.8368,
+      "step": 1312
+    },
+    {
+      "epoch": 0.48112861854159034,
+      "grad_norm": 1.9773752689361572,
+      "learning_rate": 1.065234990017585e-05,
+      "loss": 0.951,
+      "step": 1313
+    },
+    {
+      "epoch": 0.48149505313301577,
+      "grad_norm": 1.5759273767471313,
+      "learning_rate": 1.0640819857736888e-05,
+      "loss": 0.8831,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4818614877244412,
+      "grad_norm": 2.310464859008789,
+      "learning_rate": 1.062928895980338e-05,
+      "loss": 0.7709,
+      "step": 1315
+    },
+    {
+      "epoch": 0.48222792231586664,
+      "grad_norm": 1.967894196510315,
+      "learning_rate": 1.0617757221769085e-05,
+      "loss": 0.8223,
+      "step": 1316
+    },
+    {
+      "epoch": 0.482594356907292,
+      "grad_norm": 1.784820318222046,
+      "learning_rate": 1.0606224659028866e-05,
+      "loss": 0.8454,
+      "step": 1317
+    },
+    {
+      "epoch": 0.48296079149871746,
+      "grad_norm": 2.418149709701538,
+      "learning_rate": 1.0594691286978694e-05,
+      "loss": 0.8274,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4833272260901429,
+      "grad_norm": 2.046618938446045,
+      "learning_rate": 1.0583157121015624e-05,
+      "loss": 0.7996,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4836936606815683,
+      "grad_norm": 1.8235247135162354,
+      "learning_rate": 1.057162217653777e-05,
+      "loss": 0.8943,
+      "step": 1320
+    },
+    {
+      "epoch": 0.48406009527299376,
+      "grad_norm": 1.7685117721557617,
+      "learning_rate": 1.0560086468944285e-05,
+      "loss": 0.8375,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4844265298644192,
+      "grad_norm": 1.8292624950408936,
+      "learning_rate": 1.0548550013635332e-05,
+      "loss": 0.8761,
+      "step": 1322
+    },
+    {
+      "epoch": 0.48479296445584463,
+      "grad_norm": 1.636870265007019,
+      "learning_rate": 1.0537012826012085e-05,
+      "loss": 0.8491,
+      "step": 1323
+    },
+    {
+      "epoch": 0.48515939904727007,
+      "grad_norm": 1.6674764156341553,
+      "learning_rate": 1.0525474921476686e-05,
+      "loss": 0.8977,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4855258336386955,
+      "grad_norm": 1.8313924074172974,
+      "learning_rate": 1.0513936315432245e-05,
+      "loss": 0.8016,
+      "step": 1325
+    },
+    {
+      "epoch": 0.48589226823012094,
+      "grad_norm": 2.065986156463623,
+      "learning_rate": 1.0502397023282792e-05,
+      "loss": 0.8052,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4862587028215464,
+      "grad_norm": 1.9295268058776855,
+      "learning_rate": 1.0490857060433292e-05,
+      "loss": 0.9053,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4866251374129718,
+      "grad_norm": 2.224064588546753,
+      "learning_rate": 1.0479316442289594e-05,
+      "loss": 0.8122,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4869915720043972,
+      "grad_norm": 1.6870437860488892,
+      "learning_rate": 1.0467775184258418e-05,
+      "loss": 0.8663,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4873580065958226,
+      "grad_norm": 1.8607361316680908,
+      "learning_rate": 1.0456233301747348e-05,
+      "loss": 0.8146,
+      "step": 1330
+    },
+    {
+      "epoch": 0.48772444118724806,
+      "grad_norm": 1.9341589212417603,
+      "learning_rate": 1.0444690810164798e-05,
+      "loss": 0.8418,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4880908757786735,
+      "grad_norm": 1.6947089433670044,
+      "learning_rate": 1.0433147724919994e-05,
+      "loss": 0.851,
+      "step": 1332
+    },
+    {
+      "epoch": 0.48845731037009893,
+      "grad_norm": 1.8106732368469238,
+      "learning_rate": 1.0421604061422956e-05,
+      "loss": 0.8499,
+      "step": 1333
+    },
+    {
+      "epoch": 0.48882374496152436,
+      "grad_norm": 2.0022175312042236,
+      "learning_rate": 1.0410059835084473e-05,
+      "loss": 0.8259,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4891901795529498,
+      "grad_norm": 1.6601237058639526,
+      "learning_rate": 1.0398515061316095e-05,
+      "loss": 0.8095,
+      "step": 1335
+    },
+    {
+      "epoch": 0.48955661414437524,
+      "grad_norm": 1.8434045314788818,
+      "learning_rate": 1.0386969755530085e-05,
+      "loss": 0.8839,
+      "step": 1336
+    },
+    {
+      "epoch": 0.48992304873580067,
+      "grad_norm": 2.060389518737793,
+      "learning_rate": 1.0375423933139432e-05,
+      "loss": 0.8123,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4902894833272261,
+      "grad_norm": 1.8199058771133423,
+      "learning_rate": 1.0363877609557812e-05,
+      "loss": 0.8199,
+      "step": 1338
+    },
+    {
+      "epoch": 0.49065591791865154,
+      "grad_norm": 2.1421399116516113,
+      "learning_rate": 1.0352330800199567e-05,
+      "loss": 0.7497,
+      "step": 1339
+    },
+    {
+      "epoch": 0.491022352510077,
+      "grad_norm": 2.0703132152557373,
+      "learning_rate": 1.0340783520479685e-05,
+      "loss": 0.8741,
+      "step": 1340
+    },
+    {
+      "epoch": 0.49138878710150236,
+      "grad_norm": 1.9738215208053589,
+      "learning_rate": 1.0329235785813783e-05,
+      "loss": 0.89,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4917552216929278,
+      "grad_norm": 1.8550047874450684,
+      "learning_rate": 1.031768761161809e-05,
+      "loss": 0.8284,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4921216562843532,
+      "grad_norm": 1.751159429550171,
+      "learning_rate": 1.0306139013309425e-05,
+      "loss": 0.8669,
+      "step": 1343
+    },
+    {
+      "epoch": 0.49248809087577866,
+      "grad_norm": 1.841949224472046,
+      "learning_rate": 1.0294590006305158e-05,
+      "loss": 0.8763,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4928545254672041,
+      "grad_norm": 1.8933924436569214,
+      "learning_rate": 1.0283040606023223e-05,
+      "loss": 0.868,
+      "step": 1345
+    },
+    {
+      "epoch": 0.49322096005862953,
+      "grad_norm": 1.6940875053405762,
+      "learning_rate": 1.0271490827882058e-05,
+      "loss": 0.9048,
+      "step": 1346
+    },
+    {
+      "epoch": 0.49358739465005497,
+      "grad_norm": 1.802107810974121,
+      "learning_rate": 1.0259940687300627e-05,
+      "loss": 0.862,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4939538292414804,
+      "grad_norm": 1.8945834636688232,
+      "learning_rate": 1.0248390199698366e-05,
+      "loss": 0.8206,
+      "step": 1348
+    },
+    {
+      "epoch": 0.49432026383290584,
+      "grad_norm": 1.7853095531463623,
+      "learning_rate": 1.0236839380495174e-05,
+      "loss": 0.8967,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4946866984243313,
+      "grad_norm": 2.1143691539764404,
+      "learning_rate": 1.02252882451114e-05,
+      "loss": 0.839,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4950531330157567,
+      "grad_norm": 2.1969010829925537,
+      "learning_rate": 1.0213736808967809e-05,
+      "loss": 0.7853,
+      "step": 1351
+    },
+    {
+      "epoch": 0.49541956760718214,
+      "grad_norm": 1.987229347229004,
+      "learning_rate": 1.0202185087485563e-05,
+      "loss": 0.8451,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4957860021986075,
+      "grad_norm": 1.8627501726150513,
+      "learning_rate": 1.0190633096086217e-05,
+      "loss": 0.8877,
+      "step": 1353
+    },
+    {
+      "epoch": 0.49615243679003296,
+      "grad_norm": 1.9795594215393066,
+      "learning_rate": 1.0179080850191678e-05,
+      "loss": 0.8398,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4965188713814584,
+      "grad_norm": 2.168057680130005,
+      "learning_rate": 1.0167528365224197e-05,
+      "loss": 0.8705,
+      "step": 1355
+    },
+    {
+      "epoch": 0.49688530597288383,
+      "grad_norm": 2.0126590728759766,
+      "learning_rate": 1.0155975656606338e-05,
+      "loss": 0.8346,
+      "step": 1356
+    },
+    {
+      "epoch": 0.49725174056430926,
+      "grad_norm": 1.975282073020935,
+      "learning_rate": 1.014442273976097e-05,
+      "loss": 0.863,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4976181751557347,
+      "grad_norm": 2.144953489303589,
+      "learning_rate": 1.013286963011124e-05,
+      "loss": 0.9066,
+      "step": 1358
+    },
+    {
+      "epoch": 0.49798460974716013,
+      "grad_norm": 1.99014151096344,
+      "learning_rate": 1.0121316343080542e-05,
+      "loss": 0.822,
+      "step": 1359
+    },
+    {
+      "epoch": 0.49835104433858557,
+      "grad_norm": 1.8074718713760376,
+      "learning_rate": 1.0109762894092521e-05,
+      "loss": 0.856,
+      "step": 1360
+    },
+    {
+      "epoch": 0.498717478930011,
+      "grad_norm": 1.2285923957824707,
+      "learning_rate": 1.0098209298571034e-05,
+      "loss": 0.5514,
+      "step": 1361
+    },
+    {
+      "epoch": 0.49908391352143644,
+      "grad_norm": 2.1667258739471436,
+      "learning_rate": 1.0086655571940127e-05,
+      "loss": 0.827,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4994503481128619,
+      "grad_norm": 1.9957808256149292,
+      "learning_rate": 1.0075101729624025e-05,
+      "loss": 0.7647,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4998167827042873,
+      "grad_norm": 1.786232352256775,
+      "learning_rate": 1.0063547787047114e-05,
+      "loss": 0.9069,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5001832172957127,
+      "grad_norm": 1.7317724227905273,
+      "learning_rate": 1.0051993759633903e-05,
+      "loss": 0.8322,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5005496518871382,
+      "grad_norm": 2.1333634853363037,
+      "learning_rate": 1.004043966280902e-05,
+      "loss": 0.8836,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5005496518871382,
+      "eval_loss": 0.8037351369857788,
+      "eval_runtime": 799.1015,
+      "eval_samples_per_second": 3.394,
+      "eval_steps_per_second": 0.424,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5009160864785636,
+      "grad_norm": 1.9786326885223389,
+      "learning_rate": 1.0028885511997187e-05,
+      "loss": 0.8427,
+      "step": 1367
+    },
+    {
+      "epoch": 0.501282521069989,
+      "grad_norm": 1.7938628196716309,
+      "learning_rate": 1.0017331322623189e-05,
+      "loss": 0.8882,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5016489556614144,
+      "grad_norm": 1.6222481727600098,
+      "learning_rate": 1.0005777110111878e-05,
+      "loss": 0.9011,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5020153902528399,
+      "grad_norm": 1.868241786956787,
+      "learning_rate": 9.994222889888124e-06,
+      "loss": 0.8445,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5023818248442653,
+      "grad_norm": 2.0541722774505615,
+      "learning_rate": 9.982668677376811e-06,
+      "loss": 0.8564,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5027482594356907,
+      "grad_norm": 1.7385966777801514,
+      "learning_rate": 9.971114488002817e-06,
+      "loss": 0.8452,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5031146940271162,
+      "grad_norm": 1.80974280834198,
+      "learning_rate": 9.959560337190983e-06,
+      "loss": 0.8711,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5034811286185416,
+      "grad_norm": 1.864256501197815,
+      "learning_rate": 9.9480062403661e-06,
+      "loss": 0.8603,
+      "step": 1374
+    },
+    {
+      "epoch": 0.503847563209967,
+      "grad_norm": 1.8246670961380005,
+      "learning_rate": 9.936452212952888e-06,
+      "loss": 0.8189,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5042139978013924,
+      "grad_norm": 1.888363003730774,
+      "learning_rate": 9.924898270375975e-06,
+      "loss": 0.8572,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5045804323928179,
+      "grad_norm": 2.028559446334839,
+      "learning_rate": 9.913344428059876e-06,
+      "loss": 0.807,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5049468669842433,
+      "grad_norm": 1.936490535736084,
+      "learning_rate": 9.90179070142897e-06,
+      "loss": 0.8288,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5053133015756688,
+      "grad_norm": 1.8482036590576172,
+      "learning_rate": 9.890237105907482e-06,
+      "loss": 0.8411,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5056797361670942,
+      "grad_norm": 2.113351345062256,
+      "learning_rate": 9.878683656919461e-06,
+      "loss": 0.8251,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5060461707585197,
+      "grad_norm": 2.063748836517334,
+      "learning_rate": 9.867130369888764e-06,
+      "loss": 0.8169,
+      "step": 1381
+    },
+    {
+      "epoch": 0.506412605349945,
+      "grad_norm": 1.8193755149841309,
+      "learning_rate": 9.855577260239031e-06,
+      "loss": 0.9078,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5067790399413704,
+      "grad_norm": 1.823914647102356,
+      "learning_rate": 9.844024343393665e-06,
+      "loss": 0.7845,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5071454745327959,
+      "grad_norm": 1.9707849025726318,
+      "learning_rate": 9.832471634775806e-06,
+      "loss": 0.7809,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5075119091242213,
+      "grad_norm": 1.912359595298767,
+      "learning_rate": 9.820919149808324e-06,
+      "loss": 0.8774,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5078783437156468,
+      "grad_norm": 1.7348723411560059,
+      "learning_rate": 9.809366903913785e-06,
+      "loss": 0.8599,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5082447783070722,
+      "grad_norm": 1.7531144618988037,
+      "learning_rate": 9.797814912514436e-06,
+      "loss": 0.8436,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5086112128984976,
+      "grad_norm": 1.9358574151992798,
+      "learning_rate": 9.786263191032196e-06,
+      "loss": 0.8119,
+      "step": 1388
+    },
+    {
+      "epoch": 0.508977647489923,
+      "grad_norm": 2.0156936645507812,
+      "learning_rate": 9.774711754888602e-06,
+      "loss": 0.8819,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5093440820813485,
+      "grad_norm": 1.9308418035507202,
+      "learning_rate": 9.763160619504828e-06,
+      "loss": 0.9012,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5097105166727739,
+      "grad_norm": 1.8228894472122192,
+      "learning_rate": 9.751609800301638e-06,
+      "loss": 0.8609,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5100769512641994,
+      "grad_norm": 1.7600071430206299,
+      "learning_rate": 9.740059312699375e-06,
+      "loss": 0.9132,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5104433858556248,
+      "grad_norm": 1.6532927751541138,
+      "learning_rate": 9.728509172117947e-06,
+      "loss": 0.8989,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5108098204470503,
+      "grad_norm": 1.9439219236373901,
+      "learning_rate": 9.716959393976784e-06,
+      "loss": 0.8466,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5111762550384756,
+      "grad_norm": 1.8679782152175903,
+      "learning_rate": 9.705409993694844e-06,
+      "loss": 0.8169,
+      "step": 1395
+    },
+    {
+      "epoch": 0.511542689629901,
+      "grad_norm": 2.1819117069244385,
+      "learning_rate": 9.693860986690579e-06,
+      "loss": 0.7761,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5119091242213265,
+      "grad_norm": 2.5734708309173584,
+      "learning_rate": 9.68231238838191e-06,
+      "loss": 0.7716,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5122755588127519,
+      "grad_norm": 1.8313605785369873,
+      "learning_rate": 9.670764214186222e-06,
+      "loss": 0.8782,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5126419934041774,
+      "grad_norm": 2.187032461166382,
+      "learning_rate": 9.659216479520321e-06,
+      "loss": 0.7673,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5130084279956028,
+      "grad_norm": 2.0457074642181396,
+      "learning_rate": 9.647669199800437e-06,
+      "loss": 0.8126,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5133748625870282,
+      "grad_norm": 1.690920114517212,
+      "learning_rate": 9.636122390442191e-06,
+      "loss": 0.8425,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5137412971784536,
+      "grad_norm": 1.909960150718689,
+      "learning_rate": 9.62457606686057e-06,
+      "loss": 0.875,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5141077317698791,
+      "grad_norm": 2.1119649410247803,
+      "learning_rate": 9.613030244469917e-06,
+      "loss": 0.8522,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5144741663613045,
+      "grad_norm": 2.170766830444336,
+      "learning_rate": 9.601484938683912e-06,
+      "loss": 0.7812,
+      "step": 1404
+    },
+    {
+      "epoch": 0.51484060095273,
+      "grad_norm": 1.9513856172561646,
+      "learning_rate": 9.589940164915528e-06,
+      "loss": 0.8697,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5152070355441554,
+      "grad_norm": 1.9524122476577759,
+      "learning_rate": 9.578395938577047e-06,
+      "loss": 0.8508,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5155734701355807,
+      "grad_norm": 2.066345691680908,
+      "learning_rate": 9.566852275080008e-06,
+      "loss": 0.8639,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5159399047270062,
+      "grad_norm": 1.8953510522842407,
+      "learning_rate": 9.555309189835204e-06,
+      "loss": 0.8548,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5163063393184316,
+      "grad_norm": 1.7897011041641235,
+      "learning_rate": 9.543766698252657e-06,
+      "loss": 0.8487,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5166727739098571,
+      "grad_norm": 1.8420007228851318,
+      "learning_rate": 9.532224815741586e-06,
+      "loss": 0.8521,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5170392085012825,
+      "grad_norm": 2.073654890060425,
+      "learning_rate": 9.52068355771041e-06,
+      "loss": 0.7481,
+      "step": 1411
+    },
+    {
+      "epoch": 0.517405643092708,
+      "grad_norm": 1.80082106590271,
+      "learning_rate": 9.50914293956671e-06,
+      "loss": 0.8771,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5177720776841334,
+      "grad_norm": 1.9965510368347168,
+      "learning_rate": 9.497602976717207e-06,
+      "loss": 0.8244,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5181385122755588,
+      "grad_norm": 1.875783920288086,
+      "learning_rate": 9.486063684567762e-06,
+      "loss": 0.811,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5185049468669842,
+      "grad_norm": 1.7989037036895752,
+      "learning_rate": 9.474525078523317e-06,
+      "loss": 0.8373,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5188713814584097,
+      "grad_norm": 1.6763300895690918,
+      "learning_rate": 9.462987173987918e-06,
+      "loss": 0.8654,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5192378160498351,
+      "grad_norm": 1.7529571056365967,
+      "learning_rate": 9.451449986364671e-06,
+      "loss": 0.8432,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5196042506412606,
+      "grad_norm": 2.070521116256714,
+      "learning_rate": 9.439913531055719e-06,
+      "loss": 0.8319,
+      "step": 1418
+    },
+    {
+      "epoch": 0.519970685232686,
+      "grad_norm": 2.0340991020202637,
+      "learning_rate": 9.428377823462233e-06,
+      "loss": 0.8814,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5203371198241113,
+      "grad_norm": 1.8266870975494385,
+      "learning_rate": 9.41684287898438e-06,
+      "loss": 0.8147,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5207035544155368,
+      "grad_norm": 1.816367506980896,
+      "learning_rate": 9.40530871302131e-06,
+      "loss": 0.8683,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5210699890069622,
+      "grad_norm": 1.8408104181289673,
+      "learning_rate": 9.393775340971139e-06,
+      "loss": 0.7732,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5214364235983877,
+      "grad_norm": 1.931918978691101,
+      "learning_rate": 9.382242778230917e-06,
+      "loss": 0.8475,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5218028581898131,
+      "grad_norm": 2.0313661098480225,
+      "learning_rate": 9.370711040196621e-06,
+      "loss": 0.8688,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5221692927812386,
+      "grad_norm": 1.982347846031189,
+      "learning_rate": 9.359180142263116e-06,
+      "loss": 0.8364,
+      "step": 1425
+    },
+    {
+      "epoch": 0.522535727372664,
+      "grad_norm": 1.9787847995758057,
+      "learning_rate": 9.347650099824151e-06,
+      "loss": 0.9508,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5229021619640895,
+      "grad_norm": 1.7552639245986938,
+      "learning_rate": 9.336120928272335e-06,
+      "loss": 0.842,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5232685965555148,
+      "grad_norm": 1.8307520151138306,
+      "learning_rate": 9.324592642999112e-06,
+      "loss": 0.8447,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5236350311469403,
+      "grad_norm": 2.5280072689056396,
+      "learning_rate": 9.31306525939475e-06,
+      "loss": 0.7423,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5240014657383657,
+      "grad_norm": 2.190800189971924,
+      "learning_rate": 9.301538792848297e-06,
+      "loss": 0.8356,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5243679003297911,
+      "grad_norm": 1.8437484502792358,
+      "learning_rate": 9.290013258747591e-06,
+      "loss": 0.8997,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5247343349212166,
+      "grad_norm": 1.7690744400024414,
+      "learning_rate": 9.278488672479218e-06,
+      "loss": 0.8618,
+      "step": 1432
+    },
+    {
+      "epoch": 0.525100769512642,
+      "grad_norm": 2.0354177951812744,
+      "learning_rate": 9.266965049428503e-06,
+      "loss": 0.8644,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5254672041040674,
+      "grad_norm": 1.8663829565048218,
+      "learning_rate": 9.255442404979484e-06,
+      "loss": 0.8965,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5258336386954928,
+      "grad_norm": 1.9830769300460815,
+      "learning_rate": 9.243920754514895e-06,
+      "loss": 0.8205,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5262000732869183,
+      "grad_norm": 1.9793612957000732,
+      "learning_rate": 9.232400113416136e-06,
+      "loss": 0.776,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5265665078783437,
+      "grad_norm": 1.7279483079910278,
+      "learning_rate": 9.22088049706326e-06,
+      "loss": 0.8636,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5269329424697692,
+      "grad_norm": 1.8364406824111938,
+      "learning_rate": 9.20936192083496e-06,
+      "loss": 0.8017,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5272993770611946,
+      "grad_norm": 1.7913066148757935,
+      "learning_rate": 9.197844400108536e-06,
+      "loss": 0.8637,
+      "step": 1439
+    },
+    {
+      "epoch": 0.52766581165262,
+      "grad_norm": 1.9976123571395874,
+      "learning_rate": 9.18632795025988e-06,
+      "loss": 0.9093,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5280322462440454,
+      "grad_norm": 1.79649817943573,
+      "learning_rate": 9.174812586663447e-06,
+      "loss": 0.8709,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5283986808354708,
+      "grad_norm": 1.9000849723815918,
+      "learning_rate": 9.163298324692254e-06,
+      "loss": 0.8728,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5287651154268963,
+      "grad_norm": 2.158200740814209,
+      "learning_rate": 9.151785179717838e-06,
+      "loss": 0.9365,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5291315500183217,
+      "grad_norm": 2.1834192276000977,
+      "learning_rate": 9.14027316711025e-06,
+      "loss": 0.8341,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5294979846097472,
+      "grad_norm": 2.1154308319091797,
+      "learning_rate": 9.12876230223803e-06,
+      "loss": 0.8308,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5298644192011726,
+      "grad_norm": 2.185143232345581,
+      "learning_rate": 9.117252600468175e-06,
+      "loss": 0.8251,
+      "step": 1446
+    },
+    {
+      "epoch": 0.530230853792598,
+      "grad_norm": 2.063204526901245,
+      "learning_rate": 9.105744077166145e-06,
+      "loss": 0.8575,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5305972883840234,
+      "grad_norm": 1.9822335243225098,
+      "learning_rate": 9.094236747695814e-06,
+      "loss": 0.8331,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5309637229754489,
+      "grad_norm": 2.593322992324829,
+      "learning_rate": 9.08273062741947e-06,
+      "loss": 0.7768,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5313301575668743,
+      "grad_norm": 2.102851390838623,
+      "learning_rate": 9.071225731697789e-06,
+      "loss": 0.7953,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5316965921582998,
+      "grad_norm": 1.9108489751815796,
+      "learning_rate": 9.059722075889798e-06,
+      "loss": 0.7976,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5320630267497252,
+      "grad_norm": 2.1853110790252686,
+      "learning_rate": 9.048219675352882e-06,
+      "loss": 0.8989,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5324294613411507,
+      "grad_norm": 2.1810035705566406,
+      "learning_rate": 9.036718545442748e-06,
+      "loss": 0.7938,
+      "step": 1453
+    },
+    {
+      "epoch": 0.532795895932576,
+      "grad_norm": 1.8937190771102905,
+      "learning_rate": 9.025218701513404e-06,
+      "loss": 0.83,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5331623305240014,
+      "grad_norm": 1.77532160282135,
+      "learning_rate": 9.013720158917145e-06,
+      "loss": 0.8331,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5335287651154269,
+      "grad_norm": 2.0107715129852295,
+      "learning_rate": 9.002222933004518e-06,
+      "loss": 0.7664,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5338951997068523,
+      "grad_norm": 1.7541794776916504,
+      "learning_rate": 8.990727039124327e-06,
+      "loss": 0.8028,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5342616342982778,
+      "grad_norm": 1.9145309925079346,
+      "learning_rate": 8.979232492623587e-06,
+      "loss": 0.8315,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5346280688897032,
+      "grad_norm": 1.8287135362625122,
+      "learning_rate": 8.967739308847524e-06,
+      "loss": 0.8707,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5349945034811286,
+      "grad_norm": 1.9218909740447998,
+      "learning_rate": 8.956247503139537e-06,
+      "loss": 0.8504,
+      "step": 1460
+    },
+    {
+      "epoch": 0.535360938072554,
+      "grad_norm": 2.1773972511291504,
+      "learning_rate": 8.94475709084118e-06,
+      "loss": 0.8332,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5357273726639795,
+      "grad_norm": 2.0021514892578125,
+      "learning_rate": 8.93326808729216e-06,
+      "loss": 0.8127,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5360938072554049,
+      "grad_norm": 1.9204373359680176,
+      "learning_rate": 8.921780507830296e-06,
+      "loss": 0.8519,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5364602418468304,
+      "grad_norm": 2.199204206466675,
+      "learning_rate": 8.91029436779151e-06,
+      "loss": 0.7477,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5368266764382558,
+      "grad_norm": 1.8903037309646606,
+      "learning_rate": 8.898809682509796e-06,
+      "loss": 0.8401,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5371931110296811,
+      "grad_norm": 2.7648026943206787,
+      "learning_rate": 8.887326467317215e-06,
+      "loss": 0.8437,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5375595456211066,
+      "grad_norm": 2.081468105316162,
+      "learning_rate": 8.87584473754385e-06,
+      "loss": 0.8562,
+      "step": 1467
+    },
+    {
+      "epoch": 0.537925980212532,
+      "grad_norm": 2.111710786819458,
+      "learning_rate": 8.864364508517816e-06,
+      "loss": 0.8454,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5382924148039575,
+      "grad_norm": 2.428865432739258,
+      "learning_rate": 8.852885795565221e-06,
+      "loss": 0.752,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5386588493953829,
+      "grad_norm": 2.243191719055176,
+      "learning_rate": 8.841408614010148e-06,
+      "loss": 0.8432,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5390252839868084,
+      "grad_norm": 1.8198479413986206,
+      "learning_rate": 8.829932979174635e-06,
+      "loss": 0.8663,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5393917185782338,
+      "grad_norm": 1.930250883102417,
+      "learning_rate": 8.81845890637865e-06,
+      "loss": 0.8288,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5397581531696592,
+      "grad_norm": 1.8728525638580322,
+      "learning_rate": 8.80698641094008e-06,
+      "loss": 0.8411,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5401245877610846,
+      "grad_norm": 2.136594533920288,
+      "learning_rate": 8.795515508174713e-06,
+      "loss": 0.8127,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5404910223525101,
+      "grad_norm": 1.3855836391448975,
+      "learning_rate": 8.784046213396206e-06,
+      "loss": 0.5478,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5408574569439355,
+      "grad_norm": 1.8245359659194946,
+      "learning_rate": 8.772578541916063e-06,
+      "loss": 0.8099,
+      "step": 1476
+    },
+    {
+      "epoch": 0.541223891535361,
+      "grad_norm": 1.8170462846755981,
+      "learning_rate": 8.76111250904363e-06,
+      "loss": 0.8622,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5415903261267864,
+      "grad_norm": 1.8183109760284424,
+      "learning_rate": 8.749648130086054e-06,
+      "loss": 0.8705,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5419567607182117,
+      "grad_norm": 1.731231689453125,
+      "learning_rate": 8.738185420348287e-06,
+      "loss": 0.893,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5423231953096372,
+      "grad_norm": 1.8436676263809204,
+      "learning_rate": 8.726724395133048e-06,
+      "loss": 0.8448,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5426896299010626,
+      "grad_norm": 2.3991260528564453,
+      "learning_rate": 8.715265069740809e-06,
+      "loss": 0.7758,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5430560644924881,
+      "grad_norm": 1.7913585901260376,
+      "learning_rate": 8.703807459469763e-06,
+      "loss": 0.821,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5434224990839135,
+      "grad_norm": 1.7945986986160278,
+      "learning_rate": 8.692351579615822e-06,
+      "loss": 0.8389,
+      "step": 1483
+    },
+    {
+      "epoch": 0.543788933675339,
+      "grad_norm": 1.5730236768722534,
+      "learning_rate": 8.680897445472588e-06,
+      "loss": 0.5625,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5441553682667644,
+      "grad_norm": 1.9446221590042114,
+      "learning_rate": 8.669445072331327e-06,
+      "loss": 0.8495,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5445218028581899,
+      "grad_norm": 1.8746920824050903,
+      "learning_rate": 8.657994475480967e-06,
+      "loss": 0.8506,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5448882374496152,
+      "grad_norm": 2.2092230319976807,
+      "learning_rate": 8.646545670208044e-06,
+      "loss": 0.7538,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5452546720410407,
+      "grad_norm": 1.8926215171813965,
+      "learning_rate": 8.63509867179672e-06,
