diff --git "a/qwen_2_5_7b_math_instruct/limo/trainer_state.json" "b/qwen_2_5_7b_math_instruct/limo/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/qwen_2_5_7b_math_instruct/limo/trainer_state.json"
@@ -0,0 +1,14393 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 2050,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004878048780487805,
+      "grad_norm": 37.90617752075195,
+      "learning_rate": 5e-06,
+      "loss": 5.3731,
+      "step": 1
+    },
+    {
+      "epoch": 0.00975609756097561,
+      "grad_norm": 32.92101287841797,
+      "learning_rate": 4.999997064365715e-06,
+      "loss": 3.7922,
+      "step": 2
+    },
+    {
+      "epoch": 0.014634146341463415,
+      "grad_norm": 28.488866806030273,
+      "learning_rate": 4.999988257469751e-06,
+      "loss": 3.3238,
+      "step": 3
+    },
+    {
+      "epoch": 0.01951219512195122,
+      "grad_norm": 28.274654388427734,
+      "learning_rate": 4.999973579332793e-06,
+      "loss": 3.5405,
+      "step": 4
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 18.761985778808594,
+      "learning_rate": 4.999953029989312e-06,
+      "loss": 2.7492,
+      "step": 5
+    },
+    {
+      "epoch": 0.02926829268292683,
+      "grad_norm": 9.646541595458984,
+      "learning_rate": 4.999926609487568e-06,
+      "loss": 1.7146,
+      "step": 6
+    },
+    {
+      "epoch": 0.03414634146341464,
+      "grad_norm": 15.763612747192383,
+      "learning_rate": 4.9998943178896106e-06,
+      "loss": 2.8485,
+      "step": 7
+    },
+    {
+      "epoch": 0.03902439024390244,
+      "grad_norm": 13.853349685668945,
+      "learning_rate": 4.999856155271276e-06,
+      "loss": 3.0975,
+      "step": 8
+    },
+    {
+      "epoch": 0.04390243902439024,
+      "grad_norm": 16.893617630004883,
+      "learning_rate": 4.999812121722191e-06,
+      "loss": 2.8534,
+      "step": 9
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 15.47554874420166,
+      "learning_rate": 4.999762217345766e-06,
+      "loss": 3.4569,
+      "step": 10
+    },
+    {
+      "epoch": 0.05365853658536585,
+      "grad_norm": 6.277958393096924,
+      "learning_rate": 4.999706442259205e-06,
+      "loss": 1.8088,
+      "step": 11
+    },
+    {
+      "epoch": 0.05853658536585366,
+      "grad_norm": 5.451688289642334,
+      "learning_rate": 4.999644796593492e-06,
+      "loss": 1.8104,
+      "step": 12
+    },
+    {
+      "epoch": 0.06341463414634146,
+      "grad_norm": 4.333822727203369,
+      "learning_rate": 4.999577280493407e-06,
+      "loss": 1.7201,
+      "step": 13
+    },
+    {
+      "epoch": 0.06829268292682927,
+      "grad_norm": 4.488610744476318,
+      "learning_rate": 4.99950389411751e-06,
+      "loss": 1.9804,
+      "step": 14
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 1.8763024806976318,
+      "learning_rate": 4.999424637638148e-06,
+      "loss": 0.9227,
+      "step": 15
+    },
+    {
+      "epoch": 0.07804878048780488,
+      "grad_norm": 2.98995304107666,
+      "learning_rate": 4.999339511241458e-06,
+      "loss": 1.2368,
+      "step": 16
+    },
+    {
+      "epoch": 0.08292682926829269,
+      "grad_norm": 5.778397083282471,
+      "learning_rate": 4.9992485151273584e-06,
+      "loss": 2.5745,
+      "step": 17
+    },
+    {
+      "epoch": 0.08780487804878048,
+      "grad_norm": 5.222808361053467,
+      "learning_rate": 4.999151649509554e-06,
+      "loss": 2.0987,
+      "step": 18
+    },
+    {
+      "epoch": 0.09268292682926829,
+      "grad_norm": 4.075173377990723,
+      "learning_rate": 4.9990489146155356e-06,
+      "loss": 1.7732,
+      "step": 19
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 4.210625171661377,
+      "learning_rate": 4.9989403106865765e-06,
+      "loss": 1.8928,
+      "step": 20
+    },
+    {
+      "epoch": 0.1024390243902439,
+      "grad_norm": 1.7968504428863525,
+      "learning_rate": 4.9988258379777334e-06,
+      "loss": 1.0457,
+      "step": 21
+    },
+    {
+      "epoch": 0.1073170731707317,
+      "grad_norm": 2.494248151779175,
+      "learning_rate": 4.998705496757846e-06,
+      "loss": 1.3236,
+      "step": 22
+    },
+    {
+      "epoch": 0.11219512195121951,
+      "grad_norm": 4.312766075134277,
+      "learning_rate": 4.998579287309538e-06,
+      "loss": 2.4289,
+      "step": 23
+    },
+    {
+      "epoch": 0.11707317073170732,
+      "grad_norm": 5.361437797546387,
+      "learning_rate": 4.998447209929211e-06,
+      "loss": 1.7608,
+      "step": 24
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 1.4642046689987183,
+      "learning_rate": 4.998309264927053e-06,
+      "loss": 0.7571,
+      "step": 25
+    },
+    {
+      "epoch": 0.12682926829268293,
+      "grad_norm": 3.642005205154419,
+      "learning_rate": 4.998165452627025e-06,
+      "loss": 1.0634,
+      "step": 26
+    },
+    {
+      "epoch": 0.13170731707317074,
+      "grad_norm": 3.2727863788604736,
+      "learning_rate": 4.998015773366874e-06,
+      "loss": 1.4016,
+      "step": 27
+    },
+    {
+      "epoch": 0.13658536585365855,
+      "grad_norm": 1.5661338567733765,
+      "learning_rate": 4.997860227498122e-06,
+      "loss": 0.8305,
+      "step": 28
+    },
+    {
+      "epoch": 0.14146341463414633,
+      "grad_norm": 3.3989508152008057,
+      "learning_rate": 4.99769881538607e-06,
+      "loss": 1.8744,
+      "step": 29
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 3.512221574783325,
+      "learning_rate": 4.997531537409794e-06,
+      "loss": 1.7701,
+      "step": 30
+    },
+    {
+      "epoch": 0.15121951219512195,
+      "grad_norm": 2.261887311935425,
+      "learning_rate": 4.99735839396215e-06,
+      "loss": 1.0653,
+      "step": 31
+    },
+    {
+      "epoch": 0.15609756097560976,
+      "grad_norm": 2.2495696544647217,
+      "learning_rate": 4.9971793854497655e-06,
+      "loss": 1.1914,
+      "step": 32
+    },
+    {
+      "epoch": 0.16097560975609757,
+      "grad_norm": 4.262424468994141,
+      "learning_rate": 4.996994512293042e-06,
+      "loss": 1.7311,
+      "step": 33
+    },
+    {
+      "epoch": 0.16585365853658537,
+      "grad_norm": 3.361311674118042,
+      "learning_rate": 4.996803774926157e-06,
+      "loss": 1.4482,
+      "step": 34
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 1.5855786800384521,
+      "learning_rate": 4.996607173797059e-06,
+      "loss": 1.3959,
+      "step": 35
+    },
+    {
+      "epoch": 0.17560975609756097,
+      "grad_norm": 2.5001468658447266,
+      "learning_rate": 4.996404709367466e-06,
+      "loss": 1.3452,
+      "step": 36
+    },
+    {
+      "epoch": 0.18048780487804877,
+      "grad_norm": 2.7903096675872803,
+      "learning_rate": 4.996196382112868e-06,
+      "loss": 1.018,
+      "step": 37
+    },
+    {
+      "epoch": 0.18536585365853658,
+      "grad_norm": 2.7586171627044678,
+      "learning_rate": 4.9959821925225235e-06,
+      "loss": 1.291,
+      "step": 38
+    },
+    {
+      "epoch": 0.1902439024390244,
+      "grad_norm": 2.873840808868408,
+      "learning_rate": 4.995762141099456e-06,
+      "loss": 1.1418,
+      "step": 39
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 1.6896076202392578,
+      "learning_rate": 4.995536228360461e-06,
+      "loss": 1.1662,
+      "step": 40
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4597285985946655,
+      "learning_rate": 4.995304454836095e-06,
+      "loss": 1.0373,
+      "step": 41
+    },
+    {
+      "epoch": 0.2048780487804878,
+      "grad_norm": 2.4222421646118164,
+      "learning_rate": 4.9950668210706795e-06,
+      "loss": 1.1895,
+      "step": 42
+    },
+    {
+      "epoch": 0.2097560975609756,
+      "grad_norm": 3.295085906982422,
+      "learning_rate": 4.994823327622299e-06,
+      "loss": 1.4166,
+      "step": 43
+    },
+    {
+      "epoch": 0.2146341463414634,
+      "grad_norm": 1.8216716051101685,
+      "learning_rate": 4.9945739750628e-06,
+      "loss": 1.0009,
+      "step": 44
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 2.2421364784240723,
+      "learning_rate": 4.994318763977789e-06,
+      "loss": 1.1999,
+      "step": 45
+    },
+    {
+      "epoch": 0.22439024390243903,
+      "grad_norm": 5.6812744140625,
+      "learning_rate": 4.994057694966632e-06,
+      "loss": 1.3831,
+      "step": 46
+    },
+    {
+      "epoch": 0.22926829268292684,
+      "grad_norm": 1.477816104888916,
+      "learning_rate": 4.993790768642449e-06,
+      "loss": 1.1419,
+      "step": 47
+    },
+    {
+      "epoch": 0.23414634146341465,
+      "grad_norm": 4.485177993774414,
+      "learning_rate": 4.99351798563212e-06,
+      "loss": 1.9434,
+      "step": 48
+    },
+    {
+      "epoch": 0.23902439024390243,
+      "grad_norm": 2.770219326019287,
+      "learning_rate": 4.993239346576278e-06,
+      "loss": 1.3214,
+      "step": 49
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 2.721611976623535,
+      "learning_rate": 4.99295485212931e-06,
+      "loss": 1.4329,
+      "step": 50
+    },
+    {
+      "epoch": 0.24878048780487805,
+      "grad_norm": 3.714306592941284,
+      "learning_rate": 4.992664502959351e-06,
+      "loss": 1.372,
+      "step": 51
+    },
+    {
+      "epoch": 0.25365853658536586,
+      "grad_norm": 1.2679803371429443,
+      "learning_rate": 4.99236829974829e-06,
+      "loss": 0.8086,
+      "step": 52
+    },
+    {
+      "epoch": 0.25853658536585367,
+      "grad_norm": 1.6986169815063477,
+      "learning_rate": 4.992066243191762e-06,
+      "loss": 1.2012,
+      "step": 53
+    },
+    {
+      "epoch": 0.2634146341463415,
+      "grad_norm": 1.3443604707717896,
+      "learning_rate": 4.991758333999148e-06,
+      "loss": 0.8986,
+      "step": 54
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 2.0896975994110107,
+      "learning_rate": 4.991444572893575e-06,
+      "loss": 0.8819,
+      "step": 55
+    },
+    {
+      "epoch": 0.2731707317073171,
+      "grad_norm": 1.255516529083252,
+      "learning_rate": 4.991124960611916e-06,
+      "loss": 0.6255,
+      "step": 56
+    },
+    {
+      "epoch": 0.2780487804878049,
+      "grad_norm": 2.125410318374634,
+      "learning_rate": 4.99079949790478e-06,
+      "loss": 0.9983,
+      "step": 57
+    },
+    {
+      "epoch": 0.28292682926829266,
+      "grad_norm": 2.0312907695770264,
+      "learning_rate": 4.99046818553652e-06,
+      "loss": 1.1137,
+      "step": 58
+    },
+    {
+      "epoch": 0.28780487804878047,
+      "grad_norm": 2.904625177383423,
+      "learning_rate": 4.9901310242852246e-06,
+      "loss": 1.2009,
+      "step": 59
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 2.512932777404785,
+      "learning_rate": 4.9897880149427206e-06,
+      "loss": 1.2234,
+      "step": 60
+    },
+    {
+      "epoch": 0.2975609756097561,
+      "grad_norm": 2.5621752738952637,
+      "learning_rate": 4.989439158314566e-06,
+      "loss": 1.2354,
+      "step": 61
+    },
+    {
+      "epoch": 0.3024390243902439,
+      "grad_norm": 1.0051912069320679,
+      "learning_rate": 4.989084455220056e-06,
+      "loss": 0.614,
+      "step": 62
+    },
+    {
+      "epoch": 0.3073170731707317,
+      "grad_norm": 1.5008565187454224,
+      "learning_rate": 4.988723906492212e-06,
+      "loss": 1.002,
+      "step": 63
+    },
+    {
+      "epoch": 0.3121951219512195,
+      "grad_norm": 1.1436376571655273,
+      "learning_rate": 4.988357512977785e-06,
+      "loss": 0.6483,
+      "step": 64
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 1.16792893409729,
+      "learning_rate": 4.987985275537252e-06,
+      "loss": 0.638,
+      "step": 65
+    },
+    {
+      "epoch": 0.32195121951219513,
+      "grad_norm": 1.5670536756515503,
+      "learning_rate": 4.9876071950448185e-06,
+      "loss": 1.0274,
+      "step": 66
+    },
+    {
+      "epoch": 0.32682926829268294,
+      "grad_norm": 1.357515811920166,
+      "learning_rate": 4.987223272388407e-06,
+      "loss": 0.7612,
+      "step": 67
+    },
+    {
+      "epoch": 0.33170731707317075,
+      "grad_norm": 1.8383222818374634,
+      "learning_rate": 4.986833508469663e-06,
+      "loss": 1.1496,
+      "step": 68
+    },
+    {
+      "epoch": 0.33658536585365856,
+      "grad_norm": 2.8976657390594482,
+      "learning_rate": 4.98643790420395e-06,
+      "loss": 0.9924,
+      "step": 69
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 1.7679390907287598,
+      "learning_rate": 4.986036460520348e-06,
+      "loss": 0.9408,
+      "step": 70
+    },
+    {
+      "epoch": 0.3463414634146341,
+      "grad_norm": 2.3815059661865234,
+      "learning_rate": 4.98562917836165e-06,
+      "loss": 1.2923,
+      "step": 71
+    },
+    {
+      "epoch": 0.35121951219512193,
+      "grad_norm": 1.679026484489441,
+      "learning_rate": 4.985216058684362e-06,
+      "loss": 0.8476,
+      "step": 72
+    },
+    {
+      "epoch": 0.35609756097560974,
+      "grad_norm": 1.7254586219787598,
+      "learning_rate": 4.984797102458697e-06,
+      "loss": 1.1775,
+      "step": 73
+    },
+    {
+      "epoch": 0.36097560975609755,
+      "grad_norm": 1.195844292640686,
+      "learning_rate": 4.984372310668579e-06,
+      "loss": 0.6664,
+      "step": 74
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 1.754431128501892,
+      "learning_rate": 4.983941684311633e-06,
+      "loss": 1.4361,
+      "step": 75
+    },
+    {
+      "epoch": 0.37073170731707317,
+      "grad_norm": 1.8255788087844849,
+      "learning_rate": 4.983505224399188e-06,
+      "loss": 0.8977,
+      "step": 76
+    },
+    {
+      "epoch": 0.375609756097561,
+      "grad_norm": 1.340451717376709,
+      "learning_rate": 4.983062931956275e-06,
+      "loss": 0.9432,
+      "step": 77
+    },
+    {
+      "epoch": 0.3804878048780488,
+      "grad_norm": 1.7810503244400024,
+      "learning_rate": 4.9826148080216195e-06,
+      "loss": 1.0463,
+      "step": 78
+    },
+    {
+      "epoch": 0.3853658536585366,
+      "grad_norm": 1.729826807975769,
+      "learning_rate": 4.9821608536476445e-06,
+      "loss": 1.4451,
+      "step": 79
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 2.6638681888580322,
+      "learning_rate": 4.981701069900465e-06,
+      "loss": 1.0409,
+      "step": 80
+    },
+    {
+      "epoch": 0.3951219512195122,
+      "grad_norm": 1.467076301574707,
+      "learning_rate": 4.9812354578598876e-06,
+      "loss": 0.8596,
+      "step": 81
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.453657627105713,
+      "learning_rate": 4.980764018619405e-06,
+      "loss": 0.9033,
+      "step": 82
+    },
+    {
+      "epoch": 0.40487804878048783,
+      "grad_norm": 1.9025623798370361,
+      "learning_rate": 4.980286753286196e-06,
+      "loss": 1.0586,
+      "step": 83
+    },
+    {
+      "epoch": 0.4097560975609756,
+      "grad_norm": 1.4698103666305542,
+      "learning_rate": 4.97980366298112e-06,
+      "loss": 0.8611,
+      "step": 84
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 1.6086381673812866,
+      "learning_rate": 4.97931474883872e-06,
+      "loss": 0.8884,
+      "step": 85
+    },
+    {
+      "epoch": 0.4195121951219512,
+      "grad_norm": 1.6679224967956543,
+      "learning_rate": 4.978820012007213e-06,
+      "loss": 0.969,
+      "step": 86
+    },
+    {
+      "epoch": 0.424390243902439,
+      "grad_norm": 1.5308334827423096,
+      "learning_rate": 4.978319453648495e-06,
+      "loss": 1.0345,
+      "step": 87
+    },
+    {
+      "epoch": 0.4292682926829268,
+      "grad_norm": 1.7788817882537842,
+      "learning_rate": 4.977813074938128e-06,
+      "loss": 0.9649,
+      "step": 88
+    },
+    {
+      "epoch": 0.43414634146341463,
+      "grad_norm": 1.591383934020996,
+      "learning_rate": 4.977300877065347e-06,
+      "loss": 0.9299,
+      "step": 89
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 1.525985836982727,
+      "learning_rate": 4.976782861233053e-06,
+      "loss": 0.7272,
+      "step": 90
+    },
+    {
+      "epoch": 0.44390243902439025,
+      "grad_norm": 2.7333128452301025,
+      "learning_rate": 4.976259028657812e-06,
+      "loss": 0.6694,
+      "step": 91
+    },
+    {
+      "epoch": 0.44878048780487806,
+      "grad_norm": 1.2718108892440796,
+      "learning_rate": 4.975729380569845e-06,
+      "loss": 0.8037,
+      "step": 92
+    },
+    {
+      "epoch": 0.45365853658536587,
+      "grad_norm": 1.157771110534668,
+      "learning_rate": 4.975193918213035e-06,
+      "loss": 0.52,
+      "step": 93
+    },
+    {
+      "epoch": 0.4585365853658537,
+      "grad_norm": 1.0602248907089233,
+      "learning_rate": 4.974652642844921e-06,
+      "loss": 0.7135,
+      "step": 94
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 2.418555974960327,
+      "learning_rate": 4.974105555736693e-06,
+      "loss": 1.0996,
+      "step": 95
+    },
+    {
+      "epoch": 0.4682926829268293,
+      "grad_norm": 1.1728110313415527,
+      "learning_rate": 4.973552658173186e-06,
+      "loss": 0.7268,
+      "step": 96
+    },
+    {
+      "epoch": 0.47317073170731705,
+      "grad_norm": 1.943113088607788,
+      "learning_rate": 4.972993951452887e-06,
+      "loss": 0.9091,
+      "step": 97
+    },
+    {
+      "epoch": 0.47804878048780486,
+      "grad_norm": 5.063096523284912,
+      "learning_rate": 4.9724294368879214e-06,
+      "loss": 0.8242,
+      "step": 98
+    },
+    {
+      "epoch": 0.48292682926829267,
+      "grad_norm": 1.5232555866241455,
+      "learning_rate": 4.971859115804055e-06,
+      "loss": 1.0543,
+      "step": 99
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 1.1453967094421387,
+      "learning_rate": 4.9712829895406935e-06,
+      "loss": 0.8209,
+      "step": 100
+    },
+    {
+      "epoch": 0.4926829268292683,
+      "grad_norm": 2.12345814704895,
+      "learning_rate": 4.970701059450872e-06,
+      "loss": 0.5849,
+      "step": 101
+    },
+    {
+      "epoch": 0.4975609756097561,
+      "grad_norm": 1.2201842069625854,
+      "learning_rate": 4.970113326901258e-06,
+      "loss": 0.9969,
+      "step": 102
+    },
+    {
+      "epoch": 0.5024390243902439,
+      "grad_norm": 1.3047524690628052,
+      "learning_rate": 4.9695197932721455e-06,
+      "loss": 0.9339,
+      "step": 103
+    },
+    {
+      "epoch": 0.5073170731707317,
+      "grad_norm": 1.6083660125732422,
+      "learning_rate": 4.968920459957453e-06,
+      "loss": 0.9702,
+      "step": 104
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 1.5247286558151245,
+      "learning_rate": 4.968315328364719e-06,
+      "loss": 1.0449,
+      "step": 105
+    },
+    {
+      "epoch": 0.5170731707317073,
+      "grad_norm": 1.354805588722229,
+      "learning_rate": 4.9677043999151e-06,
+      "loss": 1.1431,
+      "step": 106
+    },
+    {
+      "epoch": 0.5219512195121951,
+      "grad_norm": 1.088321328163147,
+      "learning_rate": 4.967087676043366e-06,
+      "loss": 0.519,
+      "step": 107
+    },
+    {
+      "epoch": 0.526829268292683,
+      "grad_norm": 1.5086662769317627,
+      "learning_rate": 4.966465158197897e-06,
+      "loss": 0.9357,
+      "step": 108
+    },
+    {
+      "epoch": 0.5317073170731708,
+      "grad_norm": 1.3161298036575317,
+      "learning_rate": 4.965836847840681e-06,
+      "loss": 0.7234,
+      "step": 109
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 1.4465640783309937,
+      "learning_rate": 4.96520274644731e-06,
+      "loss": 0.8911,
+      "step": 110
+    },
+    {
+      "epoch": 0.5414634146341464,
+      "grad_norm": 1.0576995611190796,
+      "learning_rate": 4.964562855506976e-06,
+      "loss": 0.7254,
+      "step": 111
+    },
+    {
+      "epoch": 0.5463414634146342,
+      "grad_norm": 1.1018916368484497,
+      "learning_rate": 4.963917176522466e-06,
+      "loss": 0.6603,
+      "step": 112
+    },
+    {
+      "epoch": 0.551219512195122,
+      "grad_norm": 2.150622606277466,
+      "learning_rate": 4.963265711010164e-06,
+      "loss": 1.0451,
+      "step": 113
+    },
+    {
+      "epoch": 0.5560975609756098,
+      "grad_norm": 1.7743186950683594,
+      "learning_rate": 4.9626084605000395e-06,
+      "loss": 0.8717,
+      "step": 114
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 1.205291509628296,
+      "learning_rate": 4.961945426535652e-06,
+      "loss": 0.5802,
+      "step": 115
+    },
+    {
+      "epoch": 0.5658536585365853,
+      "grad_norm": 1.3969353437423706,
+      "learning_rate": 4.961276610674141e-06,
+      "loss": 0.9158,
+      "step": 116
+    },
+    {
+      "epoch": 0.5707317073170731,
+      "grad_norm": 1.2318240404129028,
+      "learning_rate": 4.960602014486225e-06,
+      "loss": 1.0086,
+      "step": 117
+    },
+    {
+      "epoch": 0.5756097560975609,
+      "grad_norm": 1.2202470302581787,
+      "learning_rate": 4.959921639556199e-06,
+      "loss": 0.7888,
+      "step": 118
+    },
+    {
+      "epoch": 0.5804878048780487,
+      "grad_norm": 1.1564440727233887,
+      "learning_rate": 4.959235487481928e-06,
+      "loss": 1.0053,
+      "step": 119
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 1.2278865575790405,
+      "learning_rate": 4.958543559874846e-06,
+      "loss": 0.5486,
+      "step": 120
+    },
+    {
+      "epoch": 0.5902439024390244,
+      "grad_norm": 1.5465888977050781,
+      "learning_rate": 4.9578458583599495e-06,
+      "loss": 0.8232,
+      "step": 121
+    },
+    {
+      "epoch": 0.5951219512195122,
+      "grad_norm": 1.6284047365188599,
+      "learning_rate": 4.957142384575795e-06,
+      "loss": 0.7773,
+      "step": 122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1794490814208984,
+      "learning_rate": 4.956433140174498e-06,
+      "loss": 0.8236,
+      "step": 123
+    },
+    {
+      "epoch": 0.6048780487804878,
+      "grad_norm": 1.2404091358184814,
+      "learning_rate": 4.9557181268217225e-06,
+      "loss": 0.8611,
+      "step": 124
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 1.2188373804092407,
+      "learning_rate": 4.954997346196683e-06,
+      "loss": 1.2922,
+      "step": 125
+    },
+    {
+      "epoch": 0.6146341463414634,
+      "grad_norm": 1.1514503955841064,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.7073,
+      "step": 126
+    },
+    {
+      "epoch": 0.6195121951219512,
+      "grad_norm": 1.285913109779358,
+      "learning_rate": 4.953538489914387e-06,
+      "loss": 0.729,
+      "step": 127
+    },
+    {
+      "epoch": 0.624390243902439,
+      "grad_norm": 1.8677104711532593,
+      "learning_rate": 4.9528004176832654e-06,
+      "loss": 0.6211,
+      "step": 128
+    },
+    {
+      "epoch": 0.6292682926829268,
+      "grad_norm": 1.2369698286056519,
+      "learning_rate": 4.952056585032142e-06,
+      "loss": 1.0311,
+      "step": 129
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 1.186990737915039,
+      "learning_rate": 4.951306993707913e-06,
+      "loss": 0.8207,
+      "step": 130
+    },
+    {
+      "epoch": 0.6390243902439025,
+      "grad_norm": 1.189424991607666,
+      "learning_rate": 4.950551645470998e-06,
+      "loss": 0.7853,
+      "step": 131
+    },
+    {
+      "epoch": 0.6439024390243903,
+      "grad_norm": 1.3477216958999634,
+      "learning_rate": 4.9497905420953406e-06,
+      "loss": 0.7271,
+      "step": 132
+    },
+    {
+      "epoch": 0.6487804878048781,
+      "grad_norm": 1.368592619895935,
+      "learning_rate": 4.949023685368395e-06,
+      "loss": 0.7394,
+      "step": 133
+    },
+    {
+      "epoch": 0.6536585365853659,
+      "grad_norm": 1.2769527435302734,
+      "learning_rate": 4.948251077091131e-06,
+      "loss": 1.0713,
+      "step": 134
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 1.1477972269058228,
+      "learning_rate": 4.947472719078025e-06,
+      "loss": 0.841,
+      "step": 135
+    },
+    {
+      "epoch": 0.6634146341463415,
+      "grad_norm": 1.0983368158340454,
+      "learning_rate": 4.9466886131570565e-06,
+      "loss": 0.8665,
+      "step": 136
+    },
+    {
+      "epoch": 0.6682926829268293,
+      "grad_norm": 1.2526847124099731,
+      "learning_rate": 4.945898761169704e-06,
+      "loss": 1.0613,
+      "step": 137
+    },
+    {
+      "epoch": 0.6731707317073171,
+      "grad_norm": 1.1426396369934082,
+      "learning_rate": 4.945103164970941e-06,
+      "loss": 0.6248,
+      "step": 138
+    },
+    {
+      "epoch": 0.6780487804878049,
+      "grad_norm": 1.3275880813598633,
+      "learning_rate": 4.9443018264292304e-06,
+      "loss": 0.755,
+      "step": 139
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 1.4942073822021484,
+      "learning_rate": 4.9434947474265225e-06,
+      "loss": 0.9451,
+      "step": 140
+    },
+    {
+      "epoch": 0.6878048780487804,
+      "grad_norm": 1.1777600049972534,
+      "learning_rate": 4.942681929858249e-06,
+      "loss": 1.0505,
+      "step": 141
+    },
+    {
+      "epoch": 0.6926829268292682,
+      "grad_norm": 1.2474116086959839,
+      "learning_rate": 4.941863375633315e-06,
+      "loss": 0.9174,
+      "step": 142
+    },
+    {
+      "epoch": 0.697560975609756,
+      "grad_norm": 1.5067697763442993,
+      "learning_rate": 4.9410390866741056e-06,
+      "loss": 0.7737,
+      "step": 143
+    },
+    {
+      "epoch": 0.7024390243902439,
+      "grad_norm": 1.3016574382781982,
+      "learning_rate": 4.9402090649164655e-06,
+      "loss": 0.7588,
+      "step": 144
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 1.652600884437561,
+      "learning_rate": 4.9393733123097085e-06,
+      "loss": 1.0588,
+      "step": 145
+    },
+    {
+      "epoch": 0.7121951219512195,
+      "grad_norm": 1.267997145652771,
+      "learning_rate": 4.9385318308166065e-06,
+      "loss": 0.8996,
+      "step": 146
+    },
+    {
+      "epoch": 0.7170731707317073,
+      "grad_norm": 1.8560484647750854,
+      "learning_rate": 4.937684622413385e-06,
+      "loss": 0.6286,
+      "step": 147
+    },
+    {
+      "epoch": 0.7219512195121951,
+      "grad_norm": 1.4419782161712646,
+      "learning_rate": 4.9368316890897185e-06,
+      "loss": 0.9818,
+      "step": 148
+    },
+    {
+      "epoch": 0.7268292682926829,
+      "grad_norm": 1.134084701538086,
+      "learning_rate": 4.9359730328487264e-06,
+      "loss": 0.5246,
+      "step": 149
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 1.0102615356445312,
+      "learning_rate": 4.935108655706972e-06,
+      "loss": 0.7948,
+      "step": 150
+    },
+    {
+      "epoch": 0.7365853658536585,
+      "grad_norm": 1.3480703830718994,
+      "learning_rate": 4.934238559694448e-06,
+      "loss": 1.0951,
+      "step": 151
+    },
+    {
+      "epoch": 0.7414634146341463,
+      "grad_norm": 1.4133951663970947,
+      "learning_rate": 4.9333627468545845e-06,
+      "loss": 0.6936,
+      "step": 152
+    },
+    {
+      "epoch": 0.7463414634146341,
+      "grad_norm": 1.3072413206100464,
+      "learning_rate": 4.932481219244231e-06,
+      "loss": 0.799,
+      "step": 153
+    },
+    {
+      "epoch": 0.751219512195122,
+      "grad_norm": 1.3893049955368042,
+      "learning_rate": 4.931593978933666e-06,
+      "loss": 0.7375,
+      "step": 154
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 1.2313531637191772,
+      "learning_rate": 4.930701028006577e-06,
+      "loss": 0.9487,
+      "step": 155
+    },
+    {
+      "epoch": 0.7609756097560976,
+      "grad_norm": 1.3426295518875122,
+      "learning_rate": 4.929802368560066e-06,
+      "loss": 0.7542,
+      "step": 156
+    },
+    {
+      "epoch": 0.7658536585365854,
+      "grad_norm": 1.5280004739761353,
+      "learning_rate": 4.928898002704642e-06,
+      "loss": 0.8784,
+      "step": 157
+    },
+    {
+      "epoch": 0.7707317073170732,
+      "grad_norm": 1.3149527311325073,
+      "learning_rate": 4.927987932564215e-06,
+      "loss": 0.7247,
+      "step": 158
+    },
+    {
+      "epoch": 0.775609756097561,
+      "grad_norm": 1.073188304901123,
+      "learning_rate": 4.927072160276092e-06,
+      "loss": 0.7826,
+      "step": 159
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 1.426188588142395,
+      "learning_rate": 4.926150687990969e-06,
+      "loss": 0.6129,
+      "step": 160
+    },
+    {
+      "epoch": 0.7853658536585366,
+      "grad_norm": 1.2348767518997192,
+      "learning_rate": 4.925223517872934e-06,
+      "loss": 0.9191,
+      "step": 161
+    },
+    {
+      "epoch": 0.7902439024390244,
+      "grad_norm": 1.599665641784668,
+      "learning_rate": 4.9242906520994484e-06,
+      "loss": 1.0177,
+      "step": 162
+    },
+    {
+      "epoch": 0.7951219512195122,
+      "grad_norm": 1.1841332912445068,
+      "learning_rate": 4.923352092861358e-06,
+      "loss": 0.8342,
+      "step": 163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0213048458099365,
+      "learning_rate": 4.922407842362875e-06,
+      "loss": 0.6292,
+      "step": 164
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 1.1878992319107056,
+      "learning_rate": 4.921457902821578e-06,
+      "loss": 0.9357,
+      "step": 165
+    },
+    {
+      "epoch": 0.8097560975609757,
+      "grad_norm": 1.297462821006775,
+      "learning_rate": 4.920502276468408e-06,
+      "loss": 0.8829,
+      "step": 166
+    },
+    {
+      "epoch": 0.8146341463414634,
+      "grad_norm": 1.1800835132598877,
+      "learning_rate": 4.9195409655476605e-06,
+      "loss": 0.712,
+      "step": 167
+    },
+    {
+      "epoch": 0.8195121951219512,
+      "grad_norm": 1.4254546165466309,
+      "learning_rate": 4.918573972316982e-06,
+      "loss": 0.9997,
+      "step": 168
+    },
+    {
+      "epoch": 0.824390243902439,
+      "grad_norm": 1.3243224620819092,
+      "learning_rate": 4.917601299047361e-06,
+      "loss": 0.7944,
+      "step": 169
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 1.3879033327102661,
+      "learning_rate": 4.916622948023129e-06,
+      "loss": 0.7778,
+      "step": 170
+    },
+    {
+      "epoch": 0.8341463414634146,
+      "grad_norm": 1.245430588722229,
+      "learning_rate": 4.915638921541952e-06,
+      "loss": 0.6247,
+      "step": 171
+    },
+    {
+      "epoch": 0.8390243902439024,
+      "grad_norm": 1.3728258609771729,
+      "learning_rate": 4.914649221914822e-06,
+      "loss": 0.8762,
+      "step": 172
+    },
+    {
+      "epoch": 0.8439024390243902,
+      "grad_norm": 1.3080862760543823,
+      "learning_rate": 4.913653851466057e-06,
+      "loss": 0.6381,
+      "step": 173
+    },
+    {
+      "epoch": 0.848780487804878,
+      "grad_norm": 1.5109484195709229,
+      "learning_rate": 4.912652812533291e-06,
+      "loss": 0.8127,
+      "step": 174
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 1.237879753112793,
+      "learning_rate": 4.911646107467472e-06,
+      "loss": 0.8257,
+      "step": 175
+    },
+    {
+      "epoch": 0.8585365853658536,
+      "grad_norm": 1.2725176811218262,
+      "learning_rate": 4.9106337386328524e-06,
+      "loss": 0.9758,
+      "step": 176
+    },
+    {
+      "epoch": 0.8634146341463415,
+      "grad_norm": 1.271756887435913,
+      "learning_rate": 4.909615708406991e-06,
+      "loss": 0.8436,
+      "step": 177
+    },
+    {
+      "epoch": 0.8682926829268293,
+      "grad_norm": 1.0920095443725586,
+      "learning_rate": 4.908592019180738e-06,
+      "loss": 0.6657,
+      "step": 178
+    },
+    {
+      "epoch": 0.8731707317073171,
+      "grad_norm": 0.9080491065979004,
+      "learning_rate": 4.907562673358234e-06,
+      "loss": 0.6322,
+      "step": 179
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 0.9868524074554443,
+      "learning_rate": 4.906527673356907e-06,
+      "loss": 0.5918,
+      "step": 180
+    },
+    {
+      "epoch": 0.8829268292682927,
+      "grad_norm": 1.49118173122406,
+      "learning_rate": 4.905487021607462e-06,
+      "loss": 0.4939,
+      "step": 181
+    },
+    {
+      "epoch": 0.8878048780487805,
+      "grad_norm": 1.3323795795440674,
+      "learning_rate": 4.904440720553876e-06,
+      "loss": 0.8444,
+      "step": 182
+    },
+    {
+      "epoch": 0.8926829268292683,
+      "grad_norm": 1.340374231338501,
+      "learning_rate": 4.903388772653396e-06,
+      "loss": 0.765,
+      "step": 183
+    },
+    {
+      "epoch": 0.8975609756097561,
+      "grad_norm": 1.2728081941604614,
+      "learning_rate": 4.902331180376529e-06,
+      "loss": 0.8126,
+      "step": 184
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 1.154461145401001,
+      "learning_rate": 4.901267946207038e-06,
+      "loss": 0.9256,
+      "step": 185
+    },
+    {
+      "epoch": 0.9073170731707317,
+      "grad_norm": 1.0781828165054321,
+      "learning_rate": 4.900199072641937e-06,
+      "loss": 0.7057,
+      "step": 186
+    },
+    {
+      "epoch": 0.9121951219512195,
+      "grad_norm": 1.0474439859390259,
+      "learning_rate": 4.899124562191484e-06,
+      "loss": 0.6024,
+      "step": 187
+    },
+    {
+      "epoch": 0.9170731707317074,
+      "grad_norm": 1.019089937210083,
+      "learning_rate": 4.8980444173791735e-06,
+      "loss": 0.5001,
+      "step": 188
+    },
+    {
+      "epoch": 0.9219512195121952,
+      "grad_norm": 1.3556936979293823,
+      "learning_rate": 4.896958640741735e-06,
+      "loss": 0.8636,
+      "step": 189
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 1.2974662780761719,
+      "learning_rate": 4.895867234829121e-06,
+      "loss": 0.9257,
+      "step": 190
+    },
+    {
+      "epoch": 0.9317073170731708,
+      "grad_norm": 1.3130805492401123,
+      "learning_rate": 4.894770202204509e-06,
+      "loss": 0.7026,
+      "step": 191
+    },
+    {
+      "epoch": 0.9365853658536586,
+      "grad_norm": 1.5521293878555298,
+      "learning_rate": 4.893667545444285e-06,
+      "loss": 0.8671,
+      "step": 192
+    },
+    {
+      "epoch": 0.9414634146341463,
+      "grad_norm": 1.0167640447616577,
+      "learning_rate": 4.8925592671380495e-06,
+      "loss": 0.6643,
+      "step": 193
+    },
+    {
+      "epoch": 0.9463414634146341,
+      "grad_norm": 0.8992295861244202,
+      "learning_rate": 4.891445369888601e-06,
+      "loss": 0.5294,
+      "step": 194
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 1.053918480873108,
+      "learning_rate": 4.890325856311936e-06,
+      "loss": 0.8734,
+      "step": 195
+    },
+    {
+      "epoch": 0.9560975609756097,
+      "grad_norm": 1.1650863885879517,
+      "learning_rate": 4.889200729037241e-06,
+      "loss": 0.7858,
+      "step": 196
+    },
+    {
+      "epoch": 0.9609756097560975,
+      "grad_norm": 1.0390702486038208,
+      "learning_rate": 4.888069990706884e-06,
+      "loss": 0.6428,
+      "step": 197
+    },
+    {
+      "epoch": 0.9658536585365853,
+      "grad_norm": 1.4749332666397095,
+      "learning_rate": 4.886933643976414e-06,
+      "loss": 0.55,
+      "step": 198
+    },
+    {
+      "epoch": 0.9707317073170731,
+      "grad_norm": 0.9483368992805481,
+      "learning_rate": 4.885791691514548e-06,
+      "loss": 0.4786,
+      "step": 199
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 1.0259203910827637,
+      "learning_rate": 4.884644136003172e-06,
+      "loss": 0.6137,
+      "step": 200
+    },
+    {
+      "epoch": 0.9804878048780488,
+      "grad_norm": 1.1431037187576294,
+      "learning_rate": 4.883490980137327e-06,
+      "loss": 1.2973,
+      "step": 201
+    },
+    {
+      "epoch": 0.9853658536585366,
+      "grad_norm": 1.1985281705856323,
+      "learning_rate": 4.882332226625208e-06,
+      "loss": 0.7734,
+      "step": 202
+    },
+    {
+      "epoch": 0.9902439024390244,
+      "grad_norm": 0.9396414160728455,
+      "learning_rate": 4.881167878188158e-06,
+      "loss": 0.8172,
+      "step": 203
+    },
+    {
+      "epoch": 0.9951219512195122,
+      "grad_norm": 1.3293070793151855,
+      "learning_rate": 4.8799979375606565e-06,
+      "loss": 0.727,
+      "step": 204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0133116245269775,
+      "learning_rate": 4.878822407490319e-06,
+      "loss": 0.7099,
+      "step": 205
+    },
+    {
+      "epoch": 1.0048780487804878,
+      "grad_norm": 1.0823330879211426,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.6611,
+      "step": 206
+    },
+    {
+      "epoch": 1.0097560975609756,
+      "grad_norm": 0.9297066926956177,
+      "learning_rate": 4.876454590077216e-06,
+      "loss": 0.5825,
+      "step": 207
+    },
+    {
+      "epoch": 1.0146341463414634,
+      "grad_norm": 1.0025253295898438,
+      "learning_rate": 4.875262308295289e-06,
+      "loss": 0.8154,
+      "step": 208
+    },
+    {
+      "epoch": 1.0195121951219512,
+      "grad_norm": 1.1356369256973267,
+      "learning_rate": 4.874064448192185e-06,
+      "loss": 0.7126,
+      "step": 209
+    },
+    {
+      "epoch": 1.024390243902439,
+      "grad_norm": 1.4361388683319092,
+      "learning_rate": 4.872861012581088e-06,
+      "loss": 0.6323,
+      "step": 210
+    },
+    {
+      "epoch": 1.0292682926829269,
+      "grad_norm": 1.2314097881317139,
+      "learning_rate": 4.871652004288275e-06,
+      "loss": 0.6683,
+      "step": 211
+    },
+    {
+      "epoch": 1.0341463414634147,
+      "grad_norm": 1.1820945739746094,
+      "learning_rate": 4.870437426153113e-06,
+      "loss": 0.6659,
+      "step": 212
+    },
+    {
+      "epoch": 1.0390243902439025,
+      "grad_norm": 1.1955009698867798,
+      "learning_rate": 4.869217281028045e-06,
+      "loss": 0.8964,
+      "step": 213
+    },
+    {
+      "epoch": 1.0439024390243903,
+      "grad_norm": 1.2589020729064941,
+      "learning_rate": 4.867991571778592e-06,
+      "loss": 0.894,
+      "step": 214
+    },
+    {
+      "epoch": 1.048780487804878,
+      "grad_norm": 0.8933411836624146,
+      "learning_rate": 4.866760301283342e-06,
+      "loss": 0.5199,
+      "step": 215
+    },
+    {
+      "epoch": 1.053658536585366,
+      "grad_norm": 1.126219630241394,
+      "learning_rate": 4.865523472433942e-06,
+      "loss": 0.7014,
+      "step": 216
+    },
+    {
+      "epoch": 1.0585365853658537,
+      "grad_norm": 1.1173880100250244,
+      "learning_rate": 4.8642810881350935e-06,
+      "loss": 0.6085,
+      "step": 217
+    },
+    {
+      "epoch": 1.0634146341463415,
+      "grad_norm": 0.9508454203605652,
+      "learning_rate": 4.863033151304546e-06,
+      "loss": 0.6333,
+      "step": 218
+    },
+    {
+      "epoch": 1.0682926829268293,
+      "grad_norm": 1.8204004764556885,
+      "learning_rate": 4.861779664873088e-06,
+      "loss": 0.8899,
+      "step": 219
+    },
+    {
+      "epoch": 1.0731707317073171,
+      "grad_norm": 1.5721564292907715,
+      "learning_rate": 4.8605206317845425e-06,
+      "loss": 0.8105,
+      "step": 220
+    },
+    {
+      "epoch": 1.078048780487805,
+      "grad_norm": 1.3572758436203003,
+      "learning_rate": 4.859256054995758e-06,
+      "loss": 0.8634,
+      "step": 221
+    },
+    {
+      "epoch": 1.0829268292682928,
+      "grad_norm": 0.8531430959701538,
+      "learning_rate": 4.8579859374766e-06,
+      "loss": 0.4826,
+      "step": 222
+    },
+    {
+      "epoch": 1.0878048780487806,
+      "grad_norm": 0.771088719367981,
+      "learning_rate": 4.856710282209952e-06,
+      "loss": 0.4085,
+      "step": 223
+    },
+    {
+      "epoch": 1.0926829268292684,
+      "grad_norm": 1.0056614875793457,
+      "learning_rate": 4.855429092191698e-06,
+      "loss": 0.7168,
+      "step": 224
+    },
+    {
+      "epoch": 1.0975609756097562,
+      "grad_norm": 0.9825501441955566,
+      "learning_rate": 4.854142370430725e-06,
+      "loss": 0.6276,
+      "step": 225
+    },
+    {
+      "epoch": 1.102439024390244,
+      "grad_norm": 1.0209708213806152,
+      "learning_rate": 4.8528501199489045e-06,
+      "loss": 0.6521,
+      "step": 226
+    },
+    {
+      "epoch": 1.1073170731707318,
+      "grad_norm": 1.3162622451782227,
+      "learning_rate": 4.851552343781099e-06,
+      "loss": 0.8244,
+      "step": 227
+    },
+    {
+      "epoch": 1.1121951219512196,
+      "grad_norm": 1.150299072265625,
+      "learning_rate": 4.850249044975145e-06,
+      "loss": 0.7804,
+      "step": 228
+    },
+    {
+      "epoch": 1.1170731707317074,
+      "grad_norm": 1.142996907234192,
+      "learning_rate": 4.848940226591849e-06,
+      "loss": 0.9935,
+      "step": 229
+    },
+    {
+      "epoch": 1.1219512195121952,
+      "grad_norm": 0.9757675528526306,
+      "learning_rate": 4.847625891704982e-06,
+      "loss": 0.6041,
+      "step": 230
+    },
+    {
+      "epoch": 1.126829268292683,
+      "grad_norm": 1.1686437129974365,
+      "learning_rate": 4.846306043401268e-06,
+      "loss": 0.6873,
+      "step": 231
+    },
+    {
+      "epoch": 1.1317073170731708,
+      "grad_norm": 1.4826396703720093,
+      "learning_rate": 4.844980684780381e-06,
+      "loss": 0.6412,
+      "step": 232
+    },
+    {
+      "epoch": 1.1365853658536587,
+      "grad_norm": 1.204198956489563,
+      "learning_rate": 4.8436498189549345e-06,
+      "loss": 0.5779,
+      "step": 233
+    },
+    {
+      "epoch": 1.1414634146341462,
+      "grad_norm": 0.8927454948425293,
+      "learning_rate": 4.842313449050477e-06,
+      "loss": 0.5894,
+      "step": 234
+    },
+    {
+      "epoch": 1.146341463414634,
+      "grad_norm": 0.9504485726356506,
+      "learning_rate": 4.840971578205486e-06,
+      "loss": 0.5195,
+      "step": 235
+    },
+    {
+      "epoch": 1.1512195121951219,
+      "grad_norm": 0.8682957291603088,
+      "learning_rate": 4.839624209571352e-06,
+      "loss": 0.3927,
+      "step": 236
+    },
+    {
+      "epoch": 1.1560975609756097,
+      "grad_norm": 1.0243151187896729,
+      "learning_rate": 4.838271346312381e-06,
+      "loss": 0.8549,
+      "step": 237
+    },
+    {
+      "epoch": 1.1609756097560975,
+      "grad_norm": 0.9931624531745911,
+      "learning_rate": 4.836912991605782e-06,
+      "loss": 0.8616,
+      "step": 238
+    },
+    {
+      "epoch": 1.1658536585365853,
+      "grad_norm": 1.0424671173095703,
+      "learning_rate": 4.835549148641663e-06,
+      "loss": 0.5675,
+      "step": 239
+    },
+    {
+      "epoch": 1.170731707317073,
+      "grad_norm": 0.9706572890281677,
+      "learning_rate": 4.834179820623018e-06,
+      "loss": 0.6444,
+      "step": 240
+    },
+    {
+      "epoch": 1.175609756097561,
+      "grad_norm": 0.9557904601097107,
+      "learning_rate": 4.832805010765724e-06,
+      "loss": 0.6006,
+      "step": 241
+    },
+    {