+      "loss": 0.8054,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5456211066324661,
+      "grad_norm": 1.9946545362472534,
+      "learning_rate": 8.623653495528738e-06,
+      "loss": 0.8634,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5459875412238915,
+      "grad_norm": 2.426816463470459,
+      "learning_rate": 8.612210156683407e-06,
+      "loss": 0.7489,
+      "step": 1490
+    },
+    {
+      "epoch": 0.546353975815317,
+      "grad_norm": 1.9078896045684814,
+      "learning_rate": 8.600768670537588e-06,
+      "loss": 0.835,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5467204104067424,
+      "grad_norm": 2.1838278770446777,
+      "learning_rate": 8.58932905236566e-06,
+      "loss": 0.8453,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5470868449981678,
+      "grad_norm": 2.0195322036743164,
+      "learning_rate": 8.577891317439519e-06,
+      "loss": 0.8305,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5474532795895932,
+      "grad_norm": 2.025602102279663,
+      "learning_rate": 8.56645548102854e-06,
+      "loss": 0.9082,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5478197141810187,
+      "grad_norm": 2.021904706954956,
+      "learning_rate": 8.555021558399565e-06,
+      "loss": 0.8613,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5481861487724441,
+      "grad_norm": 2.3774404525756836,
+      "learning_rate": 8.54358956481688e-06,
+      "loss": 0.7845,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5485525833638696,
+      "grad_norm": 1.6819612979888916,
+      "learning_rate": 8.532159515542204e-06,
+      "loss": 0.8947,
+      "step": 1497
+    },
+    {
+      "epoch": 0.548919017955295,
+      "grad_norm": 2.2661547660827637,
+      "learning_rate": 8.520731425834643e-06,
+      "loss": 0.7927,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5492854525467205,
+      "grad_norm": 1.8248411417007446,
+      "learning_rate": 8.509305310950705e-06,
+      "loss": 0.8244,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5496518871381458,
+      "grad_norm": 2.1215291023254395,
+      "learning_rate": 8.497881186144249e-06,
+      "loss": 0.8209,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5500183217295713,
+      "grad_norm": 2.1072847843170166,
+      "learning_rate": 8.48645906666648e-06,
+      "loss": 0.8291,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5503847563209967,
+      "grad_norm": 1.918225884437561,
+      "learning_rate": 8.475038967765942e-06,
+      "loss": 0.8143,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5507511909124221,
+      "grad_norm": 2.096271276473999,
+      "learning_rate": 8.463620904688452e-06,
+      "loss": 0.8405,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5511176255038476,
+      "grad_norm": 1.6998379230499268,
+      "learning_rate": 8.45220489267713e-06,
+      "loss": 0.8819,
+      "step": 1504
+    },
+    {
+      "epoch": 0.551484060095273,
+      "grad_norm": 2.25860333442688,
+      "learning_rate": 8.440790946972357e-06,
+      "loss": 0.8072,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5518504946866984,
+      "grad_norm": 2.0350191593170166,
+      "learning_rate": 8.429379082811743e-06,
+      "loss": 0.819,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5522169292781238,
+      "grad_norm": 1.8747131824493408,
+      "learning_rate": 8.417969315430138e-06,
+      "loss": 0.8611,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5525833638695493,
+      "grad_norm": 2.0432024002075195,
+      "learning_rate": 8.40656166005957e-06,
+      "loss": 0.8181,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5529497984609747,
+      "grad_norm": 1.9487144947052002,
+      "learning_rate": 8.395156131929266e-06,
+      "loss": 0.8138,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5533162330524002,
+      "grad_norm": 1.801957130432129,
+      "learning_rate": 8.383752746265608e-06,
+      "loss": 0.8819,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5536826676438256,
+      "grad_norm": 1.736411213874817,
+      "learning_rate": 8.372351518292112e-06,
+      "loss": 0.85,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5540491022352511,
+      "grad_norm": 2.142106056213379,
+      "learning_rate": 8.360952463229424e-06,
+      "loss": 0.8645,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5544155368266764,
+      "grad_norm": 2.392352342605591,
+      "learning_rate": 8.349555596295276e-06,
+      "loss": 0.7336,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5547819714181018,
+      "grad_norm": 1.8606106042861938,
+      "learning_rate": 8.338160932704487e-06,
+      "loss": 0.9252,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5551484060095273,
+      "grad_norm": 2.3040175437927246,
+      "learning_rate": 8.326768487668937e-06,
+      "loss": 0.7391,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5555148406009527,
+      "grad_norm": 2.4299490451812744,
+      "learning_rate": 8.315378276397542e-06,
+      "loss": 0.7129,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5558812751923782,
+      "grad_norm": 1.8836729526519775,
+      "learning_rate": 8.303990314096234e-06,
+      "loss": 0.8198,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5562477097838036,
+      "grad_norm": 1.9984331130981445,
+      "learning_rate": 8.292604615967935e-06,
+      "loss": 0.8042,
+      "step": 1518
+    },
+    {
+      "epoch": 0.556614144375229,
+      "grad_norm": 2.047516345977783,
+      "learning_rate": 8.28122119721256e-06,
+      "loss": 0.8574,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5569805789666544,
+      "grad_norm": 2.2532103061676025,
+      "learning_rate": 8.269840073026972e-06,
+      "loss": 0.857,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5573470135580799,
+      "grad_norm": 2.170428991317749,
+      "learning_rate": 8.258461258604975e-06,
+      "loss": 0.8122,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5577134481495053,
+      "grad_norm": 2.1111834049224854,
+      "learning_rate": 8.247084769137284e-06,
+      "loss": 0.7591,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5580798827409308,
+      "grad_norm": 1.7928640842437744,
+      "learning_rate": 8.235710619811515e-06,
+      "loss": 0.879,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5584463173323562,
+      "grad_norm": 1.917716383934021,
+      "learning_rate": 8.224338825812152e-06,
+      "loss": 0.8858,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5588127519237817,
+      "grad_norm": 2.1861300468444824,
+      "learning_rate": 8.212969402320546e-06,
+      "loss": 0.863,
+      "step": 1525
+    },
+    {
+      "epoch": 0.559179186515207,
+      "grad_norm": 1.7714310884475708,
+      "learning_rate": 8.201602364514878e-06,
+      "loss": 0.8418,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5595456211066324,
+      "grad_norm": 1.9631260633468628,
+      "learning_rate": 8.190237727570142e-06,
+      "loss": 0.8734,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5599120556980579,
+      "grad_norm": 2.40706205368042,
+      "learning_rate": 8.178875506658133e-06,
+      "loss": 0.728,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5602784902894833,
+      "grad_norm": 1.7918330430984497,
+      "learning_rate": 8.167515716947411e-06,
+      "loss": 0.8601,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5606449248809088,
+      "grad_norm": 1.9069151878356934,
+      "learning_rate": 8.156158373603297e-06,
+      "loss": 0.8387,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5610113594723342,
+      "grad_norm": 1.8654823303222656,
+      "learning_rate": 8.144803491787848e-06,
+      "loss": 0.884,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5613777940637596,
+      "grad_norm": 2.0031216144561768,
+      "learning_rate": 8.13345108665983e-06,
+      "loss": 0.8334,
+      "step": 1532
+    },
+    {
+      "epoch": 0.561744228655185,
+      "grad_norm": 2.1736366748809814,
+      "learning_rate": 8.122101173374704e-06,
+      "loss": 0.8301,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5621106632466105,
+      "grad_norm": 2.0205841064453125,
+      "learning_rate": 8.110753767084604e-06,
+      "loss": 0.8304,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5624770978380359,
+      "grad_norm": 1.993735432624817,
+      "learning_rate": 8.099408882938319e-06,
+      "loss": 0.8576,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5628435324294614,
+      "grad_norm": 2.1005938053131104,
+      "learning_rate": 8.088066536081269e-06,
+      "loss": 0.809,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5632099670208868,
+      "grad_norm": 1.8967087268829346,
+      "learning_rate": 8.076726741655489e-06,
+      "loss": 0.8422,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5635764016123122,
+      "grad_norm": 1.860841989517212,
+      "learning_rate": 8.065389514799609e-06,
+      "loss": 0.8539,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5639428362037376,
+      "grad_norm": 2.1330482959747314,
+      "learning_rate": 8.054054870648818e-06,
+      "loss": 0.843,
+      "step": 1539
+    },
+    {
+      "epoch": 0.564309270795163,
+      "grad_norm": 2.035189390182495,
+      "learning_rate": 8.042722824334874e-06,
+      "loss": 0.8647,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5646757053865885,
+      "grad_norm": 1.9112155437469482,
+      "learning_rate": 8.031393390986055e-06,
+      "loss": 0.7743,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5650421399780139,
+      "grad_norm": 2.167698383331299,
+      "learning_rate": 8.020066585727156e-06,
+      "loss": 0.8544,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5654085745694394,
+      "grad_norm": 2.4354660511016846,
+      "learning_rate": 8.00874242367947e-06,
+      "loss": 0.8111,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5657750091608648,
+      "grad_norm": 2.0466530323028564,
+      "learning_rate": 7.997420919960741e-06,
+      "loss": 0.8742,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5661414437522903,
+      "grad_norm": 2.8126063346862793,
+      "learning_rate": 7.986102089685189e-06,
+      "loss": 0.7313,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5665078783437156,
+      "grad_norm": 2.1306872367858887,
+      "learning_rate": 7.974785947963441e-06,
+      "loss": 0.8469,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5668743129351411,
+      "grad_norm": 1.9655848741531372,
+      "learning_rate": 7.963472509902556e-06,
+      "loss": 0.862,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5672407475265665,
+      "grad_norm": 1.8828647136688232,
+      "learning_rate": 7.952161790605975e-06,
+      "loss": 0.8262,
+      "step": 1548
+    },
+    {
+      "epoch": 0.567607182117992,
+      "grad_norm": 1.8155533075332642,
+      "learning_rate": 7.940853805173503e-06,
+      "loss": 0.8533,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5679736167094174,
+      "grad_norm": 2.217395782470703,
+      "learning_rate": 7.929548568701303e-06,
+      "loss": 0.7375,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5683400513008428,
+      "grad_norm": 2.2814624309539795,
+      "learning_rate": 7.918246096281868e-06,
+      "loss": 0.7539,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5687064858922682,
+      "grad_norm": 1.9111372232437134,
+      "learning_rate": 7.906946403003995e-06,
+      "loss": 0.8306,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5690729204836936,
+      "grad_norm": 1.7845666408538818,
+      "learning_rate": 7.89564950395278e-06,
+      "loss": 0.7968,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5694393550751191,
+      "grad_norm": 2.1907219886779785,
+      "learning_rate": 7.884355414209586e-06,
+      "loss": 0.8312,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5698057896665445,
+      "grad_norm": 2.0254149436950684,
+      "learning_rate": 7.873064148852014e-06,
+      "loss": 0.857,
+      "step": 1555
+    },
+    {
+      "epoch": 0.57017222425797,
+      "grad_norm": 2.041749954223633,
+      "learning_rate": 7.861775722953913e-06,
+      "loss": 0.9304,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5705386588493954,
+      "grad_norm": 2.1609628200531006,
+      "learning_rate": 7.850490151585326e-06,
+      "loss": 0.8433,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5709050934408209,
+      "grad_norm": 2.2469706535339355,
+      "learning_rate": 7.839207449812493e-06,
+      "loss": 0.7868,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5712715280322462,
+      "grad_norm": 1.949552059173584,
+      "learning_rate": 7.827927632697827e-06,
+      "loss": 0.8477,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5716379626236717,
+      "grad_norm": 1.771660566329956,
+      "learning_rate": 7.816650715299873e-06,
+      "loss": 0.8199,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5720043972150971,
+      "grad_norm": 1.9107954502105713,
+      "learning_rate": 7.805376712673324e-06,
+      "loss": 0.893,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5723708318065225,
+      "grad_norm": 2.2006566524505615,
+      "learning_rate": 7.794105639868976e-06,
+      "loss": 0.8021,
+      "step": 1562
+    },
+    {
+      "epoch": 0.572737266397948,
+      "grad_norm": 1.9410490989685059,
+      "learning_rate": 7.782837511933707e-06,
+      "loss": 0.8815,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5731037009893734,
+      "grad_norm": 1.9208725690841675,
+      "learning_rate": 7.771572343910471e-06,
+      "loss": 0.8115,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5734701355807988,
+      "grad_norm": 2.0697717666625977,
+      "learning_rate": 7.760310150838267e-06,
+      "loss": 0.8229,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5738365701722242,
+      "grad_norm": 2.2040605545043945,
+      "learning_rate": 7.749050947752119e-06,
+      "loss": 0.8481,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5742030047636497,
+      "grad_norm": 2.007427215576172,
+      "learning_rate": 7.737794749683074e-06,
+      "loss": 0.8856,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5745694393550751,
+      "grad_norm": 2.1302149295806885,
+      "learning_rate": 7.726541571658145e-06,
+      "loss": 0.8318,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5749358739465006,
+      "grad_norm": 1.92087721824646,
+      "learning_rate": 7.715291428700336e-06,
+      "loss": 0.8355,
+      "step": 1569
+    },
+    {
+      "epoch": 0.575302308537926,
+      "grad_norm": 2.1725895404815674,
+      "learning_rate": 7.704044335828582e-06,
+      "loss": 0.8164,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5756687431293515,
+      "grad_norm": 1.9096589088439941,
+      "learning_rate": 7.69280030805775e-06,
+      "loss": 0.8481,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5760351777207768,
+      "grad_norm": 2.3047072887420654,
+      "learning_rate": 7.681559360398623e-06,
+      "loss": 0.7868,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5764016123122023,
+      "grad_norm": 2.029869556427002,
+      "learning_rate": 7.67032150785787e-06,
+      "loss": 0.8592,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5767680469036277,
+      "grad_norm": 2.2390494346618652,
+      "learning_rate": 7.659086765438017e-06,
+      "loss": 0.7462,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5771344814950531,
+      "grad_norm": 2.311746120452881,
+      "learning_rate": 7.64785514813745e-06,
+      "loss": 0.8339,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5775009160864786,
+      "grad_norm": 2.535600185394287,
+      "learning_rate": 7.636626670950375e-06,
+      "loss": 0.7523,
+      "step": 1576
+    },
+    {
+      "epoch": 0.577867350677904,
+      "grad_norm": 2.6063930988311768,
+      "learning_rate": 7.625401348866812e-06,
+      "loss": 0.8121,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5782337852693294,
+      "grad_norm": 2.046186685562134,
+      "learning_rate": 7.614179196872566e-06,
+      "loss": 0.8722,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5786002198607548,
+      "grad_norm": 1.9237748384475708,
+      "learning_rate": 7.602960229949216e-06,
+      "loss": 0.8812,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5789666544521803,
+      "grad_norm": 2.2206146717071533,
+      "learning_rate": 7.591744463074075e-06,
+      "loss": 0.7857,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5793330890436057,
+      "grad_norm": 2.3066954612731934,
+      "learning_rate": 7.580531911220195e-06,
+      "loss": 0.8151,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5796995236350312,
+      "grad_norm": 1.853406310081482,
+      "learning_rate": 7.5693225893563326e-06,
+      "loss": 0.8598,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5800659582264566,
+      "grad_norm": 1.8438457250595093,
+      "learning_rate": 7.5581165124469355e-06,
+      "loss": 0.8669,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5804323928178821,
+      "grad_norm": 1.9874687194824219,
+      "learning_rate": 7.546913695452118e-06,
+      "loss": 0.854,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5807988274093074,
+      "grad_norm": 2.0050768852233887,
+      "learning_rate": 7.535714153327639e-06,
+      "loss": 0.7932,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5811652620007328,
+      "grad_norm": 2.3346457481384277,
+      "learning_rate": 7.5245179010248885e-06,
+      "loss": 0.785,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5815316965921583,
+      "grad_norm": 2.0014426708221436,
+      "learning_rate": 7.513324953490859e-06,
+      "loss": 0.8025,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5818981311835837,
+      "grad_norm": 1.9692394733428955,
+      "learning_rate": 7.502135325668143e-06,
+      "loss": 0.8393,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5822645657750092,
+      "grad_norm": 2.0282275676727295,
+      "learning_rate": 7.49094903249489e-06,
+      "loss": 0.8652,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5826310003664346,
+      "grad_norm": 1.8718082904815674,
+      "learning_rate": 7.47976608890481e-06,
+      "loss": 0.8978,
+      "step": 1590
+    },
+    {
+      "epoch": 0.58299743495786,
+      "grad_norm": 2.1026763916015625,
+      "learning_rate": 7.468586509827123e-06,
+      "loss": 0.8509,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5833638695492854,
+      "grad_norm": 1.8932467699050903,
+      "learning_rate": 7.457410310186568e-06,
+      "loss": 0.8365,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5837303041407109,
+      "grad_norm": 1.9677419662475586,
+      "learning_rate": 7.446237504903377e-06,
+      "loss": 0.8376,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5840967387321363,
+      "grad_norm": 2.058506965637207,
+      "learning_rate": 7.435068108893244e-06,
+      "loss": 0.8204,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5844631733235618,
+      "grad_norm": 2.407780647277832,
+      "learning_rate": 7.423902137067318e-06,
+      "loss": 0.7817,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5848296079149872,
+      "grad_norm": 1.9901889562606812,
+      "learning_rate": 7.412739604332162e-06,
+      "loss": 0.8741,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5851960425064126,
+      "grad_norm": 2.0285966396331787,
+      "learning_rate": 7.401580525589767e-06,
+      "loss": 0.8727,
+      "step": 1597
+    },
+    {
+      "epoch": 0.585562477097838,
+      "grad_norm": 2.058849334716797,
+      "learning_rate": 7.390424915737495e-06,
+      "loss": 0.8293,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5859289116892634,
+      "grad_norm": 1.9227625131607056,
+      "learning_rate": 7.379272789668093e-06,
+      "loss": 0.8764,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5862953462806889,
+      "grad_norm": 2.040282726287842,
+      "learning_rate": 7.3681241622696496e-06,
+      "loss": 0.8547,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5866617808721143,
+      "grad_norm": 2.291186571121216,
+      "learning_rate": 7.3569790484255776e-06,
+      "loss": 0.862,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5870282154635398,
+      "grad_norm": 2.185753583908081,
+      "learning_rate": 7.345837463014609e-06,
+      "loss": 0.8083,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5873946500549652,
+      "grad_norm": 2.20452618598938,
+      "learning_rate": 7.334699420910758e-06,
+      "loss": 0.8042,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5877610846463907,
+      "grad_norm": 1.9438961744308472,
+      "learning_rate": 7.323564936983312e-06,
+      "loss": 0.8873,
+      "step": 1604
+    },
+    {
+      "epoch": 0.588127519237816,
+      "grad_norm": 2.1404733657836914,
+      "learning_rate": 7.312434026096813e-06,
+      "loss": 0.8349,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5884939538292415,
+      "grad_norm": 1.8856126070022583,
+      "learning_rate": 7.301306703111017e-06,
+      "loss": 0.8748,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5888603884206669,
+      "grad_norm": 2.095350980758667,
+      "learning_rate": 7.290182982880909e-06,
+      "loss": 0.7623,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5892268230120924,
+      "grad_norm": 2.0845468044281006,
+      "learning_rate": 7.2790628802566556e-06,
+      "loss": 0.8288,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5895932576035178,
+      "grad_norm": 2.0782248973846436,
+      "learning_rate": 7.26794641008359e-06,
+      "loss": 0.7994,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5899596921949432,
+      "grad_norm": 2.1508421897888184,
+      "learning_rate": 7.256833587202208e-06,
+      "loss": 0.8597,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5903261267863686,
+      "grad_norm": 2.304708242416382,
+      "learning_rate": 7.245724426448117e-06,
+      "loss": 0.8211,
+      "step": 1611
+    },
+    {
+      "epoch": 0.590692561377794,
+      "grad_norm": 2.2719075679779053,
+      "learning_rate": 7.234618942652057e-06,
+      "loss": 0.8149,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5910589959692195,
+      "grad_norm": 1.967847466468811,
+      "learning_rate": 7.223517150639845e-06,
+      "loss": 0.8223,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5914254305606449,
+      "grad_norm": 2.4015612602233887,
+      "learning_rate": 7.2124190652323765e-06,
+      "loss": 0.8032,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5917918651520704,
+      "grad_norm": 2.075702667236328,
+      "learning_rate": 7.2013247012455935e-06,
+      "loss": 0.8662,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5921582997434958,
+      "grad_norm": 2.115882158279419,
+      "learning_rate": 7.190234073490477e-06,
+      "loss": 0.8034,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5925247343349213,
+      "grad_norm": 2.18595552444458,
+      "learning_rate": 7.179147196773008e-06,
+      "loss": 0.8682,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5928911689263466,
+      "grad_norm": 2.383286952972412,
+      "learning_rate": 7.168064085894173e-06,
+      "loss": 0.7243,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5932576035177721,
+      "grad_norm": 2.022106885910034,
+      "learning_rate": 7.156984755649925e-06,
+      "loss": 0.7914,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5936240381091975,
+      "grad_norm": 2.2780921459198,
+      "learning_rate": 7.1459092208311685e-06,
+      "loss": 0.8125,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5939904727006229,
+      "grad_norm": 2.199862003326416,
+      "learning_rate": 7.13483749622375e-06,
+      "loss": 0.805,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5943569072920484,
+      "grad_norm": 2.0567820072174072,
+      "learning_rate": 7.12376959660841e-06,
+      "loss": 0.7983,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5947233418834738,
+      "grad_norm": 2.0858473777770996,
+      "learning_rate": 7.112705536760802e-06,
+      "loss": 0.7851,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5950897764748992,
+      "grad_norm": 1.8858219385147095,
+      "learning_rate": 7.1016453314514475e-06,
+      "loss": 0.8713,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5954562110663246,
+      "grad_norm": 2.3013741970062256,
+      "learning_rate": 7.090588995445722e-06,
+      "loss": 0.8711,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5958226456577501,
+      "grad_norm": 1.9575581550598145,
+      "learning_rate": 7.079536543503833e-06,
+      "loss": 0.876,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5961890802491755,
+      "grad_norm": 1.8408575057983398,
+      "learning_rate": 7.0684879903808015e-06,
+      "loss": 0.8855,
+      "step": 1627
+    },
+    {
+      "epoch": 0.596555514840601,
+      "grad_norm": 2.0506277084350586,
+      "learning_rate": 7.057443350826447e-06,
+      "loss": 0.831,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5969219494320264,
+      "grad_norm": 2.0059189796447754,
+      "learning_rate": 7.046402639585366e-06,
+      "loss": 0.7393,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5972883840234519,
+      "grad_norm": 2.1671557426452637,
+      "learning_rate": 7.035365871396911e-06,
+      "loss": 0.8705,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5976548186148772,
+      "grad_norm": 1.984377384185791,
+      "learning_rate": 7.024333060995165e-06,
+      "loss": 0.8358,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5980212532063027,
+      "grad_norm": 2.277132034301758,
+      "learning_rate": 7.013304223108931e-06,
+      "loss": 0.7796,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5983876877977281,
+      "grad_norm": 2.3955905437469482,
+      "learning_rate": 7.002279372461706e-06,
+      "loss": 0.7446,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5987541223891535,
+      "grad_norm": 2.4661448001861572,
+      "learning_rate": 6.991258523771667e-06,
+      "loss": 0.8515,
+      "step": 1634
+    },
+    {
+      "epoch": 0.599120556980579,
+      "grad_norm": 2.060856342315674,
+      "learning_rate": 6.980241691751649e-06,
+      "loss": 0.8076,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5994869915720044,
+      "grad_norm": 2.204679250717163,
+      "learning_rate": 6.9692288911091275e-06,
+      "loss": 0.7761,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5998534261634298,
+      "grad_norm": 2.0762739181518555,
+      "learning_rate": 6.9582201365461845e-06,
+      "loss": 0.8523,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6002198607548552,
+      "grad_norm": 2.0019896030426025,
+      "learning_rate": 6.94721544275951e-06,
+      "loss": 0.8053,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6005862953462807,
+      "grad_norm": 1.9410783052444458,
+      "learning_rate": 6.936214824440369e-06,
+      "loss": 0.876,
+      "step": 1639
+    },
+    {
+      "epoch": 0.6009527299377061,
+      "grad_norm": 2.2371299266815186,
+      "learning_rate": 6.925218296274592e-06,
+      "loss": 0.7718,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6013191645291316,
+      "grad_norm": 2.1076481342315674,
+      "learning_rate": 6.914225872942546e-06,
+      "loss": 0.8437,
+      "step": 1641
+    },
+    {
+      "epoch": 0.601685599120557,
+      "grad_norm": 2.0942487716674805,
+      "learning_rate": 6.903237569119108e-06,
+      "loss": 0.8823,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6020520337119825,
+      "grad_norm": 2.637465238571167,
+      "learning_rate": 6.892253399473673e-06,
+      "loss": 0.7504,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6024184683034078,
+      "grad_norm": 2.225482702255249,
+      "learning_rate": 6.881273378670102e-06,
+      "loss": 0.8458,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6027849028948332,
+      "grad_norm": 2.0072765350341797,
+      "learning_rate": 6.870297521366725e-06,
+      "loss": 0.797,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6031513374862587,
+      "grad_norm": 2.540889024734497,
+      "learning_rate": 6.859325842216314e-06,
+      "loss": 0.7914,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6035177720776841,
+      "grad_norm": 2.056896686553955,
+      "learning_rate": 6.848358355866063e-06,
+      "loss": 0.7791,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6038842066691096,
+      "grad_norm": 2.06502103805542,
+      "learning_rate": 6.83739507695756e-06,
+      "loss": 0.8494,
+      "step": 1648
+    },
+    {
+      "epoch": 0.604250641260535,
+      "grad_norm": 2.031855821609497,
+      "learning_rate": 6.82643602012679e-06,
+      "loss": 0.856,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6046170758519605,
+      "grad_norm": 2.2233779430389404,
+      "learning_rate": 6.815481200004088e-06,
+      "loss": 0.8112,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6049835104433858,
+      "grad_norm": 2.1673338413238525,
+      "learning_rate": 6.804530631214143e-06,
+      "loss": 0.8606,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6053499450348113,
+      "grad_norm": 2.1959850788116455,
+      "learning_rate": 6.793584328375969e-06,
+      "loss": 0.7925,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6057163796262367,
+      "grad_norm": 2.0298149585723877,
+      "learning_rate": 6.782642306102871e-06,
+      "loss": 0.8058,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6060828142176622,
+      "grad_norm": 2.510085344314575,
+      "learning_rate": 6.771704579002456e-06,
+      "loss": 0.8086,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6064492488090876,
+      "grad_norm": 2.2274227142333984,
+      "learning_rate": 6.760771161676589e-06,
+      "loss": 0.8525,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6068156834005131,
+      "grad_norm": 2.153073310852051,
+      "learning_rate": 6.7498420687213806e-06,
+      "loss": 0.7919,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6071821179919384,
+      "grad_norm": 2.1709446907043457,
+      "learning_rate": 6.738917314727178e-06,
+      "loss": 0.8229,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6075485525833638,
+      "grad_norm": 2.3454272747039795,
+      "learning_rate": 6.7279969142785165e-06,
+      "loss": 0.7746,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6079149871747893,
+      "grad_norm": 2.0328891277313232,
+      "learning_rate": 6.717080881954139e-06,
+      "loss": 0.8782,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6082814217662147,
+      "grad_norm": 2.0833816528320312,
+      "learning_rate": 6.706169232326949e-06,
+      "loss": 0.8619,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6086478563576402,
+      "grad_norm": 2.4952611923217773,
+      "learning_rate": 6.695261979963995e-06,
+      "loss": 0.7782,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6090142909490656,
+      "grad_norm": 2.224849224090576,
+      "learning_rate": 6.6843591394264676e-06,
+      "loss": 0.7831,
+      "step": 1662
+    },
+    {
+      "epoch": 0.609380725540491,
+      "grad_norm": 1.9657704830169678,
+      "learning_rate": 6.673460725269649e-06,
+      "loss": 0.8668,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6097471601319164,
+      "grad_norm": 2.173832416534424,
+      "learning_rate": 6.662566752042929e-06,
+      "loss": 0.8046,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6101135947233419,
+      "grad_norm": 1.9357271194458008,
+      "learning_rate": 6.651677234289762e-06,
+      "loss": 0.8335,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6104800293147673,
+      "grad_norm": 2.003981113433838,
+      "learning_rate": 6.640792186547654e-06,
+      "loss": 0.8166,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6108464639061928,
+      "grad_norm": 2.0950145721435547,
+      "learning_rate": 6.629911623348147e-06,
+      "loss": 0.8494,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6112128984976182,
+      "grad_norm": 1.9635558128356934,
+      "learning_rate": 6.619035559216786e-06,
+      "loss": 0.8174,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6115793330890436,
+      "grad_norm": 1.9462409019470215,
+      "learning_rate": 6.608164008673121e-06,
+      "loss": 0.8574,
+      "step": 1669
+    },
+    {
+      "epoch": 0.611945767680469,
+      "grad_norm": 1.9010626077651978,
+      "learning_rate": 6.597296986230674e-06,
+      "loss": 0.8177,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6123122022718944,
+      "grad_norm": 2.2767860889434814,
+      "learning_rate": 6.586434506396924e-06,
+      "loss": 0.7865,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6126786368633199,
+      "grad_norm": 2.176220417022705,
+      "learning_rate": 6.575576583673277e-06,
+      "loss": 0.8348,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6130450714547453,
+      "grad_norm": 1.862492561340332,
+      "learning_rate": 6.5647232325550594e-06,
+      "loss": 0.8615,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6134115060461708,
+      "grad_norm": 2.045877695083618,
+      "learning_rate": 6.553874467531496e-06,
+      "loss": 0.8176,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6137779406375962,
+      "grad_norm": 2.3825926780700684,
+      "learning_rate": 6.54303030308569e-06,
+      "loss": 0.8194,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6141443752290217,
+      "grad_norm": 2.3617899417877197,
+      "learning_rate": 6.532190753694605e-06,
+      "loss": 0.7633,
+      "step": 1676
+    },
+    {
+      "epoch": 0.614510809820447,
+      "grad_norm": 2.279677391052246,
+      "learning_rate": 6.521355833829033e-06,
+      "loss": 0.8462,
+      "step": 1677
+    },
+    {
+      "epoch": 0.6148772444118725,
+      "grad_norm": 2.1296496391296387,
+      "learning_rate": 6.5105255579536e-06,
+      "loss": 0.8904,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6152436790032979,
+      "grad_norm": 2.403574228286743,
+      "learning_rate": 6.499699940526718e-06,
+      "loss": 0.8135,
+      "step": 1679
+    },
+    {
+      "epoch": 0.6156101135947234,
+      "grad_norm": 2.0573787689208984,
+      "learning_rate": 6.488878996000589e-06,
+      "loss": 0.804,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6159765481861488,
+      "grad_norm": 2.0022242069244385,
+      "learning_rate": 6.478062738821175e-06,
+      "loss": 0.8562,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6163429827775742,
+      "grad_norm": 2.484389305114746,
+      "learning_rate": 6.467251183428182e-06,
+      "loss": 0.8543,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6167094173689996,
+      "grad_norm": 2.20928692817688,
+      "learning_rate": 6.456444344255037e-06,
+      "loss": 0.8088,
+      "step": 1683
+    },
+    {
+      "epoch": 0.617075851960425,
+      "grad_norm": 2.1175923347473145,
+      "learning_rate": 6.4456422357288665e-06,
+      "loss": 0.8111,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6174422865518505,
+      "grad_norm": 2.030487298965454,
+      "learning_rate": 6.434844872270486e-06,
+      "loss": 0.8055,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6178087211432759,
+      "grad_norm": 2.0303640365600586,
+      "learning_rate": 6.42405226829438e-06,
+      "loss": 0.8603,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6181751557347014,
+      "grad_norm": 2.0851595401763916,
+      "learning_rate": 6.413264438208675e-06,
+      "loss": 0.8305,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6185415903261268,
+      "grad_norm": 2.7440991401672363,
+      "learning_rate": 6.402481396415123e-06,
+      "loss": 0.7822,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6189080249175523,
+      "grad_norm": 2.3908801078796387,
+      "learning_rate": 6.391703157309088e-06,
+      "loss": 0.7986,
+      "step": 1689
+    },
+    {
+      "epoch": 0.6192744595089776,
+      "grad_norm": 2.0651395320892334,
+      "learning_rate": 6.380929735279514e-06,
+      "loss": 0.8235,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6196408941004031,
+      "grad_norm": 2.590243339538574,
+      "learning_rate": 6.370161144708924e-06,
+      "loss": 0.6925,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6200073286918285,
+      "grad_norm": 2.1181154251098633,
+      "learning_rate": 6.359397399973386e-06,
+      "loss": 0.8766,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6203737632832539,
+      "grad_norm": 2.16276216506958,
+      "learning_rate": 6.348638515442506e-06,
+      "loss": 0.8357,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6207401978746794,
+      "grad_norm": 2.131016969680786,
+      "learning_rate": 6.337884505479383e-06,
+      "loss": 0.8125,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6211066324661048,
+      "grad_norm": 1.936014175415039,
+      "learning_rate": 6.327135384440633e-06,
+      "loss": 0.8784,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6214730670575302,
+      "grad_norm": 2.428760290145874,
+      "learning_rate": 6.316391166676323e-06,
+      "loss": 0.7639,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6218395016489556,
+      "grad_norm": 2.2411906719207764,
+      "learning_rate": 6.3056518665299895e-06,
+      "loss": 0.8401,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6222059362403811,
+      "grad_norm": 3.147825241088867,
+      "learning_rate": 6.294917498338602e-06,
+      "loss": 0.747,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6225723708318065,
+      "grad_norm": 2.061725616455078,
+      "learning_rate": 6.284188076432536e-06,
+      "loss": 0.8259,
+      "step": 1699
+    },
+    {
+      "epoch": 0.622938805423232,
+      "grad_norm": 2.19020676612854,
+      "learning_rate": 6.2734636151355735e-06,
+      "loss": 0.87,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6233052400146574,
+      "grad_norm": 2.1691603660583496,
+      "learning_rate": 6.262744128764868e-06,
+      "loss": 0.8505,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6236716746060829,
+      "grad_norm": 2.165039300918579,
+      "learning_rate": 6.252029631630938e-06,
+      "loss": 0.7862,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6240381091975082,
+      "grad_norm": 2.140118360519409,
+      "learning_rate": 6.2413201380376395e-06,
+      "loss": 0.8015,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6244045437889337,
+      "grad_norm": 1.9693007469177246,
+      "learning_rate": 6.23061566228214e-06,
+      "loss": 0.8262,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6247709783803591,
+      "grad_norm": 2.224019765853882,
+      "learning_rate": 6.21991621865492e-06,
+      "loss": 0.8128,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6251374129717845,
+      "grad_norm": 2.5416271686553955,
+      "learning_rate": 6.209221821439737e-06,
+      "loss": 0.8195,
+      "step": 1706
+    },
+    {
+      "epoch": 0.62550384756321,
+      "grad_norm": 1.9668278694152832,
+      "learning_rate": 6.1985324849136094e-06,
+      "loss": 0.8628,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6258702821546354,
+      "grad_norm": 2.0069453716278076,
+      "learning_rate": 6.187848223346804e-06,
+      "loss": 0.8441,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6262367167460609,
+      "grad_norm": 2.186694383621216,
+      "learning_rate": 6.177169051002815e-06,
+      "loss": 0.8359,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6266031513374862,
+      "grad_norm": 2.1202750205993652,
+      "learning_rate": 6.166494982138328e-06,
+      "loss": 0.813,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6269695859289117,
+      "grad_norm": 2.4543404579162598,
+      "learning_rate": 6.155826031003234e-06,
+      "loss": 0.837,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6273360205203371,
+      "grad_norm": 2.054497003555298,
+      "learning_rate": 6.145162211840576e-06,
+      "loss": 0.8348,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6277024551117626,
+      "grad_norm": 2.0176241397857666,
+      "learning_rate": 6.134503538886558e-06,
+      "loss": 0.7921,
+      "step": 1713
+    },
+    {
+      "epoch": 0.628068889703188,
+      "grad_norm": 2.011383533477783,
+      "learning_rate": 6.123850026370512e-06,
+      "loss": 0.8338,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6284353242946135,
+      "grad_norm": 2.33731746673584,
+      "learning_rate": 6.113201688514872e-06,
+      "loss": 0.785,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6288017588860388,
+      "grad_norm": 2.0671072006225586,
+      "learning_rate": 6.102558539535168e-06,
+      "loss": 0.8809,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6291681934774642,
+      "grad_norm": 2.25299334526062,
+      "learning_rate": 6.091920593640013e-06,
+      "loss": 0.8024,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6295346280688897,
+      "grad_norm": 1.911569356918335,
+      "learning_rate": 6.081287865031056e-06,
+      "loss": 0.8447,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6299010626603151,
+      "grad_norm": 1.5879853963851929,
+      "learning_rate": 6.070660367902998e-06,
+      "loss": 0.5188,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6302674972517406,
+      "grad_norm": 1.9963817596435547,
+      "learning_rate": 6.0600381164435405e-06,
+      "loss": 0.8051,
+      "step": 1720
+    },
+    {
+      "epoch": 0.630633931843166,
+      "grad_norm": 2.1238741874694824,
+      "learning_rate": 6.049421124833395e-06,
+      "loss": 0.84,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6310003664345915,
+      "grad_norm": 2.239947557449341,
+      "learning_rate": 6.0388094072462465e-06,
+      "loss": 0.8142,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6313668010260168,
+      "grad_norm": 2.5329320430755615,
+      "learning_rate": 6.028202977848734e-06,
+      "loss": 0.7447,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6317332356174423,
+      "grad_norm": 2.675870418548584,
+      "learning_rate": 6.017601850800447e-06,
+      "loss": 0.8132,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6320996702088677,
+      "grad_norm": 1.9416941404342651,
+      "learning_rate": 6.007006040253885e-06,
+      "loss": 0.824,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6324661048002932,
+      "grad_norm": 2.223480224609375,
+      "learning_rate": 5.996415560354456e-06,
+      "loss": 0.8814,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6328325393917186,
+      "grad_norm": 1.969476342201233,
+      "learning_rate": 5.985830425240455e-06,
+      "loss": 0.8341,
+      "step": 1727
+    },
+    {
+      "epoch": 0.633198973983144,
+      "grad_norm": 1.9697409868240356,
+      "learning_rate": 5.975250649043039e-06,
+      "loss": 0.915,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6335654085745694,
+      "grad_norm": 2.4558944702148438,
+      "learning_rate": 5.964676245886205e-06,
+      "loss": 0.8503,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6339318431659948,
+      "grad_norm": 1.9908795356750488,
+      "learning_rate": 5.954107229886784e-06,
+      "loss": 0.8058,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6342982777574203,
+      "grad_norm": 2.4454500675201416,
+      "learning_rate": 5.94354361515441e-06,
+      "loss": 0.7536,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6346647123488457,
+      "grad_norm": 2.3198952674865723,
+      "learning_rate": 5.932985415791513e-06,
+      "loss": 0.8192,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6350311469402712,
+      "grad_norm": 1.959184169769287,
+      "learning_rate": 5.92243264589329e-06,
+      "loss": 0.8082,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6353975815316966,
+      "grad_norm": 2.3622608184814453,
+      "learning_rate": 5.911885319547688e-06,
+      "loss": 0.8634,
+      "step": 1734
+    },
+    {
+      "epoch": 0.635764016123122,
+      "grad_norm": 2.3816590309143066,
+      "learning_rate": 5.901343450835388e-06,
+      "loss": 0.8062,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6361304507145474,
+      "grad_norm": 2.5151548385620117,
+      "learning_rate": 5.890807053829784e-06,
+      "loss": 0.757,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6364968853059729,
+      "grad_norm": 2.0787878036499023,
+      "learning_rate": 5.8802761425969654e-06,
+      "loss": 0.8356,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6368633198973983,
+      "grad_norm": 2.1201980113983154,
+      "learning_rate": 5.869750731195702e-06,
+      "loss": 0.9005,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6372297544888238,
+      "grad_norm": 2.0525357723236084,
+      "learning_rate": 5.859230833677423e-06,
+      "loss": 0.8693,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6375961890802492,
+      "grad_norm": 2.3157238960266113,
+      "learning_rate": 5.848716464086184e-06,
+      "loss": 0.8153,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6379626236716746,
+      "grad_norm": 2.4236457347869873,
+      "learning_rate": 5.838207636458674e-06,
+      "loss": 0.8364,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6383290582631,
+      "grad_norm": 2.645217180252075,
+      "learning_rate": 5.8277043648241735e-06,
+      "loss": 0.7743,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6386954928545254,
+      "grad_norm": 2.5812301635742188,
+      "learning_rate": 5.8172066632045535e-06,
+      "loss": 0.8233,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6390619274459509,
+      "grad_norm": 2.15390944480896,
+      "learning_rate": 5.8067145456142475e-06,
+      "loss": 0.8209,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6394283620373763,
+      "grad_norm": 2.439054012298584,
+      "learning_rate": 5.796228026060235e-06,
+      "loss": 0.744,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6397947966288018,
+      "grad_norm": 2.1751368045806885,
+      "learning_rate": 5.785747118542013e-06,
+      "loss": 0.7693,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6401612312202272,
+      "grad_norm": 2.5186259746551514,
+      "learning_rate": 5.7752718370515985e-06,
+      "loss": 0.7651,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6405276658116527,
+      "grad_norm": 2.056795835494995,
+      "learning_rate": 5.764802195573495e-06,
+      "loss": 0.8122,
+      "step": 1748
+    },
+    {
+      "epoch": 0.640894100403078,
+      "grad_norm": 2.7081797122955322,
+      "learning_rate": 5.7543382080846675e-06,
+      "loss": 0.6878,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6412605349945035,
+      "grad_norm": 2.0384063720703125,
+      "learning_rate": 5.7438798885545475e-06,
+      "loss": 0.8603,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6416269695859289,
+      "grad_norm": 2.258711099624634,
+      "learning_rate": 5.733427250944985e-06,
+      "loss": 0.7606,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6419934041773543,
+      "grad_norm": 2.474824905395508,
+      "learning_rate": 5.722980309210253e-06,
+      "loss": 0.7513,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6423598387687798,
+      "grad_norm": 1.9493567943572998,
+      "learning_rate": 5.712539077297023e-06,
+      "loss": 0.8404,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6427262733602052,
+      "grad_norm": 1.9995802640914917,
+      "learning_rate": 5.7021035691443375e-06,
+      "loss": 0.8871,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6430927079516306,
+      "grad_norm": 2.001211404800415,
+      "learning_rate": 5.691673798683606e-06,
+      "loss": 0.8844,
+      "step": 1755
+    },
+    {
+      "epoch": 0.643459142543056,
+      "grad_norm": 1.9877080917358398,
+      "learning_rate": 5.681249779838563e-06,
+      "loss": 0.77,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6438255771344815,
+      "grad_norm": 2.3511409759521484,
+      "learning_rate": 5.670831526525279e-06,
+      "loss": 0.7055,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6441920117259069,
+      "grad_norm": 2.2715601921081543,
+      "learning_rate": 5.660419052652123e-06,
+      "loss": 0.8774,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6445584463173324,
+      "grad_norm": 2.2003986835479736,
+      "learning_rate": 5.650012372119751e-06,
+      "loss": 0.7563,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6449248809087578,
+      "grad_norm": 2.1989638805389404,
+      "learning_rate": 5.639611498821076e-06,
+      "loss": 0.8778,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6452913155001833,
+      "grad_norm": 2.143601417541504,
+      "learning_rate": 5.62921644664127e-06,
+      "loss": 0.8044,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6456577500916086,
+      "grad_norm": 2.3783957958221436,
+      "learning_rate": 5.618827229457722e-06,
+      "loss": 0.808,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6460241846830341,
+      "grad_norm": 2.843330144882202,
+      "learning_rate": 5.60844386114004e-06,
+      "loss": 0.7242,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6463906192744595,
+      "grad_norm": 2.4551148414611816,
+      "learning_rate": 5.598066355550023e-06,
+      "loss": 0.8453,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6467570538658849,
+      "grad_norm": 2.3557980060577393,
+      "learning_rate": 5.587694726541645e-06,
+      "loss": 0.8925,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6471234884573104,
+      "grad_norm": 2.076031446456909,
+      "learning_rate": 5.577328987961022e-06,
+      "loss": 0.8178,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6474899230487358,
+      "grad_norm": 2.304417371749878,
+      "learning_rate": 5.56696915364642e-06,
+      "loss": 0.8694,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6478563576401613,
+      "grad_norm": 1.8749827146530151,
+      "learning_rate": 5.556615237428219e-06,
+      "loss": 0.8058,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6482227922315866,
+      "grad_norm": 2.451500654220581,
+      "learning_rate": 5.546267253128897e-06,
+      "loss": 0.9056,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6485892268230121,
+      "grad_norm": 2.123460531234741,
+      "learning_rate": 5.5359252145630186e-06,
+      "loss": 0.7229,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6489556614144375,
+      "grad_norm": 2.1700215339660645,
+      "learning_rate": 5.525589135537195e-06,
+      "loss": 0.8278,
+      "step": 1771
+    },
+    {
+      "epoch": 0.649322096005863,
+      "grad_norm": 2.158203125,
+      "learning_rate": 5.515259029850104e-06,
+      "loss": 0.853,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6496885305972884,
+      "grad_norm": 2.275250196456909,
+      "learning_rate": 5.504934911292428e-06,
+      "loss": 0.8514,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6500549651887139,
+      "grad_norm": 2.0677850246429443,
+      "learning_rate": 5.494616793646867e-06,
+      "loss": 0.8981,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6504213997801392,
+      "grad_norm": 2.440816640853882,
+      "learning_rate": 5.484304690688113e-06,
+      "loss": 0.8005,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6507878343715646,
+      "grad_norm": 2.1558449268341064,
+      "learning_rate": 5.473998616182825e-06,
+      "loss": 0.8014,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6511542689629901,
+      "grad_norm": 2.008288860321045,
+      "learning_rate": 5.463698583889605e-06,
+      "loss": 0.8639,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6515207035544155,
+      "grad_norm": 2.2373805046081543,
+      "learning_rate": 5.4534046075590016e-06,
+      "loss": 0.8344,
+      "step": 1778
+    },
+    {
+      "epoch": 0.651887138145841,
+      "grad_norm": 2.244774580001831,
+      "learning_rate": 5.443116700933475e-06,
+      "loss": 0.7931,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6522535727372664,
+      "grad_norm": 2.13997745513916,
+      "learning_rate": 5.432834877747377e-06,
+      "loss": 0.8223,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6526200073286919,
+      "grad_norm": 2.2929978370666504,
+      "learning_rate": 5.422559151726952e-06,
+      "loss": 0.8176,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6529864419201172,
+      "grad_norm": 2.4449853897094727,
+      "learning_rate": 5.41228953659028e-06,
+      "loss": 0.8079,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6533528765115427,
+      "grad_norm": 2.67828106880188,
+      "learning_rate": 5.402026046047311e-06,
+      "loss": 0.8899,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6537193111029681,
+      "grad_norm": 2.5562262535095215,
+      "learning_rate": 5.391768693799795e-06,
+      "loss": 0.7825,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6540857456943936,
+      "grad_norm": 2.0743002891540527,
+      "learning_rate": 5.381517493541301e-06,
+      "loss": 0.8028,
+      "step": 1785
+    },
+    {
+      "epoch": 0.654452180285819,
+      "grad_norm": 1.9827677011489868,
+      "learning_rate": 5.371272458957188e-06,
+      "loss": 0.8346,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6548186148772445,
+      "grad_norm": 2.188796281814575,
+      "learning_rate": 5.361033603724571e-06,
+      "loss": 0.8082,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6551850494686698,
+      "grad_norm": 2.5796215534210205,
+      "learning_rate": 5.350800941512321e-06,
+      "loss": 0.825,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6555514840600952,
+      "grad_norm": 2.0983974933624268,
+      "learning_rate": 5.340574485981046e-06,
+      "loss": 0.8148,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6559179186515207,
+      "grad_norm": 2.4515507221221924,
+      "learning_rate": 5.3303542507830655e-06,
+      "loss": 0.7716,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6562843532429461,
+      "grad_norm": 2.2113089561462402,
+      "learning_rate": 5.320140249562396e-06,
+      "loss": 0.8046,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6566507878343716,
+      "grad_norm": 2.609797716140747,
+      "learning_rate": 5.309932495954722e-06,
+      "loss": 0.8499,
+      "step": 1792
+    },
+    {
+      "epoch": 0.657017222425797,
+      "grad_norm": 2.038630247116089,
+      "learning_rate": 5.2997310035874e-06,
+      "loss": 0.7718,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6573836570172225,
+      "grad_norm": 1.9817092418670654,
+      "learning_rate": 5.289535786079425e-06,
+      "loss": 0.7688,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6577500916086478,
+      "grad_norm": 2.002600908279419,
+      "learning_rate": 5.279346857041408e-06,
+      "loss": 0.8807,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6581165262000733,
+      "grad_norm": 2.2284350395202637,
+      "learning_rate": 5.269164230075574e-06,
+      "loss": 0.7766,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6584829607914987,
+      "grad_norm": 2.0263407230377197,
+      "learning_rate": 5.258987918775725e-06,
+      "loss": 0.8194,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6588493953829242,
+      "grad_norm": 2.078047752380371,
+      "learning_rate": 5.248817936727242e-06,
+      "loss": 0.8122,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6592158299743496,
+      "grad_norm": 2.666020154953003,
+      "learning_rate": 5.238654297507048e-06,
+      "loss": 0.7913,
+      "step": 1799
+    },
+    {
+      "epoch": 0.659582264565775,
+      "grad_norm": 2.093247890472412,
+      "learning_rate": 5.228497014683604e-06,
+      "loss": 0.8071,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6599486991572004,
+      "grad_norm": 2.7277653217315674,
+      "learning_rate": 5.218346101816884e-06,
+      "loss": 0.733,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6603151337486258,
+      "grad_norm": 2.3742880821228027,
+      "learning_rate": 5.20820157245836e-06,
+      "loss": 0.8094,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6606815683400513,
+      "grad_norm": 2.258880615234375,
+      "learning_rate": 5.198063440150972e-06,
+      "loss": 0.8041,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6610480029314767,
+      "grad_norm": 2.546694040298462,
+      "learning_rate": 5.187931718429129e-06,
+      "loss": 0.779,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6614144375229022,
+      "grad_norm": 2.5752904415130615,
+      "learning_rate": 5.177806420818688e-06,
+      "loss": 0.8587,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6617808721143276,
+      "grad_norm": 2.5327069759368896,
+      "learning_rate": 5.167687560836908e-06,
+      "loss": 0.7962,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6621473067057531,
+      "grad_norm": 2.1464478969573975,
+      "learning_rate": 5.15757515199248e-06,
+      "loss": 0.8641,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6625137412971784,
+      "grad_norm": 2.343508005142212,
+      "learning_rate": 5.147469207785463e-06,
+      "loss": 0.842,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6628801758886039,
+      "grad_norm": 2.575166940689087,
+      "learning_rate": 5.137369741707293e-06,
+      "loss": 0.7338,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6632466104800293,
+      "grad_norm": 2.181368827819824,
+      "learning_rate": 5.127276767240761e-06,
+      "loss": 0.7625,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6636130450714548,
+      "grad_norm": 2.4319188594818115,
+      "learning_rate": 5.117190297859986e-06,
+      "loss": 0.7884,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6639794796628802,
+      "grad_norm": 2.30359148979187,
+      "learning_rate": 5.107110347030413e-06,
+      "loss": 0.755,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6643459142543056,
+      "grad_norm": 2.1688833236694336,
+      "learning_rate": 5.0970369282087654e-06,
+      "loss": 0.7901,
+      "step": 1813
+    },
+    {
+      "epoch": 0.664712348845731,
+      "grad_norm": 2.0516762733459473,
+      "learning_rate": 5.086970054843061e-06,
+      "loss": 0.8528,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6650787834371564,
+      "grad_norm": 2.297912836074829,
+      "learning_rate": 5.07690974037258e-06,
+      "loss": 0.8266,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6654452180285819,
+      "grad_norm": 2.228154420852661,
+      "learning_rate": 5.0668559982278396e-06,
+      "loss": 0.8252,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6658116526200073,
+      "grad_norm": 2.3292181491851807,
+      "learning_rate": 5.056808841830592e-06,
+      "loss": 0.8072,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6661780872114328,
+      "grad_norm": 2.572441577911377,
+      "learning_rate": 5.046768284593786e-06,
+      "loss": 0.782,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6665445218028582,
+      "grad_norm": 2.1662983894348145,
+      "learning_rate": 5.036734339921563e-06,
+      "loss": 0.7953,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6669109563942837,
+      "grad_norm": 1.9589260816574097,
+      "learning_rate": 5.026707021209243e-06,
+      "loss": 0.8491,
+      "step": 1820
+    },
+    {
+      "epoch": 0.667277390985709,
+      "grad_norm": 2.5949389934539795,
+      "learning_rate": 5.016686341843298e-06,
+      "loss": 0.7735,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6676438255771345,
+      "grad_norm": 2.1354405879974365,
+      "learning_rate": 5.0066723152013375e-06,
+      "loss": 0.8408,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6680102601685599,
+      "grad_norm": 2.4556338787078857,
+      "learning_rate": 4.996664954652082e-06,
+      "loss": 0.8595,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6683766947599853,
+      "grad_norm": 2.425250291824341,
+      "learning_rate": 4.986664273555361e-06,
+      "loss": 0.8545,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6687431293514108,
+      "grad_norm": 2.1328673362731934,
+      "learning_rate": 4.976670285262084e-06,
+      "loss": 0.84,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6691095639428362,
+      "grad_norm": 2.421682834625244,
+      "learning_rate": 4.966683003114226e-06,
+      "loss": 0.8028,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6694759985342617,
+      "grad_norm": 2.466822385787964,
+      "learning_rate": 4.956702440444815e-06,
+      "loss": 0.8076,
+      "step": 1827
+    },
+    {
+      "epoch": 0.669842433125687,
+      "grad_norm": 2.335223436355591,
+      "learning_rate": 4.9467286105778914e-06,
+      "loss": 0.8457,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6702088677171125,
+      "grad_norm": 2.1934590339660645,
+      "learning_rate": 4.936761526828528e-06,
+      "loss": 0.7996,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6705753023085379,
+      "grad_norm": 2.3251702785491943,
+      "learning_rate": 4.926801202502777e-06,
+      "loss": 0.8577,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6709417368999634,
+      "grad_norm": 1.9324642419815063,
+      "learning_rate": 4.916847650897671e-06,
+      "loss": 0.7787,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6713081714913888,
+      "grad_norm": 2.155264139175415,
+      "learning_rate": 4.906900885301204e-06,
+      "loss": 0.822,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6716746060828143,
+      "grad_norm": 2.3645036220550537,
+      "learning_rate": 4.896960918992314e-06,
+      "loss": 0.7823,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6720410406742396,
+      "grad_norm": 2.374627113342285,
+      "learning_rate": 4.887027765240847e-06,
+      "loss": 0.7844,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6724074752656651,
+      "grad_norm": 2.294753074645996,
+      "learning_rate": 4.877101437307567e-06,
+      "loss": 0.8298,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6727739098570905,
+      "grad_norm": 2.3122475147247314,
+      "learning_rate": 4.867181948444124e-06,
+      "loss": 0.7986,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6731403444485159,
+      "grad_norm": 2.2505087852478027,
+      "learning_rate": 4.8572693118930355e-06,
+      "loss": 0.8444,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6735067790399414,
+      "grad_norm": 2.7303593158721924,
+      "learning_rate": 4.847363540887676e-06,
+      "loss": 0.7972,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6738732136313668,
+      "grad_norm": 2.4003570079803467,
+      "learning_rate": 4.837464648652241e-06,
+      "loss": 0.8782,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6742396482227923,
+      "grad_norm": 2.2989132404327393,
+      "learning_rate": 4.827572648401761e-06,
+      "loss": 0.8155,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6746060828142176,
+      "grad_norm": 2.5501468181610107,
+      "learning_rate": 4.817687553342049e-06,
+      "loss": 0.8128,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6749725174056431,
+      "grad_norm": 2.325488328933716,
+      "learning_rate": 4.8078093766697135e-06,
+      "loss": 0.7826,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6753389519970685,
+      "grad_norm": 1.9862393140792847,
+      "learning_rate": 4.7979381315721206e-06,
+      "loss": 0.8583,
+      "step": 1843
+    },
+    {
+      "epoch": 0.675705386588494,
+      "grad_norm": 2.321089267730713,
+      "learning_rate": 4.788073831227377e-06,