+      "epoch": 1.1804878048780487,
+      "grad_norm": 1.2297061681747437,
+      "learning_rate": 4.831424722298531e-06,
+      "loss": 0.6897,
+      "step": 242
+    },
+    {
+      "epoch": 1.1853658536585365,
+      "grad_norm": 1.3953443765640259,
+      "learning_rate": 4.830038958463061e-06,
+      "loss": 0.772,
+      "step": 243
+    },
+    {
+      "epoch": 1.1902439024390243,
+      "grad_norm": 2.120232105255127,
+      "learning_rate": 4.828647722513785e-06,
+      "loss": 0.9361,
+      "step": 244
+    },
+    {
+      "epoch": 1.1951219512195121,
+      "grad_norm": 1.4457300901412964,
+      "learning_rate": 4.827251017718034e-06,
+      "loss": 0.8036,
+      "step": 245
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0927690267562866,
+      "learning_rate": 4.8258488473559794e-06,
+      "loss": 0.8354,
+      "step": 246
+    },
+    {
+      "epoch": 1.2048780487804878,
+      "grad_norm": 1.4953900575637817,
+      "learning_rate": 4.824441214720629e-06,
+      "loss": 0.9083,
+      "step": 247
+    },
+    {
+      "epoch": 1.2097560975609756,
+      "grad_norm": 0.9229385852813721,
+      "learning_rate": 4.823028123117818e-06,
+      "loss": 0.441,
+      "step": 248
+    },
+    {
+      "epoch": 1.2146341463414634,
+      "grad_norm": 1.8001827001571655,
+      "learning_rate": 4.8216095758662015e-06,
+      "loss": 0.7975,
+      "step": 249
+    },
+    {
+      "epoch": 1.2195121951219512,
+      "grad_norm": 1.0904715061187744,
+      "learning_rate": 4.82018557629725e-06,
+      "loss": 0.8526,
+      "step": 250
+    },
+    {
+      "epoch": 1.224390243902439,
+      "grad_norm": 1.3482294082641602,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.6198,
+      "step": 251
+    },
+    {
+      "epoch": 1.2292682926829268,
+      "grad_norm": 0.9257897138595581,
+      "learning_rate": 4.817321233597232e-06,
+      "loss": 0.8115,
+      "step": 252
+    },
+    {
+      "epoch": 1.2341463414634146,
+      "grad_norm": 1.1533488035202026,
+      "learning_rate": 4.815880897193095e-06,
+      "loss": 0.5524,
+      "step": 253
+    },
+    {
+      "epoch": 1.2390243902439024,
+      "grad_norm": 1.2778701782226562,
+      "learning_rate": 4.814435121925466e-06,
+      "loss": 0.802,
+      "step": 254
+    },
+    {
+      "epoch": 1.2439024390243902,
+      "grad_norm": 1.037063717842102,
+      "learning_rate": 4.812983911189761e-06,
+      "loss": 0.696,
+      "step": 255
+    },
+    {
+      "epoch": 1.248780487804878,
+      "grad_norm": 1.0112067461013794,
+      "learning_rate": 4.811527268394157e-06,
+      "loss": 0.5189,
+      "step": 256
+    },
+    {
+      "epoch": 1.2536585365853659,
+      "grad_norm": 1.1271675825119019,
+      "learning_rate": 4.810065196959591e-06,
+      "loss": 0.6292,
+      "step": 257
+    },
+    {
+      "epoch": 1.2585365853658537,
+      "grad_norm": 1.0987378358840942,
+      "learning_rate": 4.8085977003197496e-06,
+      "loss": 0.8452,
+      "step": 258
+    },
+    {
+      "epoch": 1.2634146341463415,
+      "grad_norm": 1.5386906862258911,
+      "learning_rate": 4.807124781921059e-06,
+      "loss": 1.1094,
+      "step": 259
+    },
+    {
+      "epoch": 1.2682926829268293,
+      "grad_norm": 0.9946249127388,
+      "learning_rate": 4.805646445222679e-06,
+      "loss": 0.6367,
+      "step": 260
+    },
+    {
+      "epoch": 1.273170731707317,
+      "grad_norm": 0.9774323105812073,
+      "learning_rate": 4.804162693696494e-06,
+      "loss": 0.6157,
+      "step": 261
+    },
+    {
+      "epoch": 1.278048780487805,
+      "grad_norm": 1.9447377920150757,
+      "learning_rate": 4.802673530827105e-06,
+      "loss": 0.6076,
+      "step": 262
+    },
+    {
+      "epoch": 1.2829268292682927,
+      "grad_norm": 0.7815698981285095,
+      "learning_rate": 4.801178960111823e-06,
+      "loss": 0.5985,
+      "step": 263
+    },
+    {
+      "epoch": 1.2878048780487805,
+      "grad_norm": 1.014426589012146,
+      "learning_rate": 4.799678985060658e-06,
+      "loss": 0.851,
+      "step": 264
+    },
+    {
+      "epoch": 1.2926829268292683,
+      "grad_norm": 0.9904977083206177,
+      "learning_rate": 4.798173609196314e-06,
+      "loss": 0.8511,
+      "step": 265
+    },
+    {
+      "epoch": 1.2975609756097561,
+      "grad_norm": 1.0354292392730713,
+      "learning_rate": 4.796662836054176e-06,
+      "loss": 0.4885,
+      "step": 266
+    },
+    {
+      "epoch": 1.302439024390244,
+      "grad_norm": 0.9485008120536804,
+      "learning_rate": 4.795146669182304e-06,
+      "loss": 0.6486,
+      "step": 267
+    },
+    {
+      "epoch": 1.3073170731707318,
+      "grad_norm": 0.9609008431434631,
+      "learning_rate": 4.793625112141431e-06,
+      "loss": 0.5446,
+      "step": 268
+    },
+    {
+      "epoch": 1.3121951219512196,
+      "grad_norm": 0.8880930542945862,
+      "learning_rate": 4.792098168504943e-06,
+      "loss": 0.5741,
+      "step": 269
+    },
+    {
+      "epoch": 1.3170731707317074,
+      "grad_norm": 0.8475173711776733,
+      "learning_rate": 4.790565841858879e-06,
+      "loss": 0.488,
+      "step": 270
+    },
+    {
+      "epoch": 1.3219512195121952,
+      "grad_norm": 1.04447603225708,
+      "learning_rate": 4.789028135801919e-06,
+      "loss": 0.7792,
+      "step": 271
+    },
+    {
+      "epoch": 1.326829268292683,
+      "grad_norm": 1.2021688222885132,
+      "learning_rate": 4.787485053945377e-06,
+      "loss": 0.8513,
+      "step": 272
+    },
+    {
+      "epoch": 1.3317073170731708,
+      "grad_norm": 0.9609706401824951,
+      "learning_rate": 4.785936599913193e-06,
+      "loss": 0.7186,
+      "step": 273
+    },
+    {
+      "epoch": 1.3365853658536586,
+      "grad_norm": 0.9922477602958679,
+      "learning_rate": 4.784382777341922e-06,
+      "loss": 0.7355,
+      "step": 274
+    },
+    {
+      "epoch": 1.3414634146341464,
+      "grad_norm": 1.2870302200317383,
+      "learning_rate": 4.782823589880729e-06,
+      "loss": 1.0985,
+      "step": 275
+    },
+    {
+      "epoch": 1.346341463414634,
+      "grad_norm": 0.9369707107543945,
+      "learning_rate": 4.7812590411913755e-06,
+      "loss": 0.7729,
+      "step": 276
+    },
+    {
+      "epoch": 1.3512195121951218,
+      "grad_norm": 1.1406941413879395,
+      "learning_rate": 4.779689134948217e-06,
+      "loss": 1.1142,
+      "step": 277
+    },
+    {
+      "epoch": 1.3560975609756096,
+      "grad_norm": 1.1078243255615234,
+      "learning_rate": 4.77811387483819e-06,
+      "loss": 0.5348,
+      "step": 278
+    },
+    {
+      "epoch": 1.3609756097560974,
+      "grad_norm": 1.1744791269302368,
+      "learning_rate": 4.776533264560804e-06,
+      "loss": 0.7202,
+      "step": 279
+    },
+    {
+      "epoch": 1.3658536585365852,
+      "grad_norm": 1.2643998861312866,
+      "learning_rate": 4.774947307828134e-06,
+      "loss": 0.9659,
+      "step": 280
+    },
+    {
+      "epoch": 1.370731707317073,
+      "grad_norm": 1.069071888923645,
+      "learning_rate": 4.773356008364812e-06,
+      "loss": 0.6257,
+      "step": 281
+    },
+    {
+      "epoch": 1.3756097560975609,
+      "grad_norm": 0.8661286234855652,
+      "learning_rate": 4.771759369908017e-06,
+      "loss": 0.4138,
+      "step": 282
+    },
+    {
+      "epoch": 1.3804878048780487,
+      "grad_norm": 0.9560813903808594,
+      "learning_rate": 4.7701573962074635e-06,
+      "loss": 0.6435,
+      "step": 283
+    },
+    {
+      "epoch": 1.3853658536585365,
+      "grad_norm": 1.0229038000106812,
+      "learning_rate": 4.7685500910254015e-06,
+      "loss": 0.5709,
+      "step": 284
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 1.7641900777816772,
+      "learning_rate": 4.766937458136598e-06,
+      "loss": 0.7815,
+      "step": 285
+    },
+    {
+      "epoch": 1.395121951219512,
+      "grad_norm": 1.011093258857727,
+      "learning_rate": 4.765319501328332e-06,
+      "loss": 0.7428,
+      "step": 286
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0194127559661865,
+      "learning_rate": 4.763696224400391e-06,
+      "loss": 0.542,
+      "step": 287
+    },
+    {
+      "epoch": 1.4048780487804877,
+      "grad_norm": 0.9469794034957886,
+      "learning_rate": 4.762067631165049e-06,
+      "loss": 0.5758,
+      "step": 288
+    },
+    {
+      "epoch": 1.4097560975609755,
+      "grad_norm": 1.1177825927734375,
+      "learning_rate": 4.760433725447071e-06,
+      "loss": 0.8141,
+      "step": 289
+    },
+    {
+      "epoch": 1.4146341463414633,
+      "grad_norm": 1.388083577156067,
+      "learning_rate": 4.758794511083697e-06,
+      "loss": 0.8286,
+      "step": 290
+    },
+    {
+      "epoch": 1.4195121951219511,
+      "grad_norm": 1.1623952388763428,
+      "learning_rate": 4.757149991924633e-06,
+      "loss": 0.6733,
+      "step": 291
+    },
+    {
+      "epoch": 1.424390243902439,
+      "grad_norm": 1.328235149383545,
+      "learning_rate": 4.755500171832045e-06,
+      "loss": 0.5397,
+      "step": 292
+    },
+    {
+      "epoch": 1.4292682926829268,
+      "grad_norm": 1.0121268033981323,
+      "learning_rate": 4.753845054680548e-06,
+      "loss": 0.6813,
+      "step": 293
+    },
+    {
+      "epoch": 1.4341463414634146,
+      "grad_norm": 1.0799837112426758,
+      "learning_rate": 4.752184644357197e-06,
+      "loss": 0.5136,
+      "step": 294
+    },
+    {
+      "epoch": 1.4390243902439024,
+      "grad_norm": 1.010602593421936,
+      "learning_rate": 4.750518944761477e-06,
+      "loss": 0.5768,
+      "step": 295
+    },
+    {
+      "epoch": 1.4439024390243902,
+      "grad_norm": 1.1127134561538696,
+      "learning_rate": 4.748847959805297e-06,
+      "loss": 0.5663,
+      "step": 296
+    },
+    {
+      "epoch": 1.448780487804878,
+      "grad_norm": 0.9182597398757935,
+      "learning_rate": 4.7471716934129774e-06,
+      "loss": 0.5599,
+      "step": 297
+    },
+    {
+      "epoch": 1.4536585365853658,
+      "grad_norm": 1.0173683166503906,
+      "learning_rate": 4.745490149521242e-06,
+      "loss": 0.5155,
+      "step": 298
+    },
+    {
+      "epoch": 1.4585365853658536,
+      "grad_norm": 0.9683080911636353,
+      "learning_rate": 4.743803332079209e-06,
+      "loss": 0.5744,
+      "step": 299
+    },
+    {
+      "epoch": 1.4634146341463414,
+      "grad_norm": 0.9615181684494019,
+      "learning_rate": 4.742111245048382e-06,
+      "loss": 0.5961,
+      "step": 300
+    },
+    {
+      "epoch": 1.4682926829268292,
+      "grad_norm": 1.1113585233688354,
+      "learning_rate": 4.740413892402639e-06,
+      "loss": 0.5751,
+      "step": 301
+    },
+    {
+      "epoch": 1.473170731707317,
+      "grad_norm": 1.1533280611038208,
+      "learning_rate": 4.738711278128228e-06,
+      "loss": 0.6668,
+      "step": 302
+    },
+    {
+      "epoch": 1.4780487804878049,
+      "grad_norm": 1.086147665977478,
+      "learning_rate": 4.7370034062237476e-06,
+      "loss": 0.4249,
+      "step": 303
+    },
+    {
+      "epoch": 1.4829268292682927,
+      "grad_norm": 3.0501999855041504,
+      "learning_rate": 4.73529028070015e-06,
+      "loss": 0.6284,
+      "step": 304
+    },
+    {
+      "epoch": 1.4878048780487805,
+      "grad_norm": 1.2545751333236694,
+      "learning_rate": 4.733571905580723e-06,
+      "loss": 0.8968,
+      "step": 305
+    },
+    {
+      "epoch": 1.4926829268292683,
+      "grad_norm": 0.9740838408470154,
+      "learning_rate": 4.731848284901082e-06,
+      "loss": 0.7402,
+      "step": 306
+    },
+    {
+      "epoch": 1.497560975609756,
+      "grad_norm": 1.0430322885513306,
+      "learning_rate": 4.730119422709165e-06,
+      "loss": 0.4697,
+      "step": 307
+    },
+    {
+      "epoch": 1.502439024390244,
+      "grad_norm": 1.2707469463348389,
+      "learning_rate": 4.728385323065215e-06,
+      "loss": 0.7548,
+      "step": 308
+    },
+    {
+      "epoch": 1.5073170731707317,
+      "grad_norm": 0.9956101775169373,
+      "learning_rate": 4.7266459900417815e-06,
+      "loss": 0.5444,
+      "step": 309
+    },
+    {
+      "epoch": 1.5121951219512195,
+      "grad_norm": 1.0116164684295654,
+      "learning_rate": 4.724901427723698e-06,
+      "loss": 0.7939,
+      "step": 310
+    },
+    {
+      "epoch": 1.5170731707317073,
+      "grad_norm": 1.0668343305587769,
+      "learning_rate": 4.723151640208084e-06,
+      "loss": 0.3966,
+      "step": 311
+    },
+    {
+      "epoch": 1.5219512195121951,
+      "grad_norm": 0.9897716045379639,
+      "learning_rate": 4.721396631604327e-06,
+      "loss": 0.4675,
+      "step": 312
+    },
+    {
+      "epoch": 1.526829268292683,
+      "grad_norm": 0.9885281920433044,
+      "learning_rate": 4.7196364060340785e-06,
+      "loss": 0.5411,
+      "step": 313
+    },
+    {
+      "epoch": 1.5317073170731708,
+      "grad_norm": 1.1385118961334229,
+      "learning_rate": 4.7178709676312416e-06,
+      "loss": 0.8041,
+      "step": 314
+    },
+    {
+      "epoch": 1.5365853658536586,
+      "grad_norm": 1.2253623008728027,
+      "learning_rate": 4.716100320541961e-06,
+      "loss": 1.0583,
+      "step": 315
+    },
+    {
+      "epoch": 1.5414634146341464,
+      "grad_norm": 1.1313822269439697,
+      "learning_rate": 4.714324468924614e-06,
+      "loss": 0.7701,
+      "step": 316
+    },
+    {
+      "epoch": 1.5463414634146342,
+      "grad_norm": 1.120343804359436,
+      "learning_rate": 4.712543416949803e-06,
+      "loss": 0.7407,
+      "step": 317
+    },
+    {
+      "epoch": 1.551219512195122,
+      "grad_norm": 1.5084882974624634,
+      "learning_rate": 4.71075716880034e-06,
+      "loss": 0.753,
+      "step": 318
+    },
+    {
+      "epoch": 1.5560975609756098,
+      "grad_norm": 1.3213189840316772,
+      "learning_rate": 4.708965728671243e-06,
+      "loss": 0.8935,
+      "step": 319
+    },
+    {
+      "epoch": 1.5609756097560976,
+      "grad_norm": 1.170746922492981,
+      "learning_rate": 4.7071691007697214e-06,
+      "loss": 0.6782,
+      "step": 320
+    },
+    {
+      "epoch": 1.5658536585365854,
+      "grad_norm": 1.2023199796676636,
+      "learning_rate": 4.705367289315172e-06,
+      "loss": 0.8011,
+      "step": 321
+    },
+    {
+      "epoch": 1.5707317073170732,
+      "grad_norm": 1.0213698148727417,
+      "learning_rate": 4.703560298539158e-06,
+      "loss": 0.5011,
+      "step": 322
+    },
+    {
+      "epoch": 1.575609756097561,
+      "grad_norm": 1.045581579208374,
+      "learning_rate": 4.701748132685415e-06,
+      "loss": 0.5501,
+      "step": 323
+    },
+    {
+      "epoch": 1.5804878048780489,
+      "grad_norm": 0.9141654372215271,
+      "learning_rate": 4.699930796009825e-06,
+      "loss": 0.5551,
+      "step": 324
+    },
+    {
+      "epoch": 1.5853658536585367,
+      "grad_norm": 1.3005549907684326,
+      "learning_rate": 4.698108292780418e-06,
+      "loss": 0.7293,
+      "step": 325
+    },
+    {
+      "epoch": 1.5902439024390245,
+      "grad_norm": 0.985907793045044,
+      "learning_rate": 4.696280627277356e-06,
+      "loss": 0.5366,
+      "step": 326
+    },
+    {
+      "epoch": 1.5951219512195123,
+      "grad_norm": 0.9095384478569031,
+      "learning_rate": 4.6944478037929255e-06,
+      "loss": 0.5508,
+      "step": 327
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.346676230430603,
+      "learning_rate": 4.692609826631525e-06,
+      "loss": 0.6719,
+      "step": 328
+    },
+    {
+      "epoch": 1.604878048780488,
+      "grad_norm": 1.088921308517456,
+      "learning_rate": 4.690766700109659e-06,
+      "loss": 0.4088,
+      "step": 329
+    },
+    {
+      "epoch": 1.6097560975609757,
+      "grad_norm": 0.8905205726623535,
+      "learning_rate": 4.6889184285559234e-06,
+      "loss": 0.4671,
+      "step": 330
+    },
+    {
+      "epoch": 1.6146341463414635,
+      "grad_norm": 1.2066144943237305,
+      "learning_rate": 4.687065016310996e-06,
+      "loss": 0.7891,
+      "step": 331
+    },
+    {
+      "epoch": 1.6195121951219513,
+      "grad_norm": 1.0449296236038208,
+      "learning_rate": 4.685206467727631e-06,
+      "loss": 0.6103,
+      "step": 332
+    },
+    {
+      "epoch": 1.6243902439024391,
+      "grad_norm": 1.15915048122406,
+      "learning_rate": 4.683342787170644e-06,
+      "loss": 0.605,
+      "step": 333
+    },
+    {
+      "epoch": 1.629268292682927,
+      "grad_norm": 1.0918726921081543,
+      "learning_rate": 4.6814739790169006e-06,
+      "loss": 0.5444,
+      "step": 334
+    },
+    {
+      "epoch": 1.6341463414634148,
+      "grad_norm": 1.0298805236816406,
+      "learning_rate": 4.679600047655313e-06,
+      "loss": 0.7902,
+      "step": 335
+    },
+    {
+      "epoch": 1.6390243902439026,
+      "grad_norm": 1.3017504215240479,
+      "learning_rate": 4.6777209974868194e-06,
+      "loss": 1.1195,
+      "step": 336
+    },
+    {
+      "epoch": 1.6439024390243904,
+      "grad_norm": 1.45652174949646,
+      "learning_rate": 4.675836832924387e-06,
+      "loss": 0.6358,
+      "step": 337
+    },
+    {
+      "epoch": 1.6487804878048782,
+      "grad_norm": 0.8610002398490906,
+      "learning_rate": 4.673947558392989e-06,
+      "loss": 0.4196,
+      "step": 338
+    },
+    {
+      "epoch": 1.653658536585366,
+      "grad_norm": 0.8891443014144897,
+      "learning_rate": 4.6720531783296e-06,
+      "loss": 0.5593,
+      "step": 339
+    },
+    {
+      "epoch": 1.6585365853658538,
+      "grad_norm": 0.9679135680198669,
+      "learning_rate": 4.670153697183185e-06,
+      "loss": 0.6149,
+      "step": 340
+    },
+    {
+      "epoch": 1.6634146341463416,
+      "grad_norm": 1.019687294960022,
+      "learning_rate": 4.668249119414692e-06,
+      "loss": 0.5855,
+      "step": 341
+    },
+    {
+      "epoch": 1.6682926829268294,
+      "grad_norm": 0.9645085334777832,
+      "learning_rate": 4.666339449497033e-06,
+      "loss": 0.6321,
+      "step": 342
+    },
+    {
+      "epoch": 1.6731707317073172,
+      "grad_norm": 1.1760913133621216,
+      "learning_rate": 4.664424691915084e-06,
+      "loss": 0.5839,
+      "step": 343
+    },
+    {
+      "epoch": 1.678048780487805,
+      "grad_norm": 1.1706181764602661,
+      "learning_rate": 4.6625048511656675e-06,
+      "loss": 0.5806,
+      "step": 344
+    },
+    {
+      "epoch": 1.6829268292682928,
+      "grad_norm": 1.1575871706008911,
+      "learning_rate": 4.660579931757543e-06,
+      "loss": 0.4945,
+      "step": 345
+    },
+    {
+      "epoch": 1.6878048780487804,
+      "grad_norm": 0.9929284453392029,
+      "learning_rate": 4.6586499382113985e-06,
+      "loss": 0.5662,
+      "step": 346
+    },
+    {
+      "epoch": 1.6926829268292682,
+      "grad_norm": 0.9940921068191528,
+      "learning_rate": 4.6567148750598375e-06,
+      "loss": 0.8166,
+      "step": 347
+    },
+    {
+      "epoch": 1.697560975609756,
+      "grad_norm": 1.1893478631973267,
+      "learning_rate": 4.6547747468473705e-06,
+      "loss": 0.9291,
+      "step": 348
+    },
+    {
+      "epoch": 1.7024390243902439,
+      "grad_norm": 0.8776846528053284,
+      "learning_rate": 4.652829558130404e-06,
+      "loss": 0.4457,
+      "step": 349
+    },
+    {
+      "epoch": 1.7073170731707317,
+      "grad_norm": 1.2133142948150635,
+      "learning_rate": 4.6508793134772265e-06,
+      "loss": 0.6364,
+      "step": 350
+    },
+    {
+      "epoch": 1.7121951219512195,
+      "grad_norm": 0.8865175247192383,
+      "learning_rate": 4.648924017468003e-06,
+      "loss": 0.5514,
+      "step": 351
+    },
+    {
+      "epoch": 1.7170731707317073,
+      "grad_norm": 1.3637226819992065,
+      "learning_rate": 4.646963674694761e-06,
+      "loss": 0.8656,
+      "step": 352
+    },
+    {
+      "epoch": 1.721951219512195,
+      "grad_norm": 0.875629723072052,
+      "learning_rate": 4.64499828976138e-06,
+      "loss": 0.3992,
+      "step": 353
+    },
+    {
+      "epoch": 1.726829268292683,
+      "grad_norm": 1.0361976623535156,
+      "learning_rate": 4.64302786728358e-06,
+      "loss": 0.5056,
+      "step": 354
+    },
+    {
+      "epoch": 1.7317073170731707,
+      "grad_norm": 0.9224256873130798,
+      "learning_rate": 4.641052411888913e-06,
+      "loss": 0.5352,
+      "step": 355
+    },
+    {
+      "epoch": 1.7365853658536585,
+      "grad_norm": 1.2034342288970947,
+      "learning_rate": 4.6390719282167515e-06,
+      "loss": 0.4753,
+      "step": 356
+    },
+    {
+      "epoch": 1.7414634146341463,
+      "grad_norm": 1.056547999382019,
+      "learning_rate": 4.637086420918276e-06,
+      "loss": 0.7975,
+      "step": 357
+    },
+    {
+      "epoch": 1.7463414634146341,
+      "grad_norm": 0.9398707151412964,
+      "learning_rate": 4.635095894656465e-06,
+      "loss": 0.6944,
+      "step": 358
+    },
+    {
+      "epoch": 1.751219512195122,
+      "grad_norm": 1.3796380758285522,
+      "learning_rate": 4.633100354106085e-06,
+      "loss": 0.377,
+      "step": 359
+    },
+    {
+      "epoch": 1.7560975609756098,
+      "grad_norm": 1.1999366283416748,
+      "learning_rate": 4.631099803953677e-06,
+      "loss": 0.8488,
+      "step": 360
+    },
+    {
+      "epoch": 1.7609756097560976,
+      "grad_norm": 1.1613068580627441,
+      "learning_rate": 4.629094248897546e-06,
+      "loss": 0.4856,
+      "step": 361
+    },
+    {
+      "epoch": 1.7658536585365854,
+      "grad_norm": 1.407758116722107,
+      "learning_rate": 4.627083693647757e-06,
+      "loss": 0.5932,
+      "step": 362
+    },
+    {
+      "epoch": 1.7707317073170732,
+      "grad_norm": 1.1262705326080322,
+      "learning_rate": 4.625068142926111e-06,
+      "loss": 0.862,
+      "step": 363
+    },
+    {
+      "epoch": 1.775609756097561,
+      "grad_norm": 1.0207730531692505,
+      "learning_rate": 4.623047601466144e-06,
+      "loss": 0.7695,
+      "step": 364
+    },
+    {
+      "epoch": 1.7804878048780488,
+      "grad_norm": 1.0084431171417236,
+      "learning_rate": 4.621022074013114e-06,
+      "loss": 0.6608,
+      "step": 365
+    },
+    {
+      "epoch": 1.7853658536585366,
+      "grad_norm": 1.044545292854309,
+      "learning_rate": 4.618991565323987e-06,
+      "loss": 0.5231,
+      "step": 366
+    },
+    {
+      "epoch": 1.7902439024390244,
+      "grad_norm": 0.9962389469146729,
+      "learning_rate": 4.616956080167426e-06,
+      "loss": 0.5736,
+      "step": 367
+    },
+    {
+      "epoch": 1.7951219512195122,
+      "grad_norm": 1.3212288618087769,
+      "learning_rate": 4.614915623323786e-06,
+      "loss": 0.9586,
+      "step": 368
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9554499983787537,
+      "learning_rate": 4.612870199585092e-06,
+      "loss": 0.4346,
+      "step": 369
+    },
+    {
+      "epoch": 1.8048780487804879,
+      "grad_norm": 1.082366943359375,
+      "learning_rate": 4.610819813755038e-06,
+      "loss": 0.505,
+      "step": 370
+    },
+    {
+      "epoch": 1.8097560975609757,
+      "grad_norm": 1.225993037223816,
+      "learning_rate": 4.608764470648971e-06,
+      "loss": 0.4801,
+      "step": 371
+    },
+    {
+      "epoch": 1.8146341463414632,
+      "grad_norm": 1.0881706476211548,
+      "learning_rate": 4.606704175093879e-06,
+      "loss": 0.4478,
+      "step": 372
+    },
+    {
+      "epoch": 1.819512195121951,
+      "grad_norm": 1.114046335220337,
+      "learning_rate": 4.604638931928383e-06,
+      "loss": 0.8015,
+      "step": 373
+    },
+    {
+      "epoch": 1.8243902439024389,
+      "grad_norm": 0.9838706851005554,
+      "learning_rate": 4.602568746002718e-06,
+      "loss": 0.5204,
+      "step": 374
+    },
+    {
+      "epoch": 1.8292682926829267,
+      "grad_norm": 1.038713812828064,
+      "learning_rate": 4.600493622178734e-06,
+      "loss": 0.8388,
+      "step": 375
+    },
+    {
+      "epoch": 1.8341463414634145,
+      "grad_norm": 1.0684245824813843,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.5744,
+      "step": 376
+    },
+    {
+      "epoch": 1.8390243902439023,
+      "grad_norm": 0.8907456994056702,
+      "learning_rate": 4.596328580341169e-06,
+      "loss": 0.5621,
+      "step": 377
+    },
+    {
+      "epoch": 1.84390243902439,
+      "grad_norm": 0.9921515583992004,
+      "learning_rate": 4.5942386721092195e-06,
+      "loss": 0.6967,
+      "step": 378
+    },
+    {
+      "epoch": 1.848780487804878,
+      "grad_norm": 1.1683647632598877,
+      "learning_rate": 4.592143845542189e-06,
+      "loss": 0.6489,
+      "step": 379
+    },
+    {
+      "epoch": 1.8536585365853657,
+      "grad_norm": 1.0409291982650757,
+      "learning_rate": 4.590044105559797e-06,
+      "loss": 0.8945,
+      "step": 380
+    },
+    {
+      "epoch": 1.8585365853658535,
+      "grad_norm": 1.0684564113616943,
+      "learning_rate": 4.587939457093296e-06,
+      "loss": 0.5476,
+      "step": 381
+    },
+    {
+      "epoch": 1.8634146341463413,
+      "grad_norm": 1.3661733865737915,
+      "learning_rate": 4.585829905085468e-06,
+      "loss": 0.6763,
+      "step": 382
+    },
+    {
+      "epoch": 1.8682926829268292,
+      "grad_norm": 1.1465227603912354,
+      "learning_rate": 4.5837154544906135e-06,
+      "loss": 0.7817,
+      "step": 383
+    },
+    {
+      "epoch": 1.873170731707317,
+      "grad_norm": 1.0232677459716797,
+      "learning_rate": 4.581596110274535e-06,
+      "loss": 0.7276,
+      "step": 384
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 1.0359809398651123,
+      "learning_rate": 4.579471877414527e-06,
+      "loss": 0.9635,
+      "step": 385
+    },
+    {
+      "epoch": 1.8829268292682926,
+      "grad_norm": 1.2472409009933472,
+      "learning_rate": 4.577342760899368e-06,
+      "loss": 0.6782,
+      "step": 386
+    },
+    {
+      "epoch": 1.8878048780487804,
+      "grad_norm": 0.9241912961006165,
+      "learning_rate": 4.575208765729302e-06,
+      "loss": 0.5327,
+      "step": 387
+    },
+    {
+      "epoch": 1.8926829268292682,
+      "grad_norm": 1.3745805025100708,
+      "learning_rate": 4.573069896916035e-06,
+      "loss": 0.8436,
+      "step": 388
+    },
+    {
+      "epoch": 1.897560975609756,
+      "grad_norm": 1.0620322227478027,
+      "learning_rate": 4.5709261594827125e-06,
+      "loss": 0.659,
+      "step": 389
+    },
+    {
+      "epoch": 1.9024390243902438,
+      "grad_norm": 1.3068687915802002,
+      "learning_rate": 4.568777558463922e-06,
+      "loss": 0.5219,
+      "step": 390
+    },
+    {
+      "epoch": 1.9073170731707316,
+      "grad_norm": 1.0368698835372925,
+      "learning_rate": 4.566624098905665e-06,
+      "loss": 0.7099,
+      "step": 391
+    },
+    {
+      "epoch": 1.9121951219512194,
+      "grad_norm": 1.0876407623291016,
+      "learning_rate": 4.564465785865359e-06,
+      "loss": 0.5276,
+      "step": 392
+    },
+    {
+      "epoch": 1.9170731707317072,
+      "grad_norm": 0.9230280518531799,
+      "learning_rate": 4.56230262441182e-06,
+      "loss": 0.4292,
+      "step": 393
+    },
+    {
+      "epoch": 1.921951219512195,
+      "grad_norm": 0.9985240697860718,
+      "learning_rate": 4.560134619625247e-06,
+      "loss": 0.6602,
+      "step": 394
+    },
+    {
+      "epoch": 1.9268292682926829,
+      "grad_norm": 0.8872730135917664,
+      "learning_rate": 4.5579617765972155e-06,
+      "loss": 0.5711,
+      "step": 395
+    },
+    {
+      "epoch": 1.9317073170731707,
+      "grad_norm": 0.9474479556083679,
+      "learning_rate": 4.555784100430662e-06,
+      "loss": 0.5299,
+      "step": 396
+    },
+    {
+      "epoch": 1.9365853658536585,
+      "grad_norm": 1.0377358198165894,
+      "learning_rate": 4.553601596239877e-06,
+      "loss": 0.4649,
+      "step": 397
+    },
+    {
+      "epoch": 1.9414634146341463,
+      "grad_norm": 1.0765342712402344,
+      "learning_rate": 4.551414269150489e-06,
+      "loss": 0.5578,
+      "step": 398
+    },
+    {
+      "epoch": 1.946341463414634,
+      "grad_norm": 1.0844486951828003,
+      "learning_rate": 4.54922212429945e-06,
+      "loss": 0.5486,
+      "step": 399
+    },
+    {
+      "epoch": 1.951219512195122,
+      "grad_norm": 1.1640657186508179,
+      "learning_rate": 4.547025166835027e-06,
+      "loss": 0.7462,
+      "step": 400
+    },
+    {
+      "epoch": 1.9560975609756097,
+      "grad_norm": 1.2733076810836792,
+      "learning_rate": 4.544823401916794e-06,
+      "loss": 0.881,
+      "step": 401
+    },
+    {
+      "epoch": 1.9609756097560975,
+      "grad_norm": 1.155869960784912,
+      "learning_rate": 4.542616834715612e-06,
+      "loss": 0.6374,
+      "step": 402
+    },
+    {
+      "epoch": 1.9658536585365853,
+      "grad_norm": 0.8623374700546265,
+      "learning_rate": 4.540405470413618e-06,
+      "loss": 0.4188,
+      "step": 403
+    },
+    {
+      "epoch": 1.9707317073170731,
+      "grad_norm": 1.0502699613571167,
+      "learning_rate": 4.53818931420422e-06,
+      "loss": 0.7233,
+      "step": 404
+    },
+    {
+      "epoch": 1.975609756097561,
+      "grad_norm": 0.8630202412605286,
+      "learning_rate": 4.535968371292076e-06,
+      "loss": 0.5896,
+      "step": 405
+    },
+    {
+      "epoch": 1.9804878048780488,
+      "grad_norm": 1.1017824411392212,
+      "learning_rate": 4.533742646893086e-06,
+      "loss": 0.6971,
+      "step": 406
+    },
+    {
+      "epoch": 1.9853658536585366,
+      "grad_norm": 0.9387734532356262,
+      "learning_rate": 4.531512146234383e-06,
+      "loss": 0.6718,
+      "step": 407
+    },
+    {
+      "epoch": 1.9902439024390244,
+      "grad_norm": 1.0347439050674438,
+      "learning_rate": 4.529276874554312e-06,
+      "loss": 0.8829,
+      "step": 408
+    },
+    {
+      "epoch": 1.9951219512195122,
+      "grad_norm": 1.0173542499542236,
+      "learning_rate": 4.527036837102426e-06,
+      "loss": 0.5154,
+      "step": 409
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1297523975372314,
+      "learning_rate": 4.524792039139471e-06,
+      "loss": 0.7721,
+      "step": 410
+    },
+    {
+      "epoch": 2.004878048780488,
+      "grad_norm": 0.9763960242271423,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.3978,
+      "step": 411
+    },
+    {
+      "epoch": 2.0097560975609756,
+      "grad_norm": 0.9650730490684509,
+      "learning_rate": 4.520288182779214e-06,
+      "loss": 0.6826,
+      "step": 412
+    },
+    {
+      "epoch": 2.0146341463414634,
+      "grad_norm": 0.7682514190673828,
+      "learning_rate": 4.518029134959253e-06,
+      "loss": 0.351,
+      "step": 413
+    },
+    {
+      "epoch": 2.0195121951219512,
+      "grad_norm": 1.0681227445602417,
+      "learning_rate": 4.515765347782878e-06,
+      "loss": 0.6467,
+      "step": 414
+    },
+    {
+      "epoch": 2.024390243902439,
+      "grad_norm": 1.2426350116729736,
+      "learning_rate": 4.5134968265666085e-06,
+      "loss": 0.8831,
+      "step": 415
+    },
+    {
+      "epoch": 2.029268292682927,
+      "grad_norm": 0.9794759154319763,
+      "learning_rate": 4.511223576638084e-06,
+      "loss": 0.6419,
+      "step": 416
+    },
+    {
+      "epoch": 2.0341463414634147,
+      "grad_norm": 1.119649887084961,
+      "learning_rate": 4.508945603336049e-06,
+      "loss": 0.8023,
+      "step": 417
+    },
+    {
+      "epoch": 2.0390243902439025,
+      "grad_norm": 0.858971118927002,
+      "learning_rate": 4.50666291201034e-06,
+      "loss": 0.4974,
+      "step": 418
+    },
+    {
+      "epoch": 2.0439024390243903,
+      "grad_norm": 1.0508594512939453,
+      "learning_rate": 4.504375508021876e-06,
+      "loss": 0.3603,
+      "step": 419
+    },
+    {
+      "epoch": 2.048780487804878,
+      "grad_norm": 0.9566763043403625,
+      "learning_rate": 4.50208339674264e-06,
+      "loss": 0.7813,
+      "step": 420
+    },
+    {
+      "epoch": 2.053658536585366,
+      "grad_norm": 1.074040412902832,
+      "learning_rate": 4.499786583555675e-06,
+      "loss": 0.8065,
+      "step": 421
+    },
+    {
+      "epoch": 2.0585365853658537,
+      "grad_norm": 0.8816580772399902,
+      "learning_rate": 4.497485073855061e-06,
+      "loss": 0.4439,
+      "step": 422
+    },
+    {
+      "epoch": 2.0634146341463415,
+      "grad_norm": 1.0733896493911743,
+      "learning_rate": 4.495178873045913e-06,
+      "loss": 0.4481,
+      "step": 423
+    },
+    {
+      "epoch": 2.0682926829268293,
+      "grad_norm": 0.9010451436042786,
+      "learning_rate": 4.4928679865443605e-06,
+      "loss": 0.4407,
+      "step": 424
+    },
+    {
+      "epoch": 2.073170731707317,
+      "grad_norm": 0.9359092712402344,
+      "learning_rate": 4.4905524197775366e-06,
+      "loss": 0.5847,
+      "step": 425
+    },
+    {
+      "epoch": 2.078048780487805,
+      "grad_norm": 0.9389141798019409,
+      "learning_rate": 4.4882321781835666e-06,
+      "loss": 0.5377,
+      "step": 426
+    },
+    {
+      "epoch": 2.0829268292682928,
+      "grad_norm": 0.89751797914505,
+      "learning_rate": 4.4859072672115565e-06,
+      "loss": 0.3879,
+      "step": 427
+    },
+    {
+      "epoch": 2.0878048780487806,
+      "grad_norm": 1.0244700908660889,
+      "learning_rate": 4.483577692321577e-06,
+      "loss": 0.8237,
+      "step": 428
+    },
+    {
+      "epoch": 2.0926829268292684,
+      "grad_norm": 1.0703409910202026,
+      "learning_rate": 4.481243458984651e-06,
+      "loss": 0.5826,
+      "step": 429
+    },
+    {
+      "epoch": 2.097560975609756,
+      "grad_norm": 0.9995833039283752,
+      "learning_rate": 4.478904572682743e-06,
+      "loss": 0.6091,
+      "step": 430
+    },
+    {
+      "epoch": 2.102439024390244,
+      "grad_norm": 0.8649471998214722,
+      "learning_rate": 4.476561038908745e-06,
+      "loss": 0.4863,
+      "step": 431
+    },
+    {
+      "epoch": 2.107317073170732,
+      "grad_norm": 1.1670926809310913,
+      "learning_rate": 4.474212863166464e-06,
+      "loss": 0.6584,
+      "step": 432
+    },
+    {
+      "epoch": 2.1121951219512196,
+      "grad_norm": 1.2743312120437622,
+      "learning_rate": 4.471860050970608e-06,
+      "loss": 0.6777,
+      "step": 433
+    },
+    {
+      "epoch": 2.1170731707317074,
+      "grad_norm": 1.2678401470184326,
+      "learning_rate": 4.469502607846774e-06,
+      "loss": 0.9609,
+      "step": 434
+    },
+    {
+      "epoch": 2.1219512195121952,
+      "grad_norm": 0.9796558618545532,
+      "learning_rate": 4.467140539331434e-06,
+      "loss": 0.4574,
+      "step": 435
+    },
+    {
+      "epoch": 2.126829268292683,
+      "grad_norm": 1.0830684900283813,
+      "learning_rate": 4.464773850971924e-06,
+      "loss": 0.3067,
+      "step": 436
+    },
+    {
+      "epoch": 2.131707317073171,
+      "grad_norm": 1.002589464187622,
+      "learning_rate": 4.46240254832643e-06,
+      "loss": 0.5383,
+      "step": 437
+    },
+    {
+      "epoch": 2.1365853658536587,
+      "grad_norm": 1.1145734786987305,
+      "learning_rate": 4.460026636963971e-06,
+      "loss": 0.6173,
+      "step": 438
+    },
+    {
+      "epoch": 2.1414634146341465,
+      "grad_norm": 0.94740891456604,
+      "learning_rate": 4.4576461224643965e-06,
+      "loss": 0.4991,
+      "step": 439
+    },
+    {
+      "epoch": 2.1463414634146343,
+      "grad_norm": 0.8613864183425903,
+      "learning_rate": 4.455261010418359e-06,
+      "loss": 0.3956,
+      "step": 440
+    },
+    {
+      "epoch": 2.151219512195122,
+      "grad_norm": 0.9509091377258301,
+      "learning_rate": 4.452871306427314e-06,
+      "loss": 0.7165,
+      "step": 441
+    },
+    {
+      "epoch": 2.15609756097561,
+      "grad_norm": 1.259364128112793,
+      "learning_rate": 4.450477016103498e-06,
+      "loss": 0.5682,
+      "step": 442
+    },
+    {
+      "epoch": 2.1609756097560977,
+      "grad_norm": 1.1716279983520508,
+      "learning_rate": 4.4480781450699205e-06,
+      "loss": 0.4917,
+      "step": 443
+    },
+    {
+      "epoch": 2.1658536585365855,
+      "grad_norm": 1.0395866632461548,
+      "learning_rate": 4.4456746989603464e-06,
+      "loss": 0.4338,
+      "step": 444
+    },
+    {
+      "epoch": 2.1707317073170733,
+      "grad_norm": 1.232602834701538,
+      "learning_rate": 4.443266683419289e-06,
+      "loss": 0.6356,
+      "step": 445
+    },
+    {
+      "epoch": 2.175609756097561,
+      "grad_norm": 1.254172921180725,
+      "learning_rate": 4.440854104101988e-06,
+      "loss": 0.472,
+      "step": 446
+    },
+    {
+      "epoch": 2.180487804878049,
+      "grad_norm": 1.2319004535675049,
+      "learning_rate": 4.438436966674406e-06,
+      "loss": 0.6408,
+      "step": 447
+    },
+    {
+      "epoch": 2.1853658536585368,
+      "grad_norm": 1.0100780725479126,
+      "learning_rate": 4.436015276813208e-06,
+      "loss": 0.4524,
+      "step": 448
+    },
+    {
+      "epoch": 2.1902439024390246,
+      "grad_norm": 0.9656887650489807,
+      "learning_rate": 4.4335890402057505e-06,
+      "loss": 0.5999,
+      "step": 449
+    },
+    {
+      "epoch": 2.1951219512195124,
+      "grad_norm": 1.1730879545211792,
+      "learning_rate": 4.431158262550067e-06,
+      "loss": 0.5953,
+      "step": 450
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9255422949790955,
+      "learning_rate": 4.428722949554858e-06,
+      "loss": 0.3487,
+      "step": 451
+    },
+    {
+      "epoch": 2.204878048780488,
+      "grad_norm": 0.9878072142601013,
+      "learning_rate": 4.426283106939474e-06,
+      "loss": 0.4937,
+      "step": 452
+    },
+    {
+      "epoch": 2.209756097560976,
+      "grad_norm": 0.982023298740387,
+      "learning_rate": 4.423838740433903e-06,
+      "loss": 0.6299,
+      "step": 453
+    },
+    {
+      "epoch": 2.2146341463414636,
+      "grad_norm": 0.7727266550064087,
+      "learning_rate": 4.4213898557787586e-06,
+      "loss": 0.2789,
+      "step": 454
+    },
+    {
+      "epoch": 2.2195121951219514,
+      "grad_norm": 1.5341951847076416,
+      "learning_rate": 4.4189364587252636e-06,
+      "loss": 0.8498,
+      "step": 455
+    },
+    {
+      "epoch": 2.2243902439024392,
+      "grad_norm": 1.1611250638961792,
+      "learning_rate": 4.416478555035241e-06,
+      "loss": 0.4075,
+      "step": 456
+    },
+    {
+      "epoch": 2.229268292682927,
+      "grad_norm": 1.0459867715835571,
+      "learning_rate": 4.4140161504810935e-06,
+      "loss": 0.4946,
+      "step": 457
+    },
+    {
+      "epoch": 2.234146341463415,
+      "grad_norm": 0.9366090297698975,
+      "learning_rate": 4.4115492508457986e-06,
+      "loss": 0.3479,
+      "step": 458
+    },
+    {
+      "epoch": 2.2390243902439027,
+      "grad_norm": 1.0325448513031006,
+      "learning_rate": 4.409077861922887e-06,
+      "loss": 0.5437,
+      "step": 459
+    },
+    {
+      "epoch": 2.2439024390243905,
+      "grad_norm": 0.9326527118682861,
+      "learning_rate": 4.406601989516435e-06,
+      "loss": 0.4594,
+      "step": 460
+    },
+    {
+      "epoch": 2.2487804878048783,
+      "grad_norm": 0.7127180099487305,
+      "learning_rate": 4.404121639441047e-06,
+      "loss": 0.3067,
+      "step": 461
+    },
+    {
+      "epoch": 2.253658536585366,
+      "grad_norm": 1.0416815280914307,
+      "learning_rate": 4.401636817521843e-06,
+      "loss": 0.5402,
+      "step": 462
+    },
+    {
+      "epoch": 2.258536585365854,
+      "grad_norm": 1.8258185386657715,
+      "learning_rate": 4.399147529594447e-06,
+      "loss": 0.3964,
+      "step": 463
+    },
+    {
+      "epoch": 2.2634146341463417,
+      "grad_norm": 0.9795071482658386,
+      "learning_rate": 4.3966537815049686e-06,
+      "loss": 0.5118,
+      "step": 464
+    },
+    {
+      "epoch": 2.2682926829268295,
+      "grad_norm": 1.1920483112335205,
+      "learning_rate": 4.394155579109994e-06,
+      "loss": 0.6511,
+      "step": 465
+    },
+    {
+      "epoch": 2.2731707317073173,
+      "grad_norm": 1.361159324645996,
+      "learning_rate": 4.391652928276572e-06,
+      "loss": 0.6874,
+      "step": 466
+    },
+    {
+      "epoch": 2.278048780487805,
+      "grad_norm": 0.9973228573799133,
+      "learning_rate": 4.389145834882195e-06,
+      "loss": 0.6057,
+      "step": 467
+    },
+    {
+      "epoch": 2.2829268292682925,
+      "grad_norm": 1.3514574766159058,
+      "learning_rate": 4.386634304814789e-06,
+      "loss": 0.4762,
+      "step": 468
+    },
+    {
+      "epoch": 2.2878048780487803,
+      "grad_norm": 1.2089687585830688,
+      "learning_rate": 4.384118343972704e-06,
+      "loss": 0.689,
+      "step": 469
+    },
+    {
+      "epoch": 2.292682926829268,
+      "grad_norm": 0.9414058327674866,
+      "learning_rate": 4.381597958264692e-06,
+      "loss": 0.7257,
+      "step": 470
+    },
+    {
+      "epoch": 2.297560975609756,
+      "grad_norm": 1.0120850801467896,
+      "learning_rate": 4.379073153609896e-06,
+      "loss": 0.7515,
+      "step": 471
+    },
+    {
+      "epoch": 2.3024390243902437,
+      "grad_norm": 1.586024522781372,
+      "learning_rate": 4.37654393593784e-06,
+      "loss": 0.64,
+      "step": 472
+    },
+    {
+      "epoch": 2.3073170731707315,
+      "grad_norm": 0.9766375422477722,
+      "learning_rate": 4.3740103111884096e-06,
+      "loss": 0.5162,
+      "step": 473
+    },
+    {
+      "epoch": 2.3121951219512193,
+      "grad_norm": 1.001685380935669,
+      "learning_rate": 4.371472285311842e-06,
+      "loss": 0.4464,
+      "step": 474
+    },
+    {
+      "epoch": 2.317073170731707,
+      "grad_norm": 0.7923868894577026,
+      "learning_rate": 4.368929864268709e-06,
+      "loss": 0.365,
+      "step": 475
+    },
+    {
+      "epoch": 2.321951219512195,
+      "grad_norm": 1.0744857788085938,
+      "learning_rate": 4.366383054029907e-06,
+      "loss": 0.6253,
+      "step": 476
+    },
+    {
+      "epoch": 2.3268292682926828,
+      "grad_norm": 0.9156374931335449,
+      "learning_rate": 4.363831860576638e-06,
+      "loss": 0.6512,
+      "step": 477
+    },
+    {
+      "epoch": 2.3317073170731706,
+      "grad_norm": 1.0944534540176392,
+      "learning_rate": 4.361276289900396e-06,
+      "loss": 0.5627,
+      "step": 478
+    },
+    {
+      "epoch": 2.3365853658536584,
+      "grad_norm": 1.1438108682632446,
+      "learning_rate": 4.358716348002962e-06,
+      "loss": 0.7402,
+      "step": 479
+    },
+    {
+      "epoch": 2.341463414634146,
+      "grad_norm": 1.1678388118743896,
+      "learning_rate": 4.356152040896376e-06,
+      "loss": 0.512,
+      "step": 480
+    },
+    {