+      "loss": 0.827,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6760718211799194,
+      "grad_norm": 2.392892360687256,
+      "learning_rate": 4.778216488804329e-06,
+      "loss": 0.8084,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6764382557713449,
+      "grad_norm": 2.309082269668579,
+      "learning_rate": 4.768366117462527e-06,
+      "loss": 0.8634,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6768046903627702,
+      "grad_norm": 2.0040342807769775,
+      "learning_rate": 4.758522730352221e-06,
+      "loss": 0.8197,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6771711249541956,
+      "grad_norm": 2.4351770877838135,
+      "learning_rate": 4.7486863406143335e-06,
+      "loss": 0.8073,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6775375595456211,
+      "grad_norm": 2.468825101852417,
+      "learning_rate": 4.738856961380438e-06,
+      "loss": 0.8616,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6779039941370465,
+      "grad_norm": 2.3825764656066895,
+      "learning_rate": 4.729034605772762e-06,
+      "loss": 0.7066,
+      "step": 1850
+    },
+    {
+      "epoch": 0.678270428728472,
+      "grad_norm": 1.9712620973587036,
+      "learning_rate": 4.719219286904154e-06,
+      "loss": 0.846,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6786368633198974,
+      "grad_norm": 2.075200319290161,
+      "learning_rate": 4.709411017878057e-06,
+      "loss": 0.8984,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6790032979113229,
+      "grad_norm": 2.2839746475219727,
+      "learning_rate": 4.699609811788518e-06,
+      "loss": 0.7861,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6793697325027482,
+      "grad_norm": 2.22563099861145,
+      "learning_rate": 4.68981568172014e-06,
+      "loss": 0.8667,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6797361670941737,
+      "grad_norm": 2.270456314086914,
+      "learning_rate": 4.6800286407480925e-06,
+      "loss": 0.7964,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6801026016855991,
+      "grad_norm": 2.176010847091675,
+      "learning_rate": 4.670248701938073e-06,
+      "loss": 0.8349,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6804690362770246,
+      "grad_norm": 2.3070919513702393,
+      "learning_rate": 4.660475878346302e-06,
+      "loss": 0.793,
+      "step": 1857
+    },
+    {
+      "epoch": 0.68083547086845,
+      "grad_norm": 2.3936715126037598,
+      "learning_rate": 4.650710183019503e-06,
+      "loss": 0.8097,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6812019054598755,
+      "grad_norm": 2.1490304470062256,
+      "learning_rate": 4.640951628994873e-06,
+      "loss": 0.7925,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6815683400513008,
+      "grad_norm": 2.260568857192993,
+      "learning_rate": 4.631200229300086e-06,
+      "loss": 0.8572,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6819347746427262,
+      "grad_norm": 2.387512683868408,
+      "learning_rate": 4.621455996953258e-06,
+      "loss": 0.7959,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6823012092341517,
+      "grad_norm": 2.167491912841797,
+      "learning_rate": 4.611718944962943e-06,
+      "loss": 0.8902,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6826676438255771,
+      "grad_norm": 2.0454623699188232,
+      "learning_rate": 4.601989086328108e-06,
+      "loss": 0.8615,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6830340784170026,
+      "grad_norm": 2.572718381881714,
+      "learning_rate": 4.592266434038113e-06,
+      "loss": 0.81,
+      "step": 1864
+    },
+    {
+      "epoch": 0.683400513008428,
+      "grad_norm": 2.236222743988037,
+      "learning_rate": 4.582551001072693e-06,
+      "loss": 0.8717,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6837669475998535,
+      "grad_norm": 2.2675299644470215,
+      "learning_rate": 4.572842800401955e-06,
+      "loss": 0.8476,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6841333821912788,
+      "grad_norm": 2.0991828441619873,
+      "learning_rate": 4.5631418449863505e-06,
+      "loss": 0.7759,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6844998167827043,
+      "grad_norm": 2.1412177085876465,
+      "learning_rate": 4.5534481477766536e-06,
+      "loss": 0.7955,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6848662513741297,
+      "grad_norm": 2.0972344875335693,
+      "learning_rate": 4.543761721713953e-06,
+      "loss": 0.8447,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6852326859655552,
+      "grad_norm": 2.348965644836426,
+      "learning_rate": 4.534082579729623e-06,
+      "loss": 0.8297,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6855991205569806,
+      "grad_norm": 2.3317277431488037,
+      "learning_rate": 4.524410734745322e-06,
+      "loss": 0.791,
+      "step": 1871
+    },
+    {
+      "epoch": 0.685965555148406,
+      "grad_norm": 2.5072484016418457,
+      "learning_rate": 4.514746199672961e-06,
+      "loss": 0.8313,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6863319897398314,
+      "grad_norm": 2.189089059829712,
+      "learning_rate": 4.5050889874146966e-06,
+      "loss": 0.8015,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6866984243312568,
+      "grad_norm": 2.1472368240356445,
+      "learning_rate": 4.495439110862912e-06,
+      "loss": 0.809,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6870648589226823,
+      "grad_norm": 2.632460832595825,
+      "learning_rate": 4.4857965829001885e-06,
+      "loss": 0.8205,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6874312935141077,
+      "grad_norm": 2.7049458026885986,
+      "learning_rate": 4.4761614163992975e-06,
+      "loss": 0.7543,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6877977281055332,
+      "grad_norm": 2.1904704570770264,
+      "learning_rate": 4.46653362422319e-06,
+      "loss": 0.8536,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6881641626969586,
+      "grad_norm": 2.2160379886627197,
+      "learning_rate": 4.4569132192249705e-06,
+      "loss": 0.7863,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6885305972883841,
+      "grad_norm": 2.341730833053589,
+      "learning_rate": 4.447300214247884e-06,
+      "loss": 0.8412,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6888970318798094,
+      "grad_norm": 2.155663013458252,
+      "learning_rate": 4.437694622125282e-06,
+      "loss": 0.8579,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6892634664712349,
+      "grad_norm": 2.2758772373199463,
+      "learning_rate": 4.428096455680638e-06,
+      "loss": 0.8193,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6896299010626603,
+      "grad_norm": 2.211446762084961,
+      "learning_rate": 4.4185057277275025e-06,
+      "loss": 0.7861,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6899963356540857,
+      "grad_norm": 2.061429977416992,
+      "learning_rate": 4.408922451069499e-06,
+      "loss": 0.8447,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6903627702455112,
+      "grad_norm": 2.347288131713867,
+      "learning_rate": 4.399346638500305e-06,
+      "loss": 0.7768,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6907292048369366,
+      "grad_norm": 2.0781826972961426,
+      "learning_rate": 4.3897783028036235e-06,
+      "loss": 0.8566,
+      "step": 1885
+    },
+    {
+      "epoch": 0.691095639428362,
+      "grad_norm": 2.022632360458374,
+      "learning_rate": 4.380217456753191e-06,
+      "loss": 0.8002,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6914620740197874,
+      "grad_norm": 2.080552101135254,
+      "learning_rate": 4.370664113112729e-06,
+      "loss": 0.87,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6918285086112129,
+      "grad_norm": 2.203587770462036,
+      "learning_rate": 4.361118284635958e-06,
+      "loss": 0.8213,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6921949432026383,
+      "grad_norm": 2.3901920318603516,
+      "learning_rate": 4.351579984066556e-06,
+      "loss": 0.7572,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6925613777940638,
+      "grad_norm": 2.557530641555786,
+      "learning_rate": 4.342049224138162e-06,
+      "loss": 0.8289,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6929278123854892,
+      "grad_norm": 2.30087947845459,
+      "learning_rate": 4.33252601757433e-06,
+      "loss": 0.8442,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6932942469769147,
+      "grad_norm": 2.3882224559783936,
+      "learning_rate": 4.323010377088548e-06,
+      "loss": 0.7472,
+      "step": 1892
+    },
+    {
+      "epoch": 0.69366068156834,
+      "grad_norm": 2.2196640968322754,
+      "learning_rate": 4.313502315384195e-06,
+      "loss": 0.826,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6940271161597655,
+      "grad_norm": 2.4228856563568115,
+      "learning_rate": 4.304001845154534e-06,
+      "loss": 0.8173,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6943935507511909,
+      "grad_norm": 2.492797374725342,
+      "learning_rate": 4.2945089790826964e-06,
+      "loss": 0.8491,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6947599853426163,
+      "grad_norm": 2.392589807510376,
+      "learning_rate": 4.285023729841653e-06,
+      "loss": 0.8966,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6951264199340418,
+      "grad_norm": 2.366020679473877,
+      "learning_rate": 4.275546110094214e-06,
+      "loss": 0.8433,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6954928545254672,
+      "grad_norm": 2.0518600940704346,
+      "learning_rate": 4.266076132493008e-06,
+      "loss": 0.7828,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6958592891168927,
+      "grad_norm": 2.632521867752075,
+      "learning_rate": 4.256613809680446e-06,
+      "loss": 0.8256,
+      "step": 1899
+    },
+    {
+      "epoch": 0.696225723708318,
+      "grad_norm": 2.2324187755584717,
+      "learning_rate": 4.247159154288737e-06,
+      "loss": 0.8586,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6965921582997435,
+      "grad_norm": 2.3237431049346924,
+      "learning_rate": 4.2377121789398415e-06,
+      "loss": 0.7983,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6969585928911689,
+      "grad_norm": 2.55372953414917,
+      "learning_rate": 4.2282728962454724e-06,
+      "loss": 0.8691,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6973250274825944,
+      "grad_norm": 2.5332720279693604,
+      "learning_rate": 4.218841318807075e-06,
+      "loss": 0.7848,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6976914620740198,
+      "grad_norm": 2.3999102115631104,
+      "learning_rate": 4.209417459215804e-06,
+      "loss": 0.7941,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6980578966654453,
+      "grad_norm": 2.0995166301727295,
+      "learning_rate": 4.2000013300525186e-06,
+      "loss": 0.7986,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6984243312568706,
+      "grad_norm": 1.9330726861953735,
+      "learning_rate": 4.190592943887741e-06,
+      "loss": 0.8,
+      "step": 1906
+    },
+    {
+      "epoch": 0.698790765848296,
+      "grad_norm": 2.1339480876922607,
+      "learning_rate": 4.181192313281674e-06,
+      "loss": 0.8436,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6991572004397215,
+      "grad_norm": 2.6887545585632324,
+      "learning_rate": 4.171799450784158e-06,
+      "loss": 0.777,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6995236350311469,
+      "grad_norm": 2.4545319080352783,
+      "learning_rate": 4.1624143689346684e-06,
+      "loss": 0.7478,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6998900696225724,
+      "grad_norm": 2.2089483737945557,
+      "learning_rate": 4.153037080262288e-06,
+      "loss": 0.8183,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7002565042139978,
+      "grad_norm": 2.125802755355835,
+      "learning_rate": 4.143667597285692e-06,
+      "loss": 0.8096,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7006229388054233,
+      "grad_norm": 2.2986371517181396,
+      "learning_rate": 4.1343059325131456e-06,
+      "loss": 0.831,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7009893733968486,
+      "grad_norm": 2.3743975162506104,
+      "learning_rate": 4.1249520984424705e-06,
+      "loss": 0.8106,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7013558079882741,
+      "grad_norm": 2.1469645500183105,
+      "learning_rate": 4.115606107561038e-06,
+      "loss": 0.8783,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7017222425796995,
+      "grad_norm": 2.0734124183654785,
+      "learning_rate": 4.106267972345748e-06,
+      "loss": 0.8672,
+      "step": 1915
+    },
+    {
+      "epoch": 0.702088677171125,
+      "grad_norm": 2.2415788173675537,
+      "learning_rate": 4.096937705263004e-06,
+      "loss": 0.8532,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7024551117625504,
+      "grad_norm": 2.446101665496826,
+      "learning_rate": 4.087615318768718e-06,
+      "loss": 0.7584,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7028215463539759,
+      "grad_norm": 2.955641269683838,
+      "learning_rate": 4.0783008253082754e-06,
+      "loss": 0.7472,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7031879809454012,
+      "grad_norm": 1.996250033378601,
+      "learning_rate": 4.068994237316527e-06,
+      "loss": 0.7776,
+      "step": 1919
+    },
+    {
+      "epoch": 0.7035544155368266,
+      "grad_norm": 2.1396939754486084,
+      "learning_rate": 4.05969556721777e-06,
+      "loss": 0.8349,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7039208501282521,
+      "grad_norm": 2.2065043449401855,
+      "learning_rate": 4.050404827425726e-06,
+      "loss": 0.8736,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7042872847196775,
+      "grad_norm": 2.3151960372924805,
+      "learning_rate": 4.041122030343531e-06,
+      "loss": 0.7911,
+      "step": 1922
+    },
+    {
+      "epoch": 0.704653719311103,
+      "grad_norm": 2.4541115760803223,
+      "learning_rate": 4.031847188363721e-06,
+      "loss": 0.8091,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7050201539025284,
+      "grad_norm": 2.3490419387817383,
+      "learning_rate": 4.022580313868213e-06,
+      "loss": 0.8117,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7053865884939539,
+      "grad_norm": 2.6550164222717285,
+      "learning_rate": 4.013321419228282e-06,
+      "loss": 0.8709,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7057530230853792,
+      "grad_norm": 2.1409599781036377,
+      "learning_rate": 4.004070516804559e-06,
+      "loss": 0.8204,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7061194576768047,
+      "grad_norm": 2.460340738296509,
+      "learning_rate": 3.994827618946992e-06,
+      "loss": 0.7722,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7064858922682301,
+      "grad_norm": 2.2745752334594727,
+      "learning_rate": 3.985592737994851e-06,
+      "loss": 0.8428,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7068523268596556,
+      "grad_norm": 2.2874555587768555,
+      "learning_rate": 3.976365886276705e-06,
+      "loss": 0.8038,
+      "step": 1929
+    },
+    {
+      "epoch": 0.707218761451081,
+      "grad_norm": 3.0215723514556885,
+      "learning_rate": 3.9671470761104e-06,
+      "loss": 0.8059,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7075851960425064,
+      "grad_norm": 2.4965901374816895,
+      "learning_rate": 3.957936319803053e-06,
+      "loss": 0.805,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7079516306339318,
+      "grad_norm": 2.127297878265381,
+      "learning_rate": 3.948733629651017e-06,
+      "loss": 0.8134,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7083180652253572,
+      "grad_norm": 2.601527214050293,
+      "learning_rate": 3.9395390179398895e-06,
+      "loss": 0.7632,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7086844998167827,
+      "grad_norm": 2.1665985584259033,
+      "learning_rate": 3.930352496944471e-06,
+      "loss": 0.8787,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7090509344082081,
+      "grad_norm": 2.3517444133758545,
+      "learning_rate": 3.921174078928771e-06,
+      "loss": 0.8207,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7094173689996336,
+      "grad_norm": 2.1779472827911377,
+      "learning_rate": 3.912003776145983e-06,
+      "loss": 0.8215,
+      "step": 1936
+    },
+    {
+      "epoch": 0.709783803591059,
+      "grad_norm": 2.2063920497894287,
+      "learning_rate": 3.902841600838453e-06,
+      "loss": 0.8076,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7101502381824845,
+      "grad_norm": 2.0973567962646484,
+      "learning_rate": 3.8936875652376895e-06,
+      "loss": 0.8224,
+      "step": 1938
+    },
+    {
+      "epoch": 0.7105166727739098,
+      "grad_norm": 2.57674503326416,
+      "learning_rate": 3.88454168156433e-06,
+      "loss": 0.8262,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7108831073653353,
+      "grad_norm": 2.488333225250244,
+      "learning_rate": 3.87540396202813e-06,
+      "loss": 0.8515,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7112495419567607,
+      "grad_norm": 2.384260892868042,
+      "learning_rate": 3.86627441882795e-06,
+      "loss": 0.7949,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7116159765481862,
+      "grad_norm": 2.0849318504333496,
+      "learning_rate": 3.8571530641517215e-06,
+      "loss": 0.8088,
+      "step": 1942
+    },
+    {
+      "epoch": 0.7119824111396116,
+      "grad_norm": 2.252814769744873,
+      "learning_rate": 3.8480399101764585e-06,
+      "loss": 0.8484,
+      "step": 1943
+    },
+    {
+      "epoch": 0.712348845731037,
+      "grad_norm": 2.517176389694214,
+      "learning_rate": 3.838934969068227e-06,
+      "loss": 0.83,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7127152803224625,
+      "grad_norm": 2.161952495574951,
+      "learning_rate": 3.829838252982113e-06,
+      "loss": 0.7282,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7130817149138878,
+      "grad_norm": 2.1794629096984863,
+      "learning_rate": 3.8207497740622426e-06,
+      "loss": 0.7889,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7134481495053133,
+      "grad_norm": 2.5052850246429443,
+      "learning_rate": 3.8116695444417284e-06,
+      "loss": 0.8533,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7138145840967387,
+      "grad_norm": 2.8699727058410645,
+      "learning_rate": 3.8025975762426804e-06,
+      "loss": 0.7708,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7141810186881642,
+      "grad_norm": 2.2950704097747803,
+      "learning_rate": 3.7935338815761757e-06,
+      "loss": 0.7936,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7145474532795896,
+      "grad_norm": 2.1499178409576416,
+      "learning_rate": 3.784478472542249e-06,
+      "loss": 0.8088,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7149138878710151,
+      "grad_norm": 2.498230457305908,
+      "learning_rate": 3.775431361229871e-06,
+      "loss": 0.8589,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7152803224624404,
+      "grad_norm": 2.3784115314483643,
+      "learning_rate": 3.766392559716939e-06,
+      "loss": 0.7791,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7156467570538659,
+      "grad_norm": 2.2865967750549316,
+      "learning_rate": 3.757362080070247e-06,
+      "loss": 0.8398,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7160131916452913,
+      "grad_norm": 2.2455546855926514,
+      "learning_rate": 3.7483399343454886e-06,
+      "loss": 0.7977,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7163796262367167,
+      "grad_norm": 2.426880121231079,
+      "learning_rate": 3.7393261345872323e-06,
+      "loss": 0.8292,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7167460608281422,
+      "grad_norm": 2.6572940349578857,
+      "learning_rate": 3.7303206928288948e-06,
+      "loss": 0.7638,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7171124954195676,
+      "grad_norm": 2.630995512008667,
+      "learning_rate": 3.7213236210927483e-06,
+      "loss": 0.7029,
+      "step": 1957
+    },
+    {
+      "epoch": 0.717478930010993,
+      "grad_norm": 2.7464239597320557,
+      "learning_rate": 3.712334931389877e-06,
+      "loss": 0.7451,
+      "step": 1958
+    },
+    {
+      "epoch": 0.7178453646024184,
+      "grad_norm": 2.4412245750427246,
+      "learning_rate": 3.7033546357201845e-06,
+      "loss": 0.8073,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7182117991938439,
+      "grad_norm": 2.5375096797943115,
+      "learning_rate": 3.6943827460723667e-06,
+      "loss": 0.8037,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7185782337852693,
+      "grad_norm": 2.4279673099517822,
+      "learning_rate": 3.685419274423898e-06,
+      "loss": 0.8134,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7189446683766948,
+      "grad_norm": 2.2107279300689697,
+      "learning_rate": 3.676464232741015e-06,
+      "loss": 0.8346,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7193111029681202,
+      "grad_norm": 2.2736778259277344,
+      "learning_rate": 3.6675176329786933e-06,
+      "loss": 0.8344,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7196775375595457,
+      "grad_norm": 2.9939663410186768,
+      "learning_rate": 3.658579487080646e-06,
+      "loss": 0.7851,
+      "step": 1964
+    },
+    {
+      "epoch": 0.720043972150971,
+      "grad_norm": 2.447913408279419,
+      "learning_rate": 3.6496498069792995e-06,
+      "loss": 0.7574,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7204104067423965,
+      "grad_norm": 2.029849052429199,
+      "learning_rate": 3.6407286045957802e-06,
+      "loss": 0.8999,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7207768413338219,
+      "grad_norm": 2.375657081604004,
+      "learning_rate": 3.6318158918398872e-06,
+      "loss": 0.8329,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7211432759252473,
+      "grad_norm": 2.311725378036499,
+      "learning_rate": 3.6229116806100993e-06,
+      "loss": 0.8484,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7215097105166728,
+      "grad_norm": 2.1953041553497314,
+      "learning_rate": 3.6140159827935316e-06,
+      "loss": 0.8258,
+      "step": 1969
+    },
+    {
+      "epoch": 0.7218761451080982,
+      "grad_norm": 2.110595464706421,
+      "learning_rate": 3.6051288102659456e-06,
+      "loss": 0.8319,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7222425796995237,
+      "grad_norm": 2.6473276615142822,
+      "learning_rate": 3.5962501748917157e-06,
+      "loss": 0.7727,
+      "step": 1971
+    },
+    {
+      "epoch": 0.722609014290949,
+      "grad_norm": 2.762803316116333,
+      "learning_rate": 3.587380088523824e-06,
+      "loss": 0.7723,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7229754488823745,
+      "grad_norm": 2.2459194660186768,
+      "learning_rate": 3.57851856300383e-06,
+      "loss": 0.8856,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7233418834737999,
+      "grad_norm": 2.0015933513641357,
+      "learning_rate": 3.5696656101618755e-06,
+      "loss": 0.8034,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7237083180652254,
+      "grad_norm": 2.0197722911834717,
+      "learning_rate": 3.56082124181665e-06,
+      "loss": 0.8149,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7240747526566508,
+      "grad_norm": 2.4105708599090576,
+      "learning_rate": 3.5519854697753885e-06,
+      "loss": 0.856,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7244411872480763,
+      "grad_norm": 2.091339111328125,
+      "learning_rate": 3.543158305833848e-06,
+      "loss": 0.7757,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7248076218395016,
+      "grad_norm": 2.361039638519287,
+      "learning_rate": 3.5343397617762887e-06,
+      "loss": 0.7726,
+      "step": 1978
+    },
+    {
+      "epoch": 0.725174056430927,
+      "grad_norm": 2.58827805519104,
+      "learning_rate": 3.525529849375472e-06,
+      "loss": 0.8109,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7255404910223525,
+      "grad_norm": 2.7834389209747314,
+      "learning_rate": 3.5167285803926252e-06,
+      "loss": 0.7179,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7259069256137779,
+      "grad_norm": 2.9694321155548096,
+      "learning_rate": 3.507935966577447e-06,
+      "loss": 0.7565,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7262733602052034,
+      "grad_norm": 2.681018114089966,
+      "learning_rate": 3.499152019668076e-06,
+      "loss": 0.7807,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7266397947966288,
+      "grad_norm": 2.498192071914673,
+      "learning_rate": 3.490376751391086e-06,
+      "loss": 0.7333,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7270062293880543,
+      "grad_norm": 2.3323285579681396,
+      "learning_rate": 3.4816101734614536e-06,
+      "loss": 0.7846,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7273726639794796,
+      "grad_norm": 2.4998133182525635,
+      "learning_rate": 3.472852297582565e-06,
+      "loss": 0.8177,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7277390985709051,
+      "grad_norm": 2.852997303009033,
+      "learning_rate": 3.464103135446183e-06,
+      "loss": 0.7623,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7281055331623305,
+      "grad_norm": 2.3916666507720947,
+      "learning_rate": 3.455362698732441e-06,
+      "loss": 0.7801,
+      "step": 1987
+    },
+    {
+      "epoch": 0.728471967753756,
+      "grad_norm": 2.727652072906494,
+      "learning_rate": 3.4466309991098255e-06,
+      "loss": 0.8518,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7288384023451814,
+      "grad_norm": 2.4952447414398193,
+      "learning_rate": 3.437908048235149e-06,
+      "loss": 0.7323,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7292048369366069,
+      "grad_norm": 2.348653554916382,
+      "learning_rate": 3.429193857753559e-06,
+      "loss": 0.8303,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7295712715280323,
+      "grad_norm": 2.3010904788970947,
+      "learning_rate": 3.4204884392984917e-06,
+      "loss": 0.8744,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7299377061194576,
+      "grad_norm": 2.9672434329986572,
+      "learning_rate": 3.4117918044916853e-06,
+      "loss": 0.6981,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7303041407108831,
+      "grad_norm": 2.3051466941833496,
+      "learning_rate": 3.4031039649431518e-06,
+      "loss": 0.8161,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7306705753023085,
+      "grad_norm": 2.262472629547119,
+      "learning_rate": 3.3944249322511502e-06,
+      "loss": 0.8048,
+      "step": 1994
+    },
+    {
+      "epoch": 0.731037009893734,
+      "grad_norm": 2.3546090126037598,
+      "learning_rate": 3.385754718002193e-06,
+      "loss": 0.8214,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7314034444851594,
+      "grad_norm": 2.283719301223755,
+      "learning_rate": 3.3770933337710165e-06,
+      "loss": 0.8198,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7317698790765849,
+      "grad_norm": 2.383819818496704,
+      "learning_rate": 3.3684407911205697e-06,
+      "loss": 0.8661,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7321363136680102,
+      "grad_norm": 2.4723551273345947,
+      "learning_rate": 3.3597971016019994e-06,
+      "loss": 0.8011,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7325027482594357,
+      "grad_norm": 2.1126275062561035,
+      "learning_rate": 3.351162276754626e-06,
+      "loss": 0.8016,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7328691828508611,
+      "grad_norm": 2.984003782272339,
+      "learning_rate": 3.342536328105944e-06,
+      "loss": 0.8045,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7332356174422866,
+      "grad_norm": 2.3070881366729736,
+      "learning_rate": 3.3339192671715992e-06,
+      "loss": 0.7834,
+      "step": 2001
+    },
+    {
+      "epoch": 0.733602052033712,
+      "grad_norm": 2.393101453781128,
+      "learning_rate": 3.325311105455361e-06,
+      "loss": 0.8523,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7339684866251374,
+      "grad_norm": 2.4970571994781494,
+      "learning_rate": 3.316711854449133e-06,
+      "loss": 0.7961,
+      "step": 2003
+    },
+    {
+      "epoch": 0.7343349212165629,
+      "grad_norm": 2.070354461669922,
+      "learning_rate": 3.308121525632908e-06,
+      "loss": 0.8765,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7347013558079882,
+      "grad_norm": 2.0780506134033203,
+      "learning_rate": 3.299540130474781e-06,
+      "loss": 0.8653,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7350677903994137,
+      "grad_norm": 2.2514584064483643,
+      "learning_rate": 3.290967680430912e-06,
+      "loss": 0.7971,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7354342249908391,
+      "grad_norm": 2.65456223487854,
+      "learning_rate": 3.282404186945525e-06,
+      "loss": 0.8195,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7358006595822646,
+      "grad_norm": 2.9336471557617188,
+      "learning_rate": 3.273849661450884e-06,
+      "loss": 0.7254,
+      "step": 2008
+    },
+    {
+      "epoch": 0.73616709417369,
+      "grad_norm": 2.372671604156494,
+      "learning_rate": 3.2653041153672782e-06,
+      "loss": 0.8261,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7365335287651155,
+      "grad_norm": 2.256636381149292,
+      "learning_rate": 3.2567675601030136e-06,
+      "loss": 0.8386,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7368999633565408,
+      "grad_norm": 2.8722164630889893,
+      "learning_rate": 3.248240007054392e-06,
+      "loss": 0.7729,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7372663979479663,
+      "grad_norm": 2.3636233806610107,
+      "learning_rate": 3.2397214676057e-06,
+      "loss": 0.835,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7376328325393917,
+      "grad_norm": 2.246462106704712,
+      "learning_rate": 3.231211953129183e-06,
+      "loss": 0.8031,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7379992671308171,
+      "grad_norm": 2.266885757446289,
+      "learning_rate": 3.2227114749850476e-06,
+      "loss": 0.856,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7383657017222426,
+      "grad_norm": 2.540126085281372,
+      "learning_rate": 3.214220044521428e-06,
+      "loss": 0.8486,
+      "step": 2015
+    },
+    {
+      "epoch": 0.738732136313668,
+      "grad_norm": 2.3920586109161377,
+      "learning_rate": 3.205737673074387e-06,
+      "loss": 0.8452,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7390985709050935,
+      "grad_norm": 2.3333163261413574,
+      "learning_rate": 3.1972643719678896e-06,
+      "loss": 0.8147,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7394650054965188,
+      "grad_norm": 2.166067600250244,
+      "learning_rate": 3.1888001525137958e-06,
+      "loss": 0.801,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7398314400879443,
+      "grad_norm": 2.6589584350585938,
+      "learning_rate": 3.1803450260118397e-06,
+      "loss": 0.7791,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7401978746793697,
+      "grad_norm": 2.284156560897827,
+      "learning_rate": 3.171899003749611e-06,
+      "loss": 0.8275,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7405643092707952,
+      "grad_norm": 2.214045286178589,
+      "learning_rate": 3.163462097002552e-06,
+      "loss": 0.7909,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7409307438622206,