+      "epoch": 2.346341463414634,
+      "grad_norm": 0.8130245208740234,
+      "learning_rate": 4.3535833746029335e-06,
+      "loss": 0.3934,
+      "step": 481
+    },
+    {
+      "epoch": 2.351219512195122,
+      "grad_norm": 1.229127287864685,
+      "learning_rate": 4.351010355155165e-06,
+      "loss": 0.4782,
+      "step": 482
+    },
+    {
+      "epoch": 2.3560975609756096,
+      "grad_norm": 0.9830904006958008,
+      "learning_rate": 4.348432988595828e-06,
+      "loss": 0.3879,
+      "step": 483
+    },
+    {
+      "epoch": 2.3609756097560974,
+      "grad_norm": 1.3584911823272705,
+      "learning_rate": 4.345851280977885e-06,
+      "loss": 0.7305,
+      "step": 484
+    },
+    {
+      "epoch": 2.3658536585365852,
+      "grad_norm": 1.0106158256530762,
+      "learning_rate": 4.343265238364496e-06,
+      "loss": 0.4247,
+      "step": 485
+    },
+    {
+      "epoch": 2.370731707317073,
+      "grad_norm": 1.1921676397323608,
+      "learning_rate": 4.340674866829001e-06,
+      "loss": 0.5446,
+      "step": 486
+    },
+    {
+      "epoch": 2.375609756097561,
+      "grad_norm": 1.1733497381210327,
+      "learning_rate": 4.338080172454908e-06,
+      "loss": 0.8386,
+      "step": 487
+    },
+    {
+      "epoch": 2.3804878048780487,
+      "grad_norm": 0.8914453387260437,
+      "learning_rate": 4.335481161335875e-06,
+      "loss": 0.5393,
+      "step": 488
+    },
+    {
+      "epoch": 2.3853658536585365,
+      "grad_norm": 1.122434377670288,
+      "learning_rate": 4.332877839575699e-06,
+      "loss": 0.4645,
+      "step": 489
+    },
+    {
+      "epoch": 2.3902439024390243,
+      "grad_norm": 1.078754186630249,
+      "learning_rate": 4.330270213288301e-06,
+      "loss": 0.6284,
+      "step": 490
+    },
+    {
+      "epoch": 2.395121951219512,
+      "grad_norm": 1.0400331020355225,
+      "learning_rate": 4.32765828859771e-06,
+      "loss": 0.8431,
+      "step": 491
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9895585775375366,
+      "learning_rate": 4.325042071638051e-06,
+      "loss": 0.6601,
+      "step": 492
+    },
+    {
+      "epoch": 2.4048780487804877,
+      "grad_norm": 1.1766878366470337,
+      "learning_rate": 4.322421568553529e-06,
+      "loss": 0.4766,
+      "step": 493
+    },
+    {
+      "epoch": 2.4097560975609755,
+      "grad_norm": 0.9032670855522156,
+      "learning_rate": 4.319796785498416e-06,
+      "loss": 0.4363,
+      "step": 494
+    },
+    {
+      "epoch": 2.4146341463414633,
+      "grad_norm": 0.9736960530281067,
+      "learning_rate": 4.317167728637032e-06,
+      "loss": 0.6109,
+      "step": 495
+    },
+    {
+      "epoch": 2.419512195121951,
+      "grad_norm": 0.7923660278320312,
+      "learning_rate": 4.314534404143738e-06,
+      "loss": 0.4987,
+      "step": 496
+    },
+    {
+      "epoch": 2.424390243902439,
+      "grad_norm": 0.9087777137756348,
+      "learning_rate": 4.3118968182029155e-06,
+      "loss": 0.6411,
+      "step": 497
+    },
+    {
+      "epoch": 2.4292682926829268,
+      "grad_norm": 0.851117730140686,
+      "learning_rate": 4.3092549770089566e-06,
+      "loss": 0.3541,
+      "step": 498
+    },
+    {
+      "epoch": 2.4341463414634146,
+      "grad_norm": 0.9581378102302551,
+      "learning_rate": 4.306608886766243e-06,
+      "loss": 0.5448,
+      "step": 499
+    },
+    {
+      "epoch": 2.4390243902439024,
+      "grad_norm": 1.081851601600647,
+      "learning_rate": 4.303958553689137e-06,
+      "loss": 0.5593,
+      "step": 500
+    },
+    {
+      "epoch": 2.44390243902439,
+      "grad_norm": 1.1111576557159424,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.7566,
+      "step": 501
+    },
+    {
+      "epoch": 2.448780487804878,
+      "grad_norm": 1.0168198347091675,
+      "learning_rate": 4.2986451839390105e-06,
+      "loss": 0.3996,
+      "step": 502
+    },
+    {
+      "epoch": 2.453658536585366,
+      "grad_norm": 0.9412428140640259,
+      "learning_rate": 4.295982159744476e-06,
+      "loss": 0.5602,
+      "step": 503
+    },
+    {
+      "epoch": 2.4585365853658536,
+      "grad_norm": 1.0618679523468018,
+      "learning_rate": 4.293314917672498e-06,
+      "loss": 0.6466,
+      "step": 504
+    },
+    {
+      "epoch": 2.4634146341463414,
+      "grad_norm": 0.8414422273635864,
+      "learning_rate": 4.290643463987114e-06,
+      "loss": 0.357,
+      "step": 505
+    },
+    {
+      "epoch": 2.4682926829268292,
+      "grad_norm": 0.93071448802948,
+      "learning_rate": 4.287967804962252e-06,
+      "loss": 0.4179,
+      "step": 506
+    },
+    {
+      "epoch": 2.473170731707317,
+      "grad_norm": 0.9793124794960022,
+      "learning_rate": 4.285287946881718e-06,
+      "loss": 0.2698,
+      "step": 507
+    },
+    {
+      "epoch": 2.478048780487805,
+      "grad_norm": 0.9545714259147644,
+      "learning_rate": 4.282603896039178e-06,
+      "loss": 0.6855,
+      "step": 508
+    },
+    {
+      "epoch": 2.4829268292682927,
+      "grad_norm": 1.1140731573104858,
+      "learning_rate": 4.279915658738145e-06,
+      "loss": 0.5114,
+      "step": 509
+    },
+    {
+      "epoch": 2.4878048780487805,
+      "grad_norm": 1.0547738075256348,
+      "learning_rate": 4.277223241291966e-06,
+      "loss": 0.7367,
+      "step": 510
+    },
+    {
+      "epoch": 2.4926829268292683,
+      "grad_norm": 0.8616530895233154,
+      "learning_rate": 4.274526650023801e-06,
+      "loss": 0.5337,
+      "step": 511
+    },
+    {
+      "epoch": 2.497560975609756,
+      "grad_norm": 1.3530237674713135,
+      "learning_rate": 4.271825891266617e-06,
+      "loss": 0.5597,
+      "step": 512
+    },
+    {
+      "epoch": 2.502439024390244,
+      "grad_norm": 1.4124853610992432,
+      "learning_rate": 4.269120971363164e-06,
+      "loss": 0.8461,
+      "step": 513
+    },
+    {
+      "epoch": 2.5073170731707317,
+      "grad_norm": 1.0308994054794312,
+      "learning_rate": 4.266411896665967e-06,
+      "loss": 0.4304,
+      "step": 514
+    },
+    {
+      "epoch": 2.5121951219512195,
+      "grad_norm": 1.0831527709960938,
+      "learning_rate": 4.263698673537309e-06,
+      "loss": 0.5428,
+      "step": 515
+    },
+    {
+      "epoch": 2.5170731707317073,
+      "grad_norm": 1.2008475065231323,
+      "learning_rate": 4.260981308349214e-06,
+      "loss": 0.6922,
+      "step": 516
+    },
+    {
+      "epoch": 2.521951219512195,
+      "grad_norm": 1.1929224729537964,
+      "learning_rate": 4.258259807483434e-06,
+      "loss": 0.5716,
+      "step": 517
+    },
+    {
+      "epoch": 2.526829268292683,
+      "grad_norm": 1.016539216041565,
+      "learning_rate": 4.255534177331435e-06,
+      "loss": 0.5806,
+      "step": 518
+    },
+    {
+      "epoch": 2.5317073170731708,
+      "grad_norm": 1.018875002861023,
+      "learning_rate": 4.252804424294378e-06,
+      "loss": 0.5581,
+      "step": 519
+    },
+    {
+      "epoch": 2.5365853658536586,
+      "grad_norm": 0.9992810487747192,
+      "learning_rate": 4.25007055478311e-06,
+      "loss": 0.6786,
+      "step": 520
+    },
+    {
+      "epoch": 2.5414634146341464,
+      "grad_norm": 1.1207003593444824,
+      "learning_rate": 4.247332575218144e-06,
+      "loss": 0.4548,
+      "step": 521
+    },
+    {
+      "epoch": 2.546341463414634,
+      "grad_norm": 1.1575409173965454,
+      "learning_rate": 4.244590492029643e-06,
+      "loss": 0.6846,
+      "step": 522
+    },
+    {
+      "epoch": 2.551219512195122,
+      "grad_norm": 0.9805243015289307,
+      "learning_rate": 4.241844311657411e-06,
+      "loss": 0.4301,
+      "step": 523
+    },
+    {
+      "epoch": 2.55609756097561,
+      "grad_norm": 0.9760981202125549,
+      "learning_rate": 4.239094040550875e-06,
+      "loss": 0.3545,
+      "step": 524
+    },
+    {
+      "epoch": 2.5609756097560976,
+      "grad_norm": 0.8702017664909363,
+      "learning_rate": 4.236339685169065e-06,
+      "loss": 0.5429,
+      "step": 525
+    },
+    {
+      "epoch": 2.5658536585365854,
+      "grad_norm": 1.0681567192077637,
+      "learning_rate": 4.233581251980604e-06,
+      "loss": 0.3289,
+      "step": 526
+    },
+    {
+      "epoch": 2.5707317073170732,
+      "grad_norm": 1.1807548999786377,
+      "learning_rate": 4.230818747463696e-06,
+      "loss": 0.4876,
+      "step": 527
+    },
+    {
+      "epoch": 2.575609756097561,
+      "grad_norm": 0.9812930226325989,
+      "learning_rate": 4.228052178106101e-06,
+      "loss": 0.5025,
+      "step": 528
+    },
+    {
+      "epoch": 2.580487804878049,
+      "grad_norm": 0.8600794672966003,
+      "learning_rate": 4.2252815504051285e-06,
+      "loss": 0.3133,
+      "step": 529
+    },
+    {
+      "epoch": 2.5853658536585367,
+      "grad_norm": 1.89119553565979,
+      "learning_rate": 4.222506870867618e-06,
+      "loss": 0.6036,
+      "step": 530
+    },
+    {
+      "epoch": 2.5902439024390245,
+      "grad_norm": 1.0424220561981201,
+      "learning_rate": 4.2197281460099245e-06,
+      "loss": 0.6877,
+      "step": 531
+    },
+    {
+      "epoch": 2.5951219512195123,
+      "grad_norm": 1.027593731880188,
+      "learning_rate": 4.216945382357905e-06,
+      "loss": 0.6352,
+      "step": 532
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.1954094171524048,
+      "learning_rate": 4.214158586446901e-06,
+      "loss": 0.966,
+      "step": 533
+    },
+    {
+      "epoch": 2.604878048780488,
+      "grad_norm": 1.2490975856781006,
+      "learning_rate": 4.211367764821722e-06,
+      "loss": 0.9059,
+      "step": 534
+    },
+    {
+      "epoch": 2.6097560975609757,
+      "grad_norm": 1.1297111511230469,
+      "learning_rate": 4.208572924036634e-06,
+      "loss": 0.5006,
+      "step": 535
+    },
+    {
+      "epoch": 2.6146341463414635,
+      "grad_norm": 0.9446660876274109,
+      "learning_rate": 4.2057740706553415e-06,
+      "loss": 0.4905,
+      "step": 536
+    },
+    {
+      "epoch": 2.6195121951219513,
+      "grad_norm": 0.916692316532135,
+      "learning_rate": 4.202971211250971e-06,
+      "loss": 0.6694,
+      "step": 537
+    },
+    {
+      "epoch": 2.624390243902439,
+      "grad_norm": 0.840929388999939,
+      "learning_rate": 4.200164352406061e-06,
+      "loss": 0.3739,
+      "step": 538
+    },
+    {
+      "epoch": 2.629268292682927,
+      "grad_norm": 1.069427728652954,
+      "learning_rate": 4.197353500712539e-06,
+      "loss": 0.6359,
+      "step": 539
+    },
+    {
+      "epoch": 2.6341463414634148,
+      "grad_norm": 1.0660371780395508,
+      "learning_rate": 4.1945386627717115e-06,
+      "loss": 0.5006,
+      "step": 540
+    },
+    {
+      "epoch": 2.6390243902439026,
+      "grad_norm": 1.9662373065948486,
+      "learning_rate": 4.191719845194246e-06,
+      "loss": 0.649,
+      "step": 541
+    },
+    {
+      "epoch": 2.6439024390243904,
+      "grad_norm": 0.9865717887878418,
+      "learning_rate": 4.188897054600156e-06,
+      "loss": 0.6179,
+      "step": 542
+    },
+    {
+      "epoch": 2.648780487804878,
+      "grad_norm": 1.0393004417419434,
+      "learning_rate": 4.186070297618787e-06,
+      "loss": 0.7156,
+      "step": 543
+    },
+    {
+      "epoch": 2.653658536585366,
+      "grad_norm": 0.9797636270523071,
+      "learning_rate": 4.183239580888799e-06,
+      "loss": 0.7249,
+      "step": 544
+    },
+    {
+      "epoch": 2.658536585365854,
+      "grad_norm": 1.180819034576416,
+      "learning_rate": 4.18040491105815e-06,
+      "loss": 0.4961,
+      "step": 545
+    },
+    {
+      "epoch": 2.6634146341463416,
+      "grad_norm": 0.9986240863800049,
+      "learning_rate": 4.177566294784085e-06,
+      "loss": 0.4397,
+      "step": 546
+    },
+    {
+      "epoch": 2.6682926829268294,
+      "grad_norm": 1.2416610717773438,
+      "learning_rate": 4.174723738733114e-06,
+      "loss": 0.7625,
+      "step": 547
+    },
+    {
+      "epoch": 2.6731707317073172,
+      "grad_norm": 0.9271990656852722,
+      "learning_rate": 4.171877249581001e-06,
+      "loss": 0.6626,
+      "step": 548
+    },
+    {
+      "epoch": 2.678048780487805,
+      "grad_norm": 0.9085447788238525,
+      "learning_rate": 4.169026834012748e-06,
+      "loss": 0.4209,
+      "step": 549
+    },
+    {
+      "epoch": 2.682926829268293,
+      "grad_norm": 0.9767999649047852,
+      "learning_rate": 4.166172498722577e-06,
+      "loss": 0.4466,
+      "step": 550
+    },
+    {
+      "epoch": 2.68780487804878,
+      "grad_norm": 1.405003309249878,
+      "learning_rate": 4.163314250413913e-06,
+      "loss": 0.8207,
+      "step": 551
+    },
+    {
+      "epoch": 2.692682926829268,
+      "grad_norm": 1.001278042793274,
+      "learning_rate": 4.160452095799378e-06,
+      "loss": 0.4949,
+      "step": 552
+    },
+    {
+      "epoch": 2.697560975609756,
+      "grad_norm": 0.8813621401786804,
+      "learning_rate": 4.157586041600759e-06,
+      "loss": 0.2726,
+      "step": 553
+    },
+    {
+      "epoch": 2.7024390243902436,
+      "grad_norm": 1.087396502494812,
+      "learning_rate": 4.154716094549008e-06,
+      "loss": 0.6358,
+      "step": 554
+    },
+    {
+      "epoch": 2.7073170731707314,
+      "grad_norm": 0.864211916923523,
+      "learning_rate": 4.151842261384217e-06,
+      "loss": 0.4146,
+      "step": 555
+    },
+    {
+      "epoch": 2.7121951219512193,
+      "grad_norm": 1.580068588256836,
+      "learning_rate": 4.148964548855603e-06,
+      "loss": 0.9695,
+      "step": 556
+    },
+    {
+      "epoch": 2.717073170731707,
+      "grad_norm": 0.7623794674873352,
+      "learning_rate": 4.146082963721496e-06,
+      "loss": 0.3429,
+      "step": 557
+    },
+    {
+      "epoch": 2.721951219512195,
+      "grad_norm": 1.2673579454421997,
+      "learning_rate": 4.143197512749322e-06,
+      "loss": 1.1916,
+      "step": 558
+    },
+    {
+      "epoch": 2.7268292682926827,
+      "grad_norm": 1.0848994255065918,
+      "learning_rate": 4.140308202715581e-06,
+      "loss": 0.8112,
+      "step": 559
+    },
+    {
+      "epoch": 2.7317073170731705,
+      "grad_norm": 0.9205752015113831,
+      "learning_rate": 4.13741504040584e-06,
+      "loss": 0.4364,
+      "step": 560
+    },
+    {
+      "epoch": 2.7365853658536583,
+      "grad_norm": 1.0304152965545654,
+      "learning_rate": 4.134518032614713e-06,
+      "loss": 0.4841,
+      "step": 561
+    },
+    {
+      "epoch": 2.741463414634146,
+      "grad_norm": 0.9854786992073059,
+      "learning_rate": 4.1316171861458445e-06,
+      "loss": 0.418,
+      "step": 562
+    },
+    {
+      "epoch": 2.746341463414634,
+      "grad_norm": 1.0625019073486328,
+      "learning_rate": 4.128712507811893e-06,
+      "loss": 0.6479,
+      "step": 563
+    },
+    {
+      "epoch": 2.7512195121951217,
+      "grad_norm": 1.0722914934158325,
+      "learning_rate": 4.125804004434517e-06,
+      "loss": 0.6462,
+      "step": 564
+    },
+    {
+      "epoch": 2.7560975609756095,
+      "grad_norm": 0.890087366104126,
+      "learning_rate": 4.12289168284436e-06,
+      "loss": 0.4462,
+      "step": 565
+    },
+    {
+      "epoch": 2.7609756097560973,
+      "grad_norm": 1.0615348815917969,
+      "learning_rate": 4.119975549881029e-06,
+      "loss": 0.6229,
+      "step": 566
+    },
+    {
+      "epoch": 2.765853658536585,
+      "grad_norm": 0.8919638395309448,
+      "learning_rate": 4.1170556123930846e-06,
+      "loss": 0.3847,
+      "step": 567
+    },
+    {
+      "epoch": 2.770731707317073,
+      "grad_norm": 0.9881047606468201,
+      "learning_rate": 4.114131877238021e-06,
+      "loss": 0.5972,
+      "step": 568
+    },
+    {
+      "epoch": 2.7756097560975608,
+      "grad_norm": 0.8665289878845215,
+      "learning_rate": 4.111204351282254e-06,
+      "loss": 0.4755,
+      "step": 569
+    },
+    {
+      "epoch": 2.7804878048780486,
+      "grad_norm": 0.7870835661888123,
+      "learning_rate": 4.108273041401098e-06,
+      "loss": 0.4341,
+      "step": 570
+    },
+    {
+      "epoch": 2.7853658536585364,
+      "grad_norm": 1.3610732555389404,
+      "learning_rate": 4.105337954478756e-06,
+      "loss": 0.8646,
+      "step": 571
+    },
+    {
+      "epoch": 2.790243902439024,
+      "grad_norm": 1.0715464353561401,
+      "learning_rate": 4.102399097408304e-06,
+      "loss": 0.7017,
+      "step": 572
+    },
+    {
+      "epoch": 2.795121951219512,
+      "grad_norm": 0.952342689037323,
+      "learning_rate": 4.099456477091667e-06,
+      "loss": 0.3381,
+      "step": 573
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.151577115058899,
+      "learning_rate": 4.096510100439611e-06,
+      "loss": 0.8217,
+      "step": 574
+    },
+    {
+      "epoch": 2.8048780487804876,
+      "grad_norm": 0.8553835153579712,
+      "learning_rate": 4.093559974371725e-06,
+      "loss": 0.3334,
+      "step": 575
+    },
+    {
+      "epoch": 2.8097560975609754,
+      "grad_norm": 1.004453420639038,
+      "learning_rate": 4.0906061058164e-06,
+      "loss": 0.8324,
+      "step": 576
+    },
+    {
+      "epoch": 2.8146341463414632,
+      "grad_norm": 0.9378971457481384,
+      "learning_rate": 4.087648501710819e-06,
+      "loss": 0.3753,
+      "step": 577
+    },
+    {
+      "epoch": 2.819512195121951,
+      "grad_norm": 0.9435027241706848,
+      "learning_rate": 4.084687169000938e-06,
+      "loss": 0.5675,
+      "step": 578
+    },
+    {
+      "epoch": 2.824390243902439,
+      "grad_norm": 0.856401801109314,
+      "learning_rate": 4.081722114641469e-06,
+      "loss": 0.5309,
+      "step": 579
+    },
+    {
+      "epoch": 2.8292682926829267,
+      "grad_norm": 1.1497118473052979,
+      "learning_rate": 4.0787533455958626e-06,
+      "loss": 0.3339,
+      "step": 580
+    },
+    {
+      "epoch": 2.8341463414634145,
+      "grad_norm": 1.0275132656097412,
+      "learning_rate": 4.075780868836296e-06,
+      "loss": 0.4303,
+      "step": 581
+    },
+    {
+      "epoch": 2.8390243902439023,
+      "grad_norm": 0.903195321559906,
+      "learning_rate": 4.072804691343653e-06,
+      "loss": 0.495,
+      "step": 582
+    },
+    {
+      "epoch": 2.84390243902439,
+      "grad_norm": 1.1491634845733643,
+      "learning_rate": 4.069824820107507e-06,
+      "loss": 0.9994,
+      "step": 583
+    },
+    {
+      "epoch": 2.848780487804878,
+      "grad_norm": 0.7706964015960693,
+      "learning_rate": 4.06684126212611e-06,
+      "loss": 0.3226,
+      "step": 584
+    },
+    {
+      "epoch": 2.8536585365853657,
+      "grad_norm": 0.9241564273834229,
+      "learning_rate": 4.063854024406369e-06,
+      "loss": 0.5793,
+      "step": 585
+    },
+    {
+      "epoch": 2.8585365853658535,
+      "grad_norm": 0.9884312152862549,
+      "learning_rate": 4.060863113963835e-06,
+      "loss": 0.4683,
+      "step": 586
+    },
+    {
+      "epoch": 2.8634146341463413,
+      "grad_norm": 0.9997614026069641,
+      "learning_rate": 4.057868537822683e-06,
+      "loss": 0.519,
+      "step": 587
+    },
+    {
+      "epoch": 2.868292682926829,
+      "grad_norm": 0.9449941515922546,
+      "learning_rate": 4.054870303015695e-06,
+      "loss": 0.381,
+      "step": 588
+    },
+    {
+      "epoch": 2.873170731707317,
+      "grad_norm": 1.0271875858306885,
+      "learning_rate": 4.05186841658425e-06,
+      "loss": 0.4554,
+      "step": 589
+    },
+    {
+      "epoch": 2.8780487804878048,
+      "grad_norm": 0.9256722331047058,
+      "learning_rate": 4.048862885578301e-06,
+      "loss": 0.5297,
+      "step": 590
+    },
+    {
+      "epoch": 2.8829268292682926,
+      "grad_norm": 1.0047836303710938,
+      "learning_rate": 4.045853717056358e-06,
+      "loss": 0.5968,
+      "step": 591
+    },
+    {
+      "epoch": 2.8878048780487804,
+      "grad_norm": 0.9485352635383606,
+      "learning_rate": 4.0428409180854775e-06,
+      "loss": 0.5042,
+      "step": 592
+    },
+    {
+      "epoch": 2.892682926829268,
+      "grad_norm": 0.9246886372566223,
+      "learning_rate": 4.039824495741238e-06,
+      "loss": 0.4622,
+      "step": 593
+    },
+    {
+      "epoch": 2.897560975609756,
+      "grad_norm": 0.8539214134216309,
+      "learning_rate": 4.036804457107733e-06,
+      "loss": 0.53,
+      "step": 594
+    },
+    {
+      "epoch": 2.902439024390244,
+      "grad_norm": 1.2358900308609009,
+      "learning_rate": 4.0337808092775435e-06,
+      "loss": 0.701,
+      "step": 595
+    },
+    {
+      "epoch": 2.9073170731707316,
+      "grad_norm": 0.8977146744728088,
+      "learning_rate": 4.030753559351728e-06,
+      "loss": 0.3942,
+      "step": 596
+    },
+    {
+      "epoch": 2.9121951219512194,
+      "grad_norm": 0.8575205206871033,
+      "learning_rate": 4.027722714439808e-06,
+      "loss": 0.3713,
+      "step": 597
+    },
+    {
+      "epoch": 2.9170731707317072,
+      "grad_norm": 1.2680315971374512,
+      "learning_rate": 4.024688281659743e-06,
+      "loss": 0.9398,
+      "step": 598
+    },
+    {
+      "epoch": 2.921951219512195,
+      "grad_norm": 1.7810138463974,
+      "learning_rate": 4.021650268137924e-06,
+      "loss": 0.6056,
+      "step": 599
+    },
+    {
+      "epoch": 2.926829268292683,
+      "grad_norm": 0.8538106083869934,
+      "learning_rate": 4.018608681009143e-06,
+      "loss": 0.4939,
+      "step": 600
+    },
+    {
+      "epoch": 2.9317073170731707,
+      "grad_norm": 1.0676621198654175,
+      "learning_rate": 4.015563527416596e-06,
+      "loss": 0.5676,
+      "step": 601
+    },
+    {
+      "epoch": 2.9365853658536585,
+      "grad_norm": 0.987902820110321,
+      "learning_rate": 4.012514814511844e-06,
+      "loss": 0.5004,
+      "step": 602
+    },
+    {
+      "epoch": 2.9414634146341463,
+      "grad_norm": 0.9196493625640869,
+      "learning_rate": 4.009462549454816e-06,
+      "loss": 0.5476,
+      "step": 603
+    },
+    {
+      "epoch": 2.946341463414634,
+      "grad_norm": 1.1288981437683105,
+      "learning_rate": 4.006406739413775e-06,
+      "loss": 0.5804,
+      "step": 604
+    },
+    {
+      "epoch": 2.951219512195122,
+      "grad_norm": 0.8905384540557861,
+      "learning_rate": 4.003347391565317e-06,
+      "loss": 0.4979,
+      "step": 605
+    },
+    {
+      "epoch": 2.9560975609756097,
+      "grad_norm": 1.000335693359375,
+      "learning_rate": 4.000284513094342e-06,
+      "loss": 0.5727,
+      "step": 606
+    },
+    {
+      "epoch": 2.9609756097560975,
+      "grad_norm": 0.9347658753395081,
+      "learning_rate": 3.997218111194042e-06,
+      "loss": 0.5235,
+      "step": 607
+    },
+    {
+      "epoch": 2.9658536585365853,
+      "grad_norm": 0.8476413488388062,
+      "learning_rate": 3.994148193065886e-06,
+      "loss": 0.3922,
+      "step": 608
+    },
+    {
+      "epoch": 2.970731707317073,
+      "grad_norm": 0.8186416029930115,
+      "learning_rate": 3.991074765919598e-06,
+      "loss": 0.3482,
+      "step": 609
+    },
+    {
+      "epoch": 2.975609756097561,
+      "grad_norm": 1.2038166522979736,
+      "learning_rate": 3.987997836973147e-06,
+      "loss": 0.4684,
+      "step": 610
+    },
+    {
+      "epoch": 2.9804878048780488,
+      "grad_norm": 1.037007212638855,
+      "learning_rate": 3.984917413452721e-06,
+      "loss": 0.4811,
+      "step": 611
+    },
+    {
+      "epoch": 2.9853658536585366,
+      "grad_norm": 1.2915143966674805,
+      "learning_rate": 3.981833502592717e-06,
+      "loss": 0.7361,
+      "step": 612
+    },
+    {
+      "epoch": 2.9902439024390244,
+      "grad_norm": 1.5253301858901978,
+      "learning_rate": 3.978746111635725e-06,
+      "loss": 0.312,
+      "step": 613
+    },
+    {
+      "epoch": 2.995121951219512,
+      "grad_norm": 0.8432179093360901,
+      "learning_rate": 3.9756552478325045e-06,
+      "loss": 0.4931,
+      "step": 614
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8148512840270996,
+      "learning_rate": 3.972560918441972e-06,
+      "loss": 0.3185,
+      "step": 615
+    },
+    {
+      "epoch": 3.004878048780488,
+      "grad_norm": 0.7907199859619141,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.3731,
+      "step": 616
+    },
+    {
+      "epoch": 3.0097560975609756,
+      "grad_norm": 0.8440088033676147,
+      "learning_rate": 3.966361891975316e-06,
+      "loss": 0.4135,
+      "step": 617
+    },
+    {
+      "epoch": 3.0146341463414634,
+      "grad_norm": 0.9237042665481567,
+      "learning_rate": 3.963257209457652e-06,
+      "loss": 0.4563,
+      "step": 618
+    },
+    {
+      "epoch": 3.0195121951219512,
+      "grad_norm": 0.7932549118995667,
+      "learning_rate": 3.960149090469561e-06,
+      "loss": 0.2539,
+      "step": 619
+    },
+    {
+      "epoch": 3.024390243902439,
+      "grad_norm": 0.9151872396469116,
+      "learning_rate": 3.957037542310484e-06,
+      "loss": 0.2685,
+      "step": 620
+    },
+    {
+      "epoch": 3.029268292682927,
+      "grad_norm": 0.8039369583129883,
+      "learning_rate": 3.953922572287915e-06,
+      "loss": 0.4401,
+      "step": 621
+    },
+    {
+      "epoch": 3.0341463414634147,
+      "grad_norm": 0.9228289723396301,
+      "learning_rate": 3.950804187717384e-06,
+      "loss": 0.5399,
+      "step": 622
+    },
+    {
+      "epoch": 3.0390243902439025,
+      "grad_norm": 1.0876184701919556,
+      "learning_rate": 3.947682395922439e-06,
+      "loss": 0.6546,
+      "step": 623
+    },
+    {
+      "epoch": 3.0439024390243903,
+      "grad_norm": 0.8611107468605042,
+      "learning_rate": 3.9445572042346346e-06,
+      "loss": 0.2007,
+      "step": 624
+    },
+    {
+      "epoch": 3.048780487804878,
+      "grad_norm": 0.8184982538223267,
+      "learning_rate": 3.941428619993505e-06,
+      "loss": 0.3513,
+      "step": 625
+    },
+    {
+      "epoch": 3.053658536585366,
+      "grad_norm": 1.0696086883544922,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.2816,
+      "step": 626
+    },
+    {
+      "epoch": 3.0585365853658537,
+      "grad_norm": 1.4805349111557007,
+      "learning_rate": 3.935161303249231e-06,
+      "loss": 0.4662,
+      "step": 627
+    },
+    {
+      "epoch": 3.0634146341463415,
+      "grad_norm": 1.0260307788848877,
+      "learning_rate": 3.932022585464928e-06,
+      "loss": 0.4517,
+      "step": 628
+    },
+    {
+      "epoch": 3.0682926829268293,
+      "grad_norm": 0.972687304019928,
+      "learning_rate": 3.928880504564943e-06,
+      "loss": 0.3406,
+      "step": 629
+    },
+    {
+      "epoch": 3.073170731707317,
+      "grad_norm": 0.9252133369445801,
+      "learning_rate": 3.92573506792848e-06,
+      "loss": 0.3444,
+      "step": 630
+    },
+    {
+      "epoch": 3.078048780487805,
+      "grad_norm": 1.0907679796218872,
+      "learning_rate": 3.9225862829426184e-06,
+      "loss": 0.3427,
+      "step": 631
+    },
+    {
+      "epoch": 3.0829268292682928,
+      "grad_norm": 1.0673291683197021,
+      "learning_rate": 3.919434157002303e-06,
+      "loss": 0.4476,
+      "step": 632
+    },
+    {
+      "epoch": 3.0878048780487806,
+      "grad_norm": 0.758525550365448,
+      "learning_rate": 3.916278697510325e-06,
+      "loss": 0.274,
+      "step": 633
+    },
+    {
+      "epoch": 3.0926829268292684,
+      "grad_norm": 0.9257010221481323,
+      "learning_rate": 3.913119911877305e-06,
+      "loss": 0.5037,
+      "step": 634
+    },
+    {
+      "epoch": 3.097560975609756,
+      "grad_norm": 1.0138721466064453,
+      "learning_rate": 3.909957807521674e-06,
+      "loss": 0.253,
+      "step": 635
+    },
+    {
+      "epoch": 3.102439024390244,
+      "grad_norm": 0.9045038819313049,
+      "learning_rate": 3.906792391869657e-06,
+      "loss": 0.381,
+      "step": 636
+    },
+    {
+      "epoch": 3.107317073170732,
+      "grad_norm": 0.9882903099060059,
+      "learning_rate": 3.903623672355258e-06,
+      "loss": 0.3979,
+      "step": 637
+    },
+    {
+      "epoch": 3.1121951219512196,
+      "grad_norm": 0.9428558945655823,
+      "learning_rate": 3.900451656420237e-06,
+      "loss": 0.3531,
+      "step": 638
+    },
+    {
+      "epoch": 3.1170731707317074,
+      "grad_norm": 0.9153510332107544,
+      "learning_rate": 3.897276351514097e-06,
+      "loss": 0.2724,
+      "step": 639
+    },
+    {
+      "epoch": 3.1219512195121952,
+      "grad_norm": 1.1453444957733154,
+      "learning_rate": 3.894097765094065e-06,
+      "loss": 0.5488,
+      "step": 640
+    },
+    {
+      "epoch": 3.126829268292683,
+      "grad_norm": 0.7873973250389099,
+      "learning_rate": 3.890915904625075e-06,
+      "loss": 0.2855,
+      "step": 641
+    },
+    {
+      "epoch": 3.131707317073171,
+      "grad_norm": 1.2348054647445679,
+      "learning_rate": 3.887730777579751e-06,
+      "loss": 0.5758,
+      "step": 642
+    },
+    {
+      "epoch": 3.1365853658536587,
+      "grad_norm": 1.1336091756820679,
+      "learning_rate": 3.884542391438387e-06,
+      "loss": 0.681,
+      "step": 643
+    },
+    {
+      "epoch": 3.1414634146341465,
+      "grad_norm": 1.1599171161651611,
+      "learning_rate": 3.88135075368893e-06,
+      "loss": 0.8581,
+      "step": 644
+    },
+    {
+      "epoch": 3.1463414634146343,
+      "grad_norm": 0.9442818760871887,
+      "learning_rate": 3.878155871826968e-06,
+      "loss": 0.4053,
+      "step": 645
+    },
+    {
+      "epoch": 3.151219512195122,
+      "grad_norm": 0.9527677297592163,
+      "learning_rate": 3.874957753355701e-06,
+      "loss": 0.3673,
+      "step": 646
+    },
+    {
+      "epoch": 3.15609756097561,
+      "grad_norm": 0.9200324416160583,
+      "learning_rate": 3.8717564057859365e-06,
+      "loss": 0.6203,
+      "step": 647
+    },
+    {
+      "epoch": 3.1609756097560977,
+      "grad_norm": 1.0606499910354614,
+      "learning_rate": 3.868551836636063e-06,
+      "loss": 0.564,
+      "step": 648
+    },
+    {
+      "epoch": 3.1658536585365855,
+      "grad_norm": 1.064496397972107,
+      "learning_rate": 3.865344053432035e-06,
+      "loss": 0.316,
+      "step": 649
+    },
+    {
+      "epoch": 3.1707317073170733,
+      "grad_norm": 1.259859323501587,
+      "learning_rate": 3.862133063707353e-06,
+      "loss": 0.4876,
+      "step": 650
+    },
+    {
+      "epoch": 3.175609756097561,
+      "grad_norm": 0.9357886910438538,
+      "learning_rate": 3.858918875003053e-06,
+      "loss": 0.3449,
+      "step": 651
+    },
+    {
+      "epoch": 3.180487804878049,
+      "grad_norm": 0.8545693755149841,
+      "learning_rate": 3.855701494867679e-06,
+      "loss": 0.3919,
+      "step": 652
+    },
+    {
+      "epoch": 3.1853658536585368,
+      "grad_norm": 0.9744400382041931,
+      "learning_rate": 3.852480930857275e-06,
+      "loss": 0.5602,
+      "step": 653
+    },
+    {
+      "epoch": 3.1902439024390246,
+      "grad_norm": 0.9290996193885803,
+      "learning_rate": 3.849257190535356e-06,
+      "loss": 0.3391,
+      "step": 654
+    },
+    {
+      "epoch": 3.1951219512195124,
+      "grad_norm": 1.150904655456543,
+      "learning_rate": 3.846030281472902e-06,
+      "loss": 0.6754,
+      "step": 655
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.9049080610275269,
+      "learning_rate": 3.842800211248333e-06,
+      "loss": 0.342,
+      "step": 656
+    },
+    {
+      "epoch": 3.204878048780488,
+      "grad_norm": 1.6837809085845947,
+      "learning_rate": 3.839566987447492e-06,
+      "loss": 0.5241,
+      "step": 657
+    },
+    {
+      "epoch": 3.209756097560976,
+      "grad_norm": 0.9485814571380615,
+      "learning_rate": 3.8363306176636296e-06,
+      "loss": 0.6228,
+      "step": 658
+    },
+    {
+      "epoch": 3.2146341463414636,
+      "grad_norm": 1.029710054397583,
+      "learning_rate": 3.833091109497384e-06,
+      "loss": 0.6792,
+      "step": 659
+    },
+    {
+      "epoch": 3.2195121951219514,
+      "grad_norm": 0.881193220615387,
+      "learning_rate": 3.829848470556765e-06,
+      "loss": 0.25,
+      "step": 660
+    },
+    {
+      "epoch": 3.2243902439024392,
+      "grad_norm": 0.8923963904380798,
+      "learning_rate": 3.8266027084571335e-06,
+      "loss": 0.4637,
+      "step": 661
+    },
+    {
+      "epoch": 3.229268292682927,
+      "grad_norm": 1.039754867553711,
+      "learning_rate": 3.823353830821187e-06,
+      "loss": 0.2432,
+      "step": 662
+    },
+    {
+      "epoch": 3.234146341463415,
+      "grad_norm": 0.8456618189811707,
+      "learning_rate": 3.820101845278937e-06,
+      "loss": 0.4113,
+      "step": 663
+    },
+    {
+      "epoch": 3.2390243902439027,
+      "grad_norm": 1.1256377696990967,
+      "learning_rate": 3.816846759467696e-06,
+      "loss": 0.4089,
+      "step": 664
+    },
+    {
+      "epoch": 3.2439024390243905,
+      "grad_norm": 1.1781972646713257,
+      "learning_rate": 3.8135885810320587e-06,
+      "loss": 0.5524,
+      "step": 665
+    },
+    {
+      "epoch": 3.2487804878048783,
+      "grad_norm": 0.8150308132171631,
+      "learning_rate": 3.810327317623881e-06,
+      "loss": 0.3115,
+      "step": 666
+    },
+    {
+      "epoch": 3.253658536585366,
+      "grad_norm": 1.0984582901000977,
+      "learning_rate": 3.8070629769022628e-06,
+      "loss": 0.4814,
+      "step": 667
+    },
+    {
+      "epoch": 3.258536585365854,
+      "grad_norm": 0.9676326513290405,
+      "learning_rate": 3.8037955665335335e-06,
+      "loss": 0.3787,
+      "step": 668
+    },
+    {
+      "epoch": 3.2634146341463417,
+      "grad_norm": 1.1023277044296265,
+      "learning_rate": 3.800525094191231e-06,
+      "loss": 0.4983,
+      "step": 669
+    },
+    {
+      "epoch": 3.2682926829268295,
+      "grad_norm": 0.8557276725769043,
+      "learning_rate": 3.797251567556083e-06,
+      "loss": 0.3505,
+      "step": 670
+    },
+    {
+      "epoch": 3.2731707317073173,
+      "grad_norm": 0.8645808100700378,
+      "learning_rate": 3.793974994315991e-06,
+      "loss": 0.2222,
+      "step": 671
+    },
+    {
+      "epoch": 3.278048780487805,
+      "grad_norm": 1.0875083208084106,
+      "learning_rate": 3.790695382166013e-06,
+      "loss": 0.454,
+      "step": 672
+    },
+    {
+      "epoch": 3.2829268292682925,
+      "grad_norm": 0.9899719953536987,
+      "learning_rate": 3.7874127388083415e-06,
+      "loss": 0.3548,
+      "step": 673
+    },
+    {
+      "epoch": 3.2878048780487803,
+      "grad_norm": 1.0319510698318481,
+      "learning_rate": 3.7841270719522895e-06,
+      "loss": 0.4269,
+      "step": 674
+    },
+    {
+      "epoch": 3.292682926829268,
+      "grad_norm": 1.1482504606246948,
+      "learning_rate": 3.7808383893142692e-06,
+      "loss": 0.2515,
+      "step": 675
+    },
+    {
+      "epoch": 3.297560975609756,
+      "grad_norm": 0.977626383304596,
+      "learning_rate": 3.7775466986177763e-06,
+      "loss": 0.3434,
+      "step": 676
+    },
+    {
+      "epoch": 3.3024390243902437,
+      "grad_norm": 0.8783503770828247,
+      "learning_rate": 3.774252007593371e-06,
+      "loss": 0.2183,
+      "step": 677
+    },
+    {
+      "epoch": 3.3073170731707315,
+      "grad_norm": 1.0656304359436035,
+      "learning_rate": 3.7709543239786593e-06,
+      "loss": 0.5429,
+      "step": 678
+    },
+    {
+      "epoch": 3.3121951219512193,
+      "grad_norm": 0.9340170621871948,
+      "learning_rate": 3.767653655518277e-06,
+      "loss": 0.3605,
+      "step": 679
+    },
+    {
+      "epoch": 3.317073170731707,
+      "grad_norm": 1.228542685508728,
+      "learning_rate": 3.7643500099638673e-06,
+      "loss": 0.3435,
+      "step": 680
+    },
+    {
+      "epoch": 3.321951219512195,
+      "grad_norm": 0.9645997881889343,
+      "learning_rate": 3.7610433950740667e-06,
+      "loss": 0.6613,
+      "step": 681
+    },
+    {
+      "epoch": 3.3268292682926828,
+      "grad_norm": 1.077845811843872,
+      "learning_rate": 3.757733818614485e-06,
+      "loss": 0.4901,
+      "step": 682
+    },
+    {
+      "epoch": 3.3317073170731706,
+      "grad_norm": 0.8504129648208618,
+      "learning_rate": 3.7544212883576856e-06,
+      "loss": 0.4069,
+      "step": 683
+    },
+    {
+      "epoch": 3.3365853658536584,
+      "grad_norm": 1.0994789600372314,
+      "learning_rate": 3.751105812083172e-06,
+      "loss": 0.3476,
+      "step": 684
+    },
+    {
+      "epoch": 3.341463414634146,
+      "grad_norm": 1.2599692344665527,
+      "learning_rate": 3.7477873975773655e-06,
+      "loss": 0.5327,
+      "step": 685
+    },
+    {
+      "epoch": 3.346341463414634,
+      "grad_norm": 1.0701467990875244,
+      "learning_rate": 3.7444660526335853e-06,
+      "loss": 0.473,
+      "step": 686
+    },
+    {
+      "epoch": 3.351219512195122,
+      "grad_norm": 1.7928602695465088,
+      "learning_rate": 3.741141785052036e-06,
+      "loss": 0.8412,
+      "step": 687
+    },
+    {
+      "epoch": 3.3560975609756096,
+      "grad_norm": 1.075492262840271,
+      "learning_rate": 3.737814602639784e-06,
+      "loss": 0.5616,
+      "step": 688
+    },
+    {
+      "epoch": 3.3609756097560974,
+      "grad_norm": 0.9147948622703552,
+      "learning_rate": 3.7344845132107427e-06,
+      "loss": 0.4454,
+      "step": 689
+    },
+    {
+      "epoch": 3.3658536585365852,
+      "grad_norm": 1.0845904350280762,
+      "learning_rate": 3.731151524585651e-06,
+      "loss": 0.4776,
+      "step": 690
+    },
+    {
+      "epoch": 3.370731707317073,
+      "grad_norm": 1.2276109457015991,
+      "learning_rate": 3.7278156445920584e-06,
+      "loss": 0.8544,
+      "step": 691
+    },
+    {
+      "epoch": 3.375609756097561,
+      "grad_norm": 0.866237998008728,
+      "learning_rate": 3.724476881064303e-06,
+      "loss": 0.297,
+      "step": 692
+    },
+    {
+      "epoch": 3.3804878048780487,
+      "grad_norm": 0.8218240737915039,
+      "learning_rate": 3.721135241843496e-06,
+      "loss": 0.3716,
+      "step": 693
+    },
+    {
+      "epoch": 3.3853658536585365,
+      "grad_norm": 1.1857250928878784,
+      "learning_rate": 3.7177907347775016e-06,
+      "loss": 0.4965,
+      "step": 694
+    },
+    {
+      "epoch": 3.3902439024390243,
+      "grad_norm": 1.726030945777893,
+      "learning_rate": 3.71444336772092e-06,
+      "loss": 0.6302,
+      "step": 695
+    },
+    {
+      "epoch": 3.395121951219512,
+      "grad_norm": 1.3070801496505737,
+      "learning_rate": 3.711093148535068e-06,
+      "loss": 0.7139,
+      "step": 696
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.210959553718567,
+      "learning_rate": 3.707740085087959e-06,
+      "loss": 0.3059,
+      "step": 697
+    },
+    {
+      "epoch": 3.4048780487804877,
+      "grad_norm": 0.9498068690299988,
+      "learning_rate": 3.7043841852542884e-06,
+      "loss": 0.4441,
+      "step": 698
+    },
+    {
+      "epoch": 3.4097560975609755,
+      "grad_norm": 0.8833593130111694,
+      "learning_rate": 3.701025456915411e-06,
+      "loss": 0.3382,
+      "step": 699
+    },
+    {
+      "epoch": 3.4146341463414633,
+      "grad_norm": 0.9266107678413391,
+      "learning_rate": 3.697663907959327e-06,
+      "loss": 0.4165,
+      "step": 700
+    },
+    {
+      "epoch": 3.419512195121951,
+      "grad_norm": 1.4498791694641113,
+      "learning_rate": 3.6942995462806574e-06,
+      "loss": 0.7236,
+      "step": 701
+    },
+    {
+      "epoch": 3.424390243902439,
+      "grad_norm": 0.8553938269615173,
+      "learning_rate": 3.6909323797806314e-06,
+      "loss": 0.3514,
+      "step": 702
+    },
+    {
+      "epoch": 3.4292682926829268,
+      "grad_norm": 1.291072130203247,
+      "learning_rate": 3.6875624163670635e-06,
+      "loss": 0.6465,
+      "step": 703
+    },
+    {
+      "epoch": 3.4341463414634146,
+      "grad_norm": 1.2313672304153442,
+      "learning_rate": 3.6841896639543394e-06,
+      "loss": 0.3876,
+      "step": 704
+    },
+    {
+      "epoch": 3.4390243902439024,
+      "grad_norm": 0.8775885105133057,
+      "learning_rate": 3.6808141304633924e-06,
+      "loss": 0.3682,
+      "step": 705
+    },
+    {
+      "epoch": 3.44390243902439,
+      "grad_norm": 0.9601542949676514,
+      "learning_rate": 3.6774358238216878e-06,
+      "loss": 0.3753,
+      "step": 706
+    },
+    {
+      "epoch": 3.448780487804878,
+      "grad_norm": 1.371579885482788,
+      "learning_rate": 3.6740547519632048e-06,
+      "loss": 0.3362,
+      "step": 707
+    },
+    {
+      "epoch": 3.453658536585366,
+      "grad_norm": 0.7805762887001038,
+      "learning_rate": 3.670670922828414e-06,
+      "loss": 0.3909,
+      "step": 708
+    },
+    {
+      "epoch": 3.4585365853658536,
+      "grad_norm": 1.2143738269805908,
+      "learning_rate": 3.667284344364264e-06,
+      "loss": 0.3859,
+      "step": 709
+    },
+    {
+      "epoch": 3.4634146341463414,
+      "grad_norm": 0.9730637073516846,
+      "learning_rate": 3.6638950245241604e-06,
+      "loss": 0.5365,
+      "step": 710
+    },
+    {
+      "epoch": 3.4682926829268292,
+      "grad_norm": 1.291444182395935,
+      "learning_rate": 3.660502971267945e-06,
+      "loss": 0.3862,
+      "step": 711
+    },
+    {
+      "epoch": 3.473170731707317,
+      "grad_norm": 0.9339273571968079,
+      "learning_rate": 3.65710819256188e-06,
+      "loss": 0.1731,
+      "step": 712
+    },
+    {
+      "epoch": 3.478048780487805,
+      "grad_norm": 0.9810922741889954,
+      "learning_rate": 3.65371069637863e-06,
+      "loss": 0.3927,
+      "step": 713
+    },
+    {
+      "epoch": 3.4829268292682927,
+      "grad_norm": 1.027161955833435,
+      "learning_rate": 3.650310490697238e-06,
+      "loss": 0.5772,
+      "step": 714
+    },
+    {
+      "epoch": 3.4878048780487805,
+      "grad_norm": 1.0883551836013794,
+      "learning_rate": 3.646907583503114e-06,
+      "loss": 0.5687,
+      "step": 715
+    },
+    {
+      "epoch": 3.4926829268292683,
+      "grad_norm": 1.0536428689956665,
+      "learning_rate": 3.6435019827880093e-06,
+      "loss": 0.3598,
+      "step": 716
+    },
+    {
+      "epoch": 3.497560975609756,
+      "grad_norm": 0.8690207004547119,
+      "learning_rate": 3.640093696550003e-06,
+      "loss": 0.4317,
+      "step": 717
+    },
+    {
+      "epoch": 3.502439024390244,
+      "grad_norm": 0.9786996245384216,
+      "learning_rate": 3.6366827327934817e-06,
+      "loss": 0.428,
+      "step": 718
+    },
+    {
+      "epoch": 3.5073170731707317,
+      "grad_norm": 1.1201248168945312,
+      "learning_rate": 3.6332690995291176e-06,
+      "loss": 0.5273,
+      "step": 719
+    },
+    {
+      "epoch": 3.5121951219512195,
+      "grad_norm": 1.447026252746582,
+      "learning_rate": 3.6298528047738545e-06,
+      "loss": 1.3048,
+      "step": 720
+    },
+    {
+      "epoch": 3.5170731707317073,
+      "grad_norm": 0.988138735294342,
+      "learning_rate": 3.626433856550886e-06,
+      "loss": 0.6072,
+      "step": 721
+    },
+    {
+      "epoch": 3.521951219512195,
+      "grad_norm": 1.019419550895691,
+      "learning_rate": 3.623012262889637e-06,
+      "loss": 0.5362,
+      "step": 722
+    },
+    {
+      "epoch": 3.526829268292683,
+      "grad_norm": 1.2538626194000244,
+      "learning_rate": 3.6195880318257465e-06,
+      "loss": 0.5864,
+      "step": 723
+    },
+    {
+      "epoch": 3.5317073170731708,
+      "grad_norm": 1.2680046558380127,
+      "learning_rate": 3.616161171401046e-06,
+      "loss": 0.7208,
+      "step": 724
+    },
+    {
+      "epoch": 3.5365853658536586,
+      "grad_norm": 1.0290186405181885,
+      "learning_rate": 3.612731689663542e-06,
+      "loss": 0.3631,
+      "step": 725
+    },
+    {
+      "epoch": 3.5414634146341464,
+      "grad_norm": 0.941155731678009,
+      "learning_rate": 3.6092995946673996e-06,
+      "loss": 0.592,
+      "step": 726
+    },
+    {
+      "epoch": 3.546341463414634,
+      "grad_norm": 1.253089427947998,
+      "learning_rate": 3.605864894472918e-06,
+      "loss": 0.388,
+      "step": 727
+    },
+    {
+      "epoch": 3.551219512195122,
+      "grad_norm": 1.1095308065414429,
+      "learning_rate": 3.602427597146516e-06,
+      "loss": 0.5986,
+      "step": 728
+    },
+    {
+      "epoch": 3.55609756097561,
+      "grad_norm": 1.3642429113388062,
+      "learning_rate": 3.5989877107607134e-06,
+      "loss": 0.5178,
+      "step": 729
+    },
+    {
+      "epoch": 3.5609756097560976,
+      "grad_norm": 1.2215455770492554,
+      "learning_rate": 3.5955452433941075e-06,
+      "loss": 0.562,
+      "step": 730
+    },
+    {
+      "epoch": 3.5658536585365854,
+      "grad_norm": 0.955337405204773,
+      "learning_rate": 3.5921002031313586e-06,
+      "loss": 0.337,
+      "step": 731
+    },
+    {
+      "epoch": 3.5707317073170732,
+      "grad_norm": 0.9776975512504578,
+      "learning_rate": 3.58865259806317e-06,
+      "loss": 0.3314,
+      "step": 732
+    },
+    {
+      "epoch": 3.575609756097561,
+      "grad_norm": 1.1856746673583984,
+      "learning_rate": 3.585202436286267e-06,
+      "loss": 0.5582,
+      "step": 733
+    },
+    {
+      "epoch": 3.580487804878049,
+      "grad_norm": 1.1398861408233643,
+      "learning_rate": 3.581749725903381e-06,
+      "loss": 0.5148,
+      "step": 734
+    },
+    {
+      "epoch": 3.5853658536585367,
+      "grad_norm": 0.9702528119087219,
+      "learning_rate": 3.5782944750232274e-06,
+      "loss": 0.4666,
+      "step": 735
+    },
+    {
+      "epoch": 3.5902439024390245,
+      "grad_norm": 0.951115608215332,
+      "learning_rate": 3.574836691760489e-06,
+      "loss": 0.1839,
+      "step": 736
+    },
+    {
+      "epoch": 3.5951219512195123,
+      "grad_norm": 1.1828500032424927,
+      "learning_rate": 3.571376384235795e-06,
+      "loss": 0.4325,
+      "step": 737
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.0837311744689941,
+      "learning_rate": 3.5679135605757035e-06,
+      "loss": 0.3254,
+      "step": 738
+    },
+    {
+      "epoch": 3.604878048780488,
+      "grad_norm": 0.9672379493713379,
+      "learning_rate": 3.564448228912682e-06,
+      "loss": 0.3019,
+      "step": 739
+    },
+    {
+      "epoch": 3.6097560975609757,
+      "grad_norm": 1.1051793098449707,
+      "learning_rate": 3.5609803973850877e-06,
+      "loss": 0.4066,
+      "step": 740
+    },
+    {
+      "epoch": 3.6146341463414635,
+      "grad_norm": 1.0380092859268188,
+      "learning_rate": 3.557510074137147e-06,
+      "loss": 0.4469,
+      "step": 741
+    },
+    {
+      "epoch": 3.6195121951219513,
+      "grad_norm": 0.8451574444770813,
+      "learning_rate": 3.554037267318942e-06,
+      "loss": 0.4371,
+      "step": 742
+    },
+    {
+      "epoch": 3.624390243902439,
+      "grad_norm": 0.9186508655548096,
+      "learning_rate": 3.5505619850863847e-06,
+      "loss": 0.3235,
+      "step": 743
+    },
+    {
+      "epoch": 3.629268292682927,
+      "grad_norm": 0.9366088509559631,
+      "learning_rate": 3.5470842356012007e-06,
+      "loss": 0.2157,
+      "step": 744
+    },
+    {
+      "epoch": 3.6341463414634148,
+      "grad_norm": 1.0610483884811401,
+      "learning_rate": 3.5436040270309113e-06,
+      "loss": 0.5039,
+      "step": 745
+    },
+    {
+      "epoch": 3.6390243902439026,
+      "grad_norm": 0.9160799980163574,
+      "learning_rate": 3.540121367548811e-06,
+      "loss": 0.2946,
+      "step": 746
+    },
+    {
+      "epoch": 3.6439024390243904,
+      "grad_norm": 1.1034797430038452,
+      "learning_rate": 3.5366362653339524e-06,
+      "loss": 0.6456,
+      "step": 747
+    },
+    {
+      "epoch": 3.648780487804878,
+      "grad_norm": 1.0821247100830078,
+      "learning_rate": 3.533148728571124e-06,
+      "loss": 0.2399,
+      "step": 748
+    },
+    {
+      "epoch": 3.653658536585366,
+      "grad_norm": 1.058131456375122,
+      "learning_rate": 3.5296587654508317e-06,
+      "loss": 0.4446,
+      "step": 749
+    },
+    {
+      "epoch": 3.658536585365854,
+      "grad_norm": 2.5837414264678955,
+      "learning_rate": 3.526166384169279e-06,
+      "loss": 0.6744,
+      "step": 750
+    },
+    {
+      "epoch": 3.6634146341463416,
+      "grad_norm": 1.1982769966125488,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.3842,
+      "step": 751
+    },
+    {
+      "epoch": 3.6682926829268294,
+      "grad_norm": 0.849543035030365,
+      "learning_rate": 3.519174399935588e-06,
+      "loss": 0.2888,
+      "step": 752
+    },
+    {
+      "epoch": 3.6731707317073172,
+      "grad_norm": 0.9241414666175842,
+      "learning_rate": 3.5156748134041767e-06,
+      "loss": 0.2104,
+      "step": 753
+    },
+    {
+      "epoch": 3.678048780487805,
+      "grad_norm": 1.1095383167266846,
+      "learning_rate": 3.5121728415529203e-06,
+      "loss": 0.7921,
+      "step": 754
+    },
+    {
+      "epoch": 3.682926829268293,
+      "grad_norm": 1.0041614770889282,
+      "learning_rate": 3.5086684926062266e-06,
+      "loss": 0.3725,
+      "step": 755
+    },
+    {
+      "epoch": 3.68780487804878,
+      "grad_norm": 1.1060123443603516,
+      "learning_rate": 3.505161774794085e-06,
+      "loss": 0.4381,
+      "step": 756
+    },
+    {
+      "epoch": 3.692682926829268,
+      "grad_norm": 0.917390763759613,
+      "learning_rate": 3.5016526963520474e-06,
+      "loss": 0.3079,
+      "step": 757
+    },
+    {
+      "epoch": 3.697560975609756,
+      "grad_norm": 1.10245943069458,
+      "learning_rate": 3.498141265521212e-06,
+      "loss": 0.8131,
+      "step": 758
+    },
+    {
+      "epoch": 3.7024390243902436,
+      "grad_norm": 0.9380331039428711,
+      "learning_rate": 3.4946274905481997e-06,
+      "loss": 0.3149,
+      "step": 759
+    },
+    {
+      "epoch": 3.7073170731707314,
+      "grad_norm": 1.1269862651824951,
+      "learning_rate": 3.4911113796851364e-06,
+      "loss": 0.4485,
+      "step": 760
+    },
+    {
+      "epoch": 3.7121951219512193,
+      "grad_norm": 0.9700295925140381,
+      "learning_rate": 3.487592941189636e-06,
+      "loss": 0.2876,
+      "step": 761
+    },
+    {
+      "epoch": 3.717073170731707,
+      "grad_norm": 1.2680857181549072,
+      "learning_rate": 3.484072183324776e-06,
+      "loss": 0.7641,
+      "step": 762
+    },
+    {
+      "epoch": 3.721951219512195,
+      "grad_norm": 1.0977661609649658,
+      "learning_rate": 3.4805491143590823e-06,
+      "loss": 0.5787,
+      "step": 763
+    },
+    {
+      "epoch": 3.7268292682926827,
+      "grad_norm": 1.1288927793502808,
+      "learning_rate": 3.4770237425665103e-06,
+      "loss": 0.5136,
+      "step": 764
+    },
+    {
+      "epoch": 3.7317073170731705,
+      "grad_norm": 1.0365726947784424,
+      "learning_rate": 3.4734960762264204e-06,
+      "loss": 0.6113,
+      "step": 765
+    },
+    {
+      "epoch": 3.7365853658536583,
+      "grad_norm": 1.0862468481063843,
+      "learning_rate": 3.469966123623563e-06,
+      "loss": 0.6132,
+      "step": 766
+    },
+    {
+      "epoch": 3.741463414634146,
+      "grad_norm": 1.078573226928711,
+      "learning_rate": 3.46643389304806e-06,
+      "loss": 0.465,
+      "step": 767
+    },
+    {
+      "epoch": 3.746341463414634,
+      "grad_norm": 1.3158698081970215,
+      "learning_rate": 3.4628993927953786e-06,
+      "loss": 0.9712,
+      "step": 768
+    },
+    {
+      "epoch": 3.7512195121951217,
+      "grad_norm": 0.9415583610534668,
+      "learning_rate": 3.45936263116632e-06,
+      "loss": 0.2983,
+      "step": 769
+    },
+    {
+      "epoch": 3.7560975609756095,
+      "grad_norm": 0.8315958380699158,
+      "learning_rate": 3.4558236164669957e-06,
+      "loss": 0.3424,
+      "step": 770
+    },
+    {
+      "epoch": 3.7609756097560973,
+      "grad_norm": 1.0617963075637817,
+      "learning_rate": 3.4522823570088073e-06,
+      "loss": 0.2894,
+      "step": 771
+    },
+    {
+      "epoch": 3.765853658536585,
+      "grad_norm": 1.1169861555099487,
+      "learning_rate": 3.4487388611084295e-06,
+      "loss": 0.4487,
+      "step": 772
+    },
+    {
+      "epoch": 3.770731707317073,
+      "grad_norm": 1.055111050605774,
+      "learning_rate": 3.445193137087788e-06,
+      "loss": 0.2635,
+      "step": 773
+    },
+    {
+      "epoch": 3.7756097560975608,
+      "grad_norm": 0.9245924949645996,
+      "learning_rate": 3.4416451932740424e-06,
+      "loss": 0.4559,
+      "step": 774
+    },
+    {
+      "epoch": 3.7804878048780486,
+      "grad_norm": 1.205375075340271,
+      "learning_rate": 3.4380950379995652e-06,
+      "loss": 0.5927,
+      "step": 775
+    },
+    {
+      "epoch": 3.7853658536585364,
+      "grad_norm": 0.9713656902313232,
+      "learning_rate": 3.434542679601922e-06,
+      "loss": 0.469,
+      "step": 776
+    },
+    {
+      "epoch": 3.790243902439024,
+      "grad_norm": 1.037654161453247,
+      "learning_rate": 3.4309881264238538e-06,
+      "loss": 0.2425,
+      "step": 777
+    },
+    {
+      "epoch": 3.795121951219512,
+      "grad_norm": 1.1205179691314697,
+      "learning_rate": 3.4274313868132547e-06,
+      "loss": 0.3137,
+      "step": 778
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.0268551111221313,
+      "learning_rate": 3.4238724691231534e-06,
+      "loss": 0.3677,
+      "step": 779
+    },
+    {
+      "epoch": 3.8048780487804876,
+      "grad_norm": 1.1247752904891968,
+      "learning_rate": 3.4203113817116955e-06,
+      "loss": 0.5795,
+      "step": 780
+    },
+    {
+      "epoch": 3.8097560975609754,
+      "grad_norm": 1.093456745147705,
+      "learning_rate": 3.4167481329421204e-06,
+      "loss": 0.3284,
+      "step": 781
+    },
+    {
+      "epoch": 3.8146341463414632,
+      "grad_norm": 1.1700347661972046,
+      "learning_rate": 3.4131827311827447e-06,
+      "loss": 0.5421,
+      "step": 782
+    },
+    {
+      "epoch": 3.819512195121951,
+      "grad_norm": 0.8392572999000549,
+      "learning_rate": 3.4096151848069416e-06,
+      "loss": 0.2708,
+      "step": 783
+    },
+    {
+      "epoch": 3.824390243902439,
+      "grad_norm": 0.8408678770065308,
+      "learning_rate": 3.4060455021931195e-06,
+      "loss": 0.3812,
+      "step": 784
+    },
+    {
+      "epoch": 3.8292682926829267,
+      "grad_norm": 1.0810067653656006,
+      "learning_rate": 3.402473691724704e-06,
+      "loss": 0.3683,
+      "step": 785
+    },
+    {
+      "epoch": 3.8341463414634145,
+      "grad_norm": 0.7941387295722961,
+      "learning_rate": 3.39889976179012e-06,
+      "loss": 0.3159,
+      "step": 786
+    },
+    {
+      "epoch": 3.8390243902439023,
+      "grad_norm": 0.9092013835906982,
+      "learning_rate": 3.3953237207827673e-06,
+      "loss": 0.488,
+      "step": 787
+    },
+    {
+      "epoch": 3.84390243902439,
+      "grad_norm": 1.1730678081512451,
+      "learning_rate": 3.391745577101005e-06,
+      "loss": 0.729,
+      "step": 788
+    },
+    {
+      "epoch": 3.848780487804878,
+      "grad_norm": 1.2216498851776123,
+      "learning_rate": 3.3881653391481306e-06,
+      "loss": 0.4317,
+      "step": 789
+    },
+    {
+      "epoch": 3.8536585365853657,
+      "grad_norm": 1.2773486375808716,
+      "learning_rate": 3.384583015332359e-06,
+      "loss": 0.4652,
+      "step": 790
+    },
+    {
+      "epoch": 3.8585365853658535,
+      "grad_norm": 1.0143038034439087,
+      "learning_rate": 3.380998614066805e-06,
+      "loss": 0.3218,
+      "step": 791
+    },
+    {
+      "epoch": 3.8634146341463413,
+      "grad_norm": 1.1280455589294434,
+      "learning_rate": 3.3774121437694606e-06,
+      "loss": 0.4218,
+      "step": 792
+    },
+    {
+      "epoch": 3.868292682926829,
+      "grad_norm": 1.0607877969741821,
+      "learning_rate": 3.3738236128631786e-06,
+      "loss": 0.2838,
+      "step": 793
+    },
+    {
+      "epoch": 3.873170731707317,
+      "grad_norm": 1.145294189453125,
+      "learning_rate": 3.3702330297756503e-06,
+      "loss": 0.5267,
+      "step": 794
+    },
+    {
+      "epoch": 3.8780487804878048,
+      "grad_norm": 1.0045257806777954,
+      "learning_rate": 3.366640402939387e-06,
+      "loss": 0.1956,
+      "step": 795
+    },
+    {
+      "epoch": 3.8829268292682926,
+      "grad_norm": 1.9531723260879517,
+      "learning_rate": 3.363045740791698e-06,
+      "loss": 0.465,
+      "step": 796
+    },
+    {
+      "epoch": 3.8878048780487804,
+      "grad_norm": 0.8978796005249023,
+      "learning_rate": 3.3594490517746774e-06,
+      "loss": 0.3693,
+      "step": 797
+    },
+    {
+      "epoch": 3.892682926829268,
+      "grad_norm": 0.9718804955482483,
+      "learning_rate": 3.3558503443351733e-06,
+      "loss": 0.4562,
+      "step": 798
+    },
+    {
+      "epoch": 3.897560975609756,
+      "grad_norm": 0.888634204864502,
+      "learning_rate": 3.352249626924777e-06,
+      "loss": 0.3944,
+      "step": 799
+    },
+    {
+      "epoch": 3.902439024390244,
+      "grad_norm": 1.355827808380127,
+      "learning_rate": 3.348646907999801e-06,
+      "loss": 0.932,
+      "step": 800
+    },
+    {
+      "epoch": 3.9073170731707316,
+      "grad_norm": 1.005395770072937,
+      "learning_rate": 3.345042196021257e-06,
+      "loss": 0.4999,
+      "step": 801
+    },
+    {
+      "epoch": 3.9121951219512194,
+      "grad_norm": 1.253622055053711,
+      "learning_rate": 3.3414354994548385e-06,
+      "loss": 0.7352,
+      "step": 802
+    },
+    {
+      "epoch": 3.9170731707317072,
+      "grad_norm": 1.058511734008789,
+      "learning_rate": 3.337826826770898e-06,
+      "loss": 0.4555,
+      "step": 803
+    },
+    {
+      "epoch": 3.921951219512195,
+      "grad_norm": 1.0374521017074585,
+      "learning_rate": 3.3342161864444312e-06,
+      "loss": 0.3344,
+      "step": 804
+    },
+    {
+      "epoch": 3.926829268292683,
+      "grad_norm": 0.8886129856109619,
+      "learning_rate": 3.3306035869550534e-06,
+      "loss": 0.2988,
+      "step": 805
+    },
+    {
+      "epoch": 3.9317073170731707,
+      "grad_norm": 1.144066572189331,
+      "learning_rate": 3.326989036786981e-06,
+      "loss": 0.2734,
+      "step": 806
+    },
+    {
+      "epoch": 3.9365853658536585,
+      "grad_norm": 1.2550209760665894,
+      "learning_rate": 3.3233725444290126e-06,
+      "loss": 0.4412,
+      "step": 807
+    },
+    {
+      "epoch": 3.9414634146341463,
+      "grad_norm": 0.9990758299827576,
+      "learning_rate": 3.3197541183745065e-06,
+      "loss": 0.6003,
+      "step": 808
+    },
+    {
+      "epoch": 3.946341463414634,
+      "grad_norm": 0.9522280097007751,
+      "learning_rate": 3.3161337671213634e-06,
+      "loss": 0.3911,
+      "step": 809
+    },
+    {
+      "epoch": 3.951219512195122,
+      "grad_norm": 0.8392520546913147,
+      "learning_rate": 3.312511499172006e-06,
+      "loss": 0.3038,
+      "step": 810
+    },
+    {
+      "epoch": 3.9560975609756097,
+      "grad_norm": 1.0443915128707886,
+      "learning_rate": 3.3088873230333562e-06,
+      "loss": 0.4902,
+      "step": 811
+    },
+    {
+      "epoch": 3.9609756097560975,
+      "grad_norm": 1.5163060426712036,
+      "learning_rate": 3.3052612472168193e-06,
+      "loss": 0.3193,
+      "step": 812
+    },
+    {
+      "epoch": 3.9658536585365853,
+      "grad_norm": 1.0339828729629517,
+      "learning_rate": 3.3016332802382618e-06,
+      "loss": 0.4381,
+      "step": 813
+    },
+    {
+      "epoch": 3.970731707317073,
+      "grad_norm": 1.1660975217819214,
+      "learning_rate": 3.2980034306179897e-06,
+      "loss": 0.3802,
+      "step": 814
+    },
+    {
+      "epoch": 3.975609756097561,
+      "grad_norm": 0.9736506938934326,
+      "learning_rate": 3.294371706880733e-06,
+      "loss": 0.4898,
+      "step": 815
+    },
+    {
+      "epoch": 3.9804878048780488,
+      "grad_norm": 0.9874656796455383,
+      "learning_rate": 3.290738117555622e-06,
+      "loss": 0.3163,
+      "step": 816
+    },
+    {
+      "epoch": 3.9853658536585366,
+      "grad_norm": 1.0968129634857178,
+      "learning_rate": 3.2871026711761666e-06,
+      "loss": 0.5979,
+      "step": 817
+    },
+    {
+      "epoch": 3.9902439024390244,
+      "grad_norm": 1.0595239400863647,
+      "learning_rate": 3.2834653762802414e-06,
+      "loss": 0.3729,
+      "step": 818
+    },
+    {
+      "epoch": 3.995121951219512,
+      "grad_norm": 1.129920482635498,
+      "learning_rate": 3.2798262414100594e-06,
+      "loss": 0.3179,
+      "step": 819
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9918786883354187,
+      "learning_rate": 3.2761852751121566e-06,
+      "loss": 0.3184,
+      "step": 820
+    },
+    {
+      "epoch": 4.004878048780488,
+      "grad_norm": 1.0081284046173096,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.4687,
+      "step": 821
+    },
+    {
+      "epoch": 4.009756097560976,
+      "grad_norm": 1.025944471359253,
+      "learning_rate": 3.2688978824408136e-06,
+      "loss": 0.2941,
+      "step": 822
+    },
+    {
+      "epoch": 4.014634146341463,
+      "grad_norm": 0.9274433255195618,
+      "learning_rate": 3.2652514731818698e-06,
+      "loss": 0.2709,
+      "step": 823
+    },
+    {
+      "epoch": 4.019512195121951,
+      "grad_norm": 0.9776120185852051,
+      "learning_rate": 3.2616032667241564e-06,
+      "loss": 0.1915,
+      "step": 824
+    },
+    {
+      "epoch": 4.024390243902439,
+      "grad_norm": 1.0483906269073486,
+      "learning_rate": 3.257953271635513e-06,
+      "loss": 0.5291,
+      "step": 825
+    },
+    {
+      "epoch": 4.029268292682927,
+      "grad_norm": 0.9861155152320862,
+      "learning_rate": 3.2543014964879814e-06,
+      "loss": 0.2661,
+      "step": 826
+    },
+    {
+      "epoch": 4.034146341463415,
+      "grad_norm": 1.320236086845398,
+      "learning_rate": 3.250647949857781e-06,
+      "loss": 0.2301,
+      "step": 827
+    },
+    {
+      "epoch": 4.0390243902439025,
+      "grad_norm": 1.469008207321167,
+      "learning_rate": 3.2469926403252932e-06,
+      "loss": 0.6312,
+      "step": 828
+    },
+    {
+      "epoch": 4.04390243902439,
+      "grad_norm": 1.1796139478683472,
+      "learning_rate": 3.2433355764750417e-06,
+      "loss": 0.1662,
+      "step": 829
+    },
+    {
+      "epoch": 4.048780487804878,
+      "grad_norm": 1.3211941719055176,
+      "learning_rate": 3.2396767668956656e-06,
+      "loss": 0.514,
+      "step": 830
+    },
+    {
+      "epoch": 4.053658536585366,
+      "grad_norm": 1.2660949230194092,
+      "learning_rate": 3.2360162201799085e-06,
+      "loss": 0.4094,
+      "step": 831
+    },
+    {
+      "epoch": 4.058536585365854,
+      "grad_norm": 1.0212953090667725,
+      "learning_rate": 3.2323539449245906e-06,
+      "loss": 0.2655,
+      "step": 832
+    },
+    {
+      "epoch": 4.0634146341463415,
+      "grad_norm": 1.0902808904647827,
+      "learning_rate": 3.2286899497305917e-06,
+      "loss": 0.2253,
+      "step": 833
+    },
+    {
+      "epoch": 4.068292682926829,
+      "grad_norm": 0.9108849167823792,
+      "learning_rate": 3.2250242432028335e-06,
+      "loss": 0.3198,
+      "step": 834
+    },
+    {
+      "epoch": 4.073170731707317,
+      "grad_norm": 1.349802851676941,
+      "learning_rate": 3.221356833950254e-06,
+      "loss": 0.7985,
+      "step": 835
+    },
+    {
+      "epoch": 4.078048780487805,
+      "grad_norm": 0.9956075549125671,
+      "learning_rate": 3.21768773058579e-06,
+      "loss": 0.2803,
+      "step": 836
+    },
+    {
+      "epoch": 4.082926829268293,
+      "grad_norm": 1.123693823814392,
+      "learning_rate": 3.21401694172636e-06,
+      "loss": 0.2803,
+      "step": 837
+    },
+    {
+      "epoch": 4.087804878048781,
+      "grad_norm": 0.7673684358596802,
+      "learning_rate": 3.2103444759928383e-06,
+      "loss": 0.1641,
+      "step": 838
+    },
+    {
+      "epoch": 4.092682926829268,
+      "grad_norm": 0.8822228312492371,
+      "learning_rate": 3.2066703420100377e-06,
+      "loss": 0.1986,
+      "step": 839
+    },
+    {
+      "epoch": 4.097560975609756,
+      "grad_norm": 1.0781583786010742,
+      "learning_rate": 3.2029945484066883e-06,
+      "loss": 0.4772,
+      "step": 840
+    },
+    {
+      "epoch": 4.102439024390244,
+      "grad_norm": 1.1041297912597656,
+      "learning_rate": 3.1993171038154203e-06,
+      "loss": 0.2845,
+      "step": 841
+    },
+    {
+      "epoch": 4.107317073170732,
+      "grad_norm": 0.8251579403877258,
+      "learning_rate": 3.1956380168727385e-06,
+      "loss": 0.267,
+      "step": 842
+    },
+    {
+      "epoch": 4.11219512195122,
+      "grad_norm": 1.3584140539169312,
+      "learning_rate": 3.191957296219007e-06,
+      "loss": 0.4082,
+      "step": 843
+    },
+    {
+      "epoch": 4.117073170731707,
+      "grad_norm": 1.0828088521957397,
+      "learning_rate": 3.1882749504984247e-06,
+      "loss": 0.235,
+      "step": 844
+    },
+    {
+      "epoch": 4.121951219512195,
+      "grad_norm": 0.7371436357498169,
+      "learning_rate": 3.1845909883590076e-06,
+      "loss": 0.2477,
+      "step": 845
+    },
+    {
+      "epoch": 4.126829268292683,
+      "grad_norm": 1.1333115100860596,
+      "learning_rate": 3.180905418452569e-06,
+      "loss": 0.5014,
+      "step": 846
+    },
+    {
+      "epoch": 4.131707317073171,
+      "grad_norm": 0.8755896687507629,
+      "learning_rate": 3.1772182494346963e-06,
+      "loss": 0.3227,
+      "step": 847
+    },
+    {
+      "epoch": 4.136585365853659,
+      "grad_norm": 1.198217749595642,
+      "learning_rate": 3.1735294899647344e-06,
+      "loss": 0.415,
+      "step": 848
+    },
+    {
+      "epoch": 4.1414634146341465,
+      "grad_norm": 1.0420305728912354,
+      "learning_rate": 3.169839148705762e-06,
+      "loss": 0.2631,
+      "step": 849
+    },
+    {
+      "epoch": 4.146341463414634,
+      "grad_norm": 1.231887698173523,
+      "learning_rate": 3.1661472343245725e-06,
+      "loss": 0.7334,
+      "step": 850
+    },
+    {
+      "epoch": 4.151219512195122,
+      "grad_norm": 1.086815357208252,
+      "learning_rate": 3.162453755491655e-06,
+      "loss": 0.4588,
+      "step": 851
+    },
+    {
+      "epoch": 4.15609756097561,
+      "grad_norm": 1.2308528423309326,
+      "learning_rate": 3.158758720881171e-06,
+      "loss": 0.431,
+      "step": 852
+    },
+    {
+      "epoch": 4.160975609756098,
+      "grad_norm": 1.3028631210327148,
+      "learning_rate": 3.155062139170937e-06,
+      "loss": 0.351,
+      "step": 853
+    },
+    {
+      "epoch": 4.1658536585365855,
+      "grad_norm": 1.2701951265335083,
+      "learning_rate": 3.1513640190424034e-06,
+      "loss": 0.2138,
+      "step": 854
+    },
+    {
+      "epoch": 4.170731707317073,
+      "grad_norm": 1.1779377460479736,
+      "learning_rate": 3.147664369180632e-06,
+      "loss": 0.2532,
+      "step": 855
+    },
+    {
+      "epoch": 4.175609756097561,
+      "grad_norm": 1.287402868270874,
+      "learning_rate": 3.143963198274278e-06,
+      "loss": 0.4129,
+      "step": 856
+    },
+    {
+      "epoch": 4.180487804878049,
+      "grad_norm": 0.9432743787765503,
+      "learning_rate": 3.140260515015569e-06,
+      "loss": 0.2941,
+      "step": 857
+    },
+    {
+      "epoch": 4.185365853658537,
+      "grad_norm": 1.0425447225570679,
+      "learning_rate": 3.136556328100284e-06,
+      "loss": 0.3346,
+      "step": 858
+    },
+    {
+      "epoch": 4.190243902439025,
+      "grad_norm": 1.240814447402954,
+      "learning_rate": 3.132850646227734e-06,
+      "loss": 0.6134,
+      "step": 859
+    },
+    {
+      "epoch": 4.195121951219512,
+      "grad_norm": 1.1277090311050415,
+      "learning_rate": 3.12914347810074e-06,
+      "loss": 0.3543,
+      "step": 860
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.275528073310852,
+      "learning_rate": 3.125434832425613e-06,
+      "loss": 0.2518,
+      "step": 861
+    },
+    {
+      "epoch": 4.204878048780488,
+      "grad_norm": 0.7928908467292786,
+      "learning_rate": 3.121724717912138e-06,
+      "loss": 0.1532,
+      "step": 862
+    },
+    {
+      "epoch": 4.209756097560976,
+      "grad_norm": 1.1028555631637573,
+      "learning_rate": 3.118013143273542e-06,
+      "loss": 0.3368,
+      "step": 863
+    },
+    {
+      "epoch": 4.214634146341464,
+      "grad_norm": 1.8106095790863037,
+      "learning_rate": 3.1143001172264893e-06,
+      "loss": 0.2334,
+      "step": 864
+    },
+    {
+      "epoch": 4.219512195121951,
+      "grad_norm": 1.2573802471160889,
+      "learning_rate": 3.1105856484910474e-06,
+      "loss": 0.3072,
+      "step": 865
+    },
+    {
+      "epoch": 4.224390243902439,
+      "grad_norm": 0.870959460735321,
+      "learning_rate": 3.1068697457906736e-06,
+      "loss": 0.2215,
+      "step": 866
+    },
+    {
+      "epoch": 4.229268292682927,
+      "grad_norm": 1.066684603691101,
+      "learning_rate": 3.1031524178521938e-06,
+      "loss": 0.4311,
+      "step": 867
+    },
+    {
+      "epoch": 4.234146341463415,
+      "grad_norm": 1.2743711471557617,
+      "learning_rate": 3.0994336734057804e-06,
+      "loss": 0.1342,
+      "step": 868
+    },
+    {
+      "epoch": 4.239024390243903,
+      "grad_norm": 0.8251144289970398,
+      "learning_rate": 3.0957135211849315e-06,
+      "loss": 0.2224,
+      "step": 869
+    },
+    {
+      "epoch": 4.2439024390243905,
+      "grad_norm": 1.1386126279830933,
+      "learning_rate": 3.0919919699264535e-06,
+      "loss": 0.2674,
+      "step": 870
+    },
+    {
+      "epoch": 4.248780487804878,
+      "grad_norm": 1.1335779428482056,
+      "learning_rate": 3.0882690283704355e-06,
+      "loss": 0.8506,
+      "step": 871
+    },
+    {
+      "epoch": 4.253658536585366,
+      "grad_norm": 0.9600904583930969,
+      "learning_rate": 3.084544705260234e-06,
+      "loss": 0.2926,
+      "step": 872
+    },
+    {
+      "epoch": 4.258536585365854,
+      "grad_norm": 0.896686851978302,
+      "learning_rate": 3.080819009342451e-06,
+      "loss": 0.1941,
+      "step": 873
+    },
+    {
+      "epoch": 4.263414634146342,
+      "grad_norm": 0.8546966314315796,
+      "learning_rate": 3.077091949366908e-06,
+      "loss": 0.2209,
+      "step": 874
+    },
+    {
+      "epoch": 4.2682926829268295,
+      "grad_norm": 0.9326870441436768,
+      "learning_rate": 3.073363534086636e-06,
+      "loss": 0.2608,
+      "step": 875
+    },
+    {
+      "epoch": 4.273170731707317,
+      "grad_norm": 0.8635675311088562,
+      "learning_rate": 3.0696337722578444e-06,
+      "loss": 0.1911,
+      "step": 876
+    },
+    {
+      "epoch": 4.278048780487805,
+      "grad_norm": 1.2039563655853271,
+      "learning_rate": 3.0659026726399072e-06,
+      "loss": 0.3195,
+      "step": 877
+    },
+    {
+      "epoch": 4.282926829268293,
+      "grad_norm": 1.041063666343689,
+      "learning_rate": 3.0621702439953393e-06,
+      "loss": 0.4438,
+      "step": 878
+    },
+    {
+      "epoch": 4.287804878048781,
+      "grad_norm": 0.9710636734962463,
+      "learning_rate": 3.0584364950897768e-06,
+      "loss": 0.1192,
+      "step": 879
+    },
+    {
+      "epoch": 4.2926829268292686,
+      "grad_norm": 1.001789927482605,
+      "learning_rate": 3.0547014346919574e-06,
+      "loss": 0.2768,
+      "step": 880
+    },
+    {
+      "epoch": 4.297560975609756,
+      "grad_norm": 1.0699266195297241,
+      "learning_rate": 3.0509650715736977e-06,
+      "loss": 0.2787,
+      "step": 881
+    },
+    {
+      "epoch": 4.302439024390244,
+      "grad_norm": 0.9404529929161072,
+      "learning_rate": 3.0472274145098744e-06,
+      "loss": 0.4201,
+      "step": 882
+    },
+    {
+      "epoch": 4.307317073170732,
+      "grad_norm": 1.296057105064392,
+      "learning_rate": 3.0434884722784026e-06,
+      "loss": 0.3718,
+      "step": 883
+    },
+    {
+      "epoch": 4.31219512195122,
+      "grad_norm": 0.9642940163612366,
+      "learning_rate": 3.0397482536602168e-06,
+      "loss": 0.2024,
+      "step": 884
+    },
+    {
+      "epoch": 4.317073170731708,
+      "grad_norm": 1.0013500452041626,
+      "learning_rate": 3.0360067674392475e-06,
+      "loss": 0.2857,
+      "step": 885
+    },
+    {
+      "epoch": 4.321951219512195,
+      "grad_norm": 1.1276105642318726,
+      "learning_rate": 3.0322640224024024e-06,
+      "loss": 0.2326,
+      "step": 886
+    },
+    {
+      "epoch": 4.326829268292683,
+      "grad_norm": 1.1642167568206787,
+      "learning_rate": 3.0285200273395478e-06,
+      "loss": 0.3897,
+      "step": 887
+    },
+    {
+      "epoch": 4.331707317073171,
+      "grad_norm": 1.1692358255386353,
+      "learning_rate": 3.024774791043481e-06,
+      "loss": 0.4172,
+      "step": 888
+    },
+    {
+      "epoch": 4.336585365853659,
+      "grad_norm": 1.2507195472717285,
+      "learning_rate": 3.021028322309921e-06,
+      "loss": 0.4637,
+      "step": 889
+    },
+    {
+      "epoch": 4.341463414634147,
+      "grad_norm": 1.2589614391326904,
+      "learning_rate": 3.0172806299374734e-06,
+      "loss": 0.4284,
+      "step": 890
+    },
+    {
+      "epoch": 4.3463414634146345,
+      "grad_norm": 1.159298300743103,
+      "learning_rate": 3.0135317227276247e-06,
+      "loss": 0.3921,
+      "step": 891
+    },
+    {
+      "epoch": 4.351219512195122,
+      "grad_norm": 1.0309828519821167,
+      "learning_rate": 3.0097816094847104e-06,
+      "loss": 0.3799,
+      "step": 892
+    },
+    {
+      "epoch": 4.35609756097561,
+      "grad_norm": 0.9921269416809082,
+      "learning_rate": 3.0060302990158984e-06,
+      "loss": 0.2218,
+      "step": 893
+    },
+    {
+      "epoch": 4.360975609756098,
+      "grad_norm": 0.9881783127784729,
+      "learning_rate": 3.002277800131171e-06,
+      "loss": 0.1882,
+      "step": 894
+    },
+    {
+      "epoch": 4.365853658536586,
+      "grad_norm": 1.0279638767242432,
+      "learning_rate": 2.998524121643298e-06,
+      "loss": 0.2855,
+      "step": 895
+    },
+    {
+      "epoch": 4.3707317073170735,
+      "grad_norm": 1.2783422470092773,
+      "learning_rate": 2.994769272367822e-06,
+      "loss": 0.3899,
+      "step": 896
+    },
+    {
+      "epoch": 4.375609756097561,
+      "grad_norm": 0.9480025172233582,
+      "learning_rate": 2.991013261123035e-06,
+      "loss": 0.2136,
+      "step": 897
+    },
+    {
+      "epoch": 4.380487804878049,
+      "grad_norm": 1.1166763305664062,
+      "learning_rate": 2.9872560967299554e-06,
+      "loss": 0.2992,
+      "step": 898
+    },
+    {
+      "epoch": 4.385365853658537,
+      "grad_norm": 1.1016902923583984,
+      "learning_rate": 2.9834977880123132e-06,
+      "loss": 0.1948,
+      "step": 899
+    },
+    {
+      "epoch": 4.390243902439025,
+      "grad_norm": 1.0177924633026123,
+      "learning_rate": 2.9797383437965243e-06,
+      "loss": 0.2447,
+      "step": 900
+    },
+    {
+      "epoch": 4.3951219512195125,
+      "grad_norm": 0.910971999168396,
+      "learning_rate": 2.975977772911671e-06,
+      "loss": 0.2198,
+      "step": 901
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.0623188018798828,
+      "learning_rate": 2.972216084189482e-06,
+      "loss": 0.2353,
+      "step": 902
+    },
+    {
+      "epoch": 4.404878048780488,
+      "grad_norm": 0.8617135286331177,
+      "learning_rate": 2.9684532864643123e-06,
+      "loss": 0.1998,
+      "step": 903
+    },
+    {
+      "epoch": 4.409756097560976,
+      "grad_norm": 1.0422427654266357,
+      "learning_rate": 2.964689388573118e-06,
+      "loss": 0.1688,
+      "step": 904
+    },
+    {
+      "epoch": 4.414634146341464,
+      "grad_norm": 1.0696886777877808,
+      "learning_rate": 2.9609243993554434e-06,
+      "loss": 0.4587,
+      "step": 905
+    },
+    {
+      "epoch": 4.419512195121952,
+      "grad_norm": 1.1068525314331055,
+      "learning_rate": 2.9571583276533923e-06,
+      "loss": 0.2282,
+      "step": 906
+    },
+    {
+      "epoch": 4.424390243902439,
+      "grad_norm": 1.363254189491272,
+      "learning_rate": 2.9533911823116124e-06,
+      "loss": 0.7017,
+      "step": 907
+    },
+    {
+      "epoch": 4.429268292682927,
+      "grad_norm": 1.504699945449829,
+      "learning_rate": 2.9496229721772734e-06,
+      "loss": 0.4154,
+      "step": 908
+    },
+    {
+      "epoch": 4.434146341463415,
+      "grad_norm": 1.0644068717956543,
+      "learning_rate": 2.9458537061000435e-06,
+      "loss": 0.3262,
+      "step": 909
+    },
+    {
+      "epoch": 4.439024390243903,
+      "grad_norm": 1.1175642013549805,
+      "learning_rate": 2.9420833929320726e-06,
+      "loss": 0.3613,
+      "step": 910
+    },
+    {
+      "epoch": 4.443902439024391,
+      "grad_norm": 1.418470859527588,
+      "learning_rate": 2.93831204152797e-06,
+      "loss": 0.5274,
+      "step": 911
+    },
+    {
+      "epoch": 4.4487804878048784,
+      "grad_norm": 0.8108303546905518,
+      "learning_rate": 2.9345396607447807e-06,
+      "loss": 0.1267,
+      "step": 912
+    },
+    {
+      "epoch": 4.453658536585366,
+      "grad_norm": 1.049116611480713,
+      "learning_rate": 2.9307662594419704e-06,
+      "loss": 0.2678,
+      "step": 913
+    },
+    {
+      "epoch": 4.458536585365854,
+      "grad_norm": 1.158342719078064,
+      "learning_rate": 2.9269918464814e-06,
+      "loss": 0.3162,
+      "step": 914
+    },
+    {
+      "epoch": 4.463414634146342,
+      "grad_norm": 1.4344679117202759,
+      "learning_rate": 2.923216430727306e-06,
+      "loss": 0.4786,
+      "step": 915
+    },
+    {
+      "epoch": 4.46829268292683,
+      "grad_norm": 1.156111478805542,
+      "learning_rate": 2.9194400210462808e-06,
+      "loss": 0.4763,
+      "step": 916
+    },
+    {
+      "epoch": 4.473170731707317,
+      "grad_norm": 1.1826188564300537,
+      "learning_rate": 2.91566262630725e-06,
+      "loss": 0.5285,
+      "step": 917
+    },
+    {
+      "epoch": 4.478048780487805,
+      "grad_norm": 1.2508891820907593,
+      "learning_rate": 2.9118842553814526e-06,
+      "loss": 0.2565,
+      "step": 918
+    },
+    {
+      "epoch": 4.482926829268292,
+      "grad_norm": 0.9462976455688477,
+      "learning_rate": 2.9081049171424223e-06,
+      "loss": 0.2208,
+      "step": 919
+    },
+    {
+      "epoch": 4.487804878048781,
+      "grad_norm": 0.9479988813400269,
+      "learning_rate": 2.9043246204659624e-06,
+      "loss": 0.2,
+      "step": 920
+    },
+    {
+      "epoch": 4.492682926829268,
+      "grad_norm": 1.0272135734558105,
+      "learning_rate": 2.9005433742301274e-06,
+      "loss": 0.4585,
+      "step": 921
+    },
+    {
+      "epoch": 4.4975609756097565,
+      "grad_norm": 1.7547694444656372,
+      "learning_rate": 2.8967611873152037e-06,
+      "loss": 0.3949,
+      "step": 922
+    },
+    {
+      "epoch": 4.5024390243902435,
+      "grad_norm": 1.0447901487350464,
+      "learning_rate": 2.892978068603683e-06,
+      "loss": 0.1928,
+      "step": 923
+    },
+    {
+      "epoch": 4.507317073170732,
+      "grad_norm": 1.0131272077560425,
+      "learning_rate": 2.889194026980249e-06,
+      "loss": 0.3458,
+      "step": 924
+    },
+    {
+      "epoch": 4.512195121951219,
+      "grad_norm": 1.1343997716903687,
+      "learning_rate": 2.8854090713317514e-06,
+      "loss": 0.1519,
+      "step": 925
+    },
+    {
+      "epoch": 4.517073170731708,
+      "grad_norm": 1.2432359457015991,
+      "learning_rate": 2.8816232105471864e-06,
+      "loss": 0.1789,
+      "step": 926
+    },
+    {
+      "epoch": 4.521951219512195,
+      "grad_norm": 1.2051594257354736,
+      "learning_rate": 2.877836453517677e-06,
+      "loss": 0.518,
+      "step": 927
+    },
+    {
+      "epoch": 4.526829268292683,
+      "grad_norm": 1.1518090963363647,
+      "learning_rate": 2.8740488091364492e-06,
+      "loss": 0.4737,
+      "step": 928
+    },
+    {
+      "epoch": 4.53170731707317,
+      "grad_norm": 1.456160068511963,
+      "learning_rate": 2.870260286298814e-06,
+      "loss": 0.6425,
+      "step": 929
+    },
+    {
+      "epoch": 4.536585365853659,
+      "grad_norm": 1.2971833944320679,
+      "learning_rate": 2.866470893902147e-06,
+      "loss": 0.3322,
+      "step": 930
+    },
+    {
+      "epoch": 4.541463414634146,
+      "grad_norm": 1.0898277759552002,
+      "learning_rate": 2.8626806408458626e-06,
+      "loss": 0.258,
+      "step": 931
+    },
+    {
+      "epoch": 4.546341463414635,
+      "grad_norm": 0.9179545044898987,
+      "learning_rate": 2.8588895360313983e-06,
+      "loss": 0.2085,
+      "step": 932
+    },
+    {
+      "epoch": 4.5512195121951216,
+      "grad_norm": 1.0029069185256958,
+      "learning_rate": 2.8550975883621935e-06,
+      "loss": 0.3558,
+      "step": 933
+    },
+    {
+      "epoch": 4.55609756097561,
+      "grad_norm": 1.0553942918777466,
+      "learning_rate": 2.8513048067436644e-06,
+      "loss": 0.3892,
+      "step": 934
+    },
+    {
+      "epoch": 4.560975609756097,
+      "grad_norm": 1.0313464403152466,
+      "learning_rate": 2.847511200083187e-06,
+      "loss": 0.2626,
+      "step": 935
+    },
+    {
+      "epoch": 4.565853658536585,
+      "grad_norm": 0.9714272618293762,
+      "learning_rate": 2.843716777290074e-06,
+      "loss": 0.2896,
+      "step": 936
+    },
+    {
+      "epoch": 4.570731707317073,
+      "grad_norm": 1.239687204360962,
+      "learning_rate": 2.839921547275556e-06,
+      "loss": 0.2688,
+      "step": 937
+    },
+    {
+      "epoch": 4.575609756097561,
+      "grad_norm": 1.451963186264038,
+      "learning_rate": 2.836125518952759e-06,
+      "loss": 0.2953,
+      "step": 938
+    },
+    {
+      "epoch": 4.580487804878048,
+      "grad_norm": 1.0113798379898071,
+      "learning_rate": 2.8323287012366845e-06,
+      "loss": 0.4247,
+      "step": 939
+    },
+    {
+      "epoch": 4.585365853658536,
+      "grad_norm": 1.1036227941513062,
+      "learning_rate": 2.828531103044186e-06,
+      "loss": 0.2388,
+      "step": 940
+    },
+    {
+      "epoch": 4.590243902439024,
+      "grad_norm": 1.0381038188934326,
+      "learning_rate": 2.8247327332939512e-06,
+      "loss": 0.3913,
+      "step": 941
+    },
+    {
+      "epoch": 4.595121951219512,
+      "grad_norm": 1.0053890943527222,
+      "learning_rate": 2.82093360090648e-06,
+      "loss": 0.4356,
+      "step": 942
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.2142903804779053,
+      "learning_rate": 2.8171337148040636e-06,
+      "loss": 0.4161,
+      "step": 943
+    },
+    {
+      "epoch": 4.6048780487804875,
+      "grad_norm": 1.8541486263275146,
+      "learning_rate": 2.813333083910761e-06,
+      "loss": 0.2803,
+      "step": 944
+    },
+    {
+      "epoch": 4.609756097560975,
+      "grad_norm": 0.9036626815795898,
+      "learning_rate": 2.8095317171523835e-06,
+      "loss": 0.1578,
+      "step": 945
+    },
+    {
+      "epoch": 4.614634146341463,
+      "grad_norm": 1.1875560283660889,
+      "learning_rate": 2.805729623456469e-06,
+      "loss": 0.4107,
+      "step": 946
+    },
+    {
+      "epoch": 4.619512195121951,
+      "grad_norm": 1.259789228439331,
+      "learning_rate": 2.8019268117522624e-06,
+      "loss": 0.5262,
+      "step": 947
+    },
+    {
+      "epoch": 4.624390243902439,
+      "grad_norm": 1.116606593132019,
+      "learning_rate": 2.798123290970695e-06,
+      "loss": 0.3334,
+      "step": 948
+    },
+    {
+      "epoch": 4.6292682926829265,
+      "grad_norm": 1.0191316604614258,
+      "learning_rate": 2.794319070044365e-06,
+      "loss": 0.395,
+      "step": 949
+    },
+    {
+      "epoch": 4.634146341463414,
+      "grad_norm": 1.1099485158920288,
+      "learning_rate": 2.790514157907512e-06,
+      "loss": 0.2684,
+      "step": 950
+    },
+    {
+      "epoch": 4.639024390243902,
+      "grad_norm": 0.9431113004684448,
+      "learning_rate": 2.786708563496002e-06,
+      "loss": 0.2696,
+      "step": 951
+    },
+    {
+      "epoch": 4.64390243902439,
+      "grad_norm": 1.19779372215271,
+      "learning_rate": 2.782902295747299e-06,
+      "loss": 0.5113,
+      "step": 952
+    },
+    {
+      "epoch": 4.648780487804878,
+      "grad_norm": 1.1742883920669556,
+      "learning_rate": 2.7790953636004536e-06,
+      "loss": 0.6055,
+      "step": 953
+    },
+    {
+      "epoch": 4.6536585365853655,
+      "grad_norm": 1.2542123794555664,
+      "learning_rate": 2.775287775996074e-06,
+      "loss": 0.5556,
+      "step": 954
+    },
+    {
+      "epoch": 4.658536585365853,
+      "grad_norm": 1.0845754146575928,
+      "learning_rate": 2.7714795418763067e-06,
+      "loss": 0.2445,
+      "step": 955
+    },
+    {
+      "epoch": 4.663414634146341,
+      "grad_norm": 1.3217185735702515,
+      "learning_rate": 2.7676706701848187e-06,
+      "loss": 0.4524,
+      "step": 956
+    },
+    {
+      "epoch": 4.668292682926829,
+      "grad_norm": 0.9640854597091675,
+      "learning_rate": 2.763861169866774e-06,
+      "loss": 0.4953,
+      "step": 957
+    },
+    {
+      "epoch": 4.673170731707317,
+      "grad_norm": 0.864771842956543,
+      "learning_rate": 2.7600510498688104e-06,
+      "loss": 0.1544,
+      "step": 958
+    },
+    {