+      "grad_norm": 2.4925358295440674,
+      "learning_rate": 3.1550343170339336e-06,
+      "loss": 0.6837,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7412971784536461,
+      "grad_norm": 2.5517702102661133,
+      "learning_rate": 3.1466156750948464e-06,
+      "loss": 0.7163,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7416636130450714,
+      "grad_norm": 2.5317089557647705,
+      "learning_rate": 3.1382061824241704e-06,
+      "loss": 0.8028,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7420300476364969,
+      "grad_norm": 2.2435755729675293,
+      "learning_rate": 3.1298058502485863e-06,
+      "loss": 0.8738,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7423964822279223,
+      "grad_norm": 2.4333622455596924,
+      "learning_rate": 3.1214146897825302e-06,
+      "loss": 0.882,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7427629168193477,
+      "grad_norm": 2.3265597820281982,
+      "learning_rate": 3.1130327122282065e-06,
+      "loss": 0.82,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7431293514107732,
+      "grad_norm": 1.9941530227661133,
+      "learning_rate": 3.1046599287755552e-06,
+      "loss": 0.4572,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7434957860021986,
+      "grad_norm": 2.354604482650757,
+      "learning_rate": 3.0962963506022447e-06,
+      "loss": 0.7864,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7438622205936241,
+      "grad_norm": 2.49794864654541,
+      "learning_rate": 3.0879419888736463e-06,
+      "loss": 0.7339,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7442286551850494,
+      "grad_norm": 2.2387988567352295,
+      "learning_rate": 3.0795968547428377e-06,
+      "loss": 0.7992,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7445950897764749,
+      "grad_norm": 2.4889345169067383,
+      "learning_rate": 3.0712609593505726e-06,
+      "loss": 0.8163,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7449615243679003,
+      "grad_norm": 2.3034846782684326,
+      "learning_rate": 3.0629343138252708e-06,
+      "loss": 0.8405,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7453279589593258,
+      "grad_norm": 2.2705564498901367,
+      "learning_rate": 3.0546169292830096e-06,
+      "loss": 0.8507,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7456943935507512,
+      "grad_norm": 2.5430757999420166,
+      "learning_rate": 3.04630881682749e-06,
+      "loss": 0.8611,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7460608281421767,
+      "grad_norm": 2.391397714614868,
+      "learning_rate": 3.0380099875500504e-06,
+      "loss": 0.8666,
+      "step": 2036
+    },
+    {
+      "epoch": 0.746427262733602,
+      "grad_norm": 3.0161325931549072,
+      "learning_rate": 3.0297204525296197e-06,
+      "loss": 0.7997,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7467936973250274,
+      "grad_norm": 2.164350748062134,
+      "learning_rate": 3.021440222832732e-06,
+      "loss": 0.8054,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7471601319164529,
+      "grad_norm": 2.2333128452301025,
+      "learning_rate": 3.0131693095134982e-06,
+      "loss": 0.8469,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7475265665078783,
+      "grad_norm": 2.2303738594055176,
+      "learning_rate": 3.0049077236135805e-06,
+      "loss": 0.803,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7478930010993038,
+      "grad_norm": 2.561152935028076,
+      "learning_rate": 2.9966554761622003e-06,
+      "loss": 0.7331,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7482594356907292,
+      "grad_norm": 2.60172700881958,
+      "learning_rate": 2.9884125781761076e-06,
+      "loss": 0.769,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7486258702821547,
+      "grad_norm": 2.1624298095703125,
+      "learning_rate": 2.980179040659572e-06,
+      "loss": 0.798,
+      "step": 2043
+    },
+    {
+      "epoch": 0.74899230487358,
+      "grad_norm": 2.607041835784912,
+      "learning_rate": 2.971954874604367e-06,
+      "loss": 0.8,
+      "step": 2044
+    },
+    {
+      "epoch": 0.7493587394650055,
+      "grad_norm": 2.153707504272461,
+      "learning_rate": 2.963740090989757e-06,
+      "loss": 0.8808,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7497251740564309,
+      "grad_norm": 3.1826159954071045,
+      "learning_rate": 2.9555347007824742e-06,
+      "loss": 0.8187,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7500916086478564,
+      "grad_norm": 2.341423273086548,
+      "learning_rate": 2.9473387149367182e-06,
+      "loss": 0.7878,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7504580432392818,
+      "grad_norm": 2.590043067932129,
+      "learning_rate": 2.9391521443941263e-06,
+      "loss": 0.8031,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7508244778307073,
+      "grad_norm": 2.0318918228149414,
+      "learning_rate": 2.930975000083773e-06,
+      "loss": 0.7952,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7508244778307073,
+      "eval_loss": 0.8080984354019165,
+      "eval_runtime": 799.0988,
+      "eval_samples_per_second": 3.394,
+      "eval_steps_per_second": 0.424,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7511909124221327,
+      "grad_norm": 2.6090524196624756,
+      "learning_rate": 2.9228072929221495e-06,
+      "loss": 0.7518,
+      "step": 2050
+    },
+    {
+      "epoch": 0.751557347013558,
+      "grad_norm": 2.003688335418701,
+      "learning_rate": 2.9146490338131405e-06,
+      "loss": 0.8436,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7519237816049835,
+      "grad_norm": 2.512913465499878,
+      "learning_rate": 2.906500233648023e-06,
+      "loss": 0.8456,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7522902161964089,
+      "grad_norm": 2.6392178535461426,
+      "learning_rate": 2.898360903305447e-06,
+      "loss": 0.7732,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7526566507878344,
+      "grad_norm": 2.2465503215789795,
+      "learning_rate": 2.890231053651419e-06,
+      "loss": 0.8681,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7530230853792598,
+      "grad_norm": 2.790282964706421,
+      "learning_rate": 2.882110695539291e-06,
+      "loss": 0.8697,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7533895199706853,
+      "grad_norm": 2.172825336456299,
+      "learning_rate": 2.873999839809736e-06,
+      "loss": 0.7808,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7537559545621106,
+      "grad_norm": 2.74893856048584,
+      "learning_rate": 2.865898497290749e-06,
+      "loss": 0.7675,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7541223891535361,
+      "grad_norm": 2.2950291633605957,
+      "learning_rate": 2.8578066787976266e-06,
+      "loss": 0.8353,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7544888237449615,
+      "grad_norm": 2.3367483615875244,
+      "learning_rate": 2.849724395132941e-06,
+      "loss": 0.8456,
+      "step": 2059
+    },
+    {
+      "epoch": 0.754855258336387,
+      "grad_norm": 2.3577630519866943,
+      "learning_rate": 2.841651657086547e-06,
+      "loss": 0.8345,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7552216929278124,
+      "grad_norm": 2.4616196155548096,
+      "learning_rate": 2.8335884754355436e-06,
+      "loss": 0.752,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7555881275192378,
+      "grad_norm": 2.157444953918457,
+      "learning_rate": 2.8255348609442835e-06,
+      "loss": 0.8047,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7559545621106633,
+      "grad_norm": 2.123469352722168,
+      "learning_rate": 2.8174908243643405e-06,
+      "loss": 0.8698,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7563209967020886,
+      "grad_norm": 2.120635986328125,
+      "learning_rate": 2.809456376434504e-06,
+      "loss": 0.815,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7566874312935141,
+      "grad_norm": 2.5298824310302734,
+      "learning_rate": 2.801431527880767e-06,
+      "loss": 0.8078,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7570538658849395,
+      "grad_norm": 2.6083366870880127,
+      "learning_rate": 2.7934162894162954e-06,
+      "loss": 0.7339,
+      "step": 2066
+    },
+    {
+      "epoch": 0.757420300476365,
+      "grad_norm": 2.328028678894043,
+      "learning_rate": 2.7854106717414354e-06,
+      "loss": 0.8201,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7577867350677904,
+      "grad_norm": 2.296004056930542,
+      "learning_rate": 2.777414685543689e-06,
+      "loss": 0.8521,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7581531696592159,
+      "grad_norm": 2.2588114738464355,
+      "learning_rate": 2.7694283414976995e-06,
+      "loss": 0.8129,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7585196042506412,
+      "grad_norm": 2.3270256519317627,
+      "learning_rate": 2.761451650265231e-06,
+      "loss": 0.8647,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7588860388420667,
+      "grad_norm": 2.4873387813568115,
+      "learning_rate": 2.753484622495173e-06,
+      "loss": 0.785,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7592524734334921,
+      "grad_norm": 2.892294406890869,
+      "learning_rate": 2.745527268823501e-06,
+      "loss": 0.7487,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7596189080249176,
+      "grad_norm": 2.315394163131714,
+      "learning_rate": 2.737579599873288e-06,
+      "loss": 0.8463,
+      "step": 2073
+    },
+    {
+      "epoch": 0.759985342616343,
+      "grad_norm": 2.721442699432373,
+      "learning_rate": 2.72964162625467e-06,
+      "loss": 0.7053,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7603517772077684,
+      "grad_norm": 2.559688091278076,
+      "learning_rate": 2.7217133585648427e-06,
+      "loss": 0.8889,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7607182117991939,
+      "grad_norm": 2.5306828022003174,
+      "learning_rate": 2.713794807388047e-06,
+      "loss": 0.8966,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7610846463906192,
+      "grad_norm": 2.15157151222229,
+      "learning_rate": 2.7058859832955432e-06,
+      "loss": 0.8052,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7614510809820447,
+      "grad_norm": 2.3579909801483154,
+      "learning_rate": 2.6979868968456147e-06,
+      "loss": 0.8203,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7618175155734701,
+      "grad_norm": 2.602248430252075,
+      "learning_rate": 2.690097558583543e-06,
+      "loss": 0.7435,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7621839501648956,
+      "grad_norm": 2.3024191856384277,
+      "learning_rate": 2.6822179790415982e-06,
+      "loss": 0.7994,
+      "step": 2080
+    },
+    {
+      "epoch": 0.762550384756321,
+      "grad_norm": 2.410738229751587,
+      "learning_rate": 2.6743481687390115e-06,
+      "loss": 0.8174,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7629168193477465,
+      "grad_norm": 2.2351765632629395,
+      "learning_rate": 2.6664881381819873e-06,
+      "loss": 0.8043,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7632832539391718,
+      "grad_norm": 2.340183973312378,
+      "learning_rate": 2.658637897863662e-06,
+      "loss": 0.8373,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7636496885305973,
+      "grad_norm": 2.4249885082244873,
+      "learning_rate": 2.650797458264107e-06,
+      "loss": 0.7883,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7640161231220227,
+      "grad_norm": 2.5286998748779297,
+      "learning_rate": 2.642966829850312e-06,
+      "loss": 0.8601,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7643825577134481,
+      "grad_norm": 2.892470359802246,
+      "learning_rate": 2.6351460230761683e-06,
+      "loss": 0.7755,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7647489923048736,
+      "grad_norm": 2.4626762866973877,
+      "learning_rate": 2.627335048382448e-06,
+      "loss": 0.7893,
+      "step": 2087
+    },
+    {
+      "epoch": 0.765115426896299,
+      "grad_norm": 2.1419596672058105,
+      "learning_rate": 2.6195339161968046e-06,
+      "loss": 0.793,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7654818614877245,
+      "grad_norm": 2.3152785301208496,
+      "learning_rate": 2.6117426369337505e-06,
+      "loss": 0.8636,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7658482960791498,
+      "grad_norm": 2.5008513927459717,
+      "learning_rate": 2.603961220994644e-06,
+      "loss": 0.8222,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7662147306705753,
+      "grad_norm": 2.367537021636963,
+      "learning_rate": 2.596189678767679e-06,
+      "loss": 0.7907,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7665811652620007,
+      "grad_norm": 2.2729032039642334,
+      "learning_rate": 2.5884280206278577e-06,
+      "loss": 0.8255,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7669475998534262,
+      "grad_norm": 2.4431867599487305,
+      "learning_rate": 2.5806762569370014e-06,
+      "loss": 0.8168,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7673140344448516,
+      "grad_norm": 2.766500234603882,
+      "learning_rate": 2.5729343980437095e-06,
+      "loss": 0.687,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7676804690362771,
+      "grad_norm": 2.3301477432250977,
+      "learning_rate": 2.565202454283364e-06,
+      "loss": 0.8608,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7680469036277024,
+      "grad_norm": 2.6721205711364746,
+      "learning_rate": 2.5574804359781167e-06,
+      "loss": 0.8265,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7684133382191279,
+      "grad_norm": 2.609799861907959,
+      "learning_rate": 2.5497683534368543e-06,
+      "loss": 0.834,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7687797728105533,
+      "grad_norm": 2.4966788291931152,
+      "learning_rate": 2.5420662169552114e-06,
+      "loss": 0.7786,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7691462074019787,
+      "grad_norm": 2.085581064224243,
+      "learning_rate": 2.53437403681554e-06,
+      "loss": 0.8449,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7695126419934042,
+      "grad_norm": 2.4107308387756348,
+      "learning_rate": 2.526691823286902e-06,
+      "loss": 0.8249,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7698790765848296,
+      "grad_norm": 2.4472155570983887,
+      "learning_rate": 2.5190195866250545e-06,
+      "loss": 0.7904,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7702455111762551,
+      "grad_norm": 2.776448965072632,
+      "learning_rate": 2.5113573370724276e-06,
+      "loss": 0.7221,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7706119457676804,
+      "grad_norm": 2.283174753189087,
+      "learning_rate": 2.503705084858129e-06,
+      "loss": 0.8141,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7709783803591059,
+      "grad_norm": 2.382694959640503,
+      "learning_rate": 2.496062840197918e-06,
+      "loss": 0.804,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7713448149505313,
+      "grad_norm": 2.3865418434143066,
+      "learning_rate": 2.488430613294187e-06,
+      "loss": 0.7391,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7717112495419568,
+      "grad_norm": 2.399531602859497,
+      "learning_rate": 2.4808084143359602e-06,
+      "loss": 0.7877,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7720776841333822,
+      "grad_norm": 2.91961932182312,
+      "learning_rate": 2.4731962534988784e-06,
+      "loss": 0.7119,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7724441187248077,
+      "grad_norm": 2.2047502994537354,
+      "learning_rate": 2.465594140945169e-06,
+      "loss": 0.8487,
+      "step": 2108
+    },
+    {
+      "epoch": 0.772810553316233,
+      "grad_norm": 3.028589963912964,
+      "learning_rate": 2.4580020868236575e-06,
+      "loss": 0.799,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7731769879076584,
+      "grad_norm": 2.747431993484497,
+      "learning_rate": 2.450420101269735e-06,
+      "loss": 0.7931,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7735434224990839,
+      "grad_norm": 2.3259732723236084,
+      "learning_rate": 2.442848194405354e-06,
+      "loss": 0.7772,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7739098570905093,
+      "grad_norm": 2.3536782264709473,
+      "learning_rate": 2.435286376339012e-06,
+      "loss": 0.7929,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7742762916819348,
+      "grad_norm": 2.4395339488983154,
+      "learning_rate": 2.427734657165731e-06,
+      "loss": 0.8416,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7746427262733602,
+      "grad_norm": 2.406297445297241,
+      "learning_rate": 2.420193046967061e-06,
+      "loss": 0.811,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7750091608647857,
+      "grad_norm": 2.6960110664367676,
+      "learning_rate": 2.412661555811052e-06,
+      "loss": 0.8142,
+      "step": 2115
+    },
+    {
+      "epoch": 0.775375595456211,
+      "grad_norm": 2.3994743824005127,
+      "learning_rate": 2.4051401937522424e-06,
+      "loss": 0.759,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7757420300476365,
+      "grad_norm": 2.3011417388916016,
+      "learning_rate": 2.3976289708316538e-06,
+      "loss": 0.7958,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7761084646390619,
+      "grad_norm": 2.311936378479004,
+      "learning_rate": 2.390127897076765e-06,
+      "loss": 0.813,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7764748992304874,
+      "grad_norm": 2.259918689727783,
+      "learning_rate": 2.382636982501513e-06,
+      "loss": 0.8316,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7768413338219128,
+      "grad_norm": 2.676295042037964,
+      "learning_rate": 2.3751562371062684e-06,
+      "loss": 0.82,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7772077684133383,
+      "grad_norm": 2.99137020111084,
+      "learning_rate": 2.3676856708778274e-06,
+      "loss": 0.7577,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7775742030047637,
+      "grad_norm": 2.3418562412261963,
+      "learning_rate": 2.3602252937893987e-06,
+      "loss": 0.8698,
+      "step": 2122
+    },
+    {
+      "epoch": 0.777940637596189,
+      "grad_norm": 2.545377492904663,
+      "learning_rate": 2.3527751158005797e-06,
+      "loss": 0.7925,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7783070721876145,
+      "grad_norm": 2.7940926551818848,
+      "learning_rate": 2.3453351468573617e-06,
+      "loss": 0.7702,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7786735067790399,
+      "grad_norm": 2.804811477661133,
+      "learning_rate": 2.3379053968921038e-06,
+      "loss": 0.7383,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7790399413704654,
+      "grad_norm": 2.2485673427581787,
+      "learning_rate": 2.330485875823526e-06,
+      "loss": 0.7616,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7794063759618908,
+      "grad_norm": 2.019737482070923,
+      "learning_rate": 2.3230765935566823e-06,
+      "loss": 0.8021,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7797728105533163,
+      "grad_norm": 1.9526029825210571,
+      "learning_rate": 2.315677559982973e-06,
+      "loss": 0.8219,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7801392451447416,
+      "grad_norm": 2.3496994972229004,
+      "learning_rate": 2.3082887849800984e-06,
+      "loss": 0.7756,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7805056797361671,
+      "grad_norm": 2.2115635871887207,
+      "learning_rate": 2.3009102784120808e-06,
+      "loss": 0.8192,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7808721143275925,
+      "grad_norm": 2.2814576625823975,
+      "learning_rate": 2.2935420501292238e-06,
+      "loss": 0.8423,
+      "step": 2131
+    },
+    {
+      "epoch": 0.781238548919018,
+      "grad_norm": 2.196955442428589,
+      "learning_rate": 2.286184109968117e-06,
+      "loss": 0.8066,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7816049835104434,
+      "grad_norm": 2.391277551651001,
+      "learning_rate": 2.278836467751604e-06,
+      "loss": 0.7948,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7819714181018688,
+      "grad_norm": 3.0919301509857178,
+      "learning_rate": 2.2714991332887916e-06,
+      "loss": 0.72,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7823378526932943,
+      "grad_norm": 2.3225972652435303,
+      "learning_rate": 2.264172116375021e-06,
+      "loss": 0.8308,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7827042872847196,
+      "grad_norm": 2.4777140617370605,
+      "learning_rate": 2.2568554267918596e-06,
+      "loss": 0.8269,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7830707218761451,
+      "grad_norm": 2.737854480743408,
+      "learning_rate": 2.2495490743070915e-06,
+      "loss": 0.8036,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7834371564675705,
+      "grad_norm": 2.548006534576416,
+      "learning_rate": 2.2422530686746936e-06,
+      "loss": 0.827,
+      "step": 2138
+    },
+    {
+      "epoch": 0.783803591058996,
+      "grad_norm": 2.565333604812622,
+      "learning_rate": 2.234967419634837e-06,
+      "loss": 0.8095,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7841700256504214,
+      "grad_norm": 2.1336276531219482,
+      "learning_rate": 2.2276921369138593e-06,
+      "loss": 0.8294,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7845364602418469,
+      "grad_norm": 2.151606321334839,
+      "learning_rate": 2.2204272302242658e-06,
+      "loss": 0.8249,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7849028948332722,
+      "grad_norm": 2.794053792953491,
+      "learning_rate": 2.2131727092647082e-06,
+      "loss": 0.7268,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7852693294246977,
+      "grad_norm": 2.621978998184204,
+      "learning_rate": 2.205928583719973e-06,
+      "loss": 0.7889,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7856357640161231,
+      "grad_norm": 2.2475435733795166,
+      "learning_rate": 2.198694863260964e-06,
+      "loss": 0.8576,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7860021986075486,
+      "grad_norm": 2.1243855953216553,
+      "learning_rate": 2.191471557544701e-06,
+      "loss": 0.7949,
+      "step": 2145
+    },
+    {
+      "epoch": 0.786368633198974,
+      "grad_norm": 2.1951136589050293,
+      "learning_rate": 2.184258676214298e-06,
+      "loss": 0.7939,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7867350677903994,
+      "grad_norm": 2.225924491882324,
+      "learning_rate": 2.1770562288989517e-06,
+      "loss": 0.8111,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7871015023818249,
+      "grad_norm": 2.1376214027404785,
+      "learning_rate": 2.169864225213931e-06,
+      "loss": 0.784,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7874679369732502,
+      "grad_norm": 3.0692696571350098,
+      "learning_rate": 2.1626826747605578e-06,
+      "loss": 0.7657,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7878343715646757,
+      "grad_norm": 2.149909257888794,
+      "learning_rate": 2.1555115871262055e-06,
+      "loss": 0.8242,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7882008061561011,
+      "grad_norm": 2.780365228652954,
+      "learning_rate": 2.1483509718842734e-06,
+      "loss": 0.7929,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7885672407475266,
+      "grad_norm": 2.1888341903686523,
+      "learning_rate": 2.141200838594184e-06,
+      "loss": 0.7964,
+      "step": 2152
+    },
+    {
+      "epoch": 0.788933675338952,
+      "grad_norm": 2.308290719985962,
+      "learning_rate": 2.1340611968013693e-06,
+      "loss": 0.7618,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7893001099303775,
+      "grad_norm": 2.4088211059570312,
+      "learning_rate": 2.1269320560372463e-06,
+      "loss": 0.831,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7896665445218028,
+      "grad_norm": 2.241452217102051,
+      "learning_rate": 2.1198134258192193e-06,
+      "loss": 0.8213,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7900329791132283,
+      "grad_norm": 2.262209177017212,
+      "learning_rate": 2.1127053156506628e-06,
+      "loss": 0.8348,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7903994137046537,
+      "grad_norm": 2.7259719371795654,
+      "learning_rate": 2.1056077350209014e-06,
+      "loss": 0.7151,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7907658482960791,
+      "grad_norm": 2.6229419708251953,
+      "learning_rate": 2.0985206934052094e-06,
+      "loss": 0.8081,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7911322828875046,
+      "grad_norm": 2.408729314804077,
+      "learning_rate": 2.091444200264784e-06,
+      "loss": 0.7861,
+      "step": 2159
+    },
+    {
+      "epoch": 0.79149871747893,
+      "grad_norm": 2.3548901081085205,
+      "learning_rate": 2.0843782650467437e-06,
+      "loss": 0.874,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7918651520703555,
+      "grad_norm": 2.125981092453003,
+      "learning_rate": 2.0773228971841165e-06,
+      "loss": 0.7882,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7922315866617808,
+      "grad_norm": 2.5681943893432617,
+      "learning_rate": 2.0702781060958133e-06,
+      "loss": 0.764,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7925980212532063,
+      "grad_norm": 3.0221588611602783,
+      "learning_rate": 2.0632439011866356e-06,
+      "loss": 0.7845,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7929644558446317,
+      "grad_norm": 2.767496109008789,
+      "learning_rate": 2.056220291847243e-06,
+      "loss": 0.8026,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7933308904360572,
+      "grad_norm": 2.1552770137786865,
+      "learning_rate": 2.049207287454156e-06,
+      "loss": 0.8342,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7936973250274826,
+      "grad_norm": 2.2508256435394287,
+      "learning_rate": 2.0422048973697362e-06,
+      "loss": 0.8133,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7940637596189081,
+      "grad_norm": 2.6234443187713623,
+      "learning_rate": 2.035213130942175e-06,
+      "loss": 0.8476,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7944301942103335,
+      "grad_norm": 2.3684444427490234,
+      "learning_rate": 2.02823199750548e-06,
+      "loss": 0.8084,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7947966288017588,
+      "grad_norm": 2.1743626594543457,
+      "learning_rate": 2.021261506379468e-06,
+      "loss": 0.8631,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7951630633931843,
+      "grad_norm": 2.5163462162017822,
+      "learning_rate": 2.01430166686974e-06,
+      "loss": 0.7819,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7955294979846097,
+      "grad_norm": 2.3898160457611084,
+      "learning_rate": 2.007352488267684e-06,
+      "loss": 0.7564,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7958959325760352,
+      "grad_norm": 2.44692063331604,
+      "learning_rate": 2.000413979850456e-06,
+      "loss": 0.8454,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7962623671674606,
+      "grad_norm": 2.379611015319824,
+      "learning_rate": 1.993486150880962e-06,
+      "loss": 0.8405,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7966288017588861,
+      "grad_norm": 2.6301681995391846,
+      "learning_rate": 1.9865690106078573e-06,
+      "loss": 0.7633,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7969952363503114,
+      "grad_norm": 3.5861284732818604,
+      "learning_rate": 1.9796625682655203e-06,
+      "loss": 0.7662,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7973616709417369,
+      "grad_norm": 2.236845016479492,
+      "learning_rate": 1.9727668330740525e-06,
+      "loss": 0.8066,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7977281055331623,
+      "grad_norm": 2.5206212997436523,
+      "learning_rate": 1.965881814239263e-06,
+      "loss": 0.8183,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7980945401245878,
+      "grad_norm": 2.225001811981201,
+      "learning_rate": 1.959007520952653e-06,
+      "loss": 0.8193,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7984609747160132,
+      "grad_norm": 2.5378408432006836,
+      "learning_rate": 1.952143962391403e-06,
+      "loss": 0.7865,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7988274093074387,
+      "grad_norm": 2.249823570251465,
+      "learning_rate": 1.9452911477183623e-06,
+      "loss": 0.7653,
+      "step": 2180
+    },
+    {
+      "epoch": 0.799193843898864,
+      "grad_norm": 2.4021379947662354,
+      "learning_rate": 1.9384490860820414e-06,
+      "loss": 0.8034,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7995602784902894,
+      "grad_norm": 2.3179380893707275,
+      "learning_rate": 1.9316177866165906e-06,
+      "loss": 0.7761,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7999267130817149,
+      "grad_norm": 2.412019729614258,
+      "learning_rate": 1.9247972584417973e-06,
+      "loss": 0.8295,
+      "step": 2183
+    },
+    {
+      "epoch": 0.8002931476731403,
+      "grad_norm": 2.0360944271087646,
+      "learning_rate": 1.9179875106630687e-06,
+      "loss": 0.8288,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8006595822645658,
+      "grad_norm": 2.4399008750915527,
+      "learning_rate": 1.9111885523714146e-06,
+      "loss": 0.8674,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8010260168559912,
+      "grad_norm": 2.487194299697876,
+      "learning_rate": 1.904400392643444e-06,
+      "loss": 0.7448,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8013924514474167,
+      "grad_norm": 2.4780404567718506,
+      "learning_rate": 1.897623040541352e-06,
+      "loss": 0.838,
+      "step": 2187
+    },
+    {
+      "epoch": 0.801758886038842,
+      "grad_norm": 2.225506544113159,
+      "learning_rate": 1.8908565051129047e-06,
+      "loss": 0.8501,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8021253206302675,
+      "grad_norm": 2.3592822551727295,
+      "learning_rate": 1.8841007953914293e-06,
+      "loss": 0.7955,
+      "step": 2189
+    },
+    {
+      "epoch": 0.8024917552216929,
+      "grad_norm": 2.3249824047088623,
+      "learning_rate": 1.8773559203957937e-06,
+      "loss": 0.8009,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8028581898131184,
+      "grad_norm": 2.304548740386963,
+      "learning_rate": 1.8706218891304084e-06,
+      "loss": 0.8716,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8032246244045438,
+      "grad_norm": 2.510841131210327,
+      "learning_rate": 1.8638987105852068e-06,
+      "loss": 0.8073,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8035910589959692,
+      "grad_norm": 2.4938032627105713,
+      "learning_rate": 1.8571863937356304e-06,
+      "loss": 0.8994,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8039574935873947,
+      "grad_norm": 2.564622163772583,
+      "learning_rate": 1.8504849475426278e-06,
+      "loss": 0.7947,
+      "step": 2194
+    },
+    {
+      "epoch": 0.80432392817882,
+      "grad_norm": 3.071754217147827,
+      "learning_rate": 1.8437943809526226e-06,
+      "loss": 0.735,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8046903627702455,
+      "grad_norm": 2.277132272720337,
+      "learning_rate": 1.8371147028975266e-06,
+      "loss": 0.8293,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8050567973616709,
+      "grad_norm": 2.702019214630127,
+      "learning_rate": 1.830445922294708e-06,
+      "loss": 0.7472,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8054232319530964,
+      "grad_norm": 2.243504047393799,