+      "epoch": 4.678048780487805,
+      "grad_norm": 2.0469486713409424,
+      "learning_rate": 2.7562403191390246e-06,
+      "loss": 0.2998,
+      "step": 959
+    },
+    {
+      "epoch": 4.682926829268292,
+      "grad_norm": 1.0981053113937378,
+      "learning_rate": 2.7524289866269467e-06,
+      "loss": 0.3188,
+      "step": 960
+    },
+    {
+      "epoch": 4.68780487804878,
+      "grad_norm": 0.959457278251648,
+      "learning_rate": 2.748617061283518e-06,
+      "loss": 0.2297,
+      "step": 961
+    },
+    {
+      "epoch": 4.692682926829268,
+      "grad_norm": 1.2737903594970703,
+      "learning_rate": 2.744804552061074e-06,
+      "loss": 0.2563,
+      "step": 962
+    },
+    {
+      "epoch": 4.697560975609756,
+      "grad_norm": 1.0287935733795166,
+      "learning_rate": 2.740991467913321e-06,
+      "loss": 0.4164,
+      "step": 963
+    },
+    {
+      "epoch": 4.702439024390244,
+      "grad_norm": 1.0602225065231323,
+      "learning_rate": 2.737177817795315e-06,
+      "loss": 0.2201,
+      "step": 964
+    },
+    {
+      "epoch": 4.7073170731707314,
+      "grad_norm": 1.0084565877914429,
+      "learning_rate": 2.7333636106634414e-06,
+      "loss": 0.283,
+      "step": 965
+    },
+    {
+      "epoch": 4.712195121951219,
+      "grad_norm": 1.2561463117599487,
+      "learning_rate": 2.7295488554753957e-06,
+      "loss": 0.4592,
+      "step": 966
+    },
+    {
+      "epoch": 4.717073170731707,
+      "grad_norm": 1.0347182750701904,
+      "learning_rate": 2.725733561190157e-06,
+      "loss": 0.2642,
+      "step": 967
+    },
+    {
+      "epoch": 4.721951219512195,
+      "grad_norm": 1.312893033027649,
+      "learning_rate": 2.721917736767973e-06,
+      "loss": 0.335,
+      "step": 968
+    },
+    {
+      "epoch": 4.726829268292683,
+      "grad_norm": 1.0608656406402588,
+      "learning_rate": 2.7181013911703357e-06,
+      "loss": 0.2288,
+      "step": 969
+    },
+    {
+      "epoch": 4.7317073170731705,
+      "grad_norm": 1.0253140926361084,
+      "learning_rate": 2.714284533359961e-06,
+      "loss": 0.2827,
+      "step": 970
+    },
+    {
+      "epoch": 4.736585365853658,
+      "grad_norm": 0.9973790049552917,
+      "learning_rate": 2.710467172300768e-06,
+      "loss": 0.4102,
+      "step": 971
+    },
+    {
+      "epoch": 4.741463414634146,
+      "grad_norm": 1.2393323183059692,
+      "learning_rate": 2.706649316957857e-06,
+      "loss": 0.4489,
+      "step": 972
+    },
+    {
+      "epoch": 4.746341463414634,
+      "grad_norm": 1.1221544742584229,
+      "learning_rate": 2.7028309762974897e-06,
+      "loss": 0.1774,
+      "step": 973
+    },
+    {
+      "epoch": 4.751219512195122,
+      "grad_norm": 1.07940673828125,
+      "learning_rate": 2.699012159287069e-06,
+      "loss": 0.3329,
+      "step": 974
+    },
+    {
+      "epoch": 4.7560975609756095,
+      "grad_norm": 1.027277946472168,
+      "learning_rate": 2.6951928748951125e-06,
+      "loss": 0.2044,
+      "step": 975
+    },
+    {
+      "epoch": 4.760975609756097,
+      "grad_norm": 1.0281989574432373,
+      "learning_rate": 2.69137313209124e-06,
+      "loss": 0.3622,
+      "step": 976
+    },
+    {
+      "epoch": 4.765853658536585,
+      "grad_norm": 1.1298357248306274,
+      "learning_rate": 2.687552939846145e-06,
+      "loss": 0.4628,
+      "step": 977
+    },
+    {
+      "epoch": 4.770731707317073,
+      "grad_norm": 1.0114378929138184,
+      "learning_rate": 2.6837323071315766e-06,
+      "loss": 0.1181,
+      "step": 978
+    },
+    {
+      "epoch": 4.775609756097561,
+      "grad_norm": 1.000329613685608,
+      "learning_rate": 2.679911242920321e-06,
+      "loss": 0.2879,
+      "step": 979
+    },
+    {
+      "epoch": 4.780487804878049,
+      "grad_norm": 1.3905553817749023,
+      "learning_rate": 2.6760897561861742e-06,
+      "loss": 0.6606,
+      "step": 980
+    },
+    {
+      "epoch": 4.785365853658536,
+      "grad_norm": 0.9317846894264221,
+      "learning_rate": 2.672267855903927e-06,
+      "loss": 0.1451,
+      "step": 981
+    },
+    {
+      "epoch": 4.790243902439024,
+      "grad_norm": 0.9949944615364075,
+      "learning_rate": 2.6684455510493413e-06,
+      "loss": 0.3478,
+      "step": 982
+    },
+    {
+      "epoch": 4.795121951219512,
+      "grad_norm": 0.9357514381408691,
+      "learning_rate": 2.6646228505991267e-06,
+      "loss": 0.3766,
+      "step": 983
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.4344565868377686,
+      "learning_rate": 2.6607997635309246e-06,
+      "loss": 0.3122,
+      "step": 984
+    },
+    {
+      "epoch": 4.804878048780488,
+      "grad_norm": 1.2585278749465942,
+      "learning_rate": 2.6569762988232838e-06,
+      "loss": 0.356,
+      "step": 985
+    },
+    {
+      "epoch": 4.809756097560975,
+      "grad_norm": 1.0323916673660278,
+      "learning_rate": 2.653152465455639e-06,
+      "loss": 0.5199,
+      "step": 986
+    },
+    {
+      "epoch": 4.814634146341463,
+      "grad_norm": 1.1831841468811035,
+      "learning_rate": 2.6493282724082913e-06,
+      "loss": 0.5155,
+      "step": 987
+    },
+    {
+      "epoch": 4.819512195121951,
+      "grad_norm": 1.177829384803772,
+      "learning_rate": 2.6455037286623864e-06,
+      "loss": 0.263,
+      "step": 988
+    },
+    {
+      "epoch": 4.824390243902439,
+      "grad_norm": 0.7737200260162354,
+      "learning_rate": 2.6416788431998935e-06,
+      "loss": 0.2256,
+      "step": 989
+    },
+    {
+      "epoch": 4.829268292682927,
+      "grad_norm": 1.0419354438781738,
+      "learning_rate": 2.637853625003585e-06,
+      "loss": 0.4914,
+      "step": 990
+    },
+    {
+      "epoch": 4.8341463414634145,
+      "grad_norm": 1.000624418258667,
+      "learning_rate": 2.6340280830570142e-06,
+      "loss": 0.3594,
+      "step": 991
+    },
+    {
+      "epoch": 4.839024390243902,
+      "grad_norm": 1.0777456760406494,
+      "learning_rate": 2.6302022263444947e-06,
+      "loss": 0.263,
+      "step": 992
+    },
+    {
+      "epoch": 4.84390243902439,
+      "grad_norm": 1.333800196647644,
+      "learning_rate": 2.6263760638510793e-06,
+      "loss": 0.5697,
+      "step": 993
+    },
+    {
+      "epoch": 4.848780487804878,
+      "grad_norm": 1.0035558938980103,
+      "learning_rate": 2.6225496045625394e-06,
+      "loss": 0.2984,
+      "step": 994
+    },
+    {
+      "epoch": 4.853658536585366,
+      "grad_norm": 1.214937448501587,
+      "learning_rate": 2.6187228574653428e-06,
+      "loss": 0.3698,
+      "step": 995
+    },
+    {
+      "epoch": 4.8585365853658535,
+      "grad_norm": 0.9136359095573425,
+      "learning_rate": 2.614895831546633e-06,
+      "loss": 0.2602,
+      "step": 996
+    },
+    {
+      "epoch": 4.863414634146341,
+      "grad_norm": 1.2129359245300293,
+      "learning_rate": 2.6110685357942096e-06,
+      "loss": 0.4928,
+      "step": 997
+    },
+    {
+      "epoch": 4.868292682926829,
+      "grad_norm": 1.2663754224777222,
+      "learning_rate": 2.6072409791965048e-06,
+      "loss": 0.5422,
+      "step": 998
+    },
+    {
+      "epoch": 4.873170731707317,
+      "grad_norm": 1.4935377836227417,
+      "learning_rate": 2.6034131707425638e-06,
+      "loss": 0.7989,
+      "step": 999
+    },
+    {
+      "epoch": 4.878048780487805,
+      "grad_norm": 0.999937891960144,
+      "learning_rate": 2.5995851194220223e-06,
+      "loss": 0.3162,
+      "step": 1000
+    },
+    {
+      "epoch": 4.882926829268293,
+      "grad_norm": 1.0912432670593262,
+      "learning_rate": 2.595756834225089e-06,
+      "loss": 0.329,
+      "step": 1001
+    },
+    {
+      "epoch": 4.88780487804878,
+      "grad_norm": 1.195001244544983,
+      "learning_rate": 2.5919283241425188e-06,
+      "loss": 0.2323,
+      "step": 1002
+    },
+    {
+      "epoch": 4.892682926829268,
+      "grad_norm": 0.7830819487571716,
+      "learning_rate": 2.5880995981655965e-06,
+      "loss": 0.2372,
+      "step": 1003
+    },
+    {
+      "epoch": 4.897560975609756,
+      "grad_norm": 1.099342703819275,
+      "learning_rate": 2.584270665286113e-06,
+      "loss": 0.1749,
+      "step": 1004
+    },
+    {
+      "epoch": 4.902439024390244,
+      "grad_norm": 1.270918846130371,
+      "learning_rate": 2.580441534496346e-06,
+      "loss": 0.3421,
+      "step": 1005
+    },
+    {
+      "epoch": 4.907317073170732,
+      "grad_norm": 0.9069335460662842,
+      "learning_rate": 2.576612214789039e-06,
+      "loss": 0.2538,
+      "step": 1006
+    },
+    {
+      "epoch": 4.912195121951219,
+      "grad_norm": 1.377794623374939,
+      "learning_rate": 2.5727827151573747e-06,
+      "loss": 0.4034,
+      "step": 1007
+    },
+    {
+      "epoch": 4.917073170731707,
+      "grad_norm": 1.1331335306167603,
+      "learning_rate": 2.568953044594964e-06,
+      "loss": 0.2865,
+      "step": 1008
+    },
+    {
+      "epoch": 4.921951219512195,
+      "grad_norm": 1.0799779891967773,
+      "learning_rate": 2.5651232120958157e-06,
+      "loss": 0.3365,
+      "step": 1009
+    },
+    {
+      "epoch": 4.926829268292683,
+      "grad_norm": 1.1052212715148926,
+      "learning_rate": 2.56129322665432e-06,
+      "loss": 0.2303,
+      "step": 1010
+    },
+    {
+      "epoch": 4.931707317073171,
+      "grad_norm": 0.913105845451355,
+      "learning_rate": 2.5574630972652263e-06,
+      "loss": 0.1745,
+      "step": 1011
+    },
+    {
+      "epoch": 4.9365853658536585,
+      "grad_norm": 1.0110231637954712,
+      "learning_rate": 2.553632832923622e-06,
+      "loss": 0.3202,
+      "step": 1012
+    },
+    {
+      "epoch": 4.941463414634146,
+      "grad_norm": 1.1608706712722778,
+      "learning_rate": 2.5498024426249107e-06,
+      "loss": 0.5757,
+      "step": 1013
+    },
+    {
+      "epoch": 4.946341463414634,
+      "grad_norm": 0.9768334627151489,
+      "learning_rate": 2.545971935364794e-06,
+      "loss": 0.1663,
+      "step": 1014
+    },
+    {
+      "epoch": 4.951219512195122,
+      "grad_norm": 1.130802035331726,
+      "learning_rate": 2.5421413201392443e-06,
+      "loss": 0.3053,
+      "step": 1015
+    },
+    {
+      "epoch": 4.95609756097561,
+      "grad_norm": 1.5004934072494507,
+      "learning_rate": 2.538310605944491e-06,
+      "loss": 0.2943,
+      "step": 1016
+    },
+    {
+      "epoch": 4.9609756097560975,
+      "grad_norm": 0.9939762353897095,
+      "learning_rate": 2.534479801776996e-06,
+      "loss": 0.2933,
+      "step": 1017
+    },
+    {
+      "epoch": 4.965853658536585,
+      "grad_norm": 0.9881106019020081,
+      "learning_rate": 2.53064891663343e-06,
+      "loss": 0.3813,
+      "step": 1018
+    },
+    {
+      "epoch": 4.970731707317073,
+      "grad_norm": 0.9731859564781189,
+      "learning_rate": 2.526817959510655e-06,
+      "loss": 0.3543,
+      "step": 1019
+    },
+    {
+      "epoch": 4.975609756097561,
+      "grad_norm": 1.3951330184936523,
+      "learning_rate": 2.5229869394057038e-06,
+      "loss": 0.4254,
+      "step": 1020
+    },
+    {
+      "epoch": 4.980487804878049,
+      "grad_norm": 0.9658120274543762,
+      "learning_rate": 2.5191558653157542e-06,
+      "loss": 0.2484,
+      "step": 1021
+    },
+    {
+      "epoch": 4.985365853658537,
+      "grad_norm": 0.8997960090637207,
+      "learning_rate": 2.515324746238113e-06,
+      "loss": 0.1693,
+      "step": 1022
+    },
+    {
+      "epoch": 4.990243902439024,
+      "grad_norm": 0.9819786548614502,
+      "learning_rate": 2.511493591170191e-06,
+      "loss": 0.3336,
+      "step": 1023
+    },
+    {
+      "epoch": 4.995121951219512,
+      "grad_norm": 1.3949546813964844,
+      "learning_rate": 2.5076624091094846e-06,
+      "loss": 0.2995,
+      "step": 1024
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.9916773438453674,
+      "learning_rate": 2.503831209053554e-06,
+      "loss": 0.3243,
+      "step": 1025
+    },
+    {
+      "epoch": 5.004878048780488,
+      "grad_norm": 0.9542043805122375,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1981,
+      "step": 1026
+    },
+    {
+      "epoch": 5.009756097560976,
+      "grad_norm": 1.2487229108810425,
+      "learning_rate": 2.4961687909464462e-06,
+      "loss": 0.3443,
+      "step": 1027
+    },
+    {
+      "epoch": 5.014634146341463,
+      "grad_norm": 1.5230785608291626,
+      "learning_rate": 2.492337590890516e-06,
+      "loss": 0.5778,
+      "step": 1028
+    },
+    {
+      "epoch": 5.019512195121951,
+      "grad_norm": 1.013128638267517,
+      "learning_rate": 2.4885064088298097e-06,
+      "loss": 0.4135,
+      "step": 1029
+    },
+    {
+      "epoch": 5.024390243902439,
+      "grad_norm": 0.944968044757843,
+      "learning_rate": 2.4846752537618875e-06,
+      "loss": 0.1586,
+      "step": 1030
+    },
+    {
+      "epoch": 5.029268292682927,
+      "grad_norm": 1.3642313480377197,
+      "learning_rate": 2.480844134684246e-06,
+      "loss": 0.407,
+      "step": 1031
+    },
+    {
+      "epoch": 5.034146341463415,
+      "grad_norm": 1.2477118968963623,
+      "learning_rate": 2.4770130605942966e-06,
+      "loss": 0.2376,
+      "step": 1032
+    },
+    {
+      "epoch": 5.0390243902439025,
+      "grad_norm": 1.0908373594284058,
+      "learning_rate": 2.4731820404893457e-06,
+      "loss": 0.2063,
+      "step": 1033
+    },
+    {
+      "epoch": 5.04390243902439,
+      "grad_norm": 1.423384666442871,
+      "learning_rate": 2.469351083366571e-06,
+      "loss": 0.255,
+      "step": 1034
+    },
+    {
+      "epoch": 5.048780487804878,
+      "grad_norm": 1.0463767051696777,
+      "learning_rate": 2.4655201982230044e-06,
+      "loss": 0.1164,
+      "step": 1035
+    },
+    {
+      "epoch": 5.053658536585366,
+      "grad_norm": 1.1025980710983276,
+      "learning_rate": 2.4616893940555094e-06,
+      "loss": 0.1895,
+      "step": 1036
+    },
+    {
+      "epoch": 5.058536585365854,
+      "grad_norm": 1.50823175907135,
+      "learning_rate": 2.457858679860757e-06,
+      "loss": 0.6323,
+      "step": 1037
+    },
+    {
+      "epoch": 5.0634146341463415,
+      "grad_norm": 1.2428679466247559,
+      "learning_rate": 2.4540280646352072e-06,
+      "loss": 0.303,
+      "step": 1038
+    },
+    {
+      "epoch": 5.068292682926829,
+      "grad_norm": 1.2156243324279785,
+      "learning_rate": 2.45019755737509e-06,
+      "loss": 0.2281,
+      "step": 1039
+    },
+    {
+      "epoch": 5.073170731707317,
+      "grad_norm": 1.9944231510162354,
+      "learning_rate": 2.4463671670763787e-06,
+      "loss": 0.3072,
+      "step": 1040
+    },
+    {
+      "epoch": 5.078048780487805,
+      "grad_norm": 1.210920810699463,
+      "learning_rate": 2.4425369027347746e-06,
+      "loss": 0.3228,
+      "step": 1041
+    },
+    {
+      "epoch": 5.082926829268293,
+      "grad_norm": 0.9305552244186401,
+      "learning_rate": 2.4387067733456804e-06,
+      "loss": 0.2808,
+      "step": 1042
+    },
+    {
+      "epoch": 5.087804878048781,
+      "grad_norm": 1.200379490852356,
+      "learning_rate": 2.4348767879041847e-06,
+      "loss": 0.2998,
+      "step": 1043
+    },
+    {
+      "epoch": 5.092682926829268,
+      "grad_norm": 1.2235733270645142,
+      "learning_rate": 2.4310469554050366e-06,
+      "loss": 0.2274,
+      "step": 1044
+    },
+    {
+      "epoch": 5.097560975609756,
+      "grad_norm": 0.9284917116165161,
+      "learning_rate": 2.4272172848426257e-06,
+      "loss": 0.3156,
+      "step": 1045
+    },
+    {
+      "epoch": 5.102439024390244,
+      "grad_norm": 1.1812609434127808,
+      "learning_rate": 2.423387785210962e-06,
+      "loss": 0.2529,
+      "step": 1046
+    },
+    {
+      "epoch": 5.107317073170732,
+      "grad_norm": 1.2920304536819458,
+      "learning_rate": 2.4195584655036544e-06,
+      "loss": 0.4466,
+      "step": 1047
+    },
+    {
+      "epoch": 5.11219512195122,
+      "grad_norm": 1.0265544652938843,
+      "learning_rate": 2.4157293347138877e-06,
+      "loss": 0.1946,
+      "step": 1048
+    },
+    {
+      "epoch": 5.117073170731707,
+      "grad_norm": 2.3565609455108643,
+      "learning_rate": 2.4119004018344043e-06,
+      "loss": 0.3177,
+      "step": 1049
+    },
+    {
+      "epoch": 5.121951219512195,
+      "grad_norm": 1.1925452947616577,
+      "learning_rate": 2.408071675857482e-06,
+      "loss": 0.2563,
+      "step": 1050
+    },
+    {
+      "epoch": 5.126829268292683,
+      "grad_norm": 1.3194458484649658,
+      "learning_rate": 2.404243165774912e-06,
+      "loss": 0.4448,
+      "step": 1051
+    },
+    {
+      "epoch": 5.131707317073171,
+      "grad_norm": 0.8365583419799805,
+      "learning_rate": 2.4004148805779785e-06,
+      "loss": 0.1032,
+      "step": 1052
+    },
+    {
+      "epoch": 5.136585365853659,
+      "grad_norm": 1.257830262184143,
+      "learning_rate": 2.3965868292574375e-06,
+      "loss": 0.2855,
+      "step": 1053
+    },
+    {
+      "epoch": 5.1414634146341465,
+      "grad_norm": 1.3698722124099731,
+      "learning_rate": 2.392759020803496e-06,
+      "loss": 0.2996,
+      "step": 1054
+    },
+    {
+      "epoch": 5.146341463414634,
+      "grad_norm": 1.1302275657653809,
+      "learning_rate": 2.3889314642057916e-06,
+      "loss": 0.2131,
+      "step": 1055
+    },
+    {
+      "epoch": 5.151219512195122,
+      "grad_norm": 1.1752660274505615,
+      "learning_rate": 2.3851041684533677e-06,
+      "loss": 0.4146,
+      "step": 1056
+    },
+    {
+      "epoch": 5.15609756097561,
+      "grad_norm": 0.8562495112419128,
+      "learning_rate": 2.381277142534658e-06,
+      "loss": 0.1059,
+      "step": 1057
+    },
+    {
+      "epoch": 5.160975609756098,
+      "grad_norm": 1.279471516609192,
+      "learning_rate": 2.3774503954374614e-06,
+      "loss": 0.1535,
+      "step": 1058
+    },
+    {
+      "epoch": 5.1658536585365855,
+      "grad_norm": 1.7923734188079834,
+      "learning_rate": 2.373623936148921e-06,
+      "loss": 0.3611,
+      "step": 1059
+    },
+    {
+      "epoch": 5.170731707317073,
+      "grad_norm": 1.5222898721694946,
+      "learning_rate": 2.369797773655506e-06,
+      "loss": 0.3535,
+      "step": 1060
+    },
+    {
+      "epoch": 5.175609756097561,
+      "grad_norm": 1.0636627674102783,
+      "learning_rate": 2.3659719169429866e-06,
+      "loss": 0.2679,
+      "step": 1061
+    },
+    {
+      "epoch": 5.180487804878049,
+      "grad_norm": 1.0380195379257202,
+      "learning_rate": 2.3621463749964153e-06,
+      "loss": 0.1407,
+      "step": 1062
+    },
+    {
+      "epoch": 5.185365853658537,
+      "grad_norm": 0.9535474181175232,
+      "learning_rate": 2.3583211568001073e-06,
+      "loss": 0.1263,
+      "step": 1063
+    },
+    {
+      "epoch": 5.190243902439025,
+      "grad_norm": 1.0084075927734375,
+      "learning_rate": 2.3544962713376144e-06,
+      "loss": 0.1556,
+      "step": 1064
+    },
+    {
+      "epoch": 5.195121951219512,
+      "grad_norm": 0.9850101470947266,
+      "learning_rate": 2.3506717275917095e-06,
+      "loss": 0.1819,
+      "step": 1065
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.9447112083435059,
+      "learning_rate": 2.346847534544362e-06,
+      "loss": 0.1648,
+      "step": 1066
+    },
+    {
+      "epoch": 5.204878048780488,
+      "grad_norm": 1.0958189964294434,
+      "learning_rate": 2.3430237011767166e-06,
+      "loss": 0.2921,
+      "step": 1067
+    },
+    {
+      "epoch": 5.209756097560976,
+      "grad_norm": 0.9113654494285583,
+      "learning_rate": 2.3392002364690762e-06,
+      "loss": 0.1123,
+      "step": 1068
+    },
+    {
+      "epoch": 5.214634146341464,
+      "grad_norm": 1.5014578104019165,
+      "learning_rate": 2.335377149400874e-06,
+      "loss": 0.2756,
+      "step": 1069
+    },
+    {
+      "epoch": 5.219512195121951,
+      "grad_norm": 1.297300100326538,
+      "learning_rate": 2.3315544489506596e-06,
+      "loss": 0.3623,
+      "step": 1070
+    },
+    {
+      "epoch": 5.224390243902439,
+      "grad_norm": 1.3639144897460938,
+      "learning_rate": 2.3277321440960733e-06,
+      "loss": 0.2713,
+      "step": 1071
+    },
+    {
+      "epoch": 5.229268292682927,
+      "grad_norm": 0.932643473148346,
+      "learning_rate": 2.323910243813826e-06,
+      "loss": 0.1127,
+      "step": 1072
+    },
+    {
+      "epoch": 5.234146341463415,
+      "grad_norm": 1.1686640977859497,
+      "learning_rate": 2.3200887570796798e-06,
+      "loss": 0.4201,
+      "step": 1073
+    },
+    {
+      "epoch": 5.239024390243903,
+      "grad_norm": 1.193697214126587,
+      "learning_rate": 2.316267692868424e-06,
+      "loss": 0.2283,
+      "step": 1074
+    },
+    {
+      "epoch": 5.2439024390243905,
+      "grad_norm": 1.6448438167572021,
+      "learning_rate": 2.312447060153856e-06,
+      "loss": 0.2204,
+      "step": 1075
+    },
+    {
+      "epoch": 5.248780487804878,
+      "grad_norm": 1.203925371170044,
+      "learning_rate": 2.308626867908761e-06,
+      "loss": 0.2119,
+      "step": 1076
+    },
+    {
+      "epoch": 5.253658536585366,
+      "grad_norm": 1.2128314971923828,
+      "learning_rate": 2.3048071251048884e-06,
+      "loss": 0.3189,
+      "step": 1077
+    },
+    {
+      "epoch": 5.258536585365854,
+      "grad_norm": 1.2930793762207031,
+      "learning_rate": 2.300987840712932e-06,
+      "loss": 0.3482,
+      "step": 1078
+    },
+    {
+      "epoch": 5.263414634146342,
+      "grad_norm": 1.1673916578292847,
+      "learning_rate": 2.297169023702511e-06,
+      "loss": 0.4532,
+      "step": 1079
+    },
+    {
+      "epoch": 5.2682926829268295,
+      "grad_norm": 1.0989543199539185,
+      "learning_rate": 2.2933506830421436e-06,
+      "loss": 0.3365,
+      "step": 1080
+    },
+    {
+      "epoch": 5.273170731707317,
+      "grad_norm": 1.0421515703201294,
+      "learning_rate": 2.2895328276992325e-06,
+      "loss": 0.1207,
+      "step": 1081
+    },
+    {
+      "epoch": 5.278048780487805,
+      "grad_norm": 1.3659073114395142,
+      "learning_rate": 2.28571546664004e-06,
+      "loss": 0.3638,
+      "step": 1082
+    },
+    {
+      "epoch": 5.282926829268293,
+      "grad_norm": 0.975866436958313,
+      "learning_rate": 2.281898608829665e-06,
+      "loss": 0.1151,
+      "step": 1083
+    },
+    {
+      "epoch": 5.287804878048781,
+      "grad_norm": 0.9325199127197266,
+      "learning_rate": 2.2780822632320273e-06,
+      "loss": 0.1909,
+      "step": 1084
+    },
+    {
+      "epoch": 5.2926829268292686,
+      "grad_norm": 1.0251420736312866,
+      "learning_rate": 2.2742664388098435e-06,
+      "loss": 0.156,
+      "step": 1085
+    },
+    {
+      "epoch": 5.297560975609756,
+      "grad_norm": 0.8853452205657959,
+      "learning_rate": 2.270451144524605e-06,
+      "loss": 0.2,
+      "step": 1086
+    },
+    {
+      "epoch": 5.302439024390244,
+      "grad_norm": 1.1478848457336426,
+      "learning_rate": 2.266636389336559e-06,
+      "loss": 0.2425,
+      "step": 1087
+    },
+    {
+      "epoch": 5.307317073170732,
+      "grad_norm": 0.9325762987136841,
+      "learning_rate": 2.262822182204686e-06,
+      "loss": 0.1064,
+      "step": 1088
+    },
+    {
+      "epoch": 5.31219512195122,
+      "grad_norm": 1.0397100448608398,
+      "learning_rate": 2.2590085320866798e-06,
+      "loss": 0.1143,
+      "step": 1089
+    },
+    {
+      "epoch": 5.317073170731708,
+      "grad_norm": 0.9676294922828674,
+      "learning_rate": 2.255195447938927e-06,
+      "loss": 0.1028,
+      "step": 1090
+    },
+    {
+      "epoch": 5.321951219512195,
+      "grad_norm": 1.4252179861068726,
+      "learning_rate": 2.251382938716482e-06,
+      "loss": 0.2209,
+      "step": 1091
+    },
+    {
+      "epoch": 5.326829268292683,
+      "grad_norm": 1.0721783638000488,
+      "learning_rate": 2.2475710133730533e-06,
+      "loss": 0.1538,
+      "step": 1092
+    },
+    {
+      "epoch": 5.331707317073171,
+      "grad_norm": 0.9011938571929932,
+      "learning_rate": 2.243759680860975e-06,
+      "loss": 0.1627,
+      "step": 1093
+    },
+    {
+      "epoch": 5.336585365853659,
+      "grad_norm": 1.1225794553756714,
+      "learning_rate": 2.2399489501311896e-06,
+      "loss": 0.2376,
+      "step": 1094
+    },
+    {
+      "epoch": 5.341463414634147,
+      "grad_norm": 1.106941819190979,
+      "learning_rate": 2.2361388301332265e-06,
+      "loss": 0.4092,
+      "step": 1095
+    },
+    {
+      "epoch": 5.3463414634146345,
+      "grad_norm": 0.7896466255187988,
+      "learning_rate": 2.2323293298151817e-06,
+      "loss": 0.1019,
+      "step": 1096
+    },
+    {
+      "epoch": 5.351219512195122,
+      "grad_norm": 1.0374608039855957,
+      "learning_rate": 2.2285204581236937e-06,
+      "loss": 0.1808,
+      "step": 1097
+    },
+    {
+      "epoch": 5.35609756097561,
+      "grad_norm": 1.07212495803833,
+      "learning_rate": 2.2247122240039268e-06,
+      "loss": 0.1773,
+      "step": 1098
+    },
+    {
+      "epoch": 5.360975609756098,
+      "grad_norm": 0.9184301495552063,
+      "learning_rate": 2.2209046363995464e-06,
+      "loss": 0.0608,
+      "step": 1099
+    },
+    {
+      "epoch": 5.365853658536586,
+      "grad_norm": 1.1575993299484253,
+      "learning_rate": 2.217097704252701e-06,
+      "loss": 0.289,
+      "step": 1100
+    },
+    {
+      "epoch": 5.3707317073170735,
+      "grad_norm": 0.869531512260437,
+      "learning_rate": 2.2132914365039993e-06,
+      "loss": 0.1568,
+      "step": 1101
+    },
+    {
+      "epoch": 5.375609756097561,
+      "grad_norm": 1.1020842790603638,
+      "learning_rate": 2.2094858420924882e-06,
+      "loss": 0.0779,
+      "step": 1102
+    },
+    {
+      "epoch": 5.380487804878049,
+      "grad_norm": 1.4796274900436401,
+      "learning_rate": 2.205680929955635e-06,
+      "loss": 0.3573,
+      "step": 1103
+    },
+    {
+      "epoch": 5.385365853658537,
+      "grad_norm": 1.0098387002944946,
+      "learning_rate": 2.201876709029305e-06,
+      "loss": 0.2391,
+      "step": 1104
+    },
+    {
+      "epoch": 5.390243902439025,
+      "grad_norm": 1.0524678230285645,
+      "learning_rate": 2.198073188247738e-06,
+      "loss": 0.1946,
+      "step": 1105
+    },
+    {
+      "epoch": 5.3951219512195125,
+      "grad_norm": 0.8759175539016724,
+      "learning_rate": 2.1942703765435317e-06,
+      "loss": 0.0508,
+      "step": 1106
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.378225564956665,
+      "learning_rate": 2.190468282847617e-06,
+      "loss": 0.4062,
+      "step": 1107
+    },
+    {
+      "epoch": 5.404878048780488,
+      "grad_norm": 1.21510910987854,
+      "learning_rate": 2.186666916089239e-06,
+      "loss": 0.539,
+      "step": 1108
+    },
+    {
+      "epoch": 5.409756097560976,
+      "grad_norm": 1.0114099979400635,
+      "learning_rate": 2.1828662851959377e-06,
+      "loss": 0.1577,
+      "step": 1109
+    },
+    {
+      "epoch": 5.414634146341464,
+      "grad_norm": 1.0858491659164429,
+      "learning_rate": 2.1790663990935203e-06,
+      "loss": 0.2497,
+      "step": 1110
+    },
+    {
+      "epoch": 5.419512195121952,
+      "grad_norm": 1.2288187742233276,
+      "learning_rate": 2.1752672667060488e-06,
+      "loss": 0.1893,
+      "step": 1111
+    },
+    {
+      "epoch": 5.424390243902439,
+      "grad_norm": 1.244839072227478,
+      "learning_rate": 2.1714688969558146e-06,
+      "loss": 0.1314,
+      "step": 1112
+    },
+    {
+      "epoch": 5.429268292682927,
+      "grad_norm": 1.169716715812683,
+      "learning_rate": 2.167671298763316e-06,
+      "loss": 0.4354,
+      "step": 1113
+    },
+    {
+      "epoch": 5.434146341463415,
+      "grad_norm": 1.0661805868148804,
+      "learning_rate": 2.1638744810472414e-06,
+      "loss": 0.3079,
+      "step": 1114
+    },
+    {
+      "epoch": 5.439024390243903,
+      "grad_norm": 0.9970805644989014,
+      "learning_rate": 2.1600784527244445e-06,
+      "loss": 0.1573,
+      "step": 1115
+    },
+    {
+      "epoch": 5.443902439024391,
+      "grad_norm": 1.458986759185791,
+      "learning_rate": 2.1562832227099266e-06,
+      "loss": 0.5079,
+      "step": 1116
+    },
+    {
+      "epoch": 5.4487804878048784,
+      "grad_norm": 1.2743233442306519,
+      "learning_rate": 2.152488799916814e-06,
+      "loss": 0.4,
+      "step": 1117
+    },
+    {
+      "epoch": 5.453658536585366,
+      "grad_norm": 1.822080135345459,
+      "learning_rate": 2.148695193256336e-06,
+      "loss": 0.5168,
+      "step": 1118
+    },
+    {
+      "epoch": 5.458536585365854,
+      "grad_norm": 0.9681462645530701,
+      "learning_rate": 2.1449024116378064e-06,
+      "loss": 0.2588,
+      "step": 1119
+    },
+    {
+      "epoch": 5.463414634146342,
+      "grad_norm": 1.0483061075210571,
+      "learning_rate": 2.1411104639686013e-06,
+      "loss": 0.1395,
+      "step": 1120
+    },
+    {
+      "epoch": 5.46829268292683,
+      "grad_norm": 1.3658936023712158,
+      "learning_rate": 2.137319359154138e-06,
+      "loss": 0.2823,
+      "step": 1121
+    },
+    {
+      "epoch": 5.473170731707317,
+      "grad_norm": 0.9887218475341797,
+      "learning_rate": 2.133529106097853e-06,
+      "loss": 0.1195,
+      "step": 1122
+    },
+    {
+      "epoch": 5.478048780487805,
+      "grad_norm": 1.1662182807922363,
+      "learning_rate": 2.1297397137011862e-06,
+      "loss": 0.2621,
+      "step": 1123
+    },
+    {
+      "epoch": 5.482926829268292,
+      "grad_norm": 1.2363375425338745,
+      "learning_rate": 2.125951190863551e-06,
+      "loss": 0.2179,
+      "step": 1124
+    },
+    {
+      "epoch": 5.487804878048781,
+      "grad_norm": 0.7923662662506104,
+      "learning_rate": 2.1221635464823237e-06,
+      "loss": 0.1064,
+      "step": 1125
+    },
+    {
+      "epoch": 5.492682926829268,
+      "grad_norm": 1.751186728477478,
+      "learning_rate": 2.1183767894528135e-06,
+      "loss": 0.5021,
+      "step": 1126
+    },
+    {
+      "epoch": 5.4975609756097565,
+      "grad_norm": 0.9320492148399353,
+      "learning_rate": 2.114590928668249e-06,
+      "loss": 0.1303,
+      "step": 1127
+    },
+    {
+      "epoch": 5.5024390243902435,
+      "grad_norm": 0.9334157109260559,
+      "learning_rate": 2.1108059730197517e-06,
+      "loss": 0.1716,
+      "step": 1128
+    },
+    {
+      "epoch": 5.507317073170732,
+      "grad_norm": 1.0579321384429932,
+      "learning_rate": 2.1070219313963173e-06,
+      "loss": 0.1431,
+      "step": 1129
+    },
+    {
+      "epoch": 5.512195121951219,
+      "grad_norm": 1.2380542755126953,
+      "learning_rate": 2.1032388126847967e-06,
+      "loss": 0.1667,
+      "step": 1130
+    },
+    {
+      "epoch": 5.517073170731708,
+      "grad_norm": 0.8622035384178162,
+      "learning_rate": 2.099456625769872e-06,
+      "loss": 0.0722,
+      "step": 1131
+    },
+    {
+      "epoch": 5.521951219512195,
+      "grad_norm": 1.125844955444336,
+      "learning_rate": 2.0956753795340376e-06,
+      "loss": 0.1991,
+      "step": 1132
+    },
+    {
+      "epoch": 5.526829268292683,
+      "grad_norm": 1.1052132844924927,
+      "learning_rate": 2.091895082857578e-06,
+      "loss": 0.357,
+      "step": 1133
+    },
+    {
+      "epoch": 5.53170731707317,
+      "grad_norm": 1.0329958200454712,
+      "learning_rate": 2.0881157446185474e-06,
+      "loss": 0.172,
+      "step": 1134
+    },
+    {
+      "epoch": 5.536585365853659,
+      "grad_norm": 1.0203710794448853,
+      "learning_rate": 2.0843373736927506e-06,
+      "loss": 0.1389,
+      "step": 1135
+    },
+    {
+      "epoch": 5.541463414634146,
+      "grad_norm": 0.7079288959503174,
+      "learning_rate": 2.08055997895372e-06,
+      "loss": 0.0853,
+      "step": 1136
+    },
+    {
+      "epoch": 5.546341463414635,
+      "grad_norm": 1.1098510026931763,
+      "learning_rate": 2.0767835692726944e-06,
+      "loss": 0.1492,
+      "step": 1137
+    },
+    {
+      "epoch": 5.5512195121951216,
+      "grad_norm": 1.1874253749847412,
+      "learning_rate": 2.0730081535186e-06,
+      "loss": 0.457,
+      "step": 1138
+    },
+    {
+      "epoch": 5.55609756097561,
+      "grad_norm": 1.0629135370254517,
+      "learning_rate": 2.06923374055803e-06,
+      "loss": 0.2759,
+      "step": 1139
+    },
+    {
+      "epoch": 5.560975609756097,
+      "grad_norm": 0.9080162048339844,
+      "learning_rate": 2.0654603392552193e-06,
+      "loss": 0.0708,
+      "step": 1140
+    },
+    {
+      "epoch": 5.565853658536585,
+      "grad_norm": 1.2526748180389404,
+      "learning_rate": 2.0616879584720305e-06,
+      "loss": 0.2727,
+      "step": 1141
+    },
+    {
+      "epoch": 5.570731707317073,
+      "grad_norm": 0.9930171966552734,
+      "learning_rate": 2.057916607067928e-06,
+      "loss": 0.1473,
+      "step": 1142
+    },
+    {
+      "epoch": 5.575609756097561,
+      "grad_norm": 0.9729718565940857,
+      "learning_rate": 2.054146293899957e-06,
+      "loss": 0.1325,
+      "step": 1143
+    },
+    {
+      "epoch": 5.580487804878048,
+      "grad_norm": 0.9204704761505127,
+      "learning_rate": 2.0503770278227274e-06,
+      "loss": 0.1246,
+      "step": 1144
+    },
+    {
+      "epoch": 5.585365853658536,
+      "grad_norm": 1.0536152124404907,
+      "learning_rate": 2.0466088176883876e-06,
+      "loss": 0.1,
+      "step": 1145
+    },
+    {
+      "epoch": 5.590243902439024,
+      "grad_norm": 1.211815595626831,
+      "learning_rate": 2.042841672346608e-06,
+      "loss": 0.2188,
+      "step": 1146
+    },
+    {
+      "epoch": 5.595121951219512,
+      "grad_norm": 1.4135674238204956,
+      "learning_rate": 2.039075600644557e-06,
+      "loss": 0.1599,
+      "step": 1147
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.0430986881256104,
+      "learning_rate": 2.0353106114268824e-06,
+      "loss": 0.1691,
+      "step": 1148
+    },
+    {
+      "epoch": 5.6048780487804875,
+      "grad_norm": 0.8738663196563721,
+      "learning_rate": 2.031546713535688e-06,
+      "loss": 0.1175,
+      "step": 1149
+    },
+    {
+      "epoch": 5.609756097560975,
+      "grad_norm": 1.015977144241333,
+      "learning_rate": 2.027783915810518e-06,
+      "loss": 0.1638,
+      "step": 1150
+    },
+    {
+      "epoch": 5.614634146341463,
+      "grad_norm": 1.184048056602478,
+      "learning_rate": 2.024022227088329e-06,
+      "loss": 0.4066,
+      "step": 1151
+    },
+    {
+      "epoch": 5.619512195121951,
+      "grad_norm": 1.1206215620040894,
+      "learning_rate": 2.020261656203476e-06,
+      "loss": 0.4061,
+      "step": 1152
+    },
+    {
+      "epoch": 5.624390243902439,
+      "grad_norm": 1.150066614151001,
+      "learning_rate": 2.016502211987687e-06,
+      "loss": 0.2951,
+      "step": 1153
+    },
+    {
+      "epoch": 5.6292682926829265,
+      "grad_norm": 1.465074896812439,
+      "learning_rate": 2.0127439032700446e-06,
+      "loss": 0.2386,
+      "step": 1154
+    },
+    {
+      "epoch": 5.634146341463414,
+      "grad_norm": 1.1083053350448608,
+      "learning_rate": 2.0089867388769664e-06,
+      "loss": 0.2346,
+      "step": 1155
+    },
+    {
+      "epoch": 5.639024390243902,
+      "grad_norm": 0.8865154981613159,
+      "learning_rate": 2.0052307276321793e-06,
+      "loss": 0.1226,
+      "step": 1156
+    },
+    {
+      "epoch": 5.64390243902439,
+      "grad_norm": 1.0869574546813965,
+      "learning_rate": 2.001475878356703e-06,
+      "loss": 0.2506,
+      "step": 1157
+    },
+    {
+      "epoch": 5.648780487804878,
+      "grad_norm": 1.066562533378601,
+      "learning_rate": 1.99772219986883e-06,
+      "loss": 0.3113,
+      "step": 1158
+    },
+    {
+      "epoch": 5.6536585365853655,
+      "grad_norm": 1.489364504814148,
+      "learning_rate": 1.9939697009841024e-06,
+      "loss": 0.1878,
+      "step": 1159
+    },
+    {
+      "epoch": 5.658536585365853,
+      "grad_norm": 1.1426478624343872,
+      "learning_rate": 1.990218390515291e-06,
+      "loss": 0.2903,
+      "step": 1160
+    },
+    {
+      "epoch": 5.663414634146341,
+      "grad_norm": 1.1296617984771729,
+      "learning_rate": 1.9864682772723757e-06,
+      "loss": 0.2505,
+      "step": 1161
+    },
+    {
+      "epoch": 5.668292682926829,
+      "grad_norm": 0.7900092005729675,
+      "learning_rate": 1.9827193700625274e-06,
+      "loss": 0.1204,
+      "step": 1162
+    },
+    {
+      "epoch": 5.673170731707317,
+      "grad_norm": 1.2168926000595093,
+      "learning_rate": 1.978971677690081e-06,
+      "loss": 0.4264,
+      "step": 1163
+    },
+    {
+      "epoch": 5.678048780487805,
+      "grad_norm": 1.0507045984268188,
+      "learning_rate": 1.97522520895652e-06,
+      "loss": 0.0916,
+      "step": 1164
+    },
+    {
+      "epoch": 5.682926829268292,
+      "grad_norm": 1.5506707429885864,
+      "learning_rate": 1.971479972660454e-06,
+      "loss": 0.2858,
+      "step": 1165
+    },
+    {
+      "epoch": 5.68780487804878,
+      "grad_norm": 0.9136821031570435,
+      "learning_rate": 1.967735977597598e-06,
+      "loss": 0.0917,
+      "step": 1166
+    },
+    {
+      "epoch": 5.692682926829268,
+      "grad_norm": 0.7896243929862976,
+      "learning_rate": 1.9639932325607538e-06,
+      "loss": 0.1089,
+      "step": 1167
+    },
+    {
+      "epoch": 5.697560975609756,
+      "grad_norm": 1.396902084350586,
+      "learning_rate": 1.9602517463397845e-06,
+      "loss": 0.1683,
+      "step": 1168
+    },
+    {
+      "epoch": 5.702439024390244,
+      "grad_norm": 1.012985348701477,
+      "learning_rate": 1.9565115277215978e-06,
+      "loss": 0.2019,
+      "step": 1169
+    },
+    {
+      "epoch": 5.7073170731707314,
+      "grad_norm": 0.9816607236862183,
+      "learning_rate": 1.952772585490127e-06,
+      "loss": 0.163,
+      "step": 1170
+    },
+    {
+      "epoch": 5.712195121951219,
+      "grad_norm": 0.9821074604988098,
+      "learning_rate": 1.9490349284263036e-06,
+      "loss": 0.0972,
+      "step": 1171
+    },
+    {
+      "epoch": 5.717073170731707,
+      "grad_norm": 0.9475919604301453,
+      "learning_rate": 1.9452985653080443e-06,
+      "loss": 0.1485,
+      "step": 1172
+    },
+    {
+      "epoch": 5.721951219512195,
+      "grad_norm": 0.921571671962738,
+      "learning_rate": 1.9415635049102245e-06,
+      "loss": 0.1658,
+      "step": 1173
+    },
+    {
+      "epoch": 5.726829268292683,
+      "grad_norm": 1.297415852546692,
+      "learning_rate": 1.937829756004662e-06,
+      "loss": 0.4056,
+      "step": 1174
+    },
+    {
+      "epoch": 5.7317073170731705,
+      "grad_norm": 1.2068060636520386,
+      "learning_rate": 1.9340973273600944e-06,
+      "loss": 0.228,
+      "step": 1175
+    },
+    {
+      "epoch": 5.736585365853658,
+      "grad_norm": 1.0924711227416992,
+      "learning_rate": 1.930366227742157e-06,
+      "loss": 0.3379,
+      "step": 1176
+    },
+    {
+      "epoch": 5.741463414634146,
+      "grad_norm": 1.2374088764190674,
+      "learning_rate": 1.9266364659133653e-06,
+      "loss": 0.2372,
+      "step": 1177
+    },
+    {
+      "epoch": 5.746341463414634,
+      "grad_norm": 0.913871705532074,
+      "learning_rate": 1.922908050633093e-06,
+      "loss": 0.113,
+      "step": 1178
+    },
+    {
+      "epoch": 5.751219512195122,
+      "grad_norm": 1.136725902557373,
+      "learning_rate": 1.919180990657551e-06,
+      "loss": 0.2227,
+      "step": 1179
+    },
+    {
+      "epoch": 5.7560975609756095,
+      "grad_norm": 0.9916288256645203,
+      "learning_rate": 1.9154552947397668e-06,
+      "loss": 0.1983,
+      "step": 1180
+    },
+    {
+      "epoch": 5.760975609756097,
+      "grad_norm": 1.3981820344924927,
+      "learning_rate": 1.9117309716295658e-06,
+      "loss": 0.3343,
+      "step": 1181
+    },
+    {
+      "epoch": 5.765853658536585,
+      "grad_norm": 1.008737325668335,
+      "learning_rate": 1.9080080300735478e-06,
+      "loss": 0.1742,
+      "step": 1182
+    },
+    {
+      "epoch": 5.770731707317073,
+      "grad_norm": 1.3809062242507935,
+      "learning_rate": 1.9042864788150695e-06,
+      "loss": 0.3167,
+      "step": 1183
+    },
+    {
+      "epoch": 5.775609756097561,
+      "grad_norm": 0.7168273329734802,
+      "learning_rate": 1.9005663265942206e-06,
+      "loss": 0.0745,
+      "step": 1184
+    },
+    {
+      "epoch": 5.780487804878049,
+      "grad_norm": 0.9125983715057373,
+      "learning_rate": 1.8968475821478066e-06,
+      "loss": 0.1352,
+      "step": 1185
+    },
+    {
+      "epoch": 5.785365853658536,
+      "grad_norm": 1.2017624378204346,
+      "learning_rate": 1.8931302542093274e-06,
+      "loss": 0.2144,
+      "step": 1186
+    },
+    {
+      "epoch": 5.790243902439024,
+      "grad_norm": 0.902391254901886,
+      "learning_rate": 1.8894143515089539e-06,
+      "loss": 0.1071,
+      "step": 1187
+    },
+    {
+      "epoch": 5.795121951219512,
+      "grad_norm": 0.9119297862052917,
+      "learning_rate": 1.8856998827735118e-06,
+      "loss": 0.141,
+      "step": 1188
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.308137059211731,
+      "learning_rate": 1.8819868567264588e-06,
+      "loss": 0.4523,
+      "step": 1189
+    },
+    {
+      "epoch": 5.804878048780488,
+      "grad_norm": 1.2294268608093262,
+      "learning_rate": 1.8782752820878636e-06,
+      "loss": 0.1783,
+      "step": 1190
+    },
+    {
+      "epoch": 5.809756097560975,
+      "grad_norm": 1.2038543224334717,
+      "learning_rate": 1.8745651675743876e-06,
+      "loss": 0.3132,
+      "step": 1191
+    },
+    {
+      "epoch": 5.814634146341463,
+      "grad_norm": 1.4764758348464966,
+      "learning_rate": 1.870856521899261e-06,
+      "loss": 0.2718,
+      "step": 1192
+    },
+    {
+      "epoch": 5.819512195121951,
+      "grad_norm": 0.9900771975517273,
+      "learning_rate": 1.867149353772267e-06,
+      "loss": 0.0695,
+      "step": 1193
+    },
+    {
+      "epoch": 5.824390243902439,
+      "grad_norm": 1.202021837234497,
+      "learning_rate": 1.863443671899717e-06,
+      "loss": 0.1107,
+      "step": 1194
+    },
+    {
+      "epoch": 5.829268292682927,
+      "grad_norm": 1.2088758945465088,
+      "learning_rate": 1.8597394849844319e-06,
+      "loss": 0.2713,
+      "step": 1195
+    },
+    {
+      "epoch": 5.8341463414634145,
+      "grad_norm": 0.9177408814430237,
+      "learning_rate": 1.8560368017257229e-06,
+      "loss": 0.1473,
+      "step": 1196
+    },
+    {
+      "epoch": 5.839024390243902,
+      "grad_norm": 1.2501375675201416,
+      "learning_rate": 1.8523356308193696e-06,
+      "loss": 0.4487,
+      "step": 1197
+    },
+    {
+      "epoch": 5.84390243902439,
+      "grad_norm": 1.118666410446167,
+      "learning_rate": 1.8486359809575977e-06,
+      "loss": 0.2547,
+      "step": 1198
+    },
+    {
+      "epoch": 5.848780487804878,
+      "grad_norm": 1.0372276306152344,
+      "learning_rate": 1.8449378608290638e-06,
+      "loss": 0.176,
+      "step": 1199
+    },
+    {
+      "epoch": 5.853658536585366,
+      "grad_norm": 1.4009885787963867,
+      "learning_rate": 1.8412412791188306e-06,
+      "loss": 0.2859,
+      "step": 1200
+    },
+    {
+      "epoch": 5.8585365853658535,
+      "grad_norm": 1.327064037322998,
+      "learning_rate": 1.8375462445083464e-06,
+      "loss": 0.282,
+      "step": 1201
+    },
+    {
+      "epoch": 5.863414634146341,
+      "grad_norm": 1.2236543893814087,
+      "learning_rate": 1.8338527656754285e-06,
+      "loss": 0.138,
+      "step": 1202
+    },
+    {
+      "epoch": 5.868292682926829,
+      "grad_norm": 1.3589495420455933,
+      "learning_rate": 1.830160851294239e-06,
+      "loss": 0.2383,
+      "step": 1203
+    },
+    {
+      "epoch": 5.873170731707317,
+      "grad_norm": 1.2593518495559692,
+      "learning_rate": 1.8264705100352662e-06,
+      "loss": 0.3166,
+      "step": 1204
+    },
+    {
+      "epoch": 5.878048780487805,
+      "grad_norm": 1.3545321226119995,
+      "learning_rate": 1.8227817505653045e-06,
+      "loss": 0.2123,
+      "step": 1205
+    },
+    {
+      "epoch": 5.882926829268293,
+      "grad_norm": 1.2606163024902344,
+      "learning_rate": 1.8190945815474323e-06,
+      "loss": 0.3234,
+      "step": 1206
+    },
+    {
+      "epoch": 5.88780487804878,
+      "grad_norm": 1.1344720125198364,
+      "learning_rate": 1.8154090116409934e-06,
+      "loss": 0.2473,
+      "step": 1207
+    },
+    {
+      "epoch": 5.892682926829268,
+      "grad_norm": 1.3704792261123657,
+      "learning_rate": 1.811725049501577e-06,
+      "loss": 0.2684,
+      "step": 1208
+    },
+    {
+      "epoch": 5.897560975609756,
+      "grad_norm": 1.3613022565841675,
+      "learning_rate": 1.8080427037809941e-06,
+      "loss": 0.4087,
+      "step": 1209
+    },
+    {
+      "epoch": 5.902439024390244,
+      "grad_norm": 1.125900387763977,
+      "learning_rate": 1.8043619831272623e-06,
+      "loss": 0.2208,
+      "step": 1210
+    },
+    {
+      "epoch": 5.907317073170732,
+      "grad_norm": 1.4554516077041626,
+      "learning_rate": 1.8006828961845807e-06,
+      "loss": 0.4104,
+      "step": 1211
+    },
+    {
+      "epoch": 5.912195121951219,
+      "grad_norm": 1.507084846496582,
+      "learning_rate": 1.7970054515933124e-06,
+      "loss": 0.4952,
+      "step": 1212
+    },
+    {
+      "epoch": 5.917073170731707,
+      "grad_norm": 1.6275372505187988,
+      "learning_rate": 1.793329657989964e-06,
+      "loss": 0.5594,
+      "step": 1213
+    },
+    {
+      "epoch": 5.921951219512195,
+      "grad_norm": 1.0565378665924072,
+      "learning_rate": 1.7896555240071627e-06,
+      "loss": 0.1108,
+      "step": 1214
+    },
+    {
+      "epoch": 5.926829268292683,
+      "grad_norm": 0.8863050937652588,
+      "learning_rate": 1.7859830582736406e-06,
+      "loss": 0.1669,
+      "step": 1215
+    },
+    {
+      "epoch": 5.931707317073171,
+      "grad_norm": 0.9882891774177551,
+      "learning_rate": 1.782312269414211e-06,
+      "loss": 0.2115,
+      "step": 1216
+    },
+    {
+      "epoch": 5.9365853658536585,
+      "grad_norm": 1.5111620426177979,
+      "learning_rate": 1.7786431660497474e-06,
+      "loss": 0.5776,
+      "step": 1217
+    },
+    {
+      "epoch": 5.941463414634146,
+      "grad_norm": 1.2369654178619385,
+      "learning_rate": 1.7749757567971678e-06,
+      "loss": 0.2491,
+      "step": 1218
+    },
+    {
+      "epoch": 5.946341463414634,
+      "grad_norm": 1.2180466651916504,
+      "learning_rate": 1.7713100502694091e-06,
+      "loss": 0.2229,
+      "step": 1219
+    },
+    {
+      "epoch": 5.951219512195122,
+      "grad_norm": 0.8823730945587158,
+      "learning_rate": 1.7676460550754104e-06,
+      "loss": 0.0664,
+      "step": 1220
+    },
+    {
+      "epoch": 5.95609756097561,
+      "grad_norm": 1.2817288637161255,
+      "learning_rate": 1.7639837798200923e-06,
+      "loss": 0.293,
+      "step": 1221
+    },
+    {
+      "epoch": 5.9609756097560975,
+      "grad_norm": 0.871067464351654,
+      "learning_rate": 1.7603232331043346e-06,
+      "loss": 0.1633,
+      "step": 1222
+    },
+    {
+      "epoch": 5.965853658536585,
+      "grad_norm": 1.4113450050354004,
+      "learning_rate": 1.7566644235249591e-06,
+      "loss": 0.6671,
+      "step": 1223
+    },
+    {
+      "epoch": 5.970731707317073,
+      "grad_norm": 1.4699230194091797,
+      "learning_rate": 1.7530073596747072e-06,
+      "loss": 0.1407,
+      "step": 1224
+    },
+    {
+      "epoch": 5.975609756097561,
+      "grad_norm": 0.8759057521820068,
+      "learning_rate": 1.74935205014222e-06,
+      "loss": 0.0762,
+      "step": 1225
+    },
+    {
+      "epoch": 5.980487804878049,
+      "grad_norm": 0.9844141006469727,
+      "learning_rate": 1.7456985035120194e-06,
+      "loss": 0.1108,
+      "step": 1226
+    },
+    {
+      "epoch": 5.985365853658537,
+      "grad_norm": 0.9976269006729126,
+      "learning_rate": 1.7420467283644877e-06,
+      "loss": 0.2162,
+      "step": 1227
+    },
+    {
+      "epoch": 5.990243902439024,
+      "grad_norm": 1.125545620918274,
+      "learning_rate": 1.738396733275844e-06,
+      "loss": 0.1998,
+      "step": 1228
+    },
+    {
+      "epoch": 5.995121951219512,
+      "grad_norm": 2.0705325603485107,
+      "learning_rate": 1.7347485268181309e-06,
+      "loss": 0.4108,
+      "step": 1229
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.8654595017433167,
+      "learning_rate": 1.7311021175591868e-06,
+      "loss": 0.1534,
+      "step": 1230
+    },
+    {
+      "epoch": 6.004878048780488,
+      "grad_norm": 1.0550576448440552,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.1718,
+      "step": 1231
+    },
+    {
+      "epoch": 6.009756097560976,
+      "grad_norm": 1.057306170463562,
+      "learning_rate": 1.7238147248878444e-06,
+      "loss": 0.1425,
+      "step": 1232
+    },
+    {
+      "epoch": 6.014634146341463,
+      "grad_norm": 0.925433337688446,
+      "learning_rate": 1.7201737585899415e-06,
+      "loss": 0.1146,
+      "step": 1233
+    },
+    {
+      "epoch": 6.019512195121951,
+      "grad_norm": 0.8379129767417908,
+      "learning_rate": 1.7165346237197594e-06,
+      "loss": 0.1446,
+      "step": 1234
+    },
+    {
+      "epoch": 6.024390243902439,
+      "grad_norm": 1.3918683528900146,
+      "learning_rate": 1.7128973288238344e-06,
+      "loss": 0.304,
+      "step": 1235
+    },
+    {
+      "epoch": 6.029268292682927,
+      "grad_norm": 0.9713860154151917,
+      "learning_rate": 1.709261882444379e-06,
+      "loss": 0.1576,
+      "step": 1236
+    },
+    {
+      "epoch": 6.034146341463415,
+      "grad_norm": 1.0337470769882202,
+      "learning_rate": 1.705628293119268e-06,
+      "loss": 0.1528,
+      "step": 1237
+    },
+    {
+      "epoch": 6.0390243902439025,
+      "grad_norm": 2.0387566089630127,
+      "learning_rate": 1.701996569382011e-06,
+      "loss": 0.5301,
+      "step": 1238
+    },
+    {
+      "epoch": 6.04390243902439,
+      "grad_norm": 1.2020694017410278,
+      "learning_rate": 1.6983667197617386e-06,
+      "loss": 0.2001,
+      "step": 1239
+    },
+    {
+      "epoch": 6.048780487804878,
+      "grad_norm": 1.1894810199737549,
+      "learning_rate": 1.6947387527831813e-06,
+      "loss": 0.068,
+      "step": 1240
+    },
+    {
+      "epoch": 6.053658536585366,
+      "grad_norm": 1.3444442749023438,
+      "learning_rate": 1.6911126769666442e-06,
+      "loss": 0.0774,
+      "step": 1241
+    },
+    {
+      "epoch": 6.058536585365854,
+      "grad_norm": 2.132267713546753,
+      "learning_rate": 1.6874885008279945e-06,
+      "loss": 0.3761,
+      "step": 1242
+    },
+    {
+      "epoch": 6.0634146341463415,
+      "grad_norm": 1.3125851154327393,
+      "learning_rate": 1.683866232878637e-06,
+      "loss": 0.0854,
+      "step": 1243
+    },
+    {
+      "epoch": 6.068292682926829,
+      "grad_norm": 1.0103996992111206,
+      "learning_rate": 1.6802458816254941e-06,
+      "loss": 0.0924,
+      "step": 1244
+    },
+    {
+      "epoch": 6.073170731707317,
+      "grad_norm": 0.8832839727401733,
+      "learning_rate": 1.676627455570988e-06,
+      "loss": 0.0909,
+      "step": 1245
+    },
+    {
+      "epoch": 6.078048780487805,
+      "grad_norm": 1.291757583618164,
+      "learning_rate": 1.6730109632130199e-06,
+      "loss": 0.2173,
+      "step": 1246
+    },
+    {
+      "epoch": 6.082926829268293,
+      "grad_norm": 0.8954195976257324,
+      "learning_rate": 1.6693964130449472e-06,
+      "loss": 0.0573,
+      "step": 1247
+    },
+    {
+      "epoch": 6.087804878048781,
+      "grad_norm": 0.8903191089630127,
+      "learning_rate": 1.6657838135555696e-06,
+      "loss": 0.0731,
+      "step": 1248
+    },
+    {
+      "epoch": 6.092682926829268,
+      "grad_norm": 1.7279207706451416,
+      "learning_rate": 1.6621731732291024e-06,
+      "loss": 0.0727,
+      "step": 1249
+    },
+    {
+      "epoch": 6.097560975609756,
+      "grad_norm": 1.3296290636062622,
+      "learning_rate": 1.6585645005451623e-06,
+      "loss": 0.1539,
+      "step": 1250
+    },
+    {
+      "epoch": 6.102439024390244,
+      "grad_norm": 0.9061964750289917,
+      "learning_rate": 1.6549578039787436e-06,
+      "loss": 0.1489,
+      "step": 1251
+    },
+    {
+      "epoch": 6.107317073170732,
+      "grad_norm": 0.7422329783439636,
+      "learning_rate": 1.6513530920001998e-06,
+      "loss": 0.0494,
+      "step": 1252
+    },
+    {
+      "epoch": 6.11219512195122,
+      "grad_norm": 0.8662224411964417,
+      "learning_rate": 1.6477503730752237e-06,
+      "loss": 0.1079,
+      "step": 1253
+    },
+    {
+      "epoch": 6.117073170731707,
+      "grad_norm": 2.585952043533325,
+      "learning_rate": 1.6441496556648278e-06,
+      "loss": 0.2317,
+      "step": 1254
+    },
+    {
+      "epoch": 6.121951219512195,
+      "grad_norm": 1.4481921195983887,
+      "learning_rate": 1.6405509482253234e-06,
+      "loss": 0.5054,
+      "step": 1255
+    },
+    {
+      "epoch": 6.126829268292683,
+      "grad_norm": 0.7095943093299866,
+      "learning_rate": 1.636954259208302e-06,
+      "loss": 0.0931,
+      "step": 1256
+    },
+    {
+      "epoch": 6.131707317073171,
+      "grad_norm": 1.049392580986023,
+      "learning_rate": 1.6333595970606143e-06,
+      "loss": 0.1507,
+      "step": 1257
+    },
+    {
+      "epoch": 6.136585365853659,
+      "grad_norm": 1.1488687992095947,
+      "learning_rate": 1.62976697022435e-06,
+      "loss": 0.2695,
+      "step": 1258
+    },
+    {
+      "epoch": 6.1414634146341465,
+      "grad_norm": 0.9358264803886414,
+      "learning_rate": 1.6261763871368225e-06,
+      "loss": 0.1395,
+      "step": 1259
+    },
+    {
+      "epoch": 6.146341463414634,
+      "grad_norm": 1.0877493619918823,
+      "learning_rate": 1.6225878562305403e-06,
+      "loss": 0.2465,
+      "step": 1260
+    },
+    {
+      "epoch": 6.151219512195122,
+      "grad_norm": 1.1635690927505493,
+      "learning_rate": 1.6190013859331958e-06,
+      "loss": 0.1768,
+      "step": 1261
+    },
+    {
+      "epoch": 6.15609756097561,
+      "grad_norm": 0.9073989391326904,
+      "learning_rate": 1.6154169846676415e-06,
+      "loss": 0.1429,
+      "step": 1262
+    },
+    {
+      "epoch": 6.160975609756098,
+      "grad_norm": 1.2221269607543945,
+      "learning_rate": 1.6118346608518698e-06,
+      "loss": 0.171,
+      "step": 1263
+    },
+    {
+      "epoch": 6.1658536585365855,
+      "grad_norm": 0.8346547484397888,
+      "learning_rate": 1.6082544228989958e-06,
+      "loss": 0.0594,
+      "step": 1264
+    },
+    {
+      "epoch": 6.170731707317073,
+      "grad_norm": 1.4623945951461792,
+      "learning_rate": 1.6046762792172336e-06,
+      "loss": 0.1183,
+      "step": 1265
+    },
+    {
+      "epoch": 6.175609756097561,
+      "grad_norm": 1.2494970560073853,
+      "learning_rate": 1.6011002382098806e-06,
+      "loss": 0.303,
+      "step": 1266
+    },
+    {
+      "epoch": 6.180487804878049,
+      "grad_norm": 0.8755192756652832,
+      "learning_rate": 1.5975263082752968e-06,
+      "loss": 0.0745,
+      "step": 1267
+    },
+    {
+      "epoch": 6.185365853658537,
+      "grad_norm": 0.8409280776977539,
+      "learning_rate": 1.5939544978068816e-06,
+      "loss": 0.0979,
+      "step": 1268
+    },
+    {
+      "epoch": 6.190243902439025,
+      "grad_norm": 1.0874266624450684,
+      "learning_rate": 1.590384815193059e-06,
+      "loss": 0.2561,
+      "step": 1269
+    },
+    {
+      "epoch": 6.195121951219512,
+      "grad_norm": 1.4487617015838623,
+      "learning_rate": 1.5868172688172559e-06,
+      "loss": 0.2705,
+      "step": 1270
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.1141877174377441,
+      "learning_rate": 1.5832518670578802e-06,
+      "loss": 0.3218,
+      "step": 1271
+    },
+    {
+      "epoch": 6.204878048780488,
+      "grad_norm": 1.1559665203094482,
+      "learning_rate": 1.5796886182883053e-06,
+      "loss": 0.3038,
+      "step": 1272
+    },
+    {
+      "epoch": 6.209756097560976,
+      "grad_norm": 0.7617695927619934,
+      "learning_rate": 1.5761275308768476e-06,
+      "loss": 0.1078,
+      "step": 1273
+    },
+    {
+      "epoch": 6.214634146341464,
+      "grad_norm": 0.7606493234634399,
+      "learning_rate": 1.5725686131867462e-06,
+      "loss": 0.0702,
+      "step": 1274
+    },
+    {
+      "epoch": 6.219512195121951,
+      "grad_norm": 1.153968334197998,
+      "learning_rate": 1.569011873576147e-06,
+      "loss": 0.1627,
+      "step": 1275
+    },
+    {
+      "epoch": 6.224390243902439,
+      "grad_norm": 1.0581713914871216,
+      "learning_rate": 1.5654573203980782e-06,
+      "loss": 0.1538,
+      "step": 1276
+    },
+    {
+      "epoch": 6.229268292682927,
+      "grad_norm": 1.156667709350586,
+      "learning_rate": 1.5619049620004354e-06,
+      "loss": 0.2563,
+      "step": 1277
+    },
+    {
+      "epoch": 6.234146341463415,
+      "grad_norm": 0.8070971369743347,
+      "learning_rate": 1.5583548067259584e-06,
+      "loss": 0.1003,
+      "step": 1278
+    },
+    {
+      "epoch": 6.239024390243903,
+      "grad_norm": 1.0131428241729736,
+      "learning_rate": 1.5548068629122126e-06,
+      "loss": 0.2334,
+      "step": 1279
+    },
+    {
+      "epoch": 6.2439024390243905,
+      "grad_norm": 0.9818571209907532,
+      "learning_rate": 1.5512611388915711e-06,
+      "loss": 0.2086,
+      "step": 1280
+    },
+    {
+      "epoch": 6.248780487804878,
+      "grad_norm": 1.344126582145691,
+      "learning_rate": 1.5477176429911934e-06,
+      "loss": 0.4732,
+      "step": 1281
+    },
+    {
+      "epoch": 6.253658536585366,
+      "grad_norm": 0.8236895203590393,
+      "learning_rate": 1.5441763835330048e-06,
+      "loss": 0.0692,
+      "step": 1282
+    },
+    {
+      "epoch": 6.258536585365854,
+      "grad_norm": 1.0422422885894775,
+      "learning_rate": 1.5406373688336807e-06,
+      "loss": 0.084,
+      "step": 1283
+    },
+    {
+      "epoch": 6.263414634146342,
+      "grad_norm": 0.9542838335037231,
+      "learning_rate": 1.5371006072046225e-06,
+      "loss": 0.1759,
+      "step": 1284
+    },
+    {
+      "epoch": 6.2682926829268295,
+      "grad_norm": 0.7214556932449341,
+      "learning_rate": 1.5335661069519408e-06,
+      "loss": 0.089,
+      "step": 1285
+    },
+    {
+      "epoch": 6.273170731707317,
+      "grad_norm": 1.2626168727874756,
+      "learning_rate": 1.5300338763764371e-06,
+      "loss": 0.0866,
+      "step": 1286
+    },
+    {
+      "epoch": 6.278048780487805,
+      "grad_norm": 1.2875781059265137,
+      "learning_rate": 1.5265039237735804e-06,
+      "loss": 0.1159,
+      "step": 1287
+    },
+    {
+      "epoch": 6.282926829268293,
+      "grad_norm": 1.00115966796875,
+      "learning_rate": 1.5229762574334903e-06,
+      "loss": 0.0688,
+      "step": 1288
+    },
+    {
+      "epoch": 6.287804878048781,
+      "grad_norm": 1.2635149955749512,
+      "learning_rate": 1.5194508856409181e-06,
+      "loss": 0.2394,
+      "step": 1289
+    },
+    {
+      "epoch": 6.2926829268292686,
+      "grad_norm": 1.0395175218582153,
+      "learning_rate": 1.515927816675225e-06,
+      "loss": 0.1104,
+      "step": 1290
+    },
+    {
+      "epoch": 6.297560975609756,
+      "grad_norm": 2.234849452972412,
+      "learning_rate": 1.5124070588103648e-06,
+      "loss": 0.0526,
+      "step": 1291
+    },
+    {
+      "epoch": 6.302439024390244,
+      "grad_norm": 0.791607677936554,
+      "learning_rate": 1.5088886203148643e-06,
+      "loss": 0.1148,
+      "step": 1292
+    },
+    {
+      "epoch": 6.307317073170732,
+      "grad_norm": 1.2630048990249634,
+      "learning_rate": 1.505372509451801e-06,
+      "loss": 0.3476,
+      "step": 1293
+    },
+    {
+      "epoch": 6.31219512195122,
+      "grad_norm": 0.9735398888587952,
+      "learning_rate": 1.5018587344787888e-06,
+      "loss": 0.0946,
+      "step": 1294
+    },
+    {
+      "epoch": 6.317073170731708,
+      "grad_norm": 1.4363349676132202,
+      "learning_rate": 1.498347303647953e-06,
+      "loss": 0.1628,
+      "step": 1295
+    },
+    {
+      "epoch": 6.321951219512195,
+      "grad_norm": 1.1442455053329468,
+      "learning_rate": 1.4948382252059158e-06,
+      "loss": 0.1958,
+      "step": 1296
+    },
+    {
+      "epoch": 6.326829268292683,
+      "grad_norm": 0.9576900601387024,
+      "learning_rate": 1.4913315073937742e-06,
+      "loss": 0.1721,
+      "step": 1297
+    },
+    {
+      "epoch": 6.331707317073171,
+      "grad_norm": 1.2679523229599,
+      "learning_rate": 1.4878271584470805e-06,
+      "loss": 0.2193,
+      "step": 1298
+    },
+    {
+      "epoch": 6.336585365853659,
+      "grad_norm": 1.2026294469833374,
+      "learning_rate": 1.4843251865958242e-06,
+      "loss": 0.1086,
+      "step": 1299
+    },
+    {
+      "epoch": 6.341463414634147,
+      "grad_norm": 1.2589906454086304,
+      "learning_rate": 1.4808256000644128e-06,
+      "loss": 0.1949,
+      "step": 1300
+    },
+    {
+      "epoch": 6.3463414634146345,
+      "grad_norm": 1.1509987115859985,
+      "learning_rate": 1.4773284070716504e-06,
+      "loss": 0.2124,
+      "step": 1301
+    },
+    {
+      "epoch": 6.351219512195122,
+      "grad_norm": 1.9435495138168335,
+      "learning_rate": 1.473833615830722e-06,
+      "loss": 0.1059,
+      "step": 1302
+    },
+    {
+      "epoch": 6.35609756097561,
+      "grad_norm": 1.1303460597991943,
+      "learning_rate": 1.4703412345491692e-06,
+      "loss": 0.1562,
+      "step": 1303
+    },
+    {
+      "epoch": 6.360975609756098,
+      "grad_norm": 1.0799435377120972,
+      "learning_rate": 1.4668512714288763e-06,
+      "loss": 0.1182,
+      "step": 1304
+    },
+    {
+      "epoch": 6.365853658536586,
+      "grad_norm": 0.8857361674308777,
+      "learning_rate": 1.4633637346660478e-06,
+      "loss": 0.0652,
+      "step": 1305
+    },
+    {
+      "epoch": 6.3707317073170735,
+      "grad_norm": 1.1127703189849854,
+      "learning_rate": 1.4598786324511892e-06,
+      "loss": 0.099,
+      "step": 1306
+    },
+    {
+      "epoch": 6.375609756097561,
+      "grad_norm": 1.0147268772125244,
+      "learning_rate": 1.456395972969089e-06,
+      "loss": 0.0864,
+      "step": 1307
+    },
+    {
+      "epoch": 6.380487804878049,
+      "grad_norm": 1.2551109790802002,
+      "learning_rate": 1.4529157643987995e-06,
+      "loss": 0.2552,
+      "step": 1308
+    },
+    {
+      "epoch": 6.385365853658537,
+      "grad_norm": 1.044129490852356,
+      "learning_rate": 1.4494380149136162e-06,
+      "loss": 0.1989,
+      "step": 1309
+    },
+    {
+      "epoch": 6.390243902439025,
+      "grad_norm": 1.366299033164978,
+      "learning_rate": 1.4459627326810576e-06,
+      "loss": 0.1694,
+      "step": 1310
+    },
+    {
+      "epoch": 6.3951219512195125,
+      "grad_norm": 1.2774457931518555,
+      "learning_rate": 1.4424899258628533e-06,
+      "loss": 0.1415,
+      "step": 1311
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.07902991771698,
+      "learning_rate": 1.439019602614914e-06,
+      "loss": 0.0497,
+      "step": 1312
+    },
+    {
+      "epoch": 6.404878048780488,
+      "grad_norm": 1.1242656707763672,
+      "learning_rate": 1.4355517710873184e-06,
+      "loss": 0.2435,
+      "step": 1313
+    },
+    {
+      "epoch": 6.409756097560976,
+      "grad_norm": 1.9656187295913696,
+      "learning_rate": 1.432086439424297e-06,
+      "loss": 0.3044,
+      "step": 1314
+    },
+    {
+      "epoch": 6.414634146341464,
+      "grad_norm": 1.5738033056259155,
+      "learning_rate": 1.428623615764206e-06,
+      "loss": 0.4497,
+      "step": 1315
+    },
+    {
+      "epoch": 6.419512195121952,
+      "grad_norm": 0.8496007919311523,
+      "learning_rate": 1.4251633082395117e-06,
+      "loss": 0.142,
+      "step": 1316
+    },
+    {
+      "epoch": 6.424390243902439,
+      "grad_norm": 1.3462088108062744,
+      "learning_rate": 1.4217055249767734e-06,
+      "loss": 0.177,
+      "step": 1317
+    },
+    {
+      "epoch": 6.429268292682927,
+      "grad_norm": 0.9757170677185059,
+      "learning_rate": 1.4182502740966203e-06,
+      "loss": 0.0472,
+      "step": 1318
+    },
+    {
+      "epoch": 6.434146341463415,
+      "grad_norm": 1.0830655097961426,
+      "learning_rate": 1.4147975637137334e-06,
+      "loss": 0.1507,
+      "step": 1319
+    },
+    {
+      "epoch": 6.439024390243903,
+      "grad_norm": 1.2118395566940308,
+      "learning_rate": 1.411347401936831e-06,
+      "loss": 0.1994,
+      "step": 1320
+    },
+    {
+      "epoch": 6.443902439024391,
+      "grad_norm": 1.5108318328857422,
+      "learning_rate": 1.4078997968686425e-06,
+      "loss": 0.1086,
+      "step": 1321
+    },
+    {
+      "epoch": 6.4487804878048784,
+      "grad_norm": 1.0448169708251953,
+      "learning_rate": 1.404454756605893e-06,
+      "loss": 0.1801,
+      "step": 1322
+    },
+    {
+      "epoch": 6.453658536585366,
+      "grad_norm": 1.2599785327911377,
+      "learning_rate": 1.4010122892392872e-06,
+      "loss": 0.3254,
+      "step": 1323
+    },
+    {
+      "epoch": 6.458536585365854,
+      "grad_norm": 1.0006107091903687,
+      "learning_rate": 1.3975724028534842e-06,
+      "loss": 0.2379,
+      "step": 1324
+    },
+    {
+      "epoch": 6.463414634146342,
+      "grad_norm": 1.205500602722168,
+      "learning_rate": 1.394135105527083e-06,
+      "loss": 0.1195,
+      "step": 1325
+    },
+    {
+      "epoch": 6.46829268292683,
+      "grad_norm": 1.282423973083496,
+      "learning_rate": 1.3907004053326006e-06,
+      "loss": 0.1627,
+      "step": 1326
+    },
+    {
+      "epoch": 6.473170731707317,
+      "grad_norm": 1.0242518186569214,
+      "learning_rate": 1.387268310336458e-06,
+      "loss": 0.1358,
+      "step": 1327
+    },
+    {
+      "epoch": 6.478048780487805,
+      "grad_norm": 1.1453756093978882,
+      "learning_rate": 1.3838388285989552e-06,
+      "loss": 0.1715,
+      "step": 1328
+    },
+    {
+      "epoch": 6.482926829268292,
+      "grad_norm": 0.9189904928207397,
+      "learning_rate": 1.380411968174254e-06,
+      "loss": 0.1148,
+      "step": 1329
+    },
+    {
+      "epoch": 6.487804878048781,
+      "grad_norm": 1.40865159034729,
+      "learning_rate": 1.3769877371103635e-06,
+      "loss": 0.2742,
+      "step": 1330
+    },
+    {
+      "epoch": 6.492682926829268,
+      "grad_norm": 1.5062743425369263,
+      "learning_rate": 1.373566143449115e-06,
+      "loss": 0.418,
+      "step": 1331
+    },
+    {
+      "epoch": 6.4975609756097565,
+      "grad_norm": 0.7628481388092041,
+      "learning_rate": 1.3701471952261457e-06,
+      "loss": 0.069,
+      "step": 1332
+    },
+    {
+      "epoch": 6.5024390243902435,
+      "grad_norm": 1.154660701751709,
+      "learning_rate": 1.3667309004708832e-06,
+      "loss": 0.1043,
+      "step": 1333
+    },
+    {
+      "epoch": 6.507317073170732,
+      "grad_norm": 1.6828221082687378,
+      "learning_rate": 1.3633172672065195e-06,
+      "loss": 0.2613,
+      "step": 1334
+    },
+    {
+      "epoch": 6.512195121951219,
+      "grad_norm": 1.1095471382141113,
+      "learning_rate": 1.359906303449997e-06,
+      "loss": 0.1142,
+      "step": 1335
+    },
+    {
+      "epoch": 6.517073170731708,
+      "grad_norm": 0.9421728849411011,
+      "learning_rate": 1.3564980172119913e-06,
+      "loss": 0.0699,
+      "step": 1336
+    },
+    {
+      "epoch": 6.521951219512195,
+      "grad_norm": 1.3236054182052612,
+      "learning_rate": 1.3530924164968873e-06,
+      "loss": 0.4213,
+      "step": 1337
+    },
+    {
+      "epoch": 6.526829268292683,
+      "grad_norm": 0.973343551158905,
+      "learning_rate": 1.3496895093027617e-06,
+      "loss": 0.1617,
+      "step": 1338
+    },
+    {
+      "epoch": 6.53170731707317,
+      "grad_norm": 1.0116113424301147,
+      "learning_rate": 1.3462893036213706e-06,
+      "loss": 0.1201,
+      "step": 1339
+    },
+    {
+      "epoch": 6.536585365853659,
+      "grad_norm": 1.2368037700653076,
+      "learning_rate": 1.3428918074381203e-06,
+      "loss": 0.127,
+      "step": 1340
+    },
+    {
+      "epoch": 6.541463414634146,
+      "grad_norm": 1.0005751848220825,
+      "learning_rate": 1.3394970287320553e-06,
+      "loss": 0.2255,
+      "step": 1341
+    },
+    {
+      "epoch": 6.546341463414635,
+      "grad_norm": 0.7819541096687317,
+      "learning_rate": 1.3361049754758404e-06,
+      "loss": 0.1052,
+      "step": 1342
+    },
+    {
+      "epoch": 6.5512195121951216,
+      "grad_norm": 0.8330570459365845,
+      "learning_rate": 1.3327156556357369e-06,
+      "loss": 0.0622,
+      "step": 1343
+    },
+    {
+      "epoch": 6.55609756097561,
+      "grad_norm": 1.335412859916687,
+      "learning_rate": 1.3293290771715875e-06,
+      "loss": 0.2166,
+      "step": 1344
+    },
+    {
+      "epoch": 6.560975609756097,
+      "grad_norm": 0.9714047312736511,
+      "learning_rate": 1.3259452480367963e-06,
+      "loss": 0.1613,
+      "step": 1345
+    },
+    {
+      "epoch": 6.565853658536585,
+      "grad_norm": 1.260328769683838,
+      "learning_rate": 1.3225641761783126e-06,
+      "loss": 0.2404,
+      "step": 1346
+    },
+    {
+      "epoch": 6.570731707317073,
+      "grad_norm": 1.480240821838379,
+      "learning_rate": 1.3191858695366084e-06,
+      "loss": 0.3174,
+      "step": 1347
+    },
+    {
+      "epoch": 6.575609756097561,
+      "grad_norm": 1.1139479875564575,
+      "learning_rate": 1.3158103360456603e-06,
+      "loss": 0.1937,
+      "step": 1348
+    },
+    {
+      "epoch": 6.580487804878048,
+      "grad_norm": 1.135933518409729,
+      "learning_rate": 1.3124375836329362e-06,
+      "loss": 0.183,
+      "step": 1349
+    },
+    {
+      "epoch": 6.585365853658536,
+      "grad_norm": 1.24773108959198,
+      "learning_rate": 1.3090676202193692e-06,
+      "loss": 0.0279,
+      "step": 1350
+    },
+    {
+      "epoch": 6.590243902439024,
+      "grad_norm": 1.0175641775131226,
+      "learning_rate": 1.3057004537193424e-06,
+      "loss": 0.109,
+      "step": 1351
+    },
+    {
+      "epoch": 6.595121951219512,
+      "grad_norm": 0.8159235119819641,
+      "learning_rate": 1.302336092040673e-06,
+      "loss": 0.0878,
+      "step": 1352
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.0272810459136963,
+      "learning_rate": 1.298974543084589e-06,
+      "loss": 0.116,
+      "step": 1353
+    },
+    {
+      "epoch": 6.6048780487804875,
+      "grad_norm": 1.0128616094589233,
+      "learning_rate": 1.2956158147457116e-06,
+      "loss": 0.1006,
+      "step": 1354
+    },
+    {
+      "epoch": 6.609756097560975,
+      "grad_norm": 0.9154186248779297,
+      "learning_rate": 1.2922599149120412e-06,
+      "loss": 0.1024,
+      "step": 1355
+    },
+    {
+      "epoch": 6.614634146341463,
+      "grad_norm": 0.941557765007019,
+      "learning_rate": 1.2889068514649328e-06,
+      "loss": 0.1577,
+      "step": 1356
+    },
+    {
+      "epoch": 6.619512195121951,
+      "grad_norm": 1.0502948760986328,
+      "learning_rate": 1.2855566322790796e-06,
+      "loss": 0.0702,
+      "step": 1357
+    },
+    {
+      "epoch": 6.624390243902439,
+      "grad_norm": 1.050806999206543,
+      "learning_rate": 1.2822092652224989e-06,
+      "loss": 0.1168,
+      "step": 1358
+    },
+    {
+      "epoch": 6.6292682926829265,
+      "grad_norm": 1.0025922060012817,
+      "learning_rate": 1.2788647581565048e-06,
+      "loss": 0.0951,
+      "step": 1359
+    },
+    {
+      "epoch": 6.634146341463414,
+      "grad_norm": 1.266708254814148,
+      "learning_rate": 1.275523118935697e-06,
+      "loss": 0.0827,
+      "step": 1360
+    },
+    {
+      "epoch": 6.639024390243902,
+      "grad_norm": 1.0859824419021606,
+      "learning_rate": 1.2721843554079418e-06,
+      "loss": 0.1783,
+      "step": 1361
+    },
+    {
+      "epoch": 6.64390243902439,
+      "grad_norm": 1.2870323657989502,
+      "learning_rate": 1.2688484754143493e-06,
+      "loss": 0.4135,
+      "step": 1362
+    },
+    {
+      "epoch": 6.648780487804878,
+      "grad_norm": 1.2668397426605225,
+      "learning_rate": 1.2655154867892577e-06,
+      "loss": 0.2108,
+      "step": 1363
+    },
+    {
+      "epoch": 6.6536585365853655,
+      "grad_norm": 1.4955235719680786,
+      "learning_rate": 1.2621853973602158e-06,
+      "loss": 0.2235,
+      "step": 1364
+    },
+    {
+      "epoch": 6.658536585365853,
+      "grad_norm": 1.0341984033584595,
+      "learning_rate": 1.2588582149479645e-06,
+      "loss": 0.0549,
+      "step": 1365
+    },
+    {
+      "epoch": 6.663414634146341,
+      "grad_norm": 1.038429617881775,
+      "learning_rate": 1.2555339473664151e-06,
+      "loss": 0.1321,
+      "step": 1366
+    },
+    {
+      "epoch": 6.668292682926829,
+      "grad_norm": 1.2604671716690063,
+      "learning_rate": 1.2522126024226347e-06,
+      "loss": 0.2104,
+      "step": 1367
+    },
+    {
+      "epoch": 6.673170731707317,
+      "grad_norm": 1.0938949584960938,
+      "learning_rate": 1.2488941879168278e-06,
+      "loss": 0.0685,
+      "step": 1368
+    },
+    {
+      "epoch": 6.678048780487805,
+      "grad_norm": 0.8830887079238892,
+      "learning_rate": 1.2455787116423148e-06,
+      "loss": 0.1789,
+      "step": 1369
+    },
+    {
+      "epoch": 6.682926829268292,
+      "grad_norm": 0.9735605120658875,
+      "learning_rate": 1.2422661813855158e-06,
+      "loss": 0.1276,
+      "step": 1370
+    },
+    {
+      "epoch": 6.68780487804878,
+      "grad_norm": 1.117775559425354,
+      "learning_rate": 1.238956604925934e-06,
+      "loss": 0.106,
+      "step": 1371
+    },
+    {
+      "epoch": 6.692682926829268,
+      "grad_norm": 1.5626012086868286,
+      "learning_rate": 1.2356499900361333e-06,
+      "loss": 0.2649,
+      "step": 1372
+    },
+    {
+      "epoch": 6.697560975609756,
+      "grad_norm": 0.8927910327911377,
+      "learning_rate": 1.2323463444817227e-06,
+      "loss": 0.1202,
+      "step": 1373
+    },
+    {
+      "epoch": 6.702439024390244,
+      "grad_norm": 1.393783450126648,
+      "learning_rate": 1.2290456760213405e-06,
+      "loss": 0.2409,
+      "step": 1374
+    },
+    {
+      "epoch": 6.7073170731707314,
+      "grad_norm": 1.1260027885437012,
+      "learning_rate": 1.2257479924066296e-06,
+      "loss": 0.229,
+      "step": 1375
+    },
+    {
+      "epoch": 6.712195121951219,
+      "grad_norm": 2.048091411590576,
+      "learning_rate": 1.2224533013822237e-06,
+      "loss": 0.2282,
+      "step": 1376
+    },
+    {
+      "epoch": 6.717073170731707,
+      "grad_norm": 1.2933448553085327,
+      "learning_rate": 1.2191616106857312e-06,
+      "loss": 0.2226,
+      "step": 1377
+    },
+    {
+      "epoch": 6.721951219512195,
+      "grad_norm": 1.0582860708236694,
+      "learning_rate": 1.2158729280477112e-06,
+      "loss": 0.1759,
+      "step": 1378
+    },
+    {
+      "epoch": 6.726829268292683,
+      "grad_norm": 1.248373031616211,
+      "learning_rate": 1.2125872611916578e-06,
+      "loss": 0.2021,
+      "step": 1379
+    },
+    {
+      "epoch": 6.7317073170731705,
+      "grad_norm": 0.9428868889808655,
+      "learning_rate": 1.2093046178339869e-06,
+      "loss": 0.0795,
+      "step": 1380
+    },
+    {
+      "epoch": 6.736585365853658,
+      "grad_norm": 0.8160692453384399,
+      "learning_rate": 1.206025005684009e-06,
+      "loss": 0.1338,
+      "step": 1381
+    },
+    {
+      "epoch": 6.741463414634146,
+      "grad_norm": 0.7535449862480164,
+      "learning_rate": 1.202748432443918e-06,
+      "loss": 0.0606,
+      "step": 1382
+    },
+    {
+      "epoch": 6.746341463414634,
+      "grad_norm": 1.2124066352844238,
+      "learning_rate": 1.1994749058087695e-06,
+      "loss": 0.2288,
+      "step": 1383
+    },
+    {
+      "epoch": 6.751219512195122,
+      "grad_norm": 1.502294659614563,
+      "learning_rate": 1.196204433466467e-06,
+      "loss": 0.3443,
+      "step": 1384
+    },
+    {
+      "epoch": 6.7560975609756095,
+      "grad_norm": 1.232467770576477,
+      "learning_rate": 1.192937023097738e-06,
+      "loss": 0.2051,
+      "step": 1385
+    },
+    {
+      "epoch": 6.760975609756097,
+      "grad_norm": 0.8221524953842163,
+      "learning_rate": 1.1896726823761195e-06,
+      "loss": 0.0482,
+      "step": 1386
+    },
+    {
+      "epoch": 6.765853658536585,
+      "grad_norm": 1.254210114479065,
+      "learning_rate": 1.1864114189679413e-06,
+      "loss": 0.0872,
+      "step": 1387
+    },
+    {
+      "epoch": 6.770731707317073,
+      "grad_norm": 1.212073802947998,
+      "learning_rate": 1.183153240532304e-06,
+      "loss": 0.0943,
+      "step": 1388
+    },
+    {
+      "epoch": 6.775609756097561,
+      "grad_norm": 1.2102311849594116,
+      "learning_rate": 1.179898154721063e-06,
+      "loss": 0.1226,
+      "step": 1389
+    },
+    {
+      "epoch": 6.780487804878049,
+      "grad_norm": 1.673811912536621,
+      "learning_rate": 1.1766461691788137e-06,
+      "loss": 0.0617,
+      "step": 1390
+    },
+    {
+      "epoch": 6.785365853658536,
+      "grad_norm": 1.4431748390197754,
+      "learning_rate": 1.1733972915428665e-06,
+      "loss": 0.2783,
+      "step": 1391
+    },
+    {
+      "epoch": 6.790243902439024,
+      "grad_norm": 1.1467320919036865,
+      "learning_rate": 1.1701515294432348e-06,
+      "loss": 0.1642,
+      "step": 1392
+    },
+    {
+      "epoch": 6.795121951219512,
+      "grad_norm": 1.7306759357452393,
+      "learning_rate": 1.1669088905026156e-06,
+      "loss": 0.3577,
+      "step": 1393
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 1.0853081941604614,
+      "learning_rate": 1.163669382336371e-06,
+      "loss": 0.1911,
+      "step": 1394
+    },
+    {
+      "epoch": 6.804878048780488,
+      "grad_norm": 0.7488344311714172,
+      "learning_rate": 1.160433012552508e-06,
+      "loss": 0.0403,
+      "step": 1395
+    },
+    {
+      "epoch": 6.809756097560975,
+      "grad_norm": 1.2400141954421997,
+      "learning_rate": 1.1571997887516672e-06,
+      "loss": 0.3637,
+      "step": 1396
+    },
+    {
+      "epoch": 6.814634146341463,
+      "grad_norm": 1.2724562883377075,
+      "learning_rate": 1.1539697185270982e-06,
+      "loss": 0.1827,
+      "step": 1397
+    },
+    {
+      "epoch": 6.819512195121951,
+      "grad_norm": 1.0190290212631226,
+      "learning_rate": 1.1507428094646448e-06,
+      "loss": 0.1489,
+      "step": 1398
+    },
+    {
+      "epoch": 6.824390243902439,
+      "grad_norm": 0.8387356400489807,
+      "learning_rate": 1.1475190691427255e-06,
+      "loss": 0.0689,
+      "step": 1399
+    },
+    {
+      "epoch": 6.829268292682927,
+      "grad_norm": 0.5320870876312256,
+      "learning_rate": 1.1442985051323205e-06,
+      "loss": 0.0203,
+      "step": 1400
+    },
+    {
+      "epoch": 6.8341463414634145,
+      "grad_norm": 1.9445618391036987,
+      "learning_rate": 1.1410811249969475e-06,
+      "loss": 0.5178,
+      "step": 1401
+    },
+    {
+      "epoch": 6.839024390243902,
+      "grad_norm": 1.3419086933135986,
+      "learning_rate": 1.1378669362926468e-06,
+      "loss": 0.284,
+      "step": 1402
+    },
+    {
+      "epoch": 6.84390243902439,
+      "grad_norm": 1.31820809841156,
+      "learning_rate": 1.1346559465679656e-06,
+      "loss": 0.2732,
+      "step": 1403
+    },
+    {
+      "epoch": 6.848780487804878,
+      "grad_norm": 0.6823004484176636,
+      "learning_rate": 1.1314481633639374e-06,
+      "loss": 0.0371,
+      "step": 1404
+    },
+    {
+      "epoch": 6.853658536585366,
+      "grad_norm": 1.3075387477874756,
+      "learning_rate": 1.1282435942140632e-06,
+      "loss": 0.2317,
+      "step": 1405
+    },
+    {
+      "epoch": 6.8585365853658535,
+      "grad_norm": 1.0345687866210938,
+      "learning_rate": 1.1250422466442992e-06,
+      "loss": 0.1354,
+      "step": 1406
+    },
+    {
+      "epoch": 6.863414634146341,
+      "grad_norm": 0.8593140840530396,
+      "learning_rate": 1.1218441281730334e-06,
+      "loss": 0.1287,
+      "step": 1407
+    },
+    {
+      "epoch": 6.868292682926829,
+      "grad_norm": 1.0140306949615479,
+      "learning_rate": 1.1186492463110696e-06,
+      "loss": 0.1075,
+      "step": 1408
+    },
+    {
+      "epoch": 6.873170731707317,
+      "grad_norm": 1.056603193283081,
+      "learning_rate": 1.1154576085616135e-06,
+      "loss": 0.0877,
+      "step": 1409
+    },
+    {
+      "epoch": 6.878048780487805,
+      "grad_norm": 0.9203841686248779,
+      "learning_rate": 1.1122692224202491e-06,
+      "loss": 0.0866,
+      "step": 1410
+    },
+    {
+      "epoch": 6.882926829268293,
+      "grad_norm": 1.1666905879974365,
+      "learning_rate": 1.1090840953749253e-06,
+      "loss": 0.2371,
+      "step": 1411
+    },
+    {
+      "epoch": 6.88780487804878,
+      "grad_norm": 0.806583046913147,
+      "learning_rate": 1.1059022349059362e-06,
+      "loss": 0.058,
+      "step": 1412
+    },
+    {
+      "epoch": 6.892682926829268,
+      "grad_norm": 1.529600977897644,
+      "learning_rate": 1.102723648485905e-06,
+      "loss": 0.4139,
+      "step": 1413
+    },
+    {
+      "epoch": 6.897560975609756,
+      "grad_norm": 1.5345758199691772,
+      "learning_rate": 1.0995483435797643e-06,
+      "loss": 0.2383,
+      "step": 1414
+    },
+    {
+      "epoch": 6.902439024390244,
+      "grad_norm": 0.9218557476997375,
+      "learning_rate": 1.0963763276447435e-06,
+      "loss": 0.0451,
+      "step": 1415
+    },
+    {
+      "epoch": 6.907317073170732,
+      "grad_norm": 1.0400121212005615,
+      "learning_rate": 1.0932076081303442e-06,
+      "loss": 0.0984,
+      "step": 1416
+    },
+    {
+      "epoch": 6.912195121951219,
+      "grad_norm": 0.9838992953300476,
+      "learning_rate": 1.0900421924783272e-06,
+      "loss": 0.1202,
+      "step": 1417
+    },
+    {
+      "epoch": 6.917073170731707,
+      "grad_norm": 0.9632418751716614,
+      "learning_rate": 1.0868800881226962e-06,
+      "loss": 0.1938,
+      "step": 1418
+    },
+    {
+      "epoch": 6.921951219512195,
+      "grad_norm": 1.308287501335144,
+      "learning_rate": 1.0837213024896764e-06,
+      "loss": 0.1664,
+      "step": 1419
+    },
+    {
+      "epoch": 6.926829268292683,
+      "grad_norm": 1.2088172435760498,
+      "learning_rate": 1.080565842997698e-06,
+      "loss": 0.2382,
+      "step": 1420
+    },
+    {
+      "epoch": 6.931707317073171,
+      "grad_norm": 0.6009799838066101,
+      "learning_rate": 1.0774137170573826e-06,
+      "loss": 0.0705,
+      "step": 1421
+    },
+    {
+      "epoch": 6.9365853658536585,
+      "grad_norm": 1.1284160614013672,
+      "learning_rate": 1.074264932071521e-06,
+      "loss": 0.2337,
+      "step": 1422
+    },
+    {
+      "epoch": 6.941463414634146,
+      "grad_norm": 1.1777656078338623,
+      "learning_rate": 1.0711194954350568e-06,
+      "loss": 0.1266,
+      "step": 1423
+    },
+    {
+      "epoch": 6.946341463414634,
+      "grad_norm": 1.0326296091079712,
+      "learning_rate": 1.0679774145350735e-06,
+      "loss": 0.1498,
+      "step": 1424
+    },
+    {
+      "epoch": 6.951219512195122,
+      "grad_norm": 1.2106353044509888,
+      "learning_rate": 1.0648386967507703e-06,
+      "loss": 0.3686,
+      "step": 1425
+    },
+    {
+      "epoch": 6.95609756097561,
+      "grad_norm": 0.9451847672462463,
+      "learning_rate": 1.0617033494534486e-06,
+      "loss": 0.1227,
+      "step": 1426
+    },
+    {
+      "epoch": 6.9609756097560975,
+      "grad_norm": 0.9187631011009216,
+      "learning_rate": 1.0585713800064964e-06,
+      "loss": 0.0725,
+      "step": 1427
+    },
+    {
+      "epoch": 6.965853658536585,
+      "grad_norm": 1.0765812397003174,
+      "learning_rate": 1.0554427957653663e-06,
+      "loss": 0.2055,
+      "step": 1428
+    },
+    {
+      "epoch": 6.970731707317073,
+      "grad_norm": 1.2303036451339722,
+      "learning_rate": 1.0523176040775615e-06,
+      "loss": 0.3308,
+      "step": 1429
+    },
+    {
+      "epoch": 6.975609756097561,
+      "grad_norm": 1.1930241584777832,
+      "learning_rate": 1.0491958122826173e-06,
+      "loss": 0.2065,
+      "step": 1430
+    },
+    {
+      "epoch": 6.980487804878049,
+      "grad_norm": 0.8994832038879395,
+      "learning_rate": 1.0460774277120866e-06,
+      "loss": 0.0904,
+      "step": 1431
+    },
+    {
+      "epoch": 6.985365853658537,
+      "grad_norm": 0.9850203990936279,
+      "learning_rate": 1.0429624576895177e-06,
+      "loss": 0.0591,
+      "step": 1432
+    },
+    {
+      "epoch": 6.990243902439024,
+      "grad_norm": 1.204479455947876,
+      "learning_rate": 1.03985090953044e-06,
+      "loss": 0.1802,
+      "step": 1433
+    },
+    {
+      "epoch": 6.995121951219512,
+      "grad_norm": 1.073361873626709,