+      "learning_rate": 1.8237880480469884e-06,
+      "loss": 0.8351,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8057896665445218,
+      "grad_norm": 2.38909649848938,
+      "learning_rate": 1.8171410890426322e-06,
+      "loss": 0.8351,
+      "step": 2199
+    },
+    {
+      "epoch": 0.8061561011359473,
+      "grad_norm": 2.293527126312256,
+      "learning_rate": 1.8105050541553327e-06,
+      "loss": 0.7708,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8065225357273726,
+      "grad_norm": 2.17563796043396,
+      "learning_rate": 1.8038799522441896e-06,
+      "loss": 0.7982,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8068889703187981,
+      "grad_norm": 2.4225881099700928,
+      "learning_rate": 1.7972657921537195e-06,
+      "loss": 0.7971,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8072554049102235,
+      "grad_norm": 2.258533477783203,
+      "learning_rate": 1.7906625827138259e-06,
+      "loss": 0.816,
+      "step": 2203
+    },
+    {
+      "epoch": 0.807621839501649,
+      "grad_norm": 2.410449504852295,
+      "learning_rate": 1.7840703327397913e-06,
+      "loss": 0.8089,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8079882740930744,
+      "grad_norm": 2.261918544769287,
+      "learning_rate": 1.777489051032275e-06,
+      "loss": 0.7624,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8083547086844998,
+      "grad_norm": 2.631765365600586,
+      "learning_rate": 1.7709187463772825e-06,
+      "loss": 0.7966,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8087211432759253,
+      "grad_norm": 2.48982834815979,
+      "learning_rate": 1.7643594275461763e-06,
+      "loss": 0.8113,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8090875778673506,
+      "grad_norm": 2.338019847869873,
+      "learning_rate": 1.7578111032956413e-06,
+      "loss": 0.8178,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8094540124587761,
+      "grad_norm": 2.7235567569732666,
+      "learning_rate": 1.7512737823676939e-06,
+      "loss": 0.7882,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8098204470502015,
+      "grad_norm": 2.395200729370117,
+      "learning_rate": 1.7447474734896608e-06,
+      "loss": 0.8164,
+      "step": 2210
+    },
+    {
+      "epoch": 0.810186881641627,
+      "grad_norm": 2.603877305984497,
+      "learning_rate": 1.7382321853741579e-06,
+      "loss": 0.7299,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8105533162330524,
+      "grad_norm": 2.3217122554779053,
+      "learning_rate": 1.7317279267191e-06,
+      "loss": 0.7892,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8109197508244779,
+      "grad_norm": 2.780245780944824,
+      "learning_rate": 1.7252347062076712e-06,
+      "loss": 0.8099,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8112861854159032,
+      "grad_norm": 2.129652976989746,
+      "learning_rate": 1.718752532508321e-06,
+      "loss": 0.7873,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8116526200073287,
+      "grad_norm": 2.375014305114746,
+      "learning_rate": 1.7122814142747557e-06,
+      "loss": 0.7613,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8120190545987541,
+      "grad_norm": 2.627427577972412,
+      "learning_rate": 1.7058213601459116e-06,
+      "loss": 0.7451,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8123854891901795,
+      "grad_norm": 2.5515501499176025,
+      "learning_rate": 1.699372378745966e-06,
+      "loss": 0.8093,
+      "step": 2217
+    },
+    {
+      "epoch": 0.812751923781605,
+      "grad_norm": 2.86523699760437,
+      "learning_rate": 1.6929344786843072e-06,
+      "loss": 0.8233,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8131183583730304,
+      "grad_norm": 2.387831687927246,
+      "learning_rate": 1.6865076685555348e-06,
+      "loss": 0.8221,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8134847929644559,
+      "grad_norm": 2.3213229179382324,
+      "learning_rate": 1.6800919569394403e-06,
+      "loss": 0.8073,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8138512275558812,
+      "grad_norm": 2.4749796390533447,
+      "learning_rate": 1.6736873524009945e-06,
+      "loss": 0.8235,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8142176621473067,
+      "grad_norm": 3.385014057159424,
+      "learning_rate": 1.667293863490347e-06,
+      "loss": 0.7252,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8145840967387321,
+      "grad_norm": 2.663835287094116,
+      "learning_rate": 1.6609114987428077e-06,
+      "loss": 0.7411,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8149505313301576,
+      "grad_norm": 2.109029769897461,
+      "learning_rate": 1.6545402666788312e-06,
+      "loss": 0.8714,
+      "step": 2224
+    },
+    {
+      "epoch": 0.815316965921583,
+      "grad_norm": 2.6866328716278076,
+      "learning_rate": 1.648180175804016e-06,
+      "loss": 0.7418,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8156834005130085,
+      "grad_norm": 2.237931966781616,
+      "learning_rate": 1.6418312346090837e-06,
+      "loss": 0.7964,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8160498351044339,
+      "grad_norm": 2.364955425262451,
+      "learning_rate": 1.635493451569866e-06,
+      "loss": 0.8698,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8164162696958593,
+      "grad_norm": 2.356473684310913,
+      "learning_rate": 1.6291668351473089e-06,
+      "loss": 0.8051,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8167827042872847,
+      "grad_norm": 2.1675374507904053,
+      "learning_rate": 1.6228513937874424e-06,
+      "loss": 0.8232,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8171491388787101,
+      "grad_norm": 2.407489538192749,
+      "learning_rate": 1.6165471359213847e-06,
+      "loss": 0.8406,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8175155734701356,
+      "grad_norm": 2.22987699508667,
+      "learning_rate": 1.610254069965318e-06,
+      "loss": 0.8384,
+      "step": 2231
+    },
+    {
+      "epoch": 0.817882008061561,
+      "grad_norm": 2.386528491973877,
+      "learning_rate": 1.6039722043204832e-06,
+      "loss": 0.8679,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8182484426529865,
+      "grad_norm": 3.053328037261963,
+      "learning_rate": 1.5977015473731739e-06,
+      "loss": 0.7288,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8186148772444118,
+      "grad_norm": 2.4391896724700928,
+      "learning_rate": 1.5914421074947173e-06,
+      "loss": 0.8059,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8189813118358373,
+      "grad_norm": 2.8696322441101074,
+      "learning_rate": 1.5851938930414657e-06,
+      "loss": 0.7281,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8193477464272627,
+      "grad_norm": 2.360123872756958,
+      "learning_rate": 1.5789569123547876e-06,
+      "loss": 0.7647,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8197141810186882,
+      "grad_norm": 2.4133071899414062,
+      "learning_rate": 1.5727311737610495e-06,
+      "loss": 0.8045,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8200806156101136,
+      "grad_norm": 2.3202567100524902,
+      "learning_rate": 1.566516685571613e-06,
+      "loss": 0.806,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8204470502015391,
+      "grad_norm": 2.883397340774536,
+      "learning_rate": 1.5603134560828227e-06,
+      "loss": 0.7688,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8208134847929645,
+      "grad_norm": 2.4643895626068115,
+      "learning_rate": 1.5541214935759875e-06,
+      "loss": 0.8164,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8211799193843898,
+      "grad_norm": 2.317261219024658,
+      "learning_rate": 1.5479408063173818e-06,
+      "loss": 0.8055,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8215463539758153,
+      "grad_norm": 2.808366060256958,
+      "learning_rate": 1.5417714025582197e-06,
+      "loss": 0.7559,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8219127885672407,
+      "grad_norm": 2.359262704849243,
+      "learning_rate": 1.5356132905346543e-06,
+      "loss": 0.8299,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8222792231586662,
+      "grad_norm": 2.450434446334839,
+      "learning_rate": 1.529466478467768e-06,
+      "loss": 0.831,
+      "step": 2244
+    },
+    {
+      "epoch": 0.8226456577500916,
+      "grad_norm": 2.4397132396698,
+      "learning_rate": 1.523330974563554e-06,
+      "loss": 0.797,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8230120923415171,
+      "grad_norm": 2.272751808166504,
+      "learning_rate": 1.5172067870129137e-06,
+      "loss": 0.8234,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8233785269329424,
+      "grad_norm": 2.673409938812256,
+      "learning_rate": 1.5110939239916334e-06,
+      "loss": 0.8879,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8237449615243679,
+      "grad_norm": 2.298658847808838,
+      "learning_rate": 1.504992393660386e-06,
+      "loss": 0.766,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8241113961157933,
+      "grad_norm": 2.376124858856201,
+      "learning_rate": 1.4989022041647171e-06,
+      "loss": 0.8162,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8244778307072188,
+      "grad_norm": 2.5253942012786865,
+      "learning_rate": 1.4928233636350287e-06,
+      "loss": 0.8094,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8248442652986442,
+      "grad_norm": 2.904883623123169,
+      "learning_rate": 1.4867558801865744e-06,
+      "loss": 0.8148,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8252106998900697,
+      "grad_norm": 2.704359531402588,
+      "learning_rate": 1.480699761919443e-06,
+      "loss": 0.8203,
+      "step": 2252
+    },
+    {
+      "epoch": 0.8255771344814951,
+      "grad_norm": 2.621049642562866,
+      "learning_rate": 1.4746550169185536e-06,
+      "loss": 0.8155,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8259435690729204,
+      "grad_norm": 2.492042303085327,
+      "learning_rate": 1.4686216532536435e-06,
+      "loss": 0.8047,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8263100036643459,
+      "grad_norm": 2.2928173542022705,
+      "learning_rate": 1.4625996789792475e-06,
+      "loss": 0.865,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8266764382557713,
+      "grad_norm": 2.7155025005340576,
+      "learning_rate": 1.4565891021347066e-06,
+      "loss": 0.7329,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8270428728471968,
+      "grad_norm": 2.393474578857422,
+      "learning_rate": 1.4505899307441418e-06,
+      "loss": 0.8134,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8274093074386222,
+      "grad_norm": 2.2670514583587646,
+      "learning_rate": 1.4446021728164417e-06,
+      "loss": 0.8262,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8277757420300477,
+      "grad_norm": 2.44132924079895,
+      "learning_rate": 1.4386258363452676e-06,
+      "loss": 0.8182,
+      "step": 2259
+    },
+    {
+      "epoch": 0.828142176621473,
+      "grad_norm": 2.589374542236328,
+      "learning_rate": 1.4326609293090289e-06,
+      "loss": 0.7247,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8285086112128985,
+      "grad_norm": 2.5723989009857178,
+      "learning_rate": 1.4267074596708752e-06,
+      "loss": 0.8103,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8288750458043239,
+      "grad_norm": 2.5136826038360596,
+      "learning_rate": 1.4207654353786926e-06,
+      "loss": 0.7336,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8292414803957494,
+      "grad_norm": 2.8288135528564453,
+      "learning_rate": 1.4148348643650778e-06,
+      "loss": 0.7746,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8296079149871748,
+      "grad_norm": 2.1983284950256348,
+      "learning_rate": 1.4089157545473476e-06,
+      "loss": 0.8172,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8299743495786002,
+      "grad_norm": 2.543551445007324,
+      "learning_rate": 1.4030081138275142e-06,
+      "loss": 0.8106,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8303407841700257,
+      "grad_norm": 2.4806594848632812,
+      "learning_rate": 1.3971119500922748e-06,
+      "loss": 0.7666,
+      "step": 2266
+    },
+    {
+      "epoch": 0.830707218761451,
+      "grad_norm": 2.7103400230407715,
+      "learning_rate": 1.3912272712130138e-06,
+      "loss": 0.8218,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8310736533528765,
+      "grad_norm": 2.2749123573303223,
+      "learning_rate": 1.3853540850457704e-06,
+      "loss": 0.834,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8314400879443019,
+      "grad_norm": 2.777505397796631,
+      "learning_rate": 1.3794923994312536e-06,
+      "loss": 0.8265,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8318065225357274,
+      "grad_norm": 2.378373622894287,
+      "learning_rate": 1.373642222194812e-06,
+      "loss": 0.8237,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8321729571271528,
+      "grad_norm": 2.2618765830993652,
+      "learning_rate": 1.3678035611464335e-06,
+      "loss": 0.8411,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8325393917185783,
+      "grad_norm": 2.407137632369995,
+      "learning_rate": 1.3619764240807331e-06,
+      "loss": 0.8035,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8329058263100036,
+      "grad_norm": 2.2529635429382324,
+      "learning_rate": 1.3561608187769348e-06,
+      "loss": 0.8084,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8332722609014291,
+      "grad_norm": 2.5161614418029785,
+      "learning_rate": 1.3503567529988725e-06,
+      "loss": 0.7589,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8336386954928545,
+      "grad_norm": 2.4014391899108887,
+      "learning_rate": 1.3445642344949771e-06,
+      "loss": 0.7823,
+      "step": 2275
+    },
+    {
+      "epoch": 0.83400513008428,
+      "grad_norm": 3.0440123081207275,
+      "learning_rate": 1.3387832709982606e-06,
+      "loss": 0.6788,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8343715646757054,
+      "grad_norm": 2.168713331222534,
+      "learning_rate": 1.3330138702263095e-06,
+      "loss": 0.8395,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8347379992671308,
+      "grad_norm": 2.9199330806732178,
+      "learning_rate": 1.3272560398812705e-06,
+      "loss": 0.7725,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8351044338585563,
+      "grad_norm": 2.082176923751831,
+      "learning_rate": 1.3215097876498507e-06,
+      "loss": 0.8181,
+      "step": 2279
+    },
+    {
+      "epoch": 0.8354708684499816,
+      "grad_norm": 2.3910794258117676,
+      "learning_rate": 1.3157751212032953e-06,
+      "loss": 0.8273,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8358373030414071,
+      "grad_norm": 2.448989152908325,
+      "learning_rate": 1.3100520481973867e-06,
+      "loss": 0.829,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8362037376328325,
+      "grad_norm": 2.466848373413086,
+      "learning_rate": 1.3043405762724271e-06,
+      "loss": 0.8387,
+      "step": 2282
+    },
+    {
+      "epoch": 0.836570172224258,
+      "grad_norm": 2.8775925636291504,
+      "learning_rate": 1.2986407130532309e-06,
+      "loss": 0.7192,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8369366068156834,
+      "grad_norm": 2.5638585090637207,
+      "learning_rate": 1.2929524661491143e-06,
+      "loss": 0.7766,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8373030414071089,
+      "grad_norm": 2.6455864906311035,
+      "learning_rate": 1.28727584315389e-06,
+      "loss": 0.7771,
+      "step": 2285
+    },
+    {
+      "epoch": 0.8376694759985343,
+      "grad_norm": 2.722618341445923,
+      "learning_rate": 1.281610851645848e-06,
+      "loss": 0.79,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8380359105899597,
+      "grad_norm": 2.9081006050109863,
+      "learning_rate": 1.2759574991877565e-06,
+      "loss": 0.7423,
+      "step": 2287
+    },
+    {
+      "epoch": 0.8384023451813851,
+      "grad_norm": 2.7987332344055176,
+      "learning_rate": 1.270315793326835e-06,
+      "loss": 0.7768,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8387687797728105,
+      "grad_norm": 2.355983257293701,
+      "learning_rate": 1.2646857415947656e-06,
+      "loss": 0.8366,
+      "step": 2289
+    },
+    {
+      "epoch": 0.839135214364236,
+      "grad_norm": 2.514294147491455,
+      "learning_rate": 1.2590673515076645e-06,
+      "loss": 0.7984,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8395016489556614,
+      "grad_norm": 3.055561065673828,
+      "learning_rate": 1.2534606305660847e-06,
+      "loss": 0.7668,
+      "step": 2291
+    },
+    {
+      "epoch": 0.8398680835470869,
+      "grad_norm": 2.742436647415161,
+      "learning_rate": 1.2478655862549993e-06,
+      "loss": 0.7446,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8402345181385122,
+      "grad_norm": 2.348378896713257,
+      "learning_rate": 1.2422822260437927e-06,
+      "loss": 0.8166,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8406009527299377,
+      "grad_norm": 2.8558847904205322,
+      "learning_rate": 1.2367105573862492e-06,
+      "loss": 0.804,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8409673873213631,
+      "grad_norm": 2.6892967224121094,
+      "learning_rate": 1.2311505877205476e-06,
+      "loss": 0.7358,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8413338219127886,
+      "grad_norm": 2.377847194671631,
+      "learning_rate": 1.2256023244692473e-06,
+      "loss": 0.8626,
+      "step": 2296
+    },
+    {
+      "epoch": 0.841700256504214,
+      "grad_norm": 2.764005422592163,
+      "learning_rate": 1.220065775039282e-06,
+      "loss": 0.7776,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8420666910956395,
+      "grad_norm": 2.4356181621551514,
+      "learning_rate": 1.214540946821945e-06,
+      "loss": 0.7624,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8424331256870649,
+      "grad_norm": 2.727997303009033,
+      "learning_rate": 1.2090278471928807e-06,
+      "loss": 0.8431,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8427995602784902,
+      "grad_norm": 2.2275822162628174,
+      "learning_rate": 1.2035264835120787e-06,
+      "loss": 0.825,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8431659948699157,
+      "grad_norm": 2.6020660400390625,
+      "learning_rate": 1.1980368631238582e-06,
+      "loss": 0.7773,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8435324294613411,
+      "grad_norm": 2.4587666988372803,
+      "learning_rate": 1.1925589933568626e-06,
+      "loss": 0.8367,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8438988640527666,
+      "grad_norm": 2.292140245437622,
+      "learning_rate": 1.1870928815240522e-06,
+      "loss": 0.7917,
+      "step": 2303
+    },
+    {
+      "epoch": 0.844265298644192,
+      "grad_norm": 2.52288556098938,
+      "learning_rate": 1.1816385349226822e-06,
+      "loss": 0.8458,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8446317332356175,
+      "grad_norm": 2.344015121459961,
+      "learning_rate": 1.1761959608343065e-06,
+      "loss": 0.8382,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8449981678270428,
+      "grad_norm": 2.7470791339874268,
+      "learning_rate": 1.170765166524762e-06,
+      "loss": 0.6318,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8453646024184683,
+      "grad_norm": 2.4314420223236084,
+      "learning_rate": 1.1653461592441606e-06,
+      "loss": 0.8237,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8457310370098937,
+      "grad_norm": 2.8293542861938477,
+      "learning_rate": 1.1599389462268783e-06,
+      "loss": 0.8543,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8460974716013192,
+      "grad_norm": 2.1980388164520264,
+      "learning_rate": 1.1545435346915413e-06,
+      "loss": 0.8001,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8464639061927446,
+      "grad_norm": 2.2609615325927734,
+      "learning_rate": 1.1491599318410251e-06,
+      "loss": 0.8064,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8468303407841701,
+      "grad_norm": 2.2475104331970215,
+      "learning_rate": 1.1437881448624432e-06,
+      "loss": 0.8738,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8471967753755955,
+      "grad_norm": 2.4149606227874756,
+      "learning_rate": 1.1384281809271268e-06,
+      "loss": 0.8969,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8475632099670208,
+      "grad_norm": 2.826284170150757,
+      "learning_rate": 1.1330800471906323e-06,
+      "loss": 0.7542,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8479296445584463,
+      "grad_norm": 2.559403419494629,
+      "learning_rate": 1.1277437507927137e-06,
+      "loss": 0.7561,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8482960791498717,
+      "grad_norm": 2.6555681228637695,
+      "learning_rate": 1.122419298857329e-06,
+      "loss": 0.82,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8486625137412972,
+      "grad_norm": 2.4660537242889404,
+      "learning_rate": 1.1171066984926238e-06,
+      "loss": 0.8118,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8490289483327226,
+      "grad_norm": 2.449301242828369,
+      "learning_rate": 1.1118059567909178e-06,
+      "loss": 0.8639,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8493953829241481,
+      "grad_norm": 2.502134084701538,
+      "learning_rate": 1.1065170808287017e-06,
+      "loss": 0.8315,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8497618175155734,
+      "grad_norm": 2.407904624938965,
+      "learning_rate": 1.1012400776666276e-06,
+      "loss": 0.8133,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8501282521069989,
+      "grad_norm": 2.5190916061401367,
+      "learning_rate": 1.0959749543494902e-06,
+      "loss": 0.8109,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8504946866984243,
+      "grad_norm": 2.5884244441986084,
+      "learning_rate": 1.0907217179062335e-06,
+      "loss": 0.7594,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8508611212898498,
+      "grad_norm": 2.1998255252838135,
+      "learning_rate": 1.0854803753499276e-06,
+      "loss": 0.7992,
+      "step": 2322
+    },
+    {
+      "epoch": 0.8512275558812752,
+      "grad_norm": 2.3266897201538086,
+      "learning_rate": 1.0802509336777634e-06,
+      "loss": 0.7921,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8515939904727006,
+      "grad_norm": 2.42293381690979,
+      "learning_rate": 1.0750333998710493e-06,
+      "loss": 0.8439,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8519604250641261,
+      "grad_norm": 2.2549338340759277,
+      "learning_rate": 1.0698277808951873e-06,
+      "loss": 0.8329,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8523268596555514,
+      "grad_norm": 2.6438047885894775,
+      "learning_rate": 1.0646340836996837e-06,
+      "loss": 0.7681,
+      "step": 2326
+    },
+    {
+      "epoch": 0.8526932942469769,
+      "grad_norm": 2.733704090118408,
+      "learning_rate": 1.0594523152181223e-06,
+      "loss": 0.8068,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8530597288384023,
+      "grad_norm": 2.45572829246521,
+      "learning_rate": 1.0542824823681663e-06,
+      "loss": 0.8057,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8534261634298278,
+      "grad_norm": 2.4164037704467773,
+      "learning_rate": 1.0491245920515435e-06,
+      "loss": 0.7422,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8537925980212532,
+      "grad_norm": 2.500162124633789,
+      "learning_rate": 1.0439786511540328e-06,
+      "loss": 0.8651,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8541590326126787,
+      "grad_norm": 2.2892117500305176,
+      "learning_rate": 1.0388446665454699e-06,
+      "loss": 0.749,
+      "step": 2331
+    },
+    {
+      "epoch": 0.854525467204104,
+      "grad_norm": 2.167318820953369,
+      "learning_rate": 1.0337226450797234e-06,
+      "loss": 0.838,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8548919017955295,
+      "grad_norm": 2.1648292541503906,
+      "learning_rate": 1.0286125935946934e-06,
+      "loss": 0.7958,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8552583363869549,
+      "grad_norm": 2.304607629776001,
+      "learning_rate": 1.0235145189122964e-06,
+      "loss": 0.7698,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8556247709783804,
+      "grad_norm": 2.7921359539031982,
+      "learning_rate": 1.018428427838465e-06,
+      "loss": 0.8214,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8559912055698058,
+      "grad_norm": 2.3064191341400146,
+      "learning_rate": 1.0133543271631275e-06,
+      "loss": 0.8311,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8563576401612312,
+      "grad_norm": 2.405390977859497,
+      "learning_rate": 1.0082922236602112e-06,
+      "loss": 0.7864,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8567240747526567,
+      "grad_norm": 2.9421746730804443,
+      "learning_rate": 1.0032421240876233e-06,
+      "loss": 0.7838,
+      "step": 2338
+    },
+    {
+      "epoch": 0.857090509344082,
+      "grad_norm": 2.6204707622528076,
+      "learning_rate": 9.982040351872501e-07,
+      "loss": 0.7925,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8574569439355075,
+      "grad_norm": 2.2351484298706055,
+      "learning_rate": 9.93177963684937e-07,
+      "loss": 0.8206,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8578233785269329,
+      "grad_norm": 2.7413156032562256,
+      "learning_rate": 9.881639162904898e-07,
+      "loss": 0.8392,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8581898131183584,
+      "grad_norm": 2.6065118312835693,
+      "learning_rate": 9.831618996976633e-07,
+      "loss": 0.7752,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8585562477097838,
+      "grad_norm": 2.428201675415039,
+      "learning_rate": 9.7817192058415e-07,
+      "loss": 0.7817,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8589226823012093,
+      "grad_norm": 2.6025755405426025,
+      "learning_rate": 9.731939856115736e-07,
+      "loss": 0.803,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8592891168926347,
+      "grad_norm": 2.837523937225342,
+      "learning_rate": 9.682281014254736e-07,
+      "loss": 0.7726,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8596555514840601,
+      "grad_norm": 2.9446375370025635,
+      "learning_rate": 9.632742746553102e-07,
+      "loss": 0.7836,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8600219860754855,
+      "grad_norm": 2.481795310974121,
+      "learning_rate": 9.583325119144382e-07,
+      "loss": 0.8244,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8603884206669109,
+      "grad_norm": 2.6143958568573,
+      "learning_rate": 9.534028198001122e-07,
+      "loss": 0.8114,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8607548552583364,
+      "grad_norm": 2.3166778087615967,
+      "learning_rate": 9.484852048934734e-07,
+      "loss": 0.8259,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8611212898497618,
+      "grad_norm": 2.3150179386138916,
+      "learning_rate": 9.435796737595382e-07,
+      "loss": 0.8087,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8614877244411873,
+      "grad_norm": 2.456609010696411,
+      "learning_rate": 9.386862329471869e-07,
+      "loss": 0.8285,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8618541590326126,
+      "grad_norm": 2.5039875507354736,
+      "learning_rate": 9.33804888989166e-07,
+      "loss": 0.8021,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8622205936240381,
+      "grad_norm": 2.201730489730835,
+      "learning_rate": 9.289356484020706e-07,
+      "loss": 0.816,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8625870282154635,
+      "grad_norm": 2.5118815898895264,
+      "learning_rate": 9.240785176863354e-07,
+      "loss": 0.8319,
+      "step": 2354
+    },
+    {
+      "epoch": 0.862953462806889,
+      "grad_norm": 2.379840850830078,
+      "learning_rate": 9.192335033262334e-07,
+      "loss": 0.8279,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8633198973983144,
+      "grad_norm": 2.3860526084899902,
+      "learning_rate": 9.144006117898541e-07,
+      "loss": 0.8034,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8636863319897399,
+      "grad_norm": 2.636835813522339,
+      "learning_rate": 9.095798495291119e-07,
+      "loss": 0.7626,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8640527665811653,
+      "grad_norm": 2.058418035507202,
+      "learning_rate": 9.047712229797224e-07,
+      "loss": 0.7966,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8644192011725907,
+      "grad_norm": 2.3849425315856934,
+      "learning_rate": 8.999747385612023e-07,
+      "loss": 0.8084,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8647856357640161,
+      "grad_norm": 2.908236026763916,
+      "learning_rate": 8.951904026768621e-07,
+      "loss": 0.7261,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8651520703554415,
+      "grad_norm": 2.460125207901001,
+      "learning_rate": 8.904182217137847e-07,
+      "loss": 0.8432,
+      "step": 2361
+    },
+    {
+      "epoch": 0.865518504946867,
+      "grad_norm": 2.6111361980438232,
+      "learning_rate": 8.856582020428362e-07,
+      "loss": 0.7894,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8658849395382924,
+      "grad_norm": 2.1920156478881836,
+      "learning_rate": 8.809103500186411e-07,
+      "loss": 0.8271,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8662513741297179,
+      "grad_norm": 2.735114574432373,
+      "learning_rate": 8.761746719795827e-07,
+      "loss": 0.7336,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8666178087211432,
+      "grad_norm": 2.2781002521514893,
+      "learning_rate": 8.714511742477927e-07,
+      "loss": 0.8652,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8669842433125687,
+      "grad_norm": 2.6231000423431396,
+      "learning_rate": 8.66739863129139e-07,
+      "loss": 0.8134,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8673506779039941,
+      "grad_norm": 2.858731746673584,
+      "learning_rate": 8.620407449132218e-07,
+      "loss": 0.7618,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8677171124954196,
+      "grad_norm": 2.5288403034210205,
+      "learning_rate": 8.573538258733671e-07,
+      "loss": 0.7656,
+      "step": 2368
+    },
+    {
+      "epoch": 0.868083547086845,
+      "grad_norm": 2.189743995666504,
+      "learning_rate": 8.52679112266609e-07,
+      "loss": 0.8046,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8684499816782705,
+      "grad_norm": 2.4853451251983643,
+      "learning_rate": 8.480166103336918e-07,
+      "loss": 0.7811,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8688164162696959,
+      "grad_norm": 2.425382614135742,
+      "learning_rate": 8.433663262990544e-07,
+      "loss": 0.7532,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8691828508611212,
+      "grad_norm": 2.3846395015716553,
+      "learning_rate": 8.387282663708285e-07,
+      "loss": 0.8075,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8695492854525467,
+      "grad_norm": 2.588106870651245,
+      "learning_rate": 8.341024367408213e-07,
+      "loss": 0.8226,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8699157200439721,
+      "grad_norm": 2.4514048099517822,
+      "learning_rate": 8.294888435845183e-07,
+      "loss": 0.8669,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8702821546353976,
+      "grad_norm": 2.1290717124938965,
+      "learning_rate": 8.248874930610674e-07,
+      "loss": 0.838,
+      "step": 2375
+    },
+    {
+      "epoch": 0.870648589226823,
+      "grad_norm": 2.3119587898254395,
+      "learning_rate": 8.20298391313269e-07,
+      "loss": 0.8382,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8710150238182485,
+      "grad_norm": 2.4327573776245117,
+      "learning_rate": 8.157215444675737e-07,
+      "loss": 0.8515,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8713814584096738,
+      "grad_norm": 2.836103677749634,
+      "learning_rate": 8.111569586340751e-07,
+      "loss": 0.7742,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8717478930010993,
+      "grad_norm": 2.512967824935913,
+      "learning_rate": 8.066046399064953e-07,
+      "loss": 0.8043,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8721143275925247,
+      "grad_norm": 2.3081233501434326,
+      "learning_rate": 8.020645943621774e-07,
+      "loss": 0.843,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8724807621839502,
+      "grad_norm": 2.9423413276672363,
+      "learning_rate": 7.975368280620865e-07,
+      "loss": 0.7945,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8728471967753756,
+      "grad_norm": 2.687516450881958,
+      "learning_rate": 7.930213470507864e-07,
+      "loss": 0.7514,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8732136313668011,
+      "grad_norm": 2.2465670108795166,
+      "learning_rate": 7.885181573564449e-07,
+      "loss": 0.7975,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8735800659582265,
+      "grad_norm": 2.3269989490509033,
+      "learning_rate": 7.840272649908232e-07,
+      "loss": 0.8088,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8739465005496518,
+      "grad_norm": 2.287290096282959,
+      "learning_rate": 7.795486759492621e-07,
+      "loss": 0.8352,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8743129351410773,
+      "grad_norm": 2.7851336002349854,
+      "learning_rate": 7.750823962106779e-07,
+      "loss": 0.7436,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8746793697325027,
+      "grad_norm": 2.395256519317627,
+      "learning_rate": 7.706284317375534e-07,
+      "loss": 0.8167,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8750458043239282,
+      "grad_norm": 2.470905303955078,
+      "learning_rate": 7.661867884759321e-07,
+      "loss": 0.7964,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8754122389153536,
+      "grad_norm": 2.294090986251831,
+      "learning_rate": 7.617574723554078e-07,
+      "loss": 0.8194,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8757786735067791,
+      "grad_norm": 2.5644469261169434,
+      "learning_rate": 7.57340489289119e-07,
+      "loss": 0.8178,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8761451080982045,
+      "grad_norm": 2.3230745792388916,
+      "learning_rate": 7.529358451737368e-07,
+      "loss": 0.8578,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8765115426896299,
+      "grad_norm": 2.285337448120117,
+      "learning_rate": 7.485435458894619e-07,
+      "loss": 0.7574,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8768779772810553,
+      "grad_norm": 2.14737606048584,
+      "learning_rate": 7.441635973000127e-07,
+      "loss": 0.8185,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8772444118724808,
+      "grad_norm": 2.1775426864624023,
+      "learning_rate": 7.397960052526221e-07,
+      "loss": 0.8213,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8776108464639062,
+      "grad_norm": 2.196223497390747,
+      "learning_rate": 7.354407755780257e-07,
+      "loss": 0.8192,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8779772810553316,
+      "grad_norm": 2.274160385131836,
+      "learning_rate": 7.310979140904572e-07,
+      "loss": 0.8638,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8783437156467571,
+      "grad_norm": 2.5875229835510254,
+      "learning_rate": 7.267674265876324e-07,
+      "loss": 0.8277,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8787101502381824,
+      "grad_norm": 2.819317579269409,
+      "learning_rate": 7.224493188507553e-07,
+      "loss": 0.7778,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8790765848296079,
+      "grad_norm": 2.390894651412964,
+      "learning_rate": 7.181435966444994e-07,
+      "loss": 0.804,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8794430194210333,
+      "grad_norm": 2.9968502521514893,
+      "learning_rate": 7.138502657170043e-07,
+      "loss": 0.7772,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8798094540124588,
+      "grad_norm": 2.596324920654297,
+      "learning_rate": 7.095693317998675e-07,
+      "loss": 0.7983,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8801758886038842,
+      "grad_norm": 2.346898078918457,
+      "learning_rate": 7.053008006081341e-07,
+      "loss": 0.8041,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8805423231953097,
+      "grad_norm": 2.4089863300323486,
+      "learning_rate": 7.010446778402968e-07,
+      "loss": 0.7913,
+      "step": 2403
+    },
+    {
+      "epoch": 0.880908757786735,
+      "grad_norm": 2.104218006134033,
+      "learning_rate": 6.968009691782763e-07,
+      "loss": 0.8543,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8812751923781605,
+      "grad_norm": 2.470045566558838,
+      "learning_rate": 6.925696802874238e-07,
+      "loss": 0.7533,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8816416269695859,
+      "grad_norm": 2.4869046211242676,
+      "learning_rate": 6.883508168165143e-07,
+      "loss": 0.8178,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8820080615610114,
+      "grad_norm": 2.484778881072998,
+      "learning_rate": 6.841443843977258e-07,
+      "loss": 0.8237,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8823744961524368,
+      "grad_norm": 2.4106252193450928,
+      "learning_rate": 6.799503886466485e-07,
+      "loss": 0.8261,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8827409307438622,
+      "grad_norm": 2.831423759460449,
+      "learning_rate": 6.75768835162266e-07,
+      "loss": 0.7781,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8831073653352877,
+      "grad_norm": 2.7222824096679688,
+      "learning_rate": 6.715997295269528e-07,
+      "loss": 0.7915,
+      "step": 2410
+    },
+    {
+      "epoch": 0.883473799926713,
+      "grad_norm": 2.098013401031494,
+      "learning_rate": 6.674430773064656e-07,
+      "loss": 0.8678,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8838402345181385,
+      "grad_norm": 2.491773843765259,
+      "learning_rate": 6.632988840499355e-07,
+      "loss": 0.7576,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8842066691095639,
+      "grad_norm": 2.345390558242798,
+      "learning_rate": 6.591671552898593e-07,
+      "loss": 0.8643,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8845731037009894,
+      "grad_norm": 2.5588772296905518,
+      "learning_rate": 6.55047896542097e-07,
+      "loss": 0.8814,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8849395382924148,
+      "grad_norm": 2.4958090782165527,
+      "learning_rate": 6.509411133058574e-07,
+      "loss": 0.827,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8853059728838403,
+      "grad_norm": 2.6431095600128174,
+      "learning_rate": 6.468468110636961e-07,
+      "loss": 0.8081,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8856724074752657,
+      "grad_norm": 2.8156349658966064,
+      "learning_rate": 6.427649952815096e-07,
+      "loss": 0.7839,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8860388420666911,
+      "grad_norm": 2.345520496368408,
+      "learning_rate": 6.386956714085191e-07,
+      "loss": 0.8431,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8864052766581165,
+      "grad_norm": 2.677828788757324,
+      "learning_rate": 6.346388448772722e-07,
+      "loss": 0.7809,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8867717112495419,
+      "grad_norm": 3.008817195892334,
+      "learning_rate": 6.305945211036347e-07,
+      "loss": 0.7315,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8871381458409674,
+      "grad_norm": 2.964496612548828,
+      "learning_rate": 6.265627054867773e-07,
+      "loss": 0.7653,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8875045804323928,
+      "grad_norm": 2.6443655490875244,
+      "learning_rate": 6.22543403409177e-07,
+      "loss": 0.7382,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8878710150238183,
+      "grad_norm": 2.525057792663574,
+      "learning_rate": 6.185366202365984e-07,
+      "loss": 0.8576,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8882374496152436,
+      "grad_norm": 2.3738057613372803,
+      "learning_rate": 6.145423613180989e-07,
+      "loss": 0.8157,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8886038842066691,
+      "grad_norm": 2.2518320083618164,
+      "learning_rate": 6.105606319860157e-07,
+      "loss": 0.8524,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8889703187980945,
+      "grad_norm": 2.599637269973755,
+      "learning_rate": 6.065914375559556e-07,
+      "loss": 0.7961,
+      "step": 2426
+    },
+    {
+      "epoch": 0.88933675338952,
+      "grad_norm": 2.3115074634552,
+      "learning_rate": 6.026347833267942e-07,
+      "loss": 0.8109,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8897031879809454,
+      "grad_norm": 2.493844509124756,
+      "learning_rate": 5.986906745806642e-07,
+      "loss": 0.7723,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8900696225723709,
+      "grad_norm": 2.113201379776001,
+      "learning_rate": 5.947591165829514e-07,
+      "loss": 0.8207,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8904360571637963,
+      "grad_norm": 2.5341408252716064,
+      "learning_rate": 5.908401145822862e-07,
+      "loss": 0.8079,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8908024917552217,
+      "grad_norm": 2.5149288177490234,
+      "learning_rate": 5.869336738105369e-07,
+      "loss": 0.7806,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8911689263466471,
+      "grad_norm": 2.887014389038086,
+      "learning_rate": 5.83039799482803e-07,
+      "loss": 0.7227,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8915353609380725,
+      "grad_norm": 2.5421342849731445,
+      "learning_rate": 5.791584967974051e-07,
+      "loss": 0.8006,
+      "step": 2433
+    },
+    {
+      "epoch": 0.891901795529498,
+      "grad_norm": 2.474050283432007,
+      "learning_rate": 5.752897709358818e-07,
+      "loss": 0.8368,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8922682301209234,
+      "grad_norm": 2.6157748699188232,
+      "learning_rate": 5.714336270629838e-07,
+      "loss": 0.8692,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8926346647123489,
+      "grad_norm": 2.3381001949310303,
+      "learning_rate": 5.675900703266646e-07,
+      "loss": 0.8414,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8930010993037742,
+      "grad_norm": 2.5202739238739014,
+      "learning_rate": 5.637591058580705e-07,
+      "loss": 0.8096,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8933675338951997,
+      "grad_norm": 2.232762575149536,
+      "learning_rate": 5.599407387715405e-07,
+      "loss": 0.7977,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8937339684866251,
+      "grad_norm": 2.746988296508789,
+      "learning_rate": 5.561349741645927e-07,
+      "loss": 0.7389,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8941004030780506,
+      "grad_norm": 2.633068561553955,
+      "learning_rate": 5.523418171179252e-07,
+      "loss": 0.789,
+      "step": 2440
+    },
+    {
+      "epoch": 0.894466837669476,
+      "grad_norm": 2.6220340728759766,
+      "learning_rate": 5.485612726954015e-07,
+      "loss": 0.7741,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8948332722609015,
+      "grad_norm": 2.605099678039551,
+      "learning_rate": 5.447933459440501e-07,
+      "loss": 0.8026,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8951997068523269,
+      "grad_norm": 2.5063672065734863,
+      "learning_rate": 5.410380418940542e-07,
+      "loss": 0.7285,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8955661414437522,
+      "grad_norm": 2.344224452972412,
+      "learning_rate": 5.37295365558741e-07,
+      "loss": 0.8282,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8959325760351777,
+      "grad_norm": 2.5959107875823975,
+      "learning_rate": 5.33565321934586e-07,
+      "loss": 0.7285,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8962990106266031,
+      "grad_norm": 2.6870009899139404,
+      "learning_rate": 5.298479160011982e-07,
+      "loss": 0.8097,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8966654452180286,
+      "grad_norm": 2.544396162033081,
+      "learning_rate": 5.261431527213157e-07,
+      "loss": 0.8394,
+      "step": 2447
+    },
+    {
+      "epoch": 0.897031879809454,
+      "grad_norm": 2.3084309101104736,
+      "learning_rate": 5.224510370407943e-07,
+      "loss": 0.7856,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8973983144008795,
+      "grad_norm": 2.6764068603515625,
+      "learning_rate": 5.187715738886112e-07,
+      "loss": 0.7997,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8977647489923049,
+      "grad_norm": 2.511139392852783,
+      "learning_rate": 5.151047681768484e-07,
+      "loss": 0.8524,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8981311835837303,
+      "grad_norm": 2.2220213413238525,
+      "learning_rate": 5.114506248006912e-07,
+      "loss": 0.8661,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8984976181751557,
+      "grad_norm": 2.3586859703063965,
+      "learning_rate": 5.078091486384241e-07,
+      "loss": 0.7999,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8988640527665812,
+      "grad_norm": 2.833298921585083,
+      "learning_rate": 5.041803445514159e-07,
+      "loss": 0.8183,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8992304873580066,
+      "grad_norm": 2.7366554737091064,
+      "learning_rate": 5.005642173841196e-07,
+      "loss": 0.7463,
+      "step": 2454
+    },
+    {
+      "epoch": 0.899596921949432,
+      "grad_norm": 2.269247531890869,
+      "learning_rate": 4.969607719640668e-07,
+      "loss": 0.8269,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8999633565408575,
+      "grad_norm": 2.2485527992248535,
+      "learning_rate": 4.93370013101856e-07,
+      "loss": 0.8307,
+      "step": 2456
+    },
+    {
+      "epoch": 0.9003297911322828,
+      "grad_norm": 2.685641288757324,
+      "learning_rate": 4.897919455911504e-07,
+      "loss": 0.8218,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9006962257237083,
+      "grad_norm": 2.3350040912628174,
+      "learning_rate": 4.862265742086736e-07,
+      "loss": 0.813,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9010626603151337,
+      "grad_norm": 2.218966245651245,
+      "learning_rate": 4.826739037141914e-07,
+      "loss": 0.8424,
+      "step": 2459
+    },
+    {
+      "epoch": 0.9014290949065592,
+      "grad_norm": 2.3188931941986084,
+      "learning_rate": 4.791339388505234e-07,
+      "loss": 0.8287,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9017955294979846,
+      "grad_norm": 2.376735210418701,
+      "learning_rate": 4.756066843435203e-07,
+      "loss": 0.8246,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9021619640894101,
+      "grad_norm": 2.435607671737671,
+      "learning_rate": 4.720921449020677e-07,
+      "loss": 0.8103,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9025283986808355,
+      "grad_norm": 2.3804492950439453,
+      "learning_rate": 4.685903252180768e-07,
+      "loss": 0.8147,
+      "step": 2463
+    },
+    {
+      "epoch": 0.9028948332722609,
+      "grad_norm": 2.6249334812164307,
+      "learning_rate": 4.6510122996647477e-07,
+      "loss": 0.8083,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9032612678636863,
+      "grad_norm": 2.683875560760498,
+      "learning_rate": 4.6162486380520676e-07,
+      "loss": 0.7687,
+      "step": 2465
+    },
+    {
+      "epoch": 0.9036277024551118,
+      "grad_norm": 2.6124651432037354,
+      "learning_rate": 4.581612313752193e-07,
+      "loss": 0.8229,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9039941370465372,
+      "grad_norm": 2.606337308883667,
+      "learning_rate": 4.5471033730046353e-07,
+      "loss": 0.7641,
+      "step": 2467
+    },
+    {
+      "epoch": 0.9043605716379626,
+      "grad_norm": 2.6244354248046875,
+      "learning_rate": 4.512721861878844e-07,
+      "loss": 0.7512,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9047270062293881,
+      "grad_norm": 2.6744701862335205,
+      "learning_rate": 4.478467826274102e-07,
+      "loss": 0.8043,
+      "step": 2469
+    },
+    {
+      "epoch": 0.9050934408208134,
+      "grad_norm": 2.865318536758423,
+      "learning_rate": 4.444341311919564e-07,
+      "loss": 0.8108,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9054598754122389,
+      "grad_norm": 2.284832715988159,
+      "learning_rate": 4.4103423643741537e-07,
+      "loss": 0.7966,
+      "step": 2471
+    },
+    {
+      "epoch": 0.9058263100036643,
+      "grad_norm": 2.674898386001587,
+      "learning_rate": 4.37647102902643e-07,
+      "loss": 0.7783,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9061927445950898,
+      "grad_norm": 2.2506966590881348,
+      "learning_rate": 4.3427273510946446e-07,
+      "loss": 0.8311,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9065591791865152,
+      "grad_norm": 2.267212152481079,
+      "learning_rate": 4.309111375626618e-07,
+      "loss": 0.8903,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9069256137779407,
+      "grad_norm": 2.3874728679656982,
+      "learning_rate": 4.2756231474996633e-07,
+      "loss": 0.8391,
+      "step": 2475
+    },
+    {
+      "epoch": 0.907292048369366,
+      "grad_norm": 2.156465768814087,
+      "learning_rate": 4.242262711420564e-07,
+      "loss": 0.8546,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9076584829607915,
+      "grad_norm": 2.561232566833496,
+      "learning_rate": 4.2090301119255164e-07,
+      "loss": 0.7859,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9080249175522169,
+      "grad_norm": 2.3131983280181885,
+      "learning_rate": 4.1759253933800423e-07,
+      "loss": 0.8125,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9083913521436423,
+      "grad_norm": 2.7867021560668945,
+      "learning_rate": 4.1429485999789463e-07,
+      "loss": 0.7831,
+      "step": 2479
+    },
+    {
+      "epoch": 0.9087577867350678,
+      "grad_norm": 2.2129106521606445,
+      "learning_rate": 4.1100997757462344e-07,
+      "loss": 0.8181,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9091242213264932,
+      "grad_norm": 2.780451536178589,
+      "learning_rate": 4.0773789645350945e-07,
+      "loss": 0.8104,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9094906559179187,
+      "grad_norm": 2.6728110313415527,
+      "learning_rate": 4.044786210027829e-07,
+      "loss": 0.7943,
+      "step": 2482
+    },
+    {
+      "epoch": 0.909857090509344,
+      "grad_norm": 2.3612685203552246,
+      "learning_rate": 4.012321555735732e-07,
+      "loss": 0.8163,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9102235251007695,
+      "grad_norm": 2.2369892597198486,
+      "learning_rate": 3.9799850449991574e-07,
+      "loss": 0.812,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9105899596921949,
+      "grad_norm": 2.4772753715515137,
+      "learning_rate": 3.9477767209873065e-07,
+      "loss": 0.8274,
+      "step": 2485
+    },
+    {
+      "epoch": 0.9109563942836204,
+      "grad_norm": 2.409552812576294,
+      "learning_rate": 3.9156966266983175e-07,
+      "loss": 0.783,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9113228288750458,
+      "grad_norm": 2.666396141052246,
+      "learning_rate": 3.883744804959122e-07,
+      "loss": 0.8067,
+      "step": 2487
+    },
+    {
+      "epoch": 0.9116892634664713,
+      "grad_norm": 2.2847819328308105,
+      "learning_rate": 3.851921298425398e-07,
+      "loss": 0.8637,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9120556980578967,
+      "grad_norm": 2.1684303283691406,
+      "learning_rate": 3.8202261495815296e-07,
+      "loss": 0.7933,
+      "step": 2489
+    },
+    {
+      "epoch": 0.9124221326493221,
+      "grad_norm": 2.689586877822876,
+      "learning_rate": 3.7886594007405354e-07,
+      "loss": 0.7986,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9127885672407475,
+      "grad_norm": 2.387988328933716,
+      "learning_rate": 3.757221094044028e-07,
+      "loss": 0.8394,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9131550018321729,
+      "grad_norm": 2.657233953475952,
+      "learning_rate": 3.725911271462157e-07,
+      "loss": 0.8021,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9135214364235984,
+      "grad_norm": 3.2616617679595947,
+      "learning_rate": 3.6947299747935427e-07,
+      "loss": 0.7529,
+      "step": 2493
+    },
+    {
+      "epoch": 0.9138878710150238,
+      "grad_norm": 2.5014402866363525,
+      "learning_rate": 3.663677245665198e-07,
+      "loss": 0.8123,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9142543056064493,
+      "grad_norm": 3.2536182403564453,
+      "learning_rate": 3.6327531255325285e-07,
+      "loss": 0.75,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9146207401978746,
+      "grad_norm": 2.8663833141326904,
+      "learning_rate": 3.6019576556792336e-07,
+      "loss": 0.7509,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9149871747893001,
+      "grad_norm": 2.2998969554901123,
+      "learning_rate": 3.5712908772172596e-07,
+      "loss": 0.8589,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9153536093807255,
+      "grad_norm": 2.583465337753296,
+      "learning_rate": 3.540752831086769e-07,
+      "loss": 0.7621,
+      "step": 2498
+    },
+    {
+      "epoch": 0.915720043972151,
+      "grad_norm": 2.637712001800537,
+      "learning_rate": 3.510343558056073e-07,
+      "loss": 0.8347,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9160864785635764,
+      "grad_norm": 2.455714225769043,
+      "learning_rate": 3.4800630987215087e-07,
+      "loss": 0.7838,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9164529131550019,
+      "grad_norm": 2.176424980163574,
+      "learning_rate": 3.4499114935075185e-07,
+      "loss": 0.8495,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9168193477464273,
+      "grad_norm": 2.26564621925354,
+      "learning_rate": 3.4198887826664805e-07,
+      "loss": 0.8049,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9171857823378526,
+      "grad_norm": 2.7091007232666016,
+      "learning_rate": 3.3899950062787343e-07,
+      "loss": 0.8091,
+      "step": 2503
+    },
+    {
+      "epoch": 0.9175522169292781,
+      "grad_norm": 2.7910430431365967,
+      "learning_rate": 3.360230204252468e-07,
+      "loss": 0.7887,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9179186515207035,
+      "grad_norm": 2.3435018062591553,
+      "learning_rate": 3.330594416323696e-07,
+      "loss": 0.8108,
+      "step": 2505
+    },
+    {
+      "epoch": 0.918285086112129,
+      "grad_norm": 2.3385210037231445,
+      "learning_rate": 3.3010876820562033e-07,
+      "loss": 0.827,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9186515207035544,
+      "grad_norm": 3.0472352504730225,
+      "learning_rate": 3.271710040841447e-07,
+      "loss": 0.6821,
+      "step": 2507
+    },
+    {
+      "epoch": 0.9190179552949799,
+      "grad_norm": 3.0347681045532227,
+      "learning_rate": 3.2424615318986e-07,
+      "loss": 0.8008,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9193843898864053,
+      "grad_norm": 2.7286577224731445,
+      "learning_rate": 3.213342194274438e-07,
+      "loss": 0.8319,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9197508244778307,
+      "grad_norm": 2.600947618484497,
+      "learning_rate": 3.184352066843266e-07,
+      "loss": 0.7789,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9201172590692561,
+      "grad_norm": 2.3278796672821045,
+      "learning_rate": 3.1554911883068915e-07,
+      "loss": 0.8465,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9204836936606816,
+      "grad_norm": 2.2521705627441406,
+      "learning_rate": 3.1267595971946176e-07,
+      "loss": 0.8016,
+      "step": 2512
+    },
+    {
+      "epoch": 0.920850128252107,
+      "grad_norm": 2.6614413261413574,
+      "learning_rate": 3.0981573318630943e-07,
+      "loss": 0.8114,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9212165628435325,
+      "grad_norm": 2.2802844047546387,
+      "learning_rate": 3.0696844304963557e-07,
+      "loss": 0.7933,
+      "step": 2514
+    },
+    {
+      "epoch": 0.9215829974349579,
+      "grad_norm": 2.396064281463623,
+      "learning_rate": 3.0413409311057405e-07,
+      "loss": 0.8554,
+      "step": 2515
+    },
+    {
+      "epoch": 0.9219494320263832,
+      "grad_norm": 2.4574921131134033,
+      "learning_rate": 3.0131268715298034e-07,
+      "loss": 0.8178,
+      "step": 2516
+    },
+    {
+      "epoch": 0.9223158666178087,
+      "grad_norm": 2.4022841453552246,
+      "learning_rate": 2.9850422894343257e-07,
+      "loss": 0.7644,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9226823012092341,
+      "grad_norm": 2.6130800247192383,
+      "learning_rate": 2.957087222312216e-07,
+      "loss": 0.8014,
+      "step": 2518
+    },
+    {
+      "epoch": 0.9230487358006596,
+      "grad_norm": 2.3932414054870605,
+      "learning_rate": 2.9292617074834995e-07,
+      "loss": 0.7709,
+      "step": 2519
+    },
+    {
+      "epoch": 0.923415170392085,
+      "grad_norm": 2.6321308612823486,
+      "learning_rate": 2.90156578209525e-07,
+      "loss": 0.7843,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9237816049835105,
+      "grad_norm": 2.5634989738464355,
+      "learning_rate": 2.8739994831214923e-07,
+      "loss": 0.7625,
+      "step": 2521
+    },
+    {
+      "epoch": 0.9241480395749359,
+      "grad_norm": 2.321744441986084,
+      "learning_rate": 2.8465628473632656e-07,
+      "loss": 0.8626,
+      "step": 2522
+    },
+    {
+      "epoch": 0.9245144741663613,
+      "grad_norm": 2.713566541671753,
+      "learning_rate": 2.8192559114484817e-07,
+      "loss": 0.7424,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9248809087577867,
+      "grad_norm": 2.113034963607788,
+      "learning_rate": 2.7920787118318917e-07,
+      "loss": 0.8691,
+      "step": 2524
+    },
+    {
+      "epoch": 0.9252473433492122,
+      "grad_norm": 2.3407957553863525,
+      "learning_rate": 2.7650312847950724e-07,
+      "loss": 0.8524,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9256137779406376,
+      "grad_norm": 2.479217767715454,
+      "learning_rate": 2.738113666446307e-07,
+      "loss": 0.8401,
+      "step": 2526
+    },
+    {
+      "epoch": 0.925980212532063,
+      "grad_norm": 3.277244806289673,
+      "learning_rate": 2.711325892720651e-07,
+      "loss": 0.7803,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9263466471234885,
+      "grad_norm": 2.369516611099243,
+      "learning_rate": 2.684667999379786e-07,
+      "loss": 0.8059,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9267130817149138,
+      "grad_norm": 2.7027928829193115,
+      "learning_rate": 2.658140022011968e-07,
+      "loss": 0.7425,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9270795163063393,
+      "grad_norm": 2.667623519897461,
+      "learning_rate": 2.6317419960320777e-07,
+      "loss": 0.7725,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9274459508977647,
+      "grad_norm": 2.241896152496338,
+      "learning_rate": 2.60547395668147e-07,
+      "loss": 0.754,
+      "step": 2531
+    },
+    {
+      "epoch": 0.9278123854891902,
+      "grad_norm": 2.408830404281616,
+      "learning_rate": 2.5793359390279714e-07,
+      "loss": 0.8083,
+      "step": 2532
+    },
+    {
+      "epoch": 0.9281788200806156,
+      "grad_norm": 2.520829439163208,
+      "learning_rate": 2.553327977965847e-07,
+      "loss": 0.8242,
+      "step": 2533
+    },
+    {
+      "epoch": 0.9285452546720411,
+      "grad_norm": 2.2251739501953125,
+      "learning_rate": 2.5274501082157234e-07,
+      "loss": 0.8117,
+      "step": 2534
+    },
+    {
+      "epoch": 0.9289116892634665,
+      "grad_norm": 2.495321750640869,
+      "learning_rate": 2.501702364324554e-07,
+      "loss": 0.8041,
+      "step": 2535
+    },
+    {
+      "epoch": 0.929278123854892,
+      "grad_norm": 2.2292368412017822,
+      "learning_rate": 2.4760847806655997e-07,
+      "loss": 0.8338,
+      "step": 2536
+    },
+    {
+      "epoch": 0.9296445584463173,
+      "grad_norm": 2.5278828144073486,
+      "learning_rate": 2.450597391438303e-07,
+      "loss": 0.8465,
+      "step": 2537
+    },
+    {
+      "epoch": 0.9300109930377428,
+      "grad_norm": 3.1347997188568115,
+      "learning_rate": 2.425240230668358e-07,
+      "loss": 0.749,
+      "step": 2538
+    },
+    {
+      "epoch": 0.9303774276291682,
+      "grad_norm": 2.1962716579437256,
+      "learning_rate": 2.4000133322075515e-07,
+      "loss": 0.874,
+      "step": 2539
+    },
+    {
+      "epoch": 0.9307438622205936,
+      "grad_norm": 2.8005564212799072,
+      "learning_rate": 2.3749167297338117e-07,
+      "loss": 0.734,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9311102968120191,
+      "grad_norm": 2.687960386276245,
+      "learning_rate": 2.3499504567511046e-07,
+      "loss": 0.8169,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9314767314034444,
+      "grad_norm": 2.2408933639526367,
+      "learning_rate": 2.325114546589391e-07,
+      "loss": 0.7654,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9318431659948699,
+      "grad_norm": 2.455065965652466,
+      "learning_rate": 2.3004090324046268e-07,
+      "loss": 0.8108,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9322096005862953,
+      "grad_norm": 2.5029056072235107,
+      "learning_rate": 2.2758339471786628e-07,
+      "loss": 0.8277,
+      "step": 2544
+    },
+    {
+      "epoch": 0.9325760351777208,
+      "grad_norm": 3.1110899448394775,
+      "learning_rate": 2.2513893237192442e-07,
+      "loss": 0.8104,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9329424697691462,
+      "grad_norm": 2.3951544761657715,
+      "learning_rate": 2.2270751946599556e-07,
+      "loss": 0.8182,
+      "step": 2546
+    },
+    {
+      "epoch": 0.9333089043605717,
+      "grad_norm": 2.4128873348236084,
+      "learning_rate": 2.2028915924601436e-07,
+      "loss": 0.833,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9336753389519971,
+      "grad_norm": 3.34865140914917,
+      "learning_rate": 2.178838549404916e-07,
+      "loss": 0.7215,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9340417735434225,
+      "grad_norm": 3.0843262672424316,
+      "learning_rate": 2.1549160976051088e-07,
+      "loss": 0.7681,
+      "step": 2549
+    },
+    {
+      "epoch": 0.9344082081348479,
+      "grad_norm": 2.328216552734375,
+      "learning_rate": 2.1311242689971757e-07,
+      "loss": 0.7778,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9347746427262733,
+      "grad_norm": 2.731987476348877,
+      "learning_rate": 2.1074630953432208e-07,
+      "loss": 0.7579,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9351410773176988,
+      "grad_norm": 2.6157193183898926,
+      "learning_rate": 2.0839326082309097e-07,
+      "loss": 0.7839,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9355075119091242,