+      "learning_rate": 1.0367427905423497e-06,
+      "loss": 0.2169,
+      "step": 1434
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.8216325044631958,
+      "learning_rate": 1.0336381080246858e-06,
+      "loss": 0.0718,
+      "step": 1435
+    },
+    {
+      "epoch": 7.004878048780488,
+      "grad_norm": 1.052810788154602,
+      "learning_rate": 1.0305368692688175e-06,
+      "loss": 0.1574,
+      "step": 1436
+    },
+    {
+      "epoch": 7.009756097560976,
+      "grad_norm": 0.9019646048545837,
+      "learning_rate": 1.027439081558029e-06,
+      "loss": 0.1102,
+      "step": 1437
+    },
+    {
+      "epoch": 7.014634146341463,
+      "grad_norm": 1.0521090030670166,
+      "learning_rate": 1.0243447521674967e-06,
+      "loss": 0.1744,
+      "step": 1438
+    },
+    {
+      "epoch": 7.019512195121951,
+      "grad_norm": 1.2154330015182495,
+      "learning_rate": 1.021253888364276e-06,
+      "loss": 0.4529,
+      "step": 1439
+    },
+    {
+      "epoch": 7.024390243902439,
+      "grad_norm": 0.750938892364502,
+      "learning_rate": 1.018166497407284e-06,
+      "loss": 0.0773,
+      "step": 1440
+    },
+    {
+      "epoch": 7.029268292682927,
+      "grad_norm": 0.6538183093070984,
+      "learning_rate": 1.0150825865472813e-06,
+      "loss": 0.0633,
+      "step": 1441
+    },
+    {
+      "epoch": 7.034146341463415,
+      "grad_norm": 0.7504788041114807,
+      "learning_rate": 1.0120021630268542e-06,
+      "loss": 0.0664,
+      "step": 1442
+    },
+    {
+      "epoch": 7.0390243902439025,
+      "grad_norm": 1.1257810592651367,
+      "learning_rate": 1.0089252340804025e-06,
+      "loss": 0.1104,
+      "step": 1443
+    },
+    {
+      "epoch": 7.04390243902439,
+      "grad_norm": 1.3814443349838257,
+      "learning_rate": 1.0058518069341152e-06,
+      "loss": 0.2337,
+      "step": 1444
+    },
+    {
+      "epoch": 7.048780487804878,
+      "grad_norm": 0.7977355718612671,
+      "learning_rate": 1.002781888805958e-06,
+      "loss": 0.0632,
+      "step": 1445
+    },
+    {
+      "epoch": 7.053658536585366,
+      "grad_norm": 0.9505071043968201,
+      "learning_rate": 9.997154869056588e-07,
+      "loss": 0.079,
+      "step": 1446
+    },
+    {
+      "epoch": 7.058536585365854,
+      "grad_norm": 1.8017224073410034,
+      "learning_rate": 9.966526084346837e-07,
+      "loss": 0.0992,
+      "step": 1447
+    },
+    {
+      "epoch": 7.0634146341463415,
+      "grad_norm": 1.6866507530212402,
+      "learning_rate": 9.935932605862258e-07,
+      "loss": 0.3274,
+      "step": 1448
+    },
+    {
+      "epoch": 7.068292682926829,
+      "grad_norm": 1.0027443170547485,
+      "learning_rate": 9.905374505451853e-07,
+      "loss": 0.152,
+      "step": 1449
+    },
+    {
+      "epoch": 7.073170731707317,
+      "grad_norm": 1.3855262994766235,
+      "learning_rate": 9.874851854881565e-07,
+      "loss": 0.2713,
+      "step": 1450
+    },
+    {
+      "epoch": 7.078048780487805,
+      "grad_norm": 1.0189032554626465,
+      "learning_rate": 9.844364725834058e-07,
+      "loss": 0.1553,
+      "step": 1451
+    },
+    {
+      "epoch": 7.082926829268293,
+      "grad_norm": 1.272226095199585,
+      "learning_rate": 9.813913189908571e-07,
+      "loss": 0.1665,
+      "step": 1452
+    },
+    {
+      "epoch": 7.087804878048781,
+      "grad_norm": 0.7456101775169373,
+      "learning_rate": 9.783497318620783e-07,
+      "loss": 0.0652,
+      "step": 1453
+    },
+    {
+      "epoch": 7.092682926829268,
+      "grad_norm": 0.7051019072532654,
+      "learning_rate": 9.75311718340258e-07,
+      "loss": 0.0751,
+      "step": 1454
+    },
+    {
+      "epoch": 7.097560975609756,
+      "grad_norm": 1.129547357559204,
+      "learning_rate": 9.722772855601927e-07,
+      "loss": 0.2222,
+      "step": 1455
+    },
+    {
+      "epoch": 7.102439024390244,
+      "grad_norm": 0.8566828370094299,
+      "learning_rate": 9.692464406482727e-07,
+      "loss": 0.0828,
+      "step": 1456
+    },
+    {
+      "epoch": 7.107317073170732,
+      "grad_norm": 0.7901312708854675,
+      "learning_rate": 9.662191907224582e-07,
+      "loss": 0.0483,
+      "step": 1457
+    },
+    {
+      "epoch": 7.11219512195122,
+      "grad_norm": 0.8374278545379639,
+      "learning_rate": 9.63195542892268e-07,
+      "loss": 0.0666,
+      "step": 1458
+    },
+    {
+      "epoch": 7.117073170731707,
+      "grad_norm": 0.8921765089035034,
+      "learning_rate": 9.601755042587624e-07,
+      "loss": 0.0972,
+      "step": 1459
+    },
+    {
+      "epoch": 7.121951219512195,
+      "grad_norm": 0.9732463955879211,
+      "learning_rate": 9.571590819145244e-07,
+      "loss": 0.0476,
+      "step": 1460
+    },
+    {
+      "epoch": 7.126829268292683,
+      "grad_norm": 1.284700632095337,
+      "learning_rate": 9.541462829436426e-07,
+      "loss": 0.1645,
+      "step": 1461
+    },
+    {
+      "epoch": 7.131707317073171,
+      "grad_norm": 1.2691322565078735,
+      "learning_rate": 9.511371144217005e-07,
+      "loss": 0.1717,
+      "step": 1462
+    },
+    {
+      "epoch": 7.136585365853659,
+      "grad_norm": 1.72730553150177,
+      "learning_rate": 9.481315834157512e-07,
+      "loss": 0.2492,
+      "step": 1463
+    },
+    {
+      "epoch": 7.1414634146341465,
+      "grad_norm": 1.1432853937149048,
+      "learning_rate": 9.451296969843058e-07,
+      "loss": 0.143,
+      "step": 1464
+    },
+    {
+      "epoch": 7.146341463414634,
+      "grad_norm": 1.0376734733581543,
+      "learning_rate": 9.42131462177319e-07,
+      "loss": 0.1186,
+      "step": 1465
+    },
+    {
+      "epoch": 7.151219512195122,
+      "grad_norm": 1.0639046430587769,
+      "learning_rate": 9.39136886036166e-07,
+      "loss": 0.1073,
+      "step": 1466
+    },
+    {
+      "epoch": 7.15609756097561,
+      "grad_norm": 0.8768961429595947,
+      "learning_rate": 9.361459755936316e-07,
+      "loss": 0.0595,
+      "step": 1467
+    },
+    {
+      "epoch": 7.160975609756098,
+      "grad_norm": 0.8885585069656372,
+      "learning_rate": 9.331587378738902e-07,
+      "loss": 0.1289,
+      "step": 1468
+    },
+    {
+      "epoch": 7.1658536585365855,
+      "grad_norm": 2.7977395057678223,
+      "learning_rate": 9.301751798924935e-07,
+      "loss": 0.1762,
+      "step": 1469
+    },
+    {
+      "epoch": 7.170731707317073,
+      "grad_norm": 0.7919803857803345,
+      "learning_rate": 9.27195308656349e-07,
+      "loss": 0.0919,
+      "step": 1470
+    },
+    {
+      "epoch": 7.175609756097561,
+      "grad_norm": 1.2924004793167114,
+      "learning_rate": 9.24219131163705e-07,
+      "loss": 0.2055,
+      "step": 1471
+    },
+    {
+      "epoch": 7.180487804878049,
+      "grad_norm": 0.9636824727058411,
+      "learning_rate": 9.212466544041385e-07,
+      "loss": 0.0541,
+      "step": 1472
+    },
+    {
+      "epoch": 7.185365853658537,
+      "grad_norm": 1.1815683841705322,
+      "learning_rate": 9.182778853585325e-07,
+      "loss": 0.1983,
+      "step": 1473
+    },
+    {
+      "epoch": 7.190243902439025,
+      "grad_norm": 1.063944935798645,
+      "learning_rate": 9.153128309990622e-07,
+      "loss": 0.19,
+      "step": 1474
+    },
+    {
+      "epoch": 7.195121951219512,
+      "grad_norm": 1.3285342454910278,
+      "learning_rate": 9.123514982891813e-07,
+      "loss": 0.1505,
+      "step": 1475
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.9774877429008484,
+      "learning_rate": 9.093938941836012e-07,
+      "loss": 0.1066,
+      "step": 1476
+    },
+    {
+      "epoch": 7.204878048780488,
+      "grad_norm": 0.9382181763648987,
+      "learning_rate": 9.064400256282757e-07,
+      "loss": 0.0863,
+      "step": 1477
+    },
+    {
+      "epoch": 7.209756097560976,
+      "grad_norm": 1.0966013669967651,
+      "learning_rate": 9.034898995603894e-07,
+      "loss": 0.1533,
+      "step": 1478
+    },
+    {
+      "epoch": 7.214634146341464,
+      "grad_norm": 0.8254204392433167,
+      "learning_rate": 9.00543522908334e-07,
+      "loss": 0.0917,
+      "step": 1479
+    },
+    {
+      "epoch": 7.219512195121951,
+      "grad_norm": 0.9348810315132141,
+      "learning_rate": 8.976009025916962e-07,
+      "loss": 0.065,
+      "step": 1480
+    },
+    {
+      "epoch": 7.224390243902439,
+      "grad_norm": 1.166670560836792,
+      "learning_rate": 8.946620455212438e-07,
+      "loss": 0.0884,
+      "step": 1481
+    },
+    {
+      "epoch": 7.229268292682927,
+      "grad_norm": 1.373329758644104,
+      "learning_rate": 8.917269585989027e-07,
+      "loss": 0.3353,
+      "step": 1482
+    },
+    {
+      "epoch": 7.234146341463415,
+      "grad_norm": 0.8657915592193604,
+      "learning_rate": 8.887956487177462e-07,
+      "loss": 0.0897,
+      "step": 1483
+    },
+    {
+      "epoch": 7.239024390243903,
+      "grad_norm": 0.8597664833068848,
+      "learning_rate": 8.858681227619789e-07,
+      "loss": 0.1181,
+      "step": 1484
+    },
+    {
+      "epoch": 7.2439024390243905,
+      "grad_norm": 0.839962899684906,
+      "learning_rate": 8.829443876069163e-07,
+      "loss": 0.0599,
+      "step": 1485
+    },
+    {
+      "epoch": 7.248780487804878,
+      "grad_norm": 0.8345217108726501,
+      "learning_rate": 8.800244501189722e-07,
+      "loss": 0.0387,
+      "step": 1486
+    },
+    {
+      "epoch": 7.253658536585366,
+      "grad_norm": 1.267763614654541,
+      "learning_rate": 8.771083171556407e-07,
+      "loss": 0.2695,
+      "step": 1487
+    },
+    {
+      "epoch": 7.258536585365854,
+      "grad_norm": 0.8838726878166199,
+      "learning_rate": 8.741959955654833e-07,
+      "loss": 0.0838,
+      "step": 1488
+    },
+    {
+      "epoch": 7.263414634146342,
+      "grad_norm": 0.7908660173416138,
+      "learning_rate": 8.712874921881082e-07,
+      "loss": 0.1355,
+      "step": 1489
+    },
+    {
+      "epoch": 7.2682926829268295,
+      "grad_norm": 1.2876062393188477,
+      "learning_rate": 8.683828138541559e-07,
+      "loss": 0.3281,
+      "step": 1490
+    },
+    {
+      "epoch": 7.273170731707317,
+      "grad_norm": 1.1086186170578003,
+      "learning_rate": 8.654819673852874e-07,
+      "loss": 0.1662,
+      "step": 1491
+    },
+    {
+      "epoch": 7.278048780487805,
+      "grad_norm": 0.7914115190505981,
+      "learning_rate": 8.625849595941608e-07,
+      "loss": 0.0565,
+      "step": 1492
+    },
+    {
+      "epoch": 7.282926829268293,
+      "grad_norm": 0.6734137535095215,
+      "learning_rate": 8.596917972844199e-07,
+      "loss": 0.0334,
+      "step": 1493
+    },
+    {
+      "epoch": 7.287804878048781,
+      "grad_norm": 0.9427139163017273,
+      "learning_rate": 8.568024872506792e-07,
+      "loss": 0.1858,
+      "step": 1494
+    },
+    {
+      "epoch": 7.2926829268292686,
+      "grad_norm": 0.8241920471191406,
+      "learning_rate": 8.539170362785043e-07,
+      "loss": 0.0419,
+      "step": 1495
+    },
+    {
+      "epoch": 7.297560975609756,
+      "grad_norm": 0.8029516935348511,
+      "learning_rate": 8.510354511443975e-07,
+      "loss": 0.0793,
+      "step": 1496
+    },
+    {
+      "epoch": 7.302439024390244,
+      "grad_norm": 0.7739546895027161,
+      "learning_rate": 8.48157738615784e-07,
+      "loss": 0.0646,
+      "step": 1497
+    },
+    {
+      "epoch": 7.307317073170732,
+      "grad_norm": 0.6402938961982727,
+      "learning_rate": 8.452839054509926e-07,
+      "loss": 0.0686,
+      "step": 1498
+    },
+    {
+      "epoch": 7.31219512195122,
+      "grad_norm": 2.102001667022705,
+      "learning_rate": 8.42413958399241e-07,
+      "loss": 0.089,
+      "step": 1499
+    },
+    {
+      "epoch": 7.317073170731708,
+      "grad_norm": 0.9000923037528992,
+      "learning_rate": 8.39547904200623e-07,
+      "loss": 0.0704,
+      "step": 1500
+    },
+    {
+      "epoch": 7.321951219512195,
+      "grad_norm": 0.7872455716133118,
+      "learning_rate": 8.366857495860869e-07,
+      "loss": 0.0642,
+      "step": 1501
+    },
+    {
+      "epoch": 7.326829268292683,
+      "grad_norm": 1.1488183736801147,
+      "learning_rate": 8.338275012774247e-07,
+      "loss": 0.0603,
+      "step": 1502
+    },
+    {
+      "epoch": 7.331707317073171,
+      "grad_norm": 0.929627001285553,
+      "learning_rate": 8.309731659872522e-07,
+      "loss": 0.1138,
+      "step": 1503
+    },
+    {
+      "epoch": 7.336585365853659,
+      "grad_norm": 0.9145927429199219,
+      "learning_rate": 8.281227504189992e-07,
+      "loss": 0.1613,
+      "step": 1504
+    },
+    {
+      "epoch": 7.341463414634147,
+      "grad_norm": 0.7761052846908569,
+      "learning_rate": 8.252762612668869e-07,
+      "loss": 0.0843,
+      "step": 1505
+    },
+    {
+      "epoch": 7.3463414634146345,
+      "grad_norm": 1.2363475561141968,
+      "learning_rate": 8.224337052159154e-07,
+      "loss": 0.0786,
+      "step": 1506
+    },
+    {
+      "epoch": 7.351219512195122,
+      "grad_norm": 1.2809274196624756,
+      "learning_rate": 8.195950889418503e-07,
+      "loss": 0.1432,
+      "step": 1507
+    },
+    {
+      "epoch": 7.35609756097561,
+      "grad_norm": 1.4146350622177124,
+      "learning_rate": 8.167604191112021e-07,
+      "loss": 0.2068,
+      "step": 1508
+    },
+    {
+      "epoch": 7.360975609756098,
+      "grad_norm": 1.1985394954681396,
+      "learning_rate": 8.139297023812131e-07,
+      "loss": 0.1531,
+      "step": 1509
+    },
+    {
+      "epoch": 7.365853658536586,
+      "grad_norm": 0.9208254218101501,
+      "learning_rate": 8.111029453998448e-07,
+      "loss": 0.0809,
+      "step": 1510
+    },
+    {
+      "epoch": 7.3707317073170735,
+      "grad_norm": 1.5169329643249512,
+      "learning_rate": 8.082801548057553e-07,
+      "loss": 0.122,
+      "step": 1511
+    },
+    {
+      "epoch": 7.375609756097561,
+      "grad_norm": 0.702643632888794,
+      "learning_rate": 8.05461337228289e-07,
+      "loss": 0.0567,
+      "step": 1512
+    },
+    {
+      "epoch": 7.380487804878049,
+      "grad_norm": 0.833108127117157,
+      "learning_rate": 8.026464992874617e-07,
+      "loss": 0.0791,
+      "step": 1513
+    },
+    {
+      "epoch": 7.385365853658537,
+      "grad_norm": 0.5722957253456116,
+      "learning_rate": 7.998356475939398e-07,
+      "loss": 0.0135,
+      "step": 1514
+    },
+    {
+      "epoch": 7.390243902439025,
+      "grad_norm": 0.6496143341064453,
+      "learning_rate": 7.970287887490289e-07,
+      "loss": 0.0336,
+      "step": 1515
+    },
+    {
+      "epoch": 7.3951219512195125,
+      "grad_norm": 0.5775876641273499,
+      "learning_rate": 7.942259293446594e-07,
+      "loss": 0.0192,
+      "step": 1516
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.8442699909210205,
+      "learning_rate": 7.914270759633669e-07,
+      "loss": 0.073,
+      "step": 1517
+    },
+    {
+      "epoch": 7.404878048780488,
+      "grad_norm": 0.5718724131584167,
+      "learning_rate": 7.886322351782782e-07,
+      "loss": 0.0201,
+      "step": 1518
+    },
+    {
+      "epoch": 7.409756097560976,
+      "grad_norm": 1.0626249313354492,
+      "learning_rate": 7.858414135530995e-07,
+      "loss": 0.138,
+      "step": 1519
+    },
+    {
+      "epoch": 7.414634146341464,
+      "grad_norm": 1.07315993309021,
+      "learning_rate": 7.83054617642095e-07,
+      "loss": 0.092,
+      "step": 1520
+    },
+    {
+      "epoch": 7.419512195121952,
+      "grad_norm": 1.1001923084259033,
+      "learning_rate": 7.802718539900761e-07,
+      "loss": 0.1063,
+      "step": 1521
+    },
+    {
+      "epoch": 7.424390243902439,
+      "grad_norm": 0.6773379445075989,
+      "learning_rate": 7.774931291323826e-07,
+      "loss": 0.0586,
+      "step": 1522
+    },
+    {
+      "epoch": 7.429268292682927,
+      "grad_norm": 1.2498868703842163,
+      "learning_rate": 7.747184495948723e-07,
+      "loss": 0.3014,
+      "step": 1523
+    },
+    {
+      "epoch": 7.434146341463415,
+      "grad_norm": 1.6956262588500977,
+      "learning_rate": 7.719478218939e-07,
+      "loss": 0.1645,
+      "step": 1524
+    },
+    {
+      "epoch": 7.439024390243903,
+      "grad_norm": 0.733025848865509,
+      "learning_rate": 7.691812525363044e-07,
+      "loss": 0.0463,
+      "step": 1525
+    },
+    {
+      "epoch": 7.443902439024391,
+      "grad_norm": 0.8861591219902039,
+      "learning_rate": 7.66418748019396e-07,
+      "loss": 0.0616,
+      "step": 1526
+    },
+    {
+      "epoch": 7.4487804878048784,
+      "grad_norm": 0.6299625039100647,
+      "learning_rate": 7.636603148309363e-07,
+      "loss": 0.0386,
+      "step": 1527
+    },
+    {
+      "epoch": 7.453658536585366,
+      "grad_norm": 1.0957764387130737,
+      "learning_rate": 7.609059594491253e-07,
+      "loss": 0.1015,
+      "step": 1528
+    },
+    {
+      "epoch": 7.458536585365854,
+      "grad_norm": 1.3109453916549683,
+      "learning_rate": 7.581556883425886e-07,
+      "loss": 0.3966,
+      "step": 1529
+    },
+    {
+      "epoch": 7.463414634146342,
+      "grad_norm": 1.3401238918304443,
+      "learning_rate": 7.55409507970358e-07,
+      "loss": 0.2687,
+      "step": 1530
+    },
+    {
+      "epoch": 7.46829268292683,
+      "grad_norm": 0.6099500060081482,
+      "learning_rate": 7.526674247818569e-07,
+      "loss": 0.0355,
+      "step": 1531
+    },
+    {
+      "epoch": 7.473170731707317,
+      "grad_norm": 1.170320987701416,
+      "learning_rate": 7.499294452168904e-07,
+      "loss": 0.1753,
+      "step": 1532
+    },
+    {
+      "epoch": 7.478048780487805,
+      "grad_norm": 0.9354384541511536,
+      "learning_rate": 7.471955757056227e-07,
+      "loss": 0.0812,
+      "step": 1533
+    },
+    {
+      "epoch": 7.482926829268292,
+      "grad_norm": 0.9029474854469299,
+      "learning_rate": 7.444658226685656e-07,
+      "loss": 0.055,
+      "step": 1534
+    },
+    {
+      "epoch": 7.487804878048781,
+      "grad_norm": 1.1110976934432983,
+      "learning_rate": 7.417401925165666e-07,
+      "loss": 0.1035,
+      "step": 1535
+    },
+    {
+      "epoch": 7.492682926829268,
+      "grad_norm": 0.7902478575706482,
+      "learning_rate": 7.390186916507869e-07,
+      "loss": 0.0745,
+      "step": 1536
+    },
+    {
+      "epoch": 7.4975609756097565,
+      "grad_norm": 0.5375012159347534,
+      "learning_rate": 7.363013264626914e-07,
+      "loss": 0.0206,
+      "step": 1537
+    },
+    {
+      "epoch": 7.5024390243902435,
+      "grad_norm": 1.016384243965149,
+      "learning_rate": 7.335881033340334e-07,
+      "loss": 0.1809,
+      "step": 1538
+    },
+    {
+      "epoch": 7.507317073170732,
+      "grad_norm": 1.161800742149353,
+      "learning_rate": 7.308790286368373e-07,
+      "loss": 0.1462,
+      "step": 1539
+    },
+    {
+      "epoch": 7.512195121951219,
+      "grad_norm": 1.2205479145050049,
+      "learning_rate": 7.281741087333846e-07,
+      "loss": 0.2044,
+      "step": 1540
+    },
+    {
+      "epoch": 7.517073170731708,
+      "grad_norm": 1.619023323059082,
+      "learning_rate": 7.254733499761993e-07,
+      "loss": 0.1539,
+      "step": 1541
+    },
+    {
+      "epoch": 7.521951219512195,
+      "grad_norm": 1.1716769933700562,
+      "learning_rate": 7.22776758708035e-07,
+      "loss": 0.2405,
+      "step": 1542
+    },
+    {
+      "epoch": 7.526829268292683,
+      "grad_norm": 1.1175401210784912,
+      "learning_rate": 7.200843412618555e-07,
+      "loss": 0.1471,
+      "step": 1543
+    },
+    {
+      "epoch": 7.53170731707317,
+      "grad_norm": 0.9348238706588745,
+      "learning_rate": 7.173961039608227e-07,
+      "loss": 0.0995,
+      "step": 1544
+    },
+    {
+      "epoch": 7.536585365853659,
+      "grad_norm": 0.9532445669174194,
+      "learning_rate": 7.147120531182828e-07,
+      "loss": 0.1762,
+      "step": 1545
+    },
+    {
+      "epoch": 7.541463414634146,
+      "grad_norm": 0.821707010269165,
+      "learning_rate": 7.120321950377487e-07,
+      "loss": 0.0363,
+      "step": 1546
+    },
+    {
+      "epoch": 7.546341463414635,
+      "grad_norm": 0.8796511292457581,
+      "learning_rate": 7.093565360128863e-07,
+      "loss": 0.0725,
+      "step": 1547
+    },
+    {
+      "epoch": 7.5512195121951216,
+      "grad_norm": 0.9664599299430847,
+      "learning_rate": 7.066850823275024e-07,
+      "loss": 0.098,
+      "step": 1548
+    },
+    {
+      "epoch": 7.55609756097561,
+      "grad_norm": 1.4284769296646118,
+      "learning_rate": 7.040178402555245e-07,
+      "loss": 0.0962,
+      "step": 1549
+    },
+    {
+      "epoch": 7.560975609756097,
+      "grad_norm": 0.8095406293869019,
+      "learning_rate": 7.013548160609901e-07,
+      "loss": 0.0726,
+      "step": 1550
+    },
+    {
+      "epoch": 7.565853658536585,
+      "grad_norm": 0.920695960521698,
+      "learning_rate": 6.986960159980327e-07,
+      "loss": 0.1442,
+      "step": 1551
+    },
+    {
+      "epoch": 7.570731707317073,
+      "grad_norm": 1.2659902572631836,
+      "learning_rate": 6.960414463108631e-07,
+      "loss": 0.2081,
+      "step": 1552
+    },
+    {
+      "epoch": 7.575609756097561,
+      "grad_norm": 0.9657885432243347,
+      "learning_rate": 6.933911132337575e-07,
+      "loss": 0.0971,
+      "step": 1553
+    },
+    {
+      "epoch": 7.580487804878048,
+      "grad_norm": 0.8895794153213501,
+      "learning_rate": 6.907450229910443e-07,
+      "loss": 0.161,
+      "step": 1554
+    },
+    {
+      "epoch": 7.585365853658536,
+      "grad_norm": 0.7729442119598389,
+      "learning_rate": 6.881031817970848e-07,
+      "loss": 0.0582,
+      "step": 1555
+    },
+    {
+      "epoch": 7.590243902439024,
+      "grad_norm": 1.2856751680374146,
+      "learning_rate": 6.854655958562625e-07,
+      "loss": 0.3543,
+      "step": 1556
+    },
+    {
+      "epoch": 7.595121951219512,
+      "grad_norm": 0.7980842590332031,
+      "learning_rate": 6.82832271362969e-07,
+      "loss": 0.0522,
+      "step": 1557
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.8392421007156372,
+      "learning_rate": 6.802032145015855e-07,
+      "loss": 0.0934,
+      "step": 1558
+    },
+    {
+      "epoch": 7.6048780487804875,
+      "grad_norm": 1.5677391290664673,
+      "learning_rate": 6.775784314464717e-07,
+      "loss": 0.1563,
+      "step": 1559
+    },
+    {
+      "epoch": 7.609756097560975,
+      "grad_norm": 0.9202232956886292,
+      "learning_rate": 6.749579283619492e-07,
+      "loss": 0.1134,
+      "step": 1560
+    },
+    {
+      "epoch": 7.614634146341463,
+      "grad_norm": 0.761900007724762,
+      "learning_rate": 6.723417114022907e-07,
+      "loss": 0.0782,
+      "step": 1561
+    },
+    {
+      "epoch": 7.619512195121951,
+      "grad_norm": 1.2099084854125977,
+      "learning_rate": 6.697297867117e-07,
+      "loss": 0.1507,
+      "step": 1562
+    },
+    {
+      "epoch": 7.624390243902439,
+      "grad_norm": 1.1177504062652588,
+      "learning_rate": 6.671221604243014e-07,
+      "loss": 0.1404,
+      "step": 1563
+    },
+    {
+      "epoch": 7.6292682926829265,
+      "grad_norm": 1.752407431602478,
+      "learning_rate": 6.645188386641257e-07,
+      "loss": 0.3834,
+      "step": 1564
+    },
+    {
+      "epoch": 7.634146341463414,
+      "grad_norm": 0.5573866367340088,
+      "learning_rate": 6.61919827545093e-07,
+      "loss": 0.0228,
+      "step": 1565
+    },
+    {
+      "epoch": 7.639024390243902,
+      "grad_norm": 1.4471404552459717,
+      "learning_rate": 6.593251331709993e-07,
+      "loss": 0.3744,
+      "step": 1566
+    },
+    {
+      "epoch": 7.64390243902439,
+      "grad_norm": 1.831053376197815,
+      "learning_rate": 6.567347616355049e-07,
+      "loss": 0.3253,
+      "step": 1567
+    },
+    {
+      "epoch": 7.648780487804878,
+      "grad_norm": 0.6713041663169861,
+      "learning_rate": 6.541487190221163e-07,
+      "loss": 0.0339,
+      "step": 1568
+    },
+    {
+      "epoch": 7.6536585365853655,
+      "grad_norm": 0.6553516387939453,
+      "learning_rate": 6.515670114041725e-07,
+      "loss": 0.0308,
+      "step": 1569
+    },
+    {
+      "epoch": 7.658536585365853,
+      "grad_norm": 0.8251855373382568,
+      "learning_rate": 6.489896448448349e-07,
+      "loss": 0.0938,
+      "step": 1570
+    },
+    {
+      "epoch": 7.663414634146341,
+      "grad_norm": 1.358507752418518,
+      "learning_rate": 6.464166253970672e-07,
+      "loss": 0.2063,
+      "step": 1571
+    },
+    {
+      "epoch": 7.668292682926829,
+      "grad_norm": 0.5880959033966064,
+      "learning_rate": 6.43847959103624e-07,
+      "loss": 0.0353,
+      "step": 1572
+    },
+    {
+      "epoch": 7.673170731707317,
+      "grad_norm": 1.2899762392044067,
+      "learning_rate": 6.412836519970383e-07,
+      "loss": 0.1629,
+      "step": 1573
+    },
+    {
+      "epoch": 7.678048780487805,
+      "grad_norm": 0.645285427570343,
+      "learning_rate": 6.387237100996041e-07,
+      "loss": 0.029,
+      "step": 1574
+    },
+    {
+      "epoch": 7.682926829268292,
+      "grad_norm": 0.8590375185012817,
+      "learning_rate": 6.361681394233631e-07,
+      "loss": 0.1296,
+      "step": 1575
+    },
+    {
+      "epoch": 7.68780487804878,
+      "grad_norm": 0.7912958264350891,
+      "learning_rate": 6.336169459700933e-07,
+      "loss": 0.0744,
+      "step": 1576
+    },
+    {
+      "epoch": 7.692682926829268,
+      "grad_norm": 1.0610671043395996,
+      "learning_rate": 6.310701357312909e-07,
+      "loss": 0.1001,
+      "step": 1577
+    },
+    {
+      "epoch": 7.697560975609756,
+      "grad_norm": 0.9412058591842651,
+      "learning_rate": 6.285277146881588e-07,
+      "loss": 0.045,
+      "step": 1578
+    },
+    {
+      "epoch": 7.702439024390244,
+      "grad_norm": 0.9780194759368896,
+      "learning_rate": 6.259896888115904e-07,
+      "loss": 0.1099,
+      "step": 1579
+    },
+    {
+      "epoch": 7.7073170731707314,
+      "grad_norm": 1.0854240655899048,
+      "learning_rate": 6.234560640621606e-07,
+      "loss": 0.1211,
+      "step": 1580
+    },
+    {
+      "epoch": 7.712195121951219,
+      "grad_norm": 0.7268218994140625,
+      "learning_rate": 6.209268463901047e-07,
+      "loss": 0.029,
+      "step": 1581
+    },
+    {
+      "epoch": 7.717073170731707,
+      "grad_norm": 0.802086353302002,
+      "learning_rate": 6.184020417353084e-07,
+      "loss": 0.0714,
+      "step": 1582
+    },
+    {
+      "epoch": 7.721951219512195,
+      "grad_norm": 1.4602060317993164,
+      "learning_rate": 6.158816560272962e-07,
+      "loss": 0.2754,
+      "step": 1583
+    },
+    {
+      "epoch": 7.726829268292683,
+      "grad_norm": 1.6787465810775757,
+      "learning_rate": 6.133656951852113e-07,
+      "loss": 0.3757,
+      "step": 1584
+    },
+    {
+      "epoch": 7.7317073170731705,
+      "grad_norm": 0.8407995700836182,
+      "learning_rate": 6.10854165117806e-07,
+      "loss": 0.0656,
+      "step": 1585
+    },
+    {
+      "epoch": 7.736585365853658,
+      "grad_norm": 0.863117516040802,
+      "learning_rate": 6.083470717234285e-07,
+      "loss": 0.059,
+      "step": 1586
+    },
+    {
+      "epoch": 7.741463414634146,
+      "grad_norm": 0.8231815099716187,
+      "learning_rate": 6.058444208900061e-07,
+      "loss": 0.0688,
+      "step": 1587
+    },
+    {
+      "epoch": 7.746341463414634,
+      "grad_norm": 0.5683103203773499,
+      "learning_rate": 6.033462184950317e-07,
+      "loss": 0.0206,
+      "step": 1588
+    },
+    {
+      "epoch": 7.751219512195122,
+      "grad_norm": 0.9569870233535767,
+      "learning_rate": 6.008524704055535e-07,
+      "loss": 0.0697,
+      "step": 1589
+    },
+    {
+      "epoch": 7.7560975609756095,
+      "grad_norm": 0.93723064661026,
+      "learning_rate": 5.983631824781572e-07,
+      "loss": 0.1364,
+      "step": 1590
+    },
+    {
+      "epoch": 7.760975609756097,
+      "grad_norm": 0.9307323098182678,
+      "learning_rate": 5.95878360558953e-07,
+      "loss": 0.0888,
+      "step": 1591
+    },
+    {
+      "epoch": 7.765853658536585,
+      "grad_norm": 0.9898451566696167,
+      "learning_rate": 5.933980104835652e-07,
+      "loss": 0.1927,
+      "step": 1592
+    },
+    {
+      "epoch": 7.770731707317073,
+      "grad_norm": 1.1425596475601196,
+      "learning_rate": 5.909221380771132e-07,
+      "loss": 0.1798,
+      "step": 1593
+    },
+    {
+      "epoch": 7.775609756097561,
+      "grad_norm": 1.2097892761230469,
+      "learning_rate": 5.884507491542024e-07,
+      "loss": 0.167,
+      "step": 1594
+    },
+    {
+      "epoch": 7.780487804878049,
+      "grad_norm": 0.7511147856712341,
+      "learning_rate": 5.859838495189068e-07,
+      "loss": 0.052,
+      "step": 1595
+    },
+    {
+      "epoch": 7.785365853658536,
+      "grad_norm": 1.0986175537109375,
+      "learning_rate": 5.835214449647602e-07,
+      "loss": 0.1336,
+      "step": 1596
+    },
+    {
+      "epoch": 7.790243902439024,
+      "grad_norm": 0.8146430850028992,
+      "learning_rate": 5.810635412747373e-07,
+      "loss": 0.0657,
+      "step": 1597
+    },
+    {
+      "epoch": 7.795121951219512,
+      "grad_norm": 0.6946802735328674,
+      "learning_rate": 5.786101442212422e-07,
+      "loss": 0.0503,
+      "step": 1598
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 1.2120875120162964,
+      "learning_rate": 5.761612595660979e-07,
+      "loss": 0.2715,
+      "step": 1599
+    },
+    {
+      "epoch": 7.804878048780488,
+      "grad_norm": 0.6051833629608154,
+      "learning_rate": 5.737168930605272e-07,
+      "loss": 0.0212,
+      "step": 1600
+    },
+    {
+      "epoch": 7.809756097560975,
+      "grad_norm": 0.7571241855621338,
+      "learning_rate": 5.712770504451426e-07,
+      "loss": 0.0686,
+      "step": 1601
+    },
+    {
+      "epoch": 7.814634146341463,
+      "grad_norm": 1.5256556272506714,
+      "learning_rate": 5.688417374499336e-07,
+      "loss": 0.1827,
+      "step": 1602
+    },
+    {
+      "epoch": 7.819512195121951,
+      "grad_norm": 1.1259573698043823,
+      "learning_rate": 5.664109597942504e-07,
+      "loss": 0.0728,
+      "step": 1603
+    },
+    {
+      "epoch": 7.824390243902439,
+      "grad_norm": 1.527320146560669,
+      "learning_rate": 5.639847231867917e-07,
+      "loss": 0.3322,
+      "step": 1604
+    },
+    {
+      "epoch": 7.829268292682927,
+      "grad_norm": 1.0056272745132446,
+      "learning_rate": 5.61563033325594e-07,
+      "loss": 0.1727,
+      "step": 1605
+    },
+    {
+      "epoch": 7.8341463414634145,
+      "grad_norm": 1.0396225452423096,
+      "learning_rate": 5.591458958980123e-07,
+      "loss": 0.0781,
+      "step": 1606
+    },
+    {
+      "epoch": 7.839024390243902,
+      "grad_norm": 1.085661768913269,
+      "learning_rate": 5.567333165807115e-07,
+      "loss": 0.144,
+      "step": 1607
+    },
+    {
+      "epoch": 7.84390243902439,
+      "grad_norm": 0.5520138740539551,
+      "learning_rate": 5.543253010396538e-07,
+      "loss": 0.038,
+      "step": 1608
+    },
+    {
+      "epoch": 7.848780487804878,
+      "grad_norm": 1.343543291091919,
+      "learning_rate": 5.519218549300806e-07,
+      "loss": 0.1562,
+      "step": 1609
+    },
+    {
+      "epoch": 7.853658536585366,
+      "grad_norm": 1.2264282703399658,
+      "learning_rate": 5.495229838965021e-07,
+      "loss": 0.1926,
+      "step": 1610
+    },
+    {
+      "epoch": 7.8585365853658535,
+      "grad_norm": 0.854395866394043,
+      "learning_rate": 5.471286935726866e-07,
+      "loss": 0.0471,
+      "step": 1611
+    },
+    {
+      "epoch": 7.863414634146341,
+      "grad_norm": 1.2901803255081177,
+      "learning_rate": 5.447389895816416e-07,
+      "loss": 0.2489,
+      "step": 1612
+    },
+    {
+      "epoch": 7.868292682926829,
+      "grad_norm": 1.7009903192520142,
+      "learning_rate": 5.423538775356049e-07,
+      "loss": 0.335,
+      "step": 1613
+    },
+    {
+      "epoch": 7.873170731707317,
+      "grad_norm": 0.8392524719238281,
+      "learning_rate": 5.399733630360287e-07,
+      "loss": 0.0951,
+      "step": 1614
+    },
+    {
+      "epoch": 7.878048780487805,
+      "grad_norm": 1.0800468921661377,
+      "learning_rate": 5.375974516735713e-07,
+      "loss": 0.1884,
+      "step": 1615
+    },
+    {
+      "epoch": 7.882926829268293,
+      "grad_norm": 0.9787384271621704,
+      "learning_rate": 5.352261490280767e-07,
+      "loss": 0.0705,
+      "step": 1616
+    },
+    {
+      "epoch": 7.88780487804878,
+      "grad_norm": 1.0949972867965698,
+      "learning_rate": 5.328594606685661e-07,
+      "loss": 0.0589,
+      "step": 1617
+    },
+    {
+      "epoch": 7.892682926829268,
+      "grad_norm": 0.8980702757835388,
+      "learning_rate": 5.304973921532264e-07,
+      "loss": 0.1445,
+      "step": 1618
+    },
+    {
+      "epoch": 7.897560975609756,
+      "grad_norm": 0.945303738117218,
+      "learning_rate": 5.281399490293923e-07,
+      "loss": 0.0972,
+      "step": 1619
+    },
+    {
+      "epoch": 7.902439024390244,
+      "grad_norm": 1.1684542894363403,
+      "learning_rate": 5.257871368335357e-07,
+      "loss": 0.123,
+      "step": 1620
+    },
+    {
+      "epoch": 7.907317073170732,
+      "grad_norm": 0.9429066181182861,
+      "learning_rate": 5.234389610912552e-07,
+      "loss": 0.0194,
+      "step": 1621
+    },
+    {
+      "epoch": 7.912195121951219,
+      "grad_norm": 1.167583703994751,
+      "learning_rate": 5.210954273172578e-07,
+      "loss": 0.1536,
+      "step": 1622
+    },
+    {
+      "epoch": 7.917073170731707,
+      "grad_norm": 0.9684360027313232,
+      "learning_rate": 5.187565410153497e-07,
+      "loss": 0.1505,
+      "step": 1623
+    },
+    {
+      "epoch": 7.921951219512195,
+      "grad_norm": 1.0447014570236206,
+      "learning_rate": 5.164223076784239e-07,
+      "loss": 0.1055,
+      "step": 1624
+    },
+    {
+      "epoch": 7.926829268292683,
+      "grad_norm": 1.1968408823013306,
+      "learning_rate": 5.14092732788444e-07,
+      "loss": 0.2419,
+      "step": 1625
+    },
+    {
+      "epoch": 7.931707317073171,
+      "grad_norm": 1.3599355220794678,
+      "learning_rate": 5.117678218164337e-07,
+      "loss": 0.0725,
+      "step": 1626
+    },
+    {
+      "epoch": 7.9365853658536585,
+      "grad_norm": 0.838928759098053,
+      "learning_rate": 5.094475802224644e-07,
+      "loss": 0.1036,
+      "step": 1627
+    },
+    {
+      "epoch": 7.941463414634146,
+      "grad_norm": 0.7429474592208862,
+      "learning_rate": 5.071320134556404e-07,
+      "loss": 0.0374,
+      "step": 1628
+    },
+    {
+      "epoch": 7.946341463414634,
+      "grad_norm": 0.8248745203018188,
+      "learning_rate": 5.048211269540868e-07,
+      "loss": 0.0732,
+      "step": 1629
+    },
+    {
+      "epoch": 7.951219512195122,
+      "grad_norm": 0.979279100894928,
+      "learning_rate": 5.025149261449391e-07,
+      "loss": 0.1127,
+      "step": 1630
+    },
+    {
+      "epoch": 7.95609756097561,
+      "grad_norm": 1.0153231620788574,
+      "learning_rate": 5.002134164443262e-07,
+      "loss": 0.0743,
+      "step": 1631
+    },
+    {
+      "epoch": 7.9609756097560975,
+      "grad_norm": 0.6215126514434814,
+      "learning_rate": 4.979166032573607e-07,
+      "loss": 0.0202,
+      "step": 1632
+    },
+    {
+      "epoch": 7.965853658536585,
+      "grad_norm": 1.2080820798873901,
+      "learning_rate": 4.956244919781247e-07,
+      "loss": 0.2897,
+      "step": 1633
+    },
+    {
+      "epoch": 7.970731707317073,
+      "grad_norm": 0.7493009567260742,
+      "learning_rate": 4.933370879896604e-07,
+      "loss": 0.0412,
+      "step": 1634
+    },
+    {
+      "epoch": 7.975609756097561,
+      "grad_norm": 0.9477584958076477,
+      "learning_rate": 4.91054396663952e-07,
+      "loss": 0.0618,
+      "step": 1635
+    },
+    {
+      "epoch": 7.980487804878049,
+      "grad_norm": 1.0992841720581055,
+      "learning_rate": 4.887764233619163e-07,
+      "loss": 0.0868,
+      "step": 1636
+    },
+    {
+      "epoch": 7.985365853658537,
+      "grad_norm": 0.7600829601287842,
+      "learning_rate": 4.865031734333919e-07,
+      "loss": 0.0511,
+      "step": 1637
+    },
+    {
+      "epoch": 7.990243902439024,
+      "grad_norm": 1.106069564819336,
+      "learning_rate": 4.842346522171226e-07,
+      "loss": 0.0927,
+      "step": 1638
+    },
+    {
+      "epoch": 7.995121951219512,
+      "grad_norm": 0.8861885070800781,
+      "learning_rate": 4.819708650407467e-07,
+      "loss": 0.1585,
+      "step": 1639
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.8401235938072205,
+      "learning_rate": 4.797118172207863e-07,
+      "loss": 0.0683,
+      "step": 1640
+    },
+    {
+      "epoch": 8.004878048780487,
+      "grad_norm": 1.089004635810852,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.1339,
+      "step": 1641
+    },
+    {
+      "epoch": 8.009756097560976,
+      "grad_norm": 1.0382062196731567,
+      "learning_rate": 4.752079608605295e-07,
+      "loss": 0.1023,
+      "step": 1642
+    },
+    {
+      "epoch": 8.014634146341463,
+      "grad_norm": 1.0378601551055908,
+      "learning_rate": 4.7296316289757366e-07,
+      "loss": 0.079,
+      "step": 1643
+    },
+    {
+      "epoch": 8.019512195121951,
+      "grad_norm": 0.7355173230171204,
+      "learning_rate": 4.7072312544568844e-07,
+      "loss": 0.0726,
+      "step": 1644
+    },
+    {
+      "epoch": 8.024390243902438,
+      "grad_norm": 0.7902737855911255,
+      "learning_rate": 4.6848785376561733e-07,
+      "loss": 0.0807,
+      "step": 1645
+    },
+    {
+      "epoch": 8.029268292682927,
+      "grad_norm": 0.6139519214630127,
+      "learning_rate": 4.6625735310691396e-07,
+      "loss": 0.0585,
+      "step": 1646
+    },
+    {
+      "epoch": 8.034146341463414,
+      "grad_norm": 0.684639036655426,
+      "learning_rate": 4.6403162870792524e-07,
+      "loss": 0.0715,
+      "step": 1647
+    },
+    {
+      "epoch": 8.039024390243902,
+      "grad_norm": 0.8094519972801208,
+      "learning_rate": 4.618106857957805e-07,
+      "loss": 0.1042,
+      "step": 1648
+    },
+    {
+      "epoch": 8.04390243902439,
+      "grad_norm": 0.4592023193836212,
+      "learning_rate": 4.5959452958638213e-07,
+      "loss": 0.0178,
+      "step": 1649
+    },
+    {
+      "epoch": 8.048780487804878,
+      "grad_norm": 0.47528451681137085,
+      "learning_rate": 4.573831652843888e-07,
+      "loss": 0.0199,
+      "step": 1650
+    },
+    {
+      "epoch": 8.053658536585365,
+      "grad_norm": 1.361240029335022,
+      "learning_rate": 4.55176598083206e-07,
+      "loss": 0.3006,
+      "step": 1651
+    },
+    {
+      "epoch": 8.058536585365854,
+      "grad_norm": 0.9548918604850769,
+      "learning_rate": 4.5297483316497276e-07,
+      "loss": 0.1356,
+      "step": 1652
+    },
+    {
+      "epoch": 8.06341463414634,
+      "grad_norm": 1.2566801309585571,
+      "learning_rate": 4.5077787570055097e-07,
+      "loss": 0.2039,
+      "step": 1653
+    },
+    {
+      "epoch": 8.06829268292683,
+      "grad_norm": 1.6015655994415283,
+      "learning_rate": 4.4858573084951173e-07,
+      "loss": 0.3508,
+      "step": 1654
+    },
+    {
+      "epoch": 8.073170731707316,
+      "grad_norm": 0.6489564776420593,
+      "learning_rate": 4.463984037601224e-07,
+      "loss": 0.04,
+      "step": 1655
+    },
+    {
+      "epoch": 8.078048780487805,
+      "grad_norm": 0.8055024147033691,
+      "learning_rate": 4.4421589956933827e-07,
+      "loss": 0.0503,
+      "step": 1656
+    },
+    {
+      "epoch": 8.082926829268292,
+      "grad_norm": 0.9149098992347717,
+      "learning_rate": 4.420382234027859e-07,
+      "loss": 0.0886,
+      "step": 1657
+    },
+    {
+      "epoch": 8.08780487804878,
+      "grad_norm": 0.6378782391548157,
+      "learning_rate": 4.398653803747532e-07,
+      "loss": 0.0531,
+      "step": 1658
+    },
+    {
+      "epoch": 8.092682926829267,
+      "grad_norm": 0.40032660961151123,
+      "learning_rate": 4.3769737558817996e-07,
+      "loss": 0.007,
+      "step": 1659
+    },
+    {
+      "epoch": 8.097560975609756,
+      "grad_norm": 1.2026172876358032,
+      "learning_rate": 4.355342141346405e-07,
+      "loss": 0.19,
+      "step": 1660
+    },
+    {
+      "epoch": 8.102439024390243,
+      "grad_norm": 0.5329281687736511,
+      "learning_rate": 4.3337590109433505e-07,
+      "loss": 0.0363,
+      "step": 1661
+    },
+    {
+      "epoch": 8.107317073170732,
+      "grad_norm": 0.8976152539253235,
+      "learning_rate": 4.3122244153607914e-07,
+      "loss": 0.0478,
+      "step": 1662
+    },
+    {
+      "epoch": 8.112195121951219,
+      "grad_norm": 0.7638136744499207,
+      "learning_rate": 4.2907384051728754e-07,
+      "loss": 0.1372,
+      "step": 1663
+    },
+    {
+      "epoch": 8.117073170731707,
+      "grad_norm": 0.9107884764671326,
+      "learning_rate": 4.2693010308396566e-07,
+      "loss": 0.0855,
+      "step": 1664
+    },
+    {
+      "epoch": 8.121951219512194,
+      "grad_norm": 0.896466076374054,
+      "learning_rate": 4.247912342706975e-07,
+      "loss": 0.0725,
+      "step": 1665
+    },
+    {
+      "epoch": 8.126829268292683,
+      "grad_norm": 0.837365984916687,
+      "learning_rate": 4.22657239100632e-07,
+      "loss": 0.057,
+      "step": 1666
+    },
+    {
+      "epoch": 8.13170731707317,
+      "grad_norm": 0.9322727918624878,
+      "learning_rate": 4.2052812258547265e-07,
+      "loss": 0.0762,
+      "step": 1667
+    },
+    {
+      "epoch": 8.136585365853659,
+      "grad_norm": 0.5818950533866882,
+      "learning_rate": 4.184038897254655e-07,
+      "loss": 0.0331,
+      "step": 1668
+    },
+    {
+      "epoch": 8.141463414634146,
+      "grad_norm": 1.5296962261199951,
+      "learning_rate": 4.1628454550938697e-07,
+      "loss": 0.304,
+      "step": 1669
+    },
+    {
+      "epoch": 8.146341463414634,
+      "grad_norm": 1.2551158666610718,
+      "learning_rate": 4.141700949145322e-07,
+      "loss": 0.2691,
+      "step": 1670
+    },
+    {
+      "epoch": 8.151219512195121,
+      "grad_norm": 0.9186856150627136,
+      "learning_rate": 4.1206054290670537e-07,
+      "loss": 0.1649,
+      "step": 1671
+    },
+    {
+      "epoch": 8.15609756097561,
+      "grad_norm": 1.2951780557632446,
+      "learning_rate": 4.0995589444020433e-07,
+      "loss": 0.2459,