+      "grad_norm": 2.2835865020751953,
+      "learning_rate": 2.060532839073437e-07,
+      "loss": 0.8243,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9358739465005497,
+      "grad_norm": 2.4783785343170166,
+      "learning_rate": 2.0372638191095028e-07,
+      "loss": 0.8011,
+      "step": 2554
+    },
+    {
+      "epoch": 0.936240381091975,
+      "grad_norm": 2.413538932800293,
+      "learning_rate": 2.0141255794032477e-07,
+      "loss": 0.8796,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9366068156834005,
+      "grad_norm": 2.5933644771575928,
+      "learning_rate": 1.991118150844229e-07,
+      "loss": 0.8133,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9369732502748259,
+      "grad_norm": 2.2094247341156006,
+      "learning_rate": 1.968241564147344e-07,
+      "loss": 0.7754,
+      "step": 2557
+    },
+    {
+      "epoch": 0.9373396848662514,
+      "grad_norm": 2.5051934719085693,
+      "learning_rate": 1.945495849852863e-07,
+      "loss": 0.7668,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9377061194576768,
+      "grad_norm": 2.1419708728790283,
+      "learning_rate": 1.9228810383262964e-07,
+      "loss": 0.8236,
+      "step": 2559
+    },
+    {
+      "epoch": 0.9380725540491023,
+      "grad_norm": 2.4711034297943115,
+      "learning_rate": 1.900397159758427e-07,
+      "loss": 0.8673,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9384389886405277,
+      "grad_norm": 2.5989861488342285,
+      "learning_rate": 1.8780442441652335e-07,
+      "loss": 0.7914,
+      "step": 2561
+    },
+    {
+      "epoch": 0.9388054232319532,
+      "grad_norm": 2.1352574825286865,
+      "learning_rate": 1.855822321387868e-07,
+      "loss": 0.8208,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9391718578233785,
+      "grad_norm": 2.5791261196136475,
+      "learning_rate": 1.8337314210925994e-07,
+      "loss": 0.8302,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9395382924148039,
+      "grad_norm": 2.4570558071136475,
+      "learning_rate": 1.8117715727707596e-07,
+      "loss": 0.8385,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9399047270062294,
+      "grad_norm": 2.640968084335327,
+      "learning_rate": 1.7899428057387536e-07,
+      "loss": 0.8168,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9402711615976548,
+      "grad_norm": 2.4151952266693115,
+      "learning_rate": 1.7682451491379927e-07,
+      "loss": 0.813,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9406375961890803,
+      "grad_norm": 2.6391801834106445,
+      "learning_rate": 1.7466786319348505e-07,
+      "loss": 0.7144,
+      "step": 2567
+    },
+    {
+      "epoch": 0.9410040307805057,
+      "grad_norm": 2.176057815551758,
+      "learning_rate": 1.72524328292063e-07,
+      "loss": 0.9107,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9413704653719311,
+      "grad_norm": 2.6201252937316895,
+      "learning_rate": 1.7039391307115184e-07,
+      "loss": 0.76,
+      "step": 2569
+    },
+    {
+      "epoch": 0.9417368999633565,
+      "grad_norm": 2.5767018795013428,
+      "learning_rate": 1.6827662037485537e-07,
+      "loss": 0.8244,
+      "step": 2570
+    },
+    {
+      "epoch": 0.942103334554782,
+      "grad_norm": 2.420722246170044,
+      "learning_rate": 1.6617245302976147e-07,
+      "loss": 0.7693,
+      "step": 2571
+    },
+    {
+      "epoch": 0.9424697691462074,
+      "grad_norm": 2.330512046813965,
+      "learning_rate": 1.6408141384493203e-07,
+      "loss": 0.8218,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9428362037376329,
+      "grad_norm": 2.759117364883423,
+      "learning_rate": 1.6200350561190404e-07,
+      "loss": 0.7241,
+      "step": 2573
+    },
+    {
+      "epoch": 0.9432026383290583,
+      "grad_norm": 2.423530101776123,
+      "learning_rate": 1.5993873110468627e-07,
+      "loss": 0.8249,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9435690729204836,
+      "grad_norm": 2.5526938438415527,
+      "learning_rate": 1.578870930797538e-07,
+      "loss": 0.809,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9439355075119091,
+      "grad_norm": 2.314013719558716,
+      "learning_rate": 1.5584859427604017e-07,
+      "loss": 0.8529,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9443019421033345,
+      "grad_norm": 2.5163190364837646,
+      "learning_rate": 1.5382323741494508e-07,
+      "loss": 0.7894,
+      "step": 2577
+    },
+    {
+      "epoch": 0.94466837669476,
+      "grad_norm": 2.736426591873169,
+      "learning_rate": 1.5181102520031688e-07,
+      "loss": 0.8482,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9450348112861854,
+      "grad_norm": 2.334437370300293,
+      "learning_rate": 1.4981196031846113e-07,
+      "loss": 0.8432,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9454012458776109,
+      "grad_norm": 2.745434284210205,
+      "learning_rate": 1.478260454381286e-07,
+      "loss": 0.8045,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9457676804690363,
+      "grad_norm": 2.711455821990967,
+      "learning_rate": 1.458532832105153e-07,
+      "loss": 0.759,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9461341150604617,
+      "grad_norm": 2.373372793197632,
+      "learning_rate": 1.438936762692611e-07,
+      "loss": 0.8101,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9465005496518871,
+      "grad_norm": 2.2710046768188477,
+      "learning_rate": 1.4194722723043785e-07,
+      "loss": 0.839,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9468669842433126,
+      "grad_norm": 2.3727807998657227,
+      "learning_rate": 1.4001393869255808e-07,
+      "loss": 0.7941,
+      "step": 2584
+    },
+    {
+      "epoch": 0.947233418834738,
+      "grad_norm": 2.5571014881134033,
+      "learning_rate": 1.3809381323656058e-07,
+      "loss": 0.761,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9475998534261634,
+      "grad_norm": 2.3163387775421143,
+      "learning_rate": 1.3618685342581484e-07,
+      "loss": 0.8199,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9479662880175889,
+      "grad_norm": 2.1683309078216553,
+      "learning_rate": 1.3429306180611e-07,
+      "loss": 0.8718,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9483327226090142,
+      "grad_norm": 2.4830777645111084,
+      "learning_rate": 1.324124409056593e-07,
+      "loss": 0.7473,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9486991572004397,
+      "grad_norm": 2.4777321815490723,
+      "learning_rate": 1.3054499323508995e-07,
+      "loss": 0.8145,
+      "step": 2589
+    },
+    {
+      "epoch": 0.9490655917918651,
+      "grad_norm": 2.565575361251831,
+      "learning_rate": 1.2869072128744663e-07,
+      "loss": 0.7228,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9494320263832906,
+      "grad_norm": 2.5673301219940186,
+      "learning_rate": 1.2684962753818252e-07,
+      "loss": 0.8402,
+      "step": 2591
+    },
+    {
+      "epoch": 0.949798460974716,
+      "grad_norm": 2.938429594039917,
+      "learning_rate": 1.250217144451571e-07,
+      "loss": 0.727,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9501648955661415,
+      "grad_norm": 2.7190816402435303,
+      "learning_rate": 1.23206984448635e-07,
+      "loss": 0.7654,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9505313301575669,
+      "grad_norm": 2.4703621864318848,
+      "learning_rate": 1.214054399712794e-07,
+      "loss": 0.8063,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9508977647489923,
+      "grad_norm": 2.362274408340454,
+      "learning_rate": 1.1961708341815316e-07,
+      "loss": 0.7834,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9512641993404177,
+      "grad_norm": 2.552286386489868,
+      "learning_rate": 1.1784191717671312e-07,
+      "loss": 0.8552,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9516306339318432,
+      "grad_norm": 2.515406847000122,
+      "learning_rate": 1.1607994361680585e-07,
+      "loss": 0.8017,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9519970685232686,
+      "grad_norm": 2.715442180633545,
+      "learning_rate": 1.1433116509066644e-07,
+      "loss": 0.7763,
+      "step": 2598
+    },
+    {
+      "epoch": 0.952363503114694,
+      "grad_norm": 2.3628156185150146,
+      "learning_rate": 1.1259558393291292e-07,
+      "loss": 0.825,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9527299377061195,
+      "grad_norm": 2.3146297931671143,
+      "learning_rate": 1.1087320246054634e-07,
+      "loss": 0.8421,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9530963722975448,
+      "grad_norm": 2.279325485229492,
+      "learning_rate": 1.0916402297294848e-07,
+      "loss": 0.8499,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9534628068889703,
+      "grad_norm": 2.527597427368164,
+      "learning_rate": 1.0746804775187081e-07,
+      "loss": 0.7355,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9538292414803957,
+      "grad_norm": 2.3891003131866455,
+      "learning_rate": 1.0578527906144332e-07,
+      "loss": 0.7888,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9541956760718212,
+      "grad_norm": 2.7313292026519775,
+      "learning_rate": 1.0411571914816121e-07,
+      "loss": 0.8209,
+      "step": 2604
+    },
+    {
+      "epoch": 0.9545621106632466,
+      "grad_norm": 2.987650156021118,
+      "learning_rate": 1.0245937024088493e-07,
+      "loss": 0.8033,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9549285452546721,
+      "grad_norm": 2.2838006019592285,
+      "learning_rate": 1.0081623455084344e-07,
+      "loss": 0.8121,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9552949798460975,
+      "grad_norm": 2.8768584728240967,
+      "learning_rate": 9.918631427162207e-08,
+      "loss": 0.725,
+      "step": 2607
+    },
+    {
+      "epoch": 0.955661414437523,
+      "grad_norm": 2.493642807006836,
+      "learning_rate": 9.756961157916355e-08,
+      "loss": 0.784,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9560278490289483,
+      "grad_norm": 2.661923885345459,
+      "learning_rate": 9.596612863176702e-08,
+      "loss": 0.7796,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9563942836203737,
+      "grad_norm": 2.2444090843200684,
+      "learning_rate": 9.437586757008121e-08,
+      "loss": 0.7572,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9567607182117992,
+      "grad_norm": 2.4184560775756836,
+      "learning_rate": 9.27988305171057e-08,
+      "loss": 0.7662,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9571271528032246,
+      "grad_norm": 2.5148234367370605,
+      "learning_rate": 9.123501957818415e-08,
+      "loss": 0.8242,
+      "step": 2612
+    },
+    {
+      "epoch": 0.9574935873946501,
+      "grad_norm": 3.0104429721832275,
+      "learning_rate": 8.968443684100659e-08,
+      "loss": 0.7282,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9578600219860754,
+      "grad_norm": 2.3206191062927246,
+      "learning_rate": 8.814708437559827e-08,
+      "loss": 0.8091,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9582264565775009,
+      "grad_norm": 2.781954288482666,
+      "learning_rate": 8.662296423432526e-08,
+      "loss": 0.8469,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9585928911689263,
+      "grad_norm": 2.618665933609009,
+      "learning_rate": 8.511207845188884e-08,
+      "loss": 0.7353,
+      "step": 2616
+    },
+    {
+      "epoch": 0.9589593257603518,
+      "grad_norm": 2.366697311401367,
+      "learning_rate": 8.361442904532114e-08,
+      "loss": 0.8014,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9593257603517772,
+      "grad_norm": 2.34576153755188,
+      "learning_rate": 8.213001801398501e-08,
+      "loss": 0.761,
+      "step": 2618
+    },
+    {
+      "epoch": 0.9596921949432027,
+      "grad_norm": 2.4065608978271484,
+      "learning_rate": 8.065884733956864e-08,
+      "loss": 0.8039,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9600586295346281,
+      "grad_norm": 2.665904998779297,
+      "learning_rate": 7.920091898608429e-08,
+      "loss": 0.7685,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9604250641260536,
+      "grad_norm": 2.423003673553467,
+      "learning_rate": 7.775623489986728e-08,
+      "loss": 0.8128,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9607914987174789,
+      "grad_norm": 2.3617613315582275,
+      "learning_rate": 7.632479700957152e-08,
+      "loss": 0.7785,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9611579333089043,
+      "grad_norm": 2.3319296836853027,
+      "learning_rate": 7.490660722616505e-08,
+      "loss": 0.8372,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9615243679003298,
+      "grad_norm": 2.333374500274658,
+      "learning_rate": 7.35016674429323e-08,
+      "loss": 0.8241,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9618908024917552,
+      "grad_norm": 2.2863078117370605,
+      "learning_rate": 7.21099795354674e-08,
+      "loss": 0.8183,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9622572370831807,
+      "grad_norm": 2.8025894165039062,
+      "learning_rate": 7.073154536167415e-08,
+      "loss": 0.7713,
+      "step": 2626
+    },
+    {
+      "epoch": 0.962623671674606,
+      "grad_norm": 2.2680001258850098,
+      "learning_rate": 6.936636676176278e-08,
+      "loss": 0.8475,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9629901062660315,
+      "grad_norm": 2.855222225189209,
+      "learning_rate": 6.801444555824543e-08,
+      "loss": 0.7444,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9633565408574569,
+      "grad_norm": 2.8795368671417236,
+      "learning_rate": 6.667578355593951e-08,
+      "loss": 0.6874,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9637229754488824,
+      "grad_norm": 2.410686731338501,
+      "learning_rate": 6.535038254195547e-08,
+      "loss": 0.7999,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9640894100403078,
+      "grad_norm": 2.366168737411499,
+      "learning_rate": 6.403824428570571e-08,
+      "loss": 0.8394,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9644558446317333,
+      "grad_norm": 2.3753418922424316,
+      "learning_rate": 6.273937053889456e-08,
+      "loss": 0.7992,
+      "step": 2632
+    },
+    {
+      "epoch": 0.9648222792231587,
+      "grad_norm": 2.5532946586608887,
+      "learning_rate": 6.145376303551942e-08,
+      "loss": 0.8015,
+      "step": 2633
+    },
+    {
+      "epoch": 0.965188713814584,
+      "grad_norm": 2.54056715965271,
+      "learning_rate": 6.018142349186518e-08,
+      "loss": 0.8013,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9655551484060095,
+      "grad_norm": 2.5784850120544434,
+      "learning_rate": 5.8922353606504224e-08,
+      "loss": 0.8266,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9659215829974349,
+      "grad_norm": 2.5065808296203613,
+      "learning_rate": 5.767655506029646e-08,
+      "loss": 0.8145,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9662880175888604,
+      "grad_norm": 2.7400426864624023,
+      "learning_rate": 5.644402951638372e-08,
+      "loss": 0.7818,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9666544521802858,
+      "grad_norm": 2.1992766857147217,
+      "learning_rate": 5.522477862018538e-08,
+      "loss": 0.8637,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9670208867717113,
+      "grad_norm": 2.5704853534698486,
+      "learning_rate": 5.401880399940385e-08,
+      "loss": 0.7717,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9673873213631367,
+      "grad_norm": 2.2856333255767822,
+      "learning_rate": 5.28261072640146e-08,
+      "loss": 0.8073,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9677537559545621,
+      "grad_norm": 2.359670877456665,
+      "learning_rate": 5.164669000626621e-08,
+      "loss": 0.8069,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9681201905459875,
+      "grad_norm": 2.493969678878784,
+      "learning_rate": 5.048055380068251e-08,
+      "loss": 0.8215,
+      "step": 2642
+    },
+    {
+      "epoch": 0.968486625137413,
+      "grad_norm": 2.170348882675171,
+      "learning_rate": 4.932770020405486e-08,
+      "loss": 0.8041,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9688530597288384,
+      "grad_norm": 2.3696866035461426,
+      "learning_rate": 4.818813075544215e-08,
+      "loss": 0.8614,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9692194943202639,
+      "grad_norm": 2.762871503829956,
+      "learning_rate": 4.706184697617189e-08,
+      "loss": 0.8588,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9695859289116893,
+      "grad_norm": 2.926147222518921,
+      "learning_rate": 4.594885036983021e-08,
+      "loss": 0.7791,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9699523635031146,
+      "grad_norm": 2.2866110801696777,
+      "learning_rate": 4.484914242226967e-08,
+      "loss": 0.7318,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9703187980945401,
+      "grad_norm": 2.318859338760376,
+      "learning_rate": 4.376272460159925e-08,
+      "loss": 0.7696,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9706852326859655,
+      "grad_norm": 2.466014862060547,
+      "learning_rate": 4.2689598358187646e-08,
+      "loss": 0.8135,
+      "step": 2649
+    },
+    {
+      "epoch": 0.971051667277391,
+      "grad_norm": 3.416146755218506,
+      "learning_rate": 4.1629765124656665e-08,
+      "loss": 0.7498,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9714181018688164,
+      "grad_norm": 3.0079779624938965,
+      "learning_rate": 4.0583226315886736e-08,
+      "loss": 0.8217,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9717845364602419,
+      "grad_norm": 2.374130964279175,
+      "learning_rate": 3.954998332900473e-08,
+      "loss": 0.7957,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9721509710516673,
+      "grad_norm": 2.347937822341919,
+      "learning_rate": 3.853003754339057e-08,
+      "loss": 0.853,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9725174056430927,
+      "grad_norm": 2.450432062149048,
+      "learning_rate": 3.752339032067287e-08,
+      "loss": 0.8669,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9728838402345181,
+      "grad_norm": 3.1508584022521973,
+      "learning_rate": 3.653004300472551e-08,
+      "loss": 0.8051,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9732502748259436,
+      "grad_norm": 2.312556505203247,
+      "learning_rate": 3.5549996921666605e-08,
+      "loss": 0.8128,
+      "step": 2656
+    },
+    {
+      "epoch": 0.973616709417369,
+      "grad_norm": 2.2593586444854736,
+      "learning_rate": 3.458325337985735e-08,
+      "loss": 0.8601,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9739831440087944,
+      "grad_norm": 2.395019054412842,
+      "learning_rate": 3.362981366990092e-08,
+      "loss": 0.7701,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9743495786002199,
+      "grad_norm": 2.3906631469726562,
+      "learning_rate": 3.2689679064640264e-08,
+      "loss": 0.8176,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9747160131916452,
+      "grad_norm": 2.3747525215148926,
+      "learning_rate": 3.176285081915365e-08,
+      "loss": 0.7992,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9750824477830707,
+      "grad_norm": 2.841897964477539,
+      "learning_rate": 3.084933017075797e-08,
+      "loss": 0.7881,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9754488823744961,
+      "grad_norm": 2.2802422046661377,
+      "learning_rate": 2.994911833900216e-08,
+      "loss": 0.888,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9758153169659216,
+      "grad_norm": 3.2933173179626465,
+      "learning_rate": 2.906221652567043e-08,
+      "loss": 0.7663,
+      "step": 2663
+    },
+    {
+      "epoch": 0.976181751557347,
+      "grad_norm": 2.538501024246216,
+      "learning_rate": 2.818862591477567e-08,
+      "loss": 0.779,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9765481861487725,
+      "grad_norm": 2.564566135406494,
+      "learning_rate": 2.732834767256276e-08,
+      "loss": 0.8072,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9769146207401979,
+      "grad_norm": 2.49530291557312,
+      "learning_rate": 2.648138294750191e-08,
+      "loss": 0.8292,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9772810553316233,
+      "grad_norm": 2.2448463439941406,
+      "learning_rate": 2.5647732870290877e-08,
+      "loss": 0.7876,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9776474899230487,
+      "grad_norm": 2.516613245010376,
+      "learning_rate": 2.4827398553852744e-08,
+      "loss": 0.7823,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9780139245144742,
+      "grad_norm": 2.4173407554626465,
+      "learning_rate": 2.402038109333482e-08,
+      "loss": 0.8346,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9783803591058996,
+      "grad_norm": 2.280996322631836,
+      "learning_rate": 2.3226681566105303e-08,
+      "loss": 0.8332,
+      "step": 2670
+    },
+    {
+      "epoch": 0.978746793697325,
+      "grad_norm": 2.479410409927368,
+      "learning_rate": 2.244630103175216e-08,
+      "loss": 0.8781,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9791132282887505,
+      "grad_norm": 2.1932010650634766,
+      "learning_rate": 2.1679240532083147e-08,
+      "loss": 0.8298,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9794796628801758,
+      "grad_norm": 2.397739887237549,
+      "learning_rate": 2.0925501091125788e-08,
+      "loss": 0.8105,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9798460974716013,
+      "grad_norm": 2.362424612045288,
+      "learning_rate": 2.0185083715121845e-08,
+      "loss": 0.8627,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9802125320630267,
+      "grad_norm": 2.419423818588257,
+      "learning_rate": 1.9457989392527298e-08,
+      "loss": 0.8018,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9805789666544522,
+      "grad_norm": 2.3025901317596436,
+      "learning_rate": 1.8744219094013473e-08,
+      "loss": 0.805,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9809454012458776,
+      "grad_norm": 2.2877860069274902,
+      "learning_rate": 1.8043773772464802e-08,
+      "loss": 0.7481,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9813118358373031,
+      "grad_norm": 2.5664076805114746,
+      "learning_rate": 1.73566543629744e-08,
+      "loss": 0.8295,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9816782704287285,
+      "grad_norm": 2.4994051456451416,
+      "learning_rate": 1.6682861782848503e-08,
+      "loss": 0.8511,
+      "step": 2679
+    },
+    {
+      "epoch": 0.982044705020154,
+      "grad_norm": 2.296548366546631,
+      "learning_rate": 1.6022396931598682e-08,
+      "loss": 0.8148,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9824111396115793,
+      "grad_norm": 2.2606639862060547,
+      "learning_rate": 1.5375260690945194e-08,
+      "loss": 0.8016,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9827775742030047,
+      "grad_norm": 2.6477155685424805,
+      "learning_rate": 1.4741453924816962e-08,
+      "loss": 0.7658,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9831440087944302,
+      "grad_norm": 2.2549641132354736,
+      "learning_rate": 1.4120977479344933e-08,
+      "loss": 0.8707,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9835104433858556,
+      "grad_norm": 2.463927745819092,
+      "learning_rate": 1.3513832182864284e-08,
+      "loss": 0.7888,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9838768779772811,
+      "grad_norm": 2.3416614532470703,
+      "learning_rate": 1.2920018845915539e-08,
+      "loss": 0.7944,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9842433125687065,
+      "grad_norm": 2.617210626602173,
+      "learning_rate": 1.2339538261239015e-08,
+      "loss": 0.8209,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9846097471601319,
+      "grad_norm": 2.4247403144836426,
+      "learning_rate": 1.1772391203775934e-08,
+      "loss": 0.7562,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9849761817515573,
+      "grad_norm": 2.344165325164795,
+      "learning_rate": 1.1218578430667315e-08,
+      "loss": 0.8384,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9853426163429828,
+      "grad_norm": 2.7740414142608643,
+      "learning_rate": 1.0678100681255077e-08,
+      "loss": 0.8237,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9857090509344082,
+      "grad_norm": 3.037654399871826,
+      "learning_rate": 1.0150958677075384e-08,
+      "loss": 0.6686,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9860754855258337,
+      "grad_norm": 2.509370803833008,
+      "learning_rate": 9.637153121863085e-09,
+      "loss": 0.7722,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9864419201172591,
+      "grad_norm": 2.465951442718506,
+      "learning_rate": 9.136684701548381e-09,
+      "loss": 0.7896,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9868083547086846,
+      "grad_norm": 2.5230894088745117,
+      "learning_rate": 8.649554084256829e-09,
+      "loss": 0.8261,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9871747893001099,
+      "grad_norm": 3.178312301635742,
+      "learning_rate": 8.175761920309332e-09,
+      "loss": 0.724,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9875412238915353,
+      "grad_norm": 2.954385757446289,
+      "learning_rate": 7.715308842215496e-09,
+      "loss": 0.7312,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9879076584829608,
+      "grad_norm": 2.2207462787628174,
+      "learning_rate": 7.268195464683603e-09,
+      "loss": 0.832,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9882740930743862,
+      "grad_norm": 2.27506947517395,
+      "learning_rate": 6.834422384607298e-09,
+      "loss": 0.7862,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9886405276658117,
+      "grad_norm": 2.9832675457000732,
+      "learning_rate": 6.41399018107447e-09,
+      "loss": 0.8149,
+      "step": 2698
+    },
+    {
+      "epoch": 0.989006962257237,
+      "grad_norm": 2.37428617477417,
+      "learning_rate": 6.00689941536281e-09,
+      "loss": 0.8005,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9893733968486625,
+      "grad_norm": 2.8593597412109375,
+      "learning_rate": 5.613150630938702e-09,
+      "loss": 0.7504,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9897398314400879,
+      "grad_norm": 2.6248154640197754,
+      "learning_rate": 5.2327443534549994e-09,
+      "loss": 0.8944,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9901062660315134,
+      "grad_norm": 2.416533946990967,
+      "learning_rate": 4.86568109075547e-09,
+      "loss": 0.8241,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9904727006229388,
+      "grad_norm": 2.352708578109741,
+      "learning_rate": 4.511961332869241e-09,
+      "loss": 0.8209,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9908391352143643,
+      "grad_norm": 2.675865411758423,
+      "learning_rate": 4.1715855520130246e-09,
+      "loss": 0.7851,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9912055698057897,
+      "grad_norm": 2.2010722160339355,
+      "learning_rate": 3.84455420258667e-09,
+      "loss": 0.8281,
+      "step": 2705
+    },
+    {
+      "epoch": 0.991572004397215,
+      "grad_norm": 2.4950497150421143,
+      "learning_rate": 3.5308677211798313e-09,
+      "loss": 0.8054,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9919384389886405,
+      "grad_norm": 2.262019157409668,
+      "learning_rate": 3.2305265265608622e-09,
+      "loss": 0.7843,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9923048735800659,
+      "grad_norm": 2.4214706420898438,
+      "learning_rate": 2.943531019686807e-09,
+      "loss": 0.8217,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9926713081714914,
+      "grad_norm": 2.6333794593811035,
+      "learning_rate": 2.669881583696743e-09,
+      "loss": 0.7976,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9930377427629168,
+      "grad_norm": 2.6949565410614014,
+      "learning_rate": 2.409578583912886e-09,
+      "loss": 0.813,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9934041773543423,
+      "grad_norm": 2.509366750717163,
+      "learning_rate": 2.162622367839484e-09,
+      "loss": 0.7697,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9937706119457677,
+      "grad_norm": 2.343609094619751,
+      "learning_rate": 1.9290132651628157e-09,
+      "loss": 0.781,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9941370465371931,
+      "grad_norm": 2.2084622383117676,
+      "learning_rate": 1.7087515877511895e-09,
+      "loss": 0.8529,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9945034811286185,
+      "grad_norm": 2.560805320739746,
+      "learning_rate": 1.501837629653835e-09,
+      "loss": 0.7243,
+      "step": 2714
+    },
+    {
+      "epoch": 0.994869915720044,
+      "grad_norm": 2.592674970626831,
+      "learning_rate": 1.3082716671020123e-09,
+      "loss": 0.8383,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9952363503114694,
+      "grad_norm": 2.773082733154297,
+      "learning_rate": 1.1280539585045713e-09,
+      "loss": 0.8462,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9956027849028949,
+      "grad_norm": 2.683271646499634,
+      "learning_rate": 9.611847444523925e-10,
+      "loss": 0.7387,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9959692194943203,
+      "grad_norm": 3.2599453926086426,
+      "learning_rate": 8.076642477172769e-10,
+      "loss": 0.7307,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9963356540857456,
+      "grad_norm": 2.323328971862793,
+      "learning_rate": 6.674926732475051e-10,
+      "loss": 0.8463,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9967020886771711,
+      "grad_norm": 2.6469409465789795,
+      "learning_rate": 5.406702081733883e-10,
+      "loss": 0.7977,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9970685232685965,
+      "grad_norm": 2.4954774379730225,
+      "learning_rate": 4.2719702180282763e-10,
+      "loss": 0.8084,
+      "step": 2721
+    },
+    {
+      "epoch": 0.997434957860022,
+      "grad_norm": 2.8149211406707764,
+      "learning_rate": 3.270732656213138e-10,
+      "loss": 0.8145,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9978013924514474,
+      "grad_norm": 2.791076421737671,
+      "learning_rate": 2.4029907329525815e-10,
+      "loss": 0.7137,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9981678270428729,
+      "grad_norm": 2.6285765171051025,
+      "learning_rate": 1.668745606675515e-10,
+      "loss": 0.7838,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9985342616342983,
+      "grad_norm": 2.47336483001709,
+      "learning_rate": 1.0679982576089487e-10,
+      "loss": 0.813,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9989006962257237,
+      "grad_norm": 2.5160090923309326,
+      "learning_rate": 6.007494877335873e-11,
+      "loss": 0.7567,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9992671308171491,
+      "grad_norm": 2.384516954421997,
+      "learning_rate": 2.6699992085044146e-11,
+      "loss": 0.8212,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9996335654085746,
+      "grad_norm": 3.125704050064087,
+      "learning_rate": 6.675000249201091e-12,
+      "loss": 0.7671,
+      "step": 2728
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.552747964859009,
+      "learning_rate": 0.0,
+      "loss": 0.7764,
+      "step": 2729
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2729,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.426020372786971e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}