+      "step": 1672
+    },
+    {
+      "epoch": 8.160975609756097,
+      "grad_norm": 1.0069646835327148,
+      "learning_rate": 4.0785615445781106e-07,
+      "loss": 0.1252,
+      "step": 1673
+    },
+    {
+      "epoch": 8.165853658536586,
+      "grad_norm": 1.1953704357147217,
+      "learning_rate": 4.057613278907818e-07,
+      "loss": 0.1741,
+      "step": 1674
+    },
+    {
+      "epoch": 8.170731707317072,
+      "grad_norm": 1.102447748184204,
+      "learning_rate": 4.036714196588318e-07,
+      "loss": 0.0362,
+      "step": 1675
+    },
+    {
+      "epoch": 8.175609756097561,
+      "grad_norm": 0.892144501209259,
+      "learning_rate": 4.015864346701251e-07,
+      "loss": 0.0607,
+      "step": 1676
+    },
+    {
+      "epoch": 8.180487804878048,
+      "grad_norm": 0.531515896320343,
+      "learning_rate": 3.99506377821266e-07,
+      "loss": 0.0129,
+      "step": 1677
+    },
+    {
+      "epoch": 8.185365853658537,
+      "grad_norm": 0.8729956150054932,
+      "learning_rate": 3.97431253997283e-07,
+      "loss": 0.052,
+      "step": 1678
+    },
+    {
+      "epoch": 8.190243902439024,
+      "grad_norm": 0.63507080078125,
+      "learning_rate": 3.9536106807161857e-07,
+      "loss": 0.0365,
+      "step": 1679
+    },
+    {
+      "epoch": 8.195121951219512,
+      "grad_norm": 0.7739672064781189,
+      "learning_rate": 3.932958249061214e-07,
+      "loss": 0.05,
+      "step": 1680
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.7827106714248657,
+      "learning_rate": 3.9123552935102976e-07,
+      "loss": 0.0492,
+      "step": 1681
+    },
+    {
+      "epoch": 8.204878048780488,
+      "grad_norm": 1.0987051725387573,
+      "learning_rate": 3.891801862449629e-07,
+      "loss": 0.1276,
+      "step": 1682
+    },
+    {
+      "epoch": 8.209756097560975,
+      "grad_norm": 0.8080296516418457,
+      "learning_rate": 3.8712980041490905e-07,
+      "loss": 0.0743,
+      "step": 1683
+    },
+    {
+      "epoch": 8.214634146341464,
+      "grad_norm": 0.8196918964385986,
+      "learning_rate": 3.850843766762155e-07,
+      "loss": 0.08,
+      "step": 1684
+    },
+    {
+      "epoch": 8.21951219512195,
+      "grad_norm": 0.5697962641716003,
+      "learning_rate": 3.830439198325747e-07,
+      "loss": 0.0241,
+      "step": 1685
+    },
+    {
+      "epoch": 8.22439024390244,
+      "grad_norm": 0.5778628587722778,
+      "learning_rate": 3.81008434676014e-07,
+      "loss": 0.0433,
+      "step": 1686
+    },
+    {
+      "epoch": 8.229268292682926,
+      "grad_norm": 0.8138613104820251,
+      "learning_rate": 3.789779259868864e-07,
+      "loss": 0.042,
+      "step": 1687
+    },
+    {
+      "epoch": 8.234146341463415,
+      "grad_norm": 1.0728322267532349,
+      "learning_rate": 3.769523985338566e-07,
+      "loss": 0.1505,
+      "step": 1688
+    },
+    {
+      "epoch": 8.239024390243902,
+      "grad_norm": 0.9158832430839539,
+      "learning_rate": 3.749318570738897e-07,
+      "loss": 0.1193,
+      "step": 1689
+    },
+    {
+      "epoch": 8.24390243902439,
+      "grad_norm": 0.8731831908226013,
+      "learning_rate": 3.7291630635224397e-07,
+      "loss": 0.0953,
+      "step": 1690
+    },
+    {
+      "epoch": 8.248780487804877,
+      "grad_norm": 0.8154242634773254,
+      "learning_rate": 3.709057511024541e-07,
+      "loss": 0.1017,
+      "step": 1691
+    },
+    {
+      "epoch": 8.253658536585366,
+      "grad_norm": 0.870970606803894,
+      "learning_rate": 3.689001960463243e-07,
+      "loss": 0.0561,
+      "step": 1692
+    },
+    {
+      "epoch": 8.258536585365853,
+      "grad_norm": 0.7816699147224426,
+      "learning_rate": 3.668996458939156e-07,
+      "loss": 0.054,
+      "step": 1693
+    },
+    {
+      "epoch": 8.263414634146342,
+      "grad_norm": 0.8930821418762207,
+      "learning_rate": 3.649041053435354e-07,
+      "loss": 0.0871,
+      "step": 1694
+    },
+    {
+      "epoch": 8.268292682926829,
+      "grad_norm": 0.4522107243537903,
+      "learning_rate": 3.62913579081724e-07,
+      "loss": 0.0201,
+      "step": 1695
+    },
+    {
+      "epoch": 8.273170731707317,
+      "grad_norm": 1.1511738300323486,
+      "learning_rate": 3.609280717832489e-07,
+      "loss": 0.1985,
+      "step": 1696
+    },
+    {
+      "epoch": 8.278048780487804,
+      "grad_norm": 0.6477423310279846,
+      "learning_rate": 3.5894758811108795e-07,
+      "loss": 0.0346,
+      "step": 1697
+    },
+    {
+      "epoch": 8.282926829268293,
+      "grad_norm": 0.5493444204330444,
+      "learning_rate": 3.5697213271642164e-07,
+      "loss": 0.029,
+      "step": 1698
+    },
+    {
+      "epoch": 8.28780487804878,
+      "grad_norm": 0.7595999240875244,
+      "learning_rate": 3.5500171023862136e-07,
+      "loss": 0.0569,
+      "step": 1699
+    },
+    {
+      "epoch": 8.292682926829269,
+      "grad_norm": 1.1431372165679932,
+      "learning_rate": 3.530363253052399e-07,
+      "loss": 0.1001,
+      "step": 1700
+    },
+    {
+      "epoch": 8.297560975609755,
+      "grad_norm": 1.1419728994369507,
+      "learning_rate": 3.510759825319976e-07,
+      "loss": 0.1865,
+      "step": 1701
+    },
+    {
+      "epoch": 8.302439024390244,
+      "grad_norm": 1.0514655113220215,
+      "learning_rate": 3.491206865227739e-07,
+      "loss": 0.085,
+      "step": 1702
+    },
+    {
+      "epoch": 8.307317073170731,
+      "grad_norm": 1.9193140268325806,
+      "learning_rate": 3.4717044186959676e-07,
+      "loss": 0.2772,
+      "step": 1703
+    },
+    {
+      "epoch": 8.31219512195122,
+      "grad_norm": 1.0900132656097412,
+      "learning_rate": 3.452252531526301e-07,
+      "loss": 0.1614,
+      "step": 1704
+    },
+    {
+      "epoch": 8.317073170731707,
+      "grad_norm": 1.0719544887542725,
+      "learning_rate": 3.432851249401628e-07,
+      "loss": 0.1899,
+      "step": 1705
+    },
+    {
+      "epoch": 8.321951219512195,
+      "grad_norm": 0.6381435990333557,
+      "learning_rate": 3.413500617886023e-07,
+      "loss": 0.0417,
+      "step": 1706
+    },
+    {
+      "epoch": 8.326829268292682,
+      "grad_norm": 1.7609957456588745,
+      "learning_rate": 3.394200682424578e-07,
+      "loss": 0.0614,
+      "step": 1707
+    },
+    {
+      "epoch": 8.331707317073171,
+      "grad_norm": 0.9982513189315796,
+      "learning_rate": 3.374951488343328e-07,
+      "loss": 0.1509,
+      "step": 1708
+    },
+    {
+      "epoch": 8.336585365853658,
+      "grad_norm": 0.41625937819480896,
+      "learning_rate": 3.355753080849164e-07,
+      "loss": 0.0206,
+      "step": 1709
+    },
+    {
+      "epoch": 8.341463414634147,
+      "grad_norm": 0.6567705869674683,
+      "learning_rate": 3.3366055050296776e-07,
+      "loss": 0.0374,
+      "step": 1710
+    },
+    {
+      "epoch": 8.346341463414634,
+      "grad_norm": 0.6061232686042786,
+      "learning_rate": 3.3175088058530925e-07,
+      "loss": 0.0354,
+      "step": 1711
+    },
+    {
+      "epoch": 8.351219512195122,
+      "grad_norm": 1.4496053457260132,
+      "learning_rate": 3.2984630281681556e-07,
+      "loss": 0.1438,
+      "step": 1712
+    },
+    {
+      "epoch": 8.35609756097561,
+      "grad_norm": 0.7385711669921875,
+      "learning_rate": 3.2794682167040125e-07,
+      "loss": 0.0759,
+      "step": 1713
+    },
+    {
+      "epoch": 8.360975609756098,
+      "grad_norm": 0.9990761280059814,
+      "learning_rate": 3.2605244160701155e-07,
+      "loss": 0.2105,
+      "step": 1714
+    },
+    {
+      "epoch": 8.365853658536585,
+      "grad_norm": 0.8232401013374329,
+      "learning_rate": 3.2416316707561316e-07,
+      "loss": 0.0997,
+      "step": 1715
+    },
+    {
+      "epoch": 8.370731707317074,
+      "grad_norm": 0.8837026357650757,
+      "learning_rate": 3.2227900251318055e-07,
+      "loss": 0.0905,
+      "step": 1716
+    },
+    {
+      "epoch": 8.37560975609756,
+      "grad_norm": 0.7026121020317078,
+      "learning_rate": 3.2039995234468854e-07,
+      "loss": 0.0235,
+      "step": 1717
+    },
+    {
+      "epoch": 8.38048780487805,
+      "grad_norm": 0.24160675704479218,
+      "learning_rate": 3.1852602098309984e-07,
+      "loss": 0.0083,
+      "step": 1718
+    },
+    {
+      "epoch": 8.385365853658536,
+      "grad_norm": 0.7885335683822632,
+      "learning_rate": 3.1665721282935683e-07,
+      "loss": 0.0823,
+      "step": 1719
+    },
+    {
+      "epoch": 8.390243902439025,
+      "grad_norm": 0.6479135751724243,
+      "learning_rate": 3.147935322723694e-07,
+      "loss": 0.0456,
+      "step": 1720
+    },
+    {
+      "epoch": 8.395121951219512,
+      "grad_norm": 1.1197822093963623,
+      "learning_rate": 3.1293498368900414e-07,
+      "loss": 0.0558,
+      "step": 1721
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.5560716986656189,
+      "learning_rate": 3.1108157144407765e-07,
+      "loss": 0.0218,
+      "step": 1722
+    },
+    {
+      "epoch": 8.404878048780487,
+      "grad_norm": 0.954613983631134,
+      "learning_rate": 3.092332998903416e-07,
+      "loss": 0.1019,
+      "step": 1723
+    },
+    {
+      "epoch": 8.409756097560976,
+      "grad_norm": 0.9680867791175842,
+      "learning_rate": 3.073901733684748e-07,
+      "loss": 0.1924,
+      "step": 1724
+    },
+    {
+      "epoch": 8.414634146341463,
+      "grad_norm": 0.7575647830963135,
+      "learning_rate": 3.055521962070751e-07,
+      "loss": 0.1066,
+      "step": 1725
+    },
+    {
+      "epoch": 8.419512195121952,
+      "grad_norm": 1.102574110031128,
+      "learning_rate": 3.0371937272264454e-07,
+      "loss": 0.0838,
+      "step": 1726
+    },
+    {
+      "epoch": 8.424390243902439,
+      "grad_norm": 0.576019823551178,
+      "learning_rate": 3.0189170721958234e-07,
+      "loss": 0.0299,
+      "step": 1727
+    },
+    {
+      "epoch": 8.429268292682927,
+      "grad_norm": 0.9913473725318909,
+      "learning_rate": 3.000692039901756e-07,
+      "loss": 0.1202,
+      "step": 1728
+    },
+    {
+      "epoch": 8.434146341463414,
+      "grad_norm": 1.163718819618225,
+      "learning_rate": 2.982518673145862e-07,
+      "loss": 0.2031,
+      "step": 1729
+    },
+    {
+      "epoch": 8.439024390243903,
+      "grad_norm": 0.6798272728919983,
+      "learning_rate": 2.9643970146084193e-07,
+      "loss": 0.0429,
+      "step": 1730
+    },
+    {
+      "epoch": 8.44390243902439,
+      "grad_norm": 1.6693507432937622,
+      "learning_rate": 2.9463271068482955e-07,
+      "loss": 0.2449,
+      "step": 1731
+    },
+    {
+      "epoch": 8.448780487804878,
+      "grad_norm": 0.6570068001747131,
+      "learning_rate": 2.928308992302792e-07,
+      "loss": 0.0668,
+      "step": 1732
+    },
+    {
+      "epoch": 8.453658536585365,
+      "grad_norm": 0.926525890827179,
+      "learning_rate": 2.9103427132875785e-07,
+      "loss": 0.0891,
+      "step": 1733
+    },
+    {
+      "epoch": 8.458536585365854,
+      "grad_norm": 1.095659852027893,
+      "learning_rate": 2.892428311996609e-07,
+      "loss": 0.1997,
+      "step": 1734
+    },
+    {
+      "epoch": 8.463414634146341,
+      "grad_norm": 0.9572970867156982,
+      "learning_rate": 2.8745658305019824e-07,
+      "loss": 0.1081,
+      "step": 1735
+    },
+    {
+      "epoch": 8.46829268292683,
+      "grad_norm": 0.7929927706718445,
+      "learning_rate": 2.856755310753867e-07,
+      "loss": 0.0515,
+      "step": 1736
+    },
+    {
+      "epoch": 8.473170731707317,
+      "grad_norm": 0.8479703664779663,
+      "learning_rate": 2.8389967945803984e-07,
+      "loss": 0.0624,
+      "step": 1737
+    },
+    {
+      "epoch": 8.478048780487805,
+      "grad_norm": 0.5886752009391785,
+      "learning_rate": 2.821290323687592e-07,
+      "loss": 0.0297,
+      "step": 1738
+    },
+    {
+      "epoch": 8.482926829268292,
+      "grad_norm": 0.7627184987068176,
+      "learning_rate": 2.803635939659222e-07,
+      "loss": 0.067,
+      "step": 1739
+    },
+    {
+      "epoch": 8.487804878048781,
+      "grad_norm": 1.017645001411438,
+      "learning_rate": 2.786033683956732e-07,
+      "loss": 0.1739,
+      "step": 1740
+    },
+    {
+      "epoch": 8.492682926829268,
+      "grad_norm": 1.4689565896987915,
+      "learning_rate": 2.7684835979191664e-07,
+      "loss": 0.4571,
+      "step": 1741
+    },
+    {
+      "epoch": 8.497560975609757,
+      "grad_norm": 0.554760217666626,
+      "learning_rate": 2.7509857227630223e-07,
+      "loss": 0.0359,
+      "step": 1742
+    },
+    {
+      "epoch": 8.502439024390243,
+      "grad_norm": 0.9449183344841003,
+      "learning_rate": 2.733540099582188e-07,
+      "loss": 0.1137,
+      "step": 1743
+    },
+    {
+      "epoch": 8.507317073170732,
+      "grad_norm": 1.2766910791397095,
+      "learning_rate": 2.7161467693478493e-07,
+      "loss": 0.2038,
+      "step": 1744
+    },
+    {
+      "epoch": 8.512195121951219,
+      "grad_norm": 1.0461146831512451,
+      "learning_rate": 2.6988057729083613e-07,
+      "loss": 0.1233,
+      "step": 1745
+    },
+    {
+      "epoch": 8.517073170731708,
+      "grad_norm": 0.8558463454246521,
+      "learning_rate": 2.681517150989185e-07,
+      "loss": 0.1122,
+      "step": 1746
+    },
+    {
+      "epoch": 8.521951219512195,
+      "grad_norm": 0.7195177674293518,
+      "learning_rate": 2.664280944192782e-07,
+      "loss": 0.0467,
+      "step": 1747
+    },
+    {
+      "epoch": 8.526829268292683,
+      "grad_norm": 0.927327573299408,
+      "learning_rate": 2.64709719299851e-07,
+      "loss": 0.0954,
+      "step": 1748
+    },
+    {
+      "epoch": 8.53170731707317,
+      "grad_norm": 0.5121705532073975,
+      "learning_rate": 2.6299659377625296e-07,
+      "loss": 0.028,
+      "step": 1749
+    },
+    {
+      "epoch": 8.536585365853659,
+      "grad_norm": 1.0753885507583618,
+      "learning_rate": 2.612887218717733e-07,
+      "loss": 0.2088,
+      "step": 1750
+    },
+    {
+      "epoch": 8.541463414634146,
+      "grad_norm": 0.6652311086654663,
+      "learning_rate": 2.5958610759736133e-07,
+      "loss": 0.0685,
+      "step": 1751
+    },
+    {
+      "epoch": 8.546341463414635,
+      "grad_norm": 0.4434679448604584,
+      "learning_rate": 2.5788875495161846e-07,
+      "loss": 0.0193,
+      "step": 1752
+    },
+    {
+      "epoch": 8.551219512195122,
+      "grad_norm": 0.7911991477012634,
+      "learning_rate": 2.561966679207917e-07,
+      "loss": 0.0968,
+      "step": 1753
+    },
+    {
+      "epoch": 8.55609756097561,
+      "grad_norm": 0.9316043853759766,
+      "learning_rate": 2.545098504787588e-07,
+      "loss": 0.1939,
+      "step": 1754
+    },
+    {
+      "epoch": 8.560975609756097,
+      "grad_norm": 0.48057541251182556,
+      "learning_rate": 2.5282830658702323e-07,
+      "loss": 0.0119,
+      "step": 1755
+    },
+    {
+      "epoch": 8.565853658536586,
+      "grad_norm": 0.8290639519691467,
+      "learning_rate": 2.511520401947032e-07,
+      "loss": 0.092,
+      "step": 1756
+    },
+    {
+      "epoch": 8.570731707317073,
+      "grad_norm": 0.6231652498245239,
+      "learning_rate": 2.494810552385232e-07,
+      "loss": 0.0357,
+      "step": 1757
+    },
+    {
+      "epoch": 8.575609756097561,
+      "grad_norm": 0.6425905227661133,
+      "learning_rate": 2.47815355642804e-07,
+      "loss": 0.0379,
+      "step": 1758
+    },
+    {
+      "epoch": 8.580487804878048,
+      "grad_norm": 0.9716305136680603,
+      "learning_rate": 2.461549453194523e-07,
+      "loss": 0.2117,
+      "step": 1759
+    },
+    {
+      "epoch": 8.585365853658537,
+      "grad_norm": 0.9776792526245117,
+      "learning_rate": 2.444998281679553e-07,
+      "loss": 0.2307,
+      "step": 1760
+    },
+    {
+      "epoch": 8.590243902439024,
+      "grad_norm": 0.9863300919532776,
+      "learning_rate": 2.428500080753676e-07,
+      "loss": 0.171,
+      "step": 1761
+    },
+    {
+      "epoch": 8.595121951219513,
+      "grad_norm": 0.47316545248031616,
+      "learning_rate": 2.412054889163035e-07,
+      "loss": 0.0159,
+      "step": 1762
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.7214704751968384,
+      "learning_rate": 2.3956627455292924e-07,
+      "loss": 0.0661,
+      "step": 1763
+    },
+    {
+      "epoch": 8.604878048780488,
+      "grad_norm": 0.6974095702171326,
+      "learning_rate": 2.3793236883495164e-07,
+      "loss": 0.034,
+      "step": 1764
+    },
+    {
+      "epoch": 8.609756097560975,
+      "grad_norm": 0.8801345229148865,
+      "learning_rate": 2.363037755996095e-07,
+      "loss": 0.0898,
+      "step": 1765
+    },
+    {
+      "epoch": 8.614634146341464,
+      "grad_norm": 1.1015623807907104,
+      "learning_rate": 2.3468049867166747e-07,
+      "loss": 0.1701,
+      "step": 1766
+    },
+    {
+      "epoch": 8.61951219512195,
+      "grad_norm": 0.8162565231323242,
+      "learning_rate": 2.3306254186340305e-07,
+      "loss": 0.0569,
+      "step": 1767
+    },
+    {
+      "epoch": 8.62439024390244,
+      "grad_norm": 1.2020095586776733,
+      "learning_rate": 2.314499089745989e-07,
+      "loss": 0.2277,
+      "step": 1768
+    },
+    {
+      "epoch": 8.629268292682926,
+      "grad_norm": 1.1717698574066162,
+      "learning_rate": 2.2984260379253693e-07,
+      "loss": 0.3308,
+      "step": 1769
+    },
+    {
+      "epoch": 8.634146341463415,
+      "grad_norm": 0.6847605109214783,
+      "learning_rate": 2.2824063009198428e-07,
+      "loss": 0.0608,
+      "step": 1770
+    },
+    {
+      "epoch": 8.639024390243902,
+      "grad_norm": 0.7636549472808838,
+      "learning_rate": 2.2664399163518786e-07,
+      "loss": 0.0876,
+      "step": 1771
+    },
+    {
+      "epoch": 8.64390243902439,
+      "grad_norm": 1.0829453468322754,
+      "learning_rate": 2.25052692171866e-07,
+      "loss": 0.0987,
+      "step": 1772
+    },
+    {
+      "epoch": 8.648780487804878,
+      "grad_norm": 0.8043027520179749,
+      "learning_rate": 2.2346673543919645e-07,
+      "loss": 0.0682,
+      "step": 1773
+    },
+    {
+      "epoch": 8.653658536585366,
+      "grad_norm": 0.8870478868484497,
+      "learning_rate": 2.2188612516181067e-07,
+      "loss": 0.1148,
+      "step": 1774
+    },
+    {
+      "epoch": 8.658536585365853,
+      "grad_norm": 0.6538479924201965,
+      "learning_rate": 2.203108650517835e-07,
+      "loss": 0.0627,
+      "step": 1775
+    },
+    {
+      "epoch": 8.663414634146342,
+      "grad_norm": 1.0483466386795044,
+      "learning_rate": 2.1874095880862505e-07,
+      "loss": 0.0897,
+      "step": 1776
+    },
+    {
+      "epoch": 8.668292682926829,
+      "grad_norm": 0.8733198642730713,
+      "learning_rate": 2.171764101192722e-07,
+      "loss": 0.027,
+      "step": 1777
+    },
+    {
+      "epoch": 8.673170731707318,
+      "grad_norm": 0.6368119716644287,
+      "learning_rate": 2.1561722265807827e-07,
+      "loss": 0.0322,
+      "step": 1778
+    },
+    {
+      "epoch": 8.678048780487805,
+      "grad_norm": 0.7868894338607788,
+      "learning_rate": 2.1406340008680748e-07,
+      "loss": 0.0512,
+      "step": 1779
+    },
+    {
+      "epoch": 8.682926829268293,
+      "grad_norm": 0.8871926069259644,
+      "learning_rate": 2.1251494605462358e-07,
+      "loss": 0.0725,
+      "step": 1780
+    },
+    {
+      "epoch": 8.68780487804878,
+      "grad_norm": 0.7351980209350586,
+      "learning_rate": 2.1097186419808151e-07,
+      "loss": 0.0316,
+      "step": 1781
+    },
+    {
+      "epoch": 8.692682926829269,
+      "grad_norm": 0.7175747156143188,
+      "learning_rate": 2.094341581411216e-07,
+      "loss": 0.0619,
+      "step": 1782
+    },
+    {
+      "epoch": 8.697560975609756,
+      "grad_norm": 0.5884055495262146,
+      "learning_rate": 2.0790183149505733e-07,
+      "loss": 0.0401,
+      "step": 1783
+    },
+    {
+      "epoch": 8.702439024390245,
+      "grad_norm": 0.8539692163467407,
+      "learning_rate": 2.063748878585689e-07,
+      "loss": 0.0745,
+      "step": 1784
+    },
+    {
+      "epoch": 8.707317073170731,
+      "grad_norm": 0.7265436053276062,
+      "learning_rate": 2.0485333081769588e-07,
+      "loss": 0.0546,
+      "step": 1785
+    },
+    {
+      "epoch": 8.71219512195122,
+      "grad_norm": 1.1087415218353271,
+      "learning_rate": 2.0333716394582536e-07,
+      "loss": 0.2784,
+      "step": 1786
+    },
+    {
+      "epoch": 8.717073170731707,
+      "grad_norm": 1.1902683973312378,
+      "learning_rate": 2.0182639080368634e-07,
+      "loss": 0.2145,
+      "step": 1787
+    },
+    {
+      "epoch": 8.721951219512196,
+      "grad_norm": 0.5206601023674011,
+      "learning_rate": 2.003210149393417e-07,
+      "loss": 0.035,
+      "step": 1788
+    },
+    {
+      "epoch": 8.726829268292683,
+      "grad_norm": 1.3805254697799683,
+      "learning_rate": 1.9882103988817735e-07,
+      "loss": 0.2117,
+      "step": 1789
+    },
+    {
+      "epoch": 8.731707317073171,
+      "grad_norm": 0.6062557697296143,
+      "learning_rate": 1.9732646917289545e-07,
+      "loss": 0.0575,
+      "step": 1790
+    },
+    {
+      "epoch": 8.736585365853658,
+      "grad_norm": 0.5106092691421509,
+      "learning_rate": 1.958373063035071e-07,
+      "loss": 0.0334,
+      "step": 1791
+    },
+    {
+      "epoch": 8.741463414634147,
+      "grad_norm": 1.0108541250228882,
+      "learning_rate": 1.9435355477732205e-07,
+      "loss": 0.1085,
+      "step": 1792
+    },
+    {
+      "epoch": 8.746341463414634,
+      "grad_norm": 0.6246529817581177,
+      "learning_rate": 1.928752180789417e-07,
+      "loss": 0.0593,
+      "step": 1793
+    },
+    {
+      "epoch": 8.751219512195123,
+      "grad_norm": 0.8448231220245361,
+      "learning_rate": 1.9140229968025058e-07,
+      "loss": 0.1019,
+      "step": 1794
+    },
+    {
+      "epoch": 8.75609756097561,
+      "grad_norm": 1.2745120525360107,
+      "learning_rate": 1.8993480304040912e-07,
+      "loss": 0.2614,
+      "step": 1795
+    },
+    {
+      "epoch": 8.760975609756098,
+      "grad_norm": 0.8491898775100708,
+      "learning_rate": 1.8847273160584378e-07,
+      "loss": 0.1049,
+      "step": 1796
+    },
+    {
+      "epoch": 8.765853658536585,
+      "grad_norm": 0.8189971446990967,
+      "learning_rate": 1.8701608881023957e-07,
+      "loss": 0.0441,
+      "step": 1797
+    },
+    {
+      "epoch": 8.770731707317074,
+      "grad_norm": 0.949904203414917,
+      "learning_rate": 1.855648780745342e-07,
+      "loss": 0.0516,
+      "step": 1798
+    },
+    {
+      "epoch": 8.77560975609756,
+      "grad_norm": 1.0584059953689575,
+      "learning_rate": 1.8411910280690588e-07,
+      "loss": 0.0636,
+      "step": 1799
+    },
+    {
+      "epoch": 8.78048780487805,
+      "grad_norm": 1.1480365991592407,
+      "learning_rate": 1.826787664027685e-07,
+      "loss": 0.1341,
+      "step": 1800
+    },
+    {
+      "epoch": 8.785365853658536,
+      "grad_norm": 0.8752851486206055,
+      "learning_rate": 1.8124387224476347e-07,
+      "loss": 0.1199,
+      "step": 1801
+    },
+    {
+      "epoch": 8.790243902439025,
+      "grad_norm": 0.6299836039543152,
+      "learning_rate": 1.7981442370274993e-07,
+      "loss": 0.0307,
+      "step": 1802
+    },
+    {
+      "epoch": 8.795121951219512,
+      "grad_norm": 1.1905652284622192,
+      "learning_rate": 1.783904241337983e-07,
+      "loss": 0.1483,
+      "step": 1803
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 1.9020190238952637,
+      "learning_rate": 1.7697187688218291e-07,
+      "loss": 0.0777,
+      "step": 1804
+    },
+    {
+      "epoch": 8.804878048780488,
+      "grad_norm": 0.6106300950050354,
+      "learning_rate": 1.7555878527937164e-07,
+      "loss": 0.0282,
+      "step": 1805
+    },
+    {
+      "epoch": 8.809756097560976,
+      "grad_norm": 0.8685973286628723,
+      "learning_rate": 1.7415115264402065e-07,
+      "loss": 0.125,
+      "step": 1806
+    },
+    {
+      "epoch": 8.814634146341463,
+      "grad_norm": 0.9931190013885498,
+      "learning_rate": 1.727489822819664e-07,
+      "loss": 0.0508,
+      "step": 1807
+    },
+    {
+      "epoch": 8.819512195121952,
+      "grad_norm": 0.5819611549377441,
+      "learning_rate": 1.7135227748621585e-07,
+      "loss": 0.0263,
+      "step": 1808
+    },
+    {
+      "epoch": 8.824390243902439,
+      "grad_norm": 1.1228915452957153,
+      "learning_rate": 1.699610415369407e-07,
+      "loss": 0.1991,
+      "step": 1809
+    },
+    {
+      "epoch": 8.829268292682928,
+      "grad_norm": 0.8106056451797485,
+      "learning_rate": 1.6857527770146876e-07,
+      "loss": 0.0627,
+      "step": 1810
+    },
+    {
+      "epoch": 8.834146341463414,
+      "grad_norm": 0.8355074524879456,
+      "learning_rate": 1.6719498923427697e-07,
+      "loss": 0.1007,
+      "step": 1811
+    },
+    {
+      "epoch": 8.839024390243903,
+      "grad_norm": 0.9377742409706116,
+      "learning_rate": 1.6582017937698287e-07,
+      "loss": 0.1788,
+      "step": 1812
+    },
+    {
+      "epoch": 8.84390243902439,
+      "grad_norm": 0.4790495038032532,
+      "learning_rate": 1.6445085135833732e-07,
+      "loss": 0.0259,
+      "step": 1813
+    },
+    {
+      "epoch": 8.848780487804879,
+      "grad_norm": 1.1777715682983398,
+      "learning_rate": 1.6308700839421793e-07,
+      "loss": 0.0931,
+      "step": 1814
+    },
+    {
+      "epoch": 8.853658536585366,
+      "grad_norm": 0.8541386127471924,
+      "learning_rate": 1.6172865368762004e-07,
+      "loss": 0.0733,
+      "step": 1815
+    },
+    {
+      "epoch": 8.858536585365854,
+      "grad_norm": 0.7406986355781555,
+      "learning_rate": 1.6037579042864876e-07,
+      "loss": 0.0353,
+      "step": 1816
+    },
+    {
+      "epoch": 8.863414634146341,
+      "grad_norm": 0.8613910675048828,
+      "learning_rate": 1.5902842179451482e-07,
+      "loss": 0.0849,
+      "step": 1817
+    },
+    {
+      "epoch": 8.86829268292683,
+      "grad_norm": 1.0476374626159668,
+      "learning_rate": 1.576865509495229e-07,
+      "loss": 0.1157,
+      "step": 1818
+    },
+    {
+      "epoch": 8.873170731707317,
+      "grad_norm": 1.2026175260543823,
+      "learning_rate": 1.5635018104506627e-07,
+      "loss": 0.1882,
+      "step": 1819
+    },
+    {
+      "epoch": 8.878048780487806,
+      "grad_norm": 0.8143272399902344,
+      "learning_rate": 1.5501931521962055e-07,
+      "loss": 0.0842,
+      "step": 1820
+    },
+    {
+      "epoch": 8.882926829268293,
+      "grad_norm": 0.9126399159431458,
+      "learning_rate": 1.5369395659873305e-07,
+      "loss": 0.0758,
+      "step": 1821
+    },
+    {
+      "epoch": 8.887804878048781,
+      "grad_norm": 0.6952263116836548,
+      "learning_rate": 1.5237410829501864e-07,
+      "loss": 0.0885,
+      "step": 1822
+    },
+    {
+      "epoch": 8.892682926829268,
+      "grad_norm": 1.0045150518417358,
+      "learning_rate": 1.510597734081512e-07,
+      "loss": 0.1717,
+      "step": 1823
+    },
+    {
+      "epoch": 8.897560975609757,
+      "grad_norm": 0.6330220699310303,
+      "learning_rate": 1.497509550248555e-07,
+      "loss": 0.0391,
+      "step": 1824
+    },
+    {
+      "epoch": 8.902439024390244,
+      "grad_norm": 0.8720522522926331,
+      "learning_rate": 1.4844765621890135e-07,
+      "loss": 0.1003,
+      "step": 1825
+    },
+    {
+      "epoch": 8.907317073170733,
+      "grad_norm": 1.2669109106063843,
+      "learning_rate": 1.471498800510962e-07,
+      "loss": 0.12,
+      "step": 1826
+    },
+    {
+      "epoch": 8.91219512195122,
+      "grad_norm": 0.6498632431030273,
+      "learning_rate": 1.4585762956927624e-07,
+      "loss": 0.0545,
+      "step": 1827
+    },
+    {
+      "epoch": 8.917073170731708,
+      "grad_norm": 0.9127230644226074,
+      "learning_rate": 1.4457090780830185e-07,
+      "loss": 0.0715,
+      "step": 1828
+    },
+    {
+      "epoch": 8.921951219512195,
+      "grad_norm": 1.036094307899475,
+      "learning_rate": 1.432897177900483e-07,
+      "loss": 0.155,
+      "step": 1829
+    },
+    {
+      "epoch": 8.926829268292684,
+      "grad_norm": 1.3150917291641235,
+      "learning_rate": 1.4201406252340038e-07,
+      "loss": 0.1964,
+      "step": 1830
+    },
+    {
+      "epoch": 8.93170731707317,
+      "grad_norm": 0.9457276463508606,
+      "learning_rate": 1.407439450042433e-07,
+      "loss": 0.1162,
+      "step": 1831
+    },
+    {
+      "epoch": 8.93658536585366,
+      "grad_norm": 1.353790283203125,
+      "learning_rate": 1.3947936821545772e-07,
+      "loss": 0.0864,
+      "step": 1832
+    },
+    {
+      "epoch": 8.941463414634146,
+      "grad_norm": 0.9871190786361694,
+      "learning_rate": 1.3822033512691209e-07,
+      "loss": 0.0381,
+      "step": 1833
+    },
+    {
+      "epoch": 8.946341463414633,
+      "grad_norm": 0.956032395362854,
+      "learning_rate": 1.369668486954545e-07,
+      "loss": 0.0998,
+      "step": 1834
+    },
+    {
+      "epoch": 8.951219512195122,
+      "grad_norm": 0.660033643245697,
+      "learning_rate": 1.3571891186490687e-07,
+      "loss": 0.0296,
+      "step": 1835
+    },
+    {
+      "epoch": 8.95609756097561,
+      "grad_norm": 0.8637049794197083,
+      "learning_rate": 1.3447652756605894e-07,
+      "loss": 0.0962,
+      "step": 1836
+    },
+    {
+      "epoch": 8.960975609756098,
+      "grad_norm": 0.9472467303276062,
+      "learning_rate": 1.3323969871665897e-07,
+      "loss": 0.1026,
+      "step": 1837
+    },
+    {
+      "epoch": 8.965853658536584,
+      "grad_norm": 0.5771154165267944,
+      "learning_rate": 1.3200842822140818e-07,
+      "loss": 0.0225,
+      "step": 1838
+    },
+    {
+      "epoch": 8.970731707317073,
+      "grad_norm": 0.9080767631530762,
+      "learning_rate": 1.3078271897195572e-07,
+      "loss": 0.0729,
+      "step": 1839
+    },
+    {
+      "epoch": 8.975609756097562,
+      "grad_norm": 0.9956149458885193,
+      "learning_rate": 1.2956257384688807e-07,
+      "loss": 0.2064,
+      "step": 1840
+    },
+    {
+      "epoch": 8.980487804878049,
+      "grad_norm": 0.721361517906189,
+      "learning_rate": 1.283479957117248e-07,
+      "loss": 0.0659,
+      "step": 1841
+    },
+    {
+      "epoch": 8.985365853658536,
+      "grad_norm": 1.3004629611968994,
+      "learning_rate": 1.2713898741891244e-07,
+      "loss": 0.2073,
+      "step": 1842
+    },
+    {
+      "epoch": 8.990243902439024,
+      "grad_norm": 0.5133698582649231,
+      "learning_rate": 1.2593555180781591e-07,
+      "loss": 0.0136,
+      "step": 1843
+    },
+    {
+      "epoch": 8.995121951219513,
+      "grad_norm": 1.134415864944458,
+      "learning_rate": 1.2473769170471188e-07,
+      "loss": 0.2574,
+      "step": 1844
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.9587790369987488,
+      "learning_rate": 1.2354540992278452e-07,
+      "loss": 0.1117,
+      "step": 1845
+    },
+    {
+      "epoch": 9.004878048780487,
+      "grad_norm": 0.3712022304534912,
+      "learning_rate": 1.223587092621162e-07,
+      "loss": 0.0107,
+      "step": 1846
+    },
+    {
+      "epoch": 9.009756097560976,
+      "grad_norm": 1.515165090560913,
+      "learning_rate": 1.2117759250968225e-07,
+      "loss": 0.4207,
+      "step": 1847
+    },
+    {
+      "epoch": 9.014634146341463,
+      "grad_norm": 1.2430503368377686,
+      "learning_rate": 1.2000206243934358e-07,
+      "loss": 0.1597,
+      "step": 1848
+    },
+    {
+      "epoch": 9.019512195121951,
+      "grad_norm": 0.41518303751945496,
+      "learning_rate": 1.1883212181184212e-07,
+      "loss": 0.0265,
+      "step": 1849
+    },
+    {
+      "epoch": 9.024390243902438,
+      "grad_norm": 0.7908434867858887,
+      "learning_rate": 1.176677733747919e-07,
+      "loss": 0.1019,
+      "step": 1850
+    },
+    {
+      "epoch": 9.029268292682927,
+      "grad_norm": 0.682651698589325,
+      "learning_rate": 1.1650901986267365e-07,
+      "loss": 0.0484,
+      "step": 1851
+    },
+    {
+      "epoch": 9.034146341463414,
+      "grad_norm": 0.5793932676315308,
+      "learning_rate": 1.1535586399682885e-07,
+      "loss": 0.0438,
+      "step": 1852
+    },
+    {
+      "epoch": 9.039024390243902,
+      "grad_norm": 1.060425043106079,
+      "learning_rate": 1.1420830848545256e-07,
+      "loss": 0.0312,
+      "step": 1853
+    },
+    {
+      "epoch": 9.04390243902439,
+      "grad_norm": 0.7476591467857361,
+      "learning_rate": 1.1306635602358673e-07,
+      "loss": 0.1354,
+      "step": 1854
+    },
+    {
+      "epoch": 9.048780487804878,
+      "grad_norm": 0.8792941570281982,
+      "learning_rate": 1.1193000929311638e-07,
+      "loss": 0.1339,
+      "step": 1855
+    },
+    {
+      "epoch": 9.053658536585365,
+      "grad_norm": 1.3680088520050049,
+      "learning_rate": 1.1079927096275978e-07,
+      "loss": 0.2752,
+      "step": 1856
+    },
+    {
+      "epoch": 9.058536585365854,
+      "grad_norm": 0.8391672968864441,
+      "learning_rate": 1.0967414368806384e-07,
+      "loss": 0.0975,
+      "step": 1857
+    },
+    {
+      "epoch": 9.06341463414634,
+      "grad_norm": 0.4497470557689667,
+      "learning_rate": 1.0855463011139905e-07,
+      "loss": 0.0136,
+      "step": 1858
+    },
+    {
+      "epoch": 9.06829268292683,
+      "grad_norm": 0.724907636642456,
+      "learning_rate": 1.0744073286195089e-07,
+      "loss": 0.086,
+      "step": 1859
+    },
+    {
+      "epoch": 9.073170731707316,
+      "grad_norm": 1.0157090425491333,
+      "learning_rate": 1.0633245455571511e-07,
+      "loss": 0.1302,
+      "step": 1860
+    },
+    {
+      "epoch": 9.078048780487805,
+      "grad_norm": 0.6860391497612,
+      "learning_rate": 1.052297977954922e-07,
+      "loss": 0.0535,
+      "step": 1861
+    },
+    {
+      "epoch": 9.082926829268292,
+      "grad_norm": 0.3641066551208496,
+      "learning_rate": 1.0413276517087956e-07,
+      "loss": 0.015,
+      "step": 1862
+    },
+    {
+      "epoch": 9.08780487804878,
+      "grad_norm": 0.9463160037994385,
+      "learning_rate": 1.0304135925826603e-07,
+      "loss": 0.1115,
+      "step": 1863
+    },
+    {
+      "epoch": 9.092682926829267,
+      "grad_norm": 0.8962271213531494,
+      "learning_rate": 1.0195558262082683e-07,
+      "loss": 0.1314,
+      "step": 1864
+    },
+    {
+      "epoch": 9.097560975609756,
+      "grad_norm": 0.6248987913131714,
+      "learning_rate": 1.0087543780851666e-07,
+      "loss": 0.0501,
+      "step": 1865
+    },
+    {
+      "epoch": 9.102439024390243,
+      "grad_norm": 0.6669026613235474,
+      "learning_rate": 9.98009273580633e-08,
+      "loss": 0.0597,
+      "step": 1866
+    },
+    {
+      "epoch": 9.107317073170732,
+      "grad_norm": 0.3283768892288208,
+      "learning_rate": 9.87320537929623e-08,
+      "loss": 0.0083,
+      "step": 1867
+    },
+    {
+      "epoch": 9.112195121951219,
+      "grad_norm": 0.8086330890655518,
+      "learning_rate": 9.766881962347208e-08,
+      "loss": 0.0443,
+      "step": 1868
+    },
+    {
+      "epoch": 9.117073170731707,
+      "grad_norm": 1.0572131872177124,
+      "learning_rate": 9.661122734660521e-08,
+      "loss": 0.1289,
+      "step": 1869
+    },
+    {
+      "epoch": 9.121951219512194,
+      "grad_norm": 0.7925761938095093,
+      "learning_rate": 9.555927944612492e-08,
+      "loss": 0.0776,
+      "step": 1870
+    },
+    {
+      "epoch": 9.126829268292683,
+      "grad_norm": 1.1056650876998901,
+      "learning_rate": 9.451297839253915e-08,
+      "loss": 0.2814,
+      "step": 1871
+    },
+    {
+      "epoch": 9.13170731707317,
+      "grad_norm": 0.9038887619972229,
+      "learning_rate": 9.34723266430937e-08,
+      "loss": 0.1219,
+      "step": 1872
+    },
+    {
+      "epoch": 9.136585365853659,
+      "grad_norm": 1.1971497535705566,
+      "learning_rate": 9.243732664176636e-08,
+      "loss": 0.1139,
+      "step": 1873
+    },
+    {
+      "epoch": 9.141463414634146,
+      "grad_norm": 0.9862858057022095,
+      "learning_rate": 9.140798081926277e-08,
+      "loss": 0.1679,
+      "step": 1874
+    },
+    {
+      "epoch": 9.146341463414634,
+      "grad_norm": 0.4423198103904724,
+      "learning_rate": 9.03842915930095e-08,
+      "loss": 0.0311,
+      "step": 1875
+    },
+    {
+      "epoch": 9.151219512195121,
+      "grad_norm": 0.967447817325592,
+      "learning_rate": 8.936626136714754e-08,
+      "loss": 0.1105,
+      "step": 1876
+    },
+    {
+      "epoch": 9.15609756097561,
+      "grad_norm": 0.864639163017273,
+      "learning_rate": 8.835389253252918e-08,
+      "loss": 0.1232,
+      "step": 1877
+    },
+    {
+      "epoch": 9.160975609756097,
+      "grad_norm": 0.9332835674285889,
+      "learning_rate": 8.734718746670978e-08,
+      "loss": 0.2209,
+      "step": 1878
+    },
+    {
+      "epoch": 9.165853658536586,
+      "grad_norm": 0.6977441310882568,
+      "learning_rate": 8.634614853394341e-08,
+      "loss": 0.0373,
+      "step": 1879
+    },
+    {
+      "epoch": 9.170731707317072,
+      "grad_norm": 1.3148572444915771,
+      "learning_rate": 8.53507780851781e-08,
+      "loss": 0.1774,
+      "step": 1880
+    },
+    {
+      "epoch": 9.175609756097561,
+      "grad_norm": 1.3219395875930786,
+      "learning_rate": 8.436107845804842e-08,
+      "loss": 0.1643,
+      "step": 1881
+    },
+    {
+      "epoch": 9.180487804878048,
+      "grad_norm": 0.6054883599281311,
+      "learning_rate": 8.33770519768709e-08,
+      "loss": 0.0291,
+      "step": 1882
+    },
+    {
+      "epoch": 9.185365853658537,
+      "grad_norm": 0.9713771939277649,
+      "learning_rate": 8.239870095263974e-08,
+      "loss": 0.1288,
+      "step": 1883
+    },
+    {
+      "epoch": 9.190243902439024,
+      "grad_norm": 0.8614907264709473,
+      "learning_rate": 8.142602768301921e-08,
+      "loss": 0.0904,
+      "step": 1884
+    },
+    {
+      "epoch": 9.195121951219512,
+      "grad_norm": 1.0839091539382935,
+      "learning_rate": 8.045903445233982e-08,
+      "loss": 0.1746,
+      "step": 1885
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 0.9268199801445007,
+      "learning_rate": 7.949772353159191e-08,
+      "loss": 0.1187,
+      "step": 1886
+    },
+    {
+      "epoch": 9.204878048780488,
+      "grad_norm": 0.6150853633880615,
+      "learning_rate": 7.854209717842231e-08,
+      "loss": 0.0338,
+      "step": 1887
+    },
+    {
+      "epoch": 9.209756097560975,
+      "grad_norm": 0.6899910569190979,
+      "learning_rate": 7.759215763712579e-08,
+      "loss": 0.0408,
+      "step": 1888
+    },
+    {
+      "epoch": 9.214634146341464,
+      "grad_norm": 0.813095211982727,
+      "learning_rate": 7.664790713864223e-08,
+      "loss": 0.1579,
+      "step": 1889
+    },
+    {
+      "epoch": 9.21951219512195,
+      "grad_norm": 1.0909091234207153,
+      "learning_rate": 7.57093479005519e-08,
+      "loss": 0.2084,
+      "step": 1890
+    },
+    {
+      "epoch": 9.22439024390244,
+      "grad_norm": 0.5029876828193665,
+      "learning_rate": 7.477648212706746e-08,
+      "loss": 0.0214,
+      "step": 1891
+    },
+    {
+      "epoch": 9.229268292682926,
+      "grad_norm": 0.6630915403366089,
+      "learning_rate": 7.384931200903084e-08,
+      "loss": 0.0792,
+      "step": 1892
+    },
+    {
+      "epoch": 9.234146341463415,
+      "grad_norm": 0.7246444821357727,
+      "learning_rate": 7.29278397239086e-08,
+      "loss": 0.0563,
+      "step": 1893
+    },
+    {
+      "epoch": 9.239024390243902,
+      "grad_norm": 0.4730098247528076,
+      "learning_rate": 7.20120674357852e-08,
+      "loss": 0.0183,
+      "step": 1894
+    },
+    {
+      "epoch": 9.24390243902439,
+      "grad_norm": 0.6306093335151672,
+      "learning_rate": 7.110199729535805e-08,
+      "loss": 0.0664,
+      "step": 1895
+    },
+    {
+      "epoch": 9.248780487804877,
+      "grad_norm": 0.7287396192550659,
+      "learning_rate": 7.019763143993441e-08,
+      "loss": 0.0968,
+      "step": 1896
+    },
+    {
+      "epoch": 9.253658536585366,
+      "grad_norm": 0.5095722079277039,
+      "learning_rate": 6.929897199342395e-08,
+      "loss": 0.0305,
+      "step": 1897
+    },
+    {
+      "epoch": 9.258536585365853,
+      "grad_norm": 0.5435238480567932,
+      "learning_rate": 6.840602106633425e-08,
+      "loss": 0.0142,
+      "step": 1898
+    },
+    {
+      "epoch": 9.263414634146342,
+      "grad_norm": 1.1777113676071167,
+      "learning_rate": 6.751878075576867e-08,
+      "loss": 0.0887,
+      "step": 1899
+    },
+    {
+      "epoch": 9.268292682926829,
+      "grad_norm": 1.3694193363189697,
+      "learning_rate": 6.663725314541652e-08,
+      "loss": 0.2028,
+      "step": 1900
+    },
+    {
+      "epoch": 9.273170731707317,
+      "grad_norm": 0.7819920182228088,
+      "learning_rate": 6.576144030555259e-08,
+      "loss": 0.0621,
+      "step": 1901
+    },
+    {
+      "epoch": 9.278048780487804,
+      "grad_norm": 1.0538229942321777,
+      "learning_rate": 6.489134429302906e-08,
+      "loss": 0.1227,
+      "step": 1902
+    },
+    {
+      "epoch": 9.282926829268293,
+      "grad_norm": 0.4868879020214081,
+      "learning_rate": 6.402696715127387e-08,
+      "loss": 0.0275,
+      "step": 1903
+    },
+    {
+      "epoch": 9.28780487804878,
+      "grad_norm": 0.506305456161499,
+      "learning_rate": 6.316831091028237e-08,
+      "loss": 0.0222,
+      "step": 1904
+    },
+    {
+      "epoch": 9.292682926829269,
+      "grad_norm": 0.46229302883148193,
+      "learning_rate": 6.23153775866156e-08,
+      "loss": 0.0134,
+      "step": 1905
+    },
+    {
+      "epoch": 9.297560975609755,
+      "grad_norm": 0.8209920525550842,
+      "learning_rate": 6.14681691833935e-08,
+      "loss": 0.1029,
+      "step": 1906
+    },
+    {
+      "epoch": 9.302439024390244,
+      "grad_norm": 0.6274682283401489,
+      "learning_rate": 6.062668769029168e-08,
+      "loss": 0.0353,
+      "step": 1907
+    },
+    {
+      "epoch": 9.307317073170731,
+      "grad_norm": 4.185459136962891,
+      "learning_rate": 5.979093508353489e-08,
+      "loss": 0.1436,
+      "step": 1908
+    },
+    {
+      "epoch": 9.31219512195122,
+      "grad_norm": 1.0531197786331177,
+      "learning_rate": 5.896091332589532e-08,
+      "loss": 0.205,
+      "step": 1909
+    },
+    {
+      "epoch": 9.317073170731707,
+      "grad_norm": 0.6232990622520447,
+      "learning_rate": 5.813662436668477e-08,
+      "loss": 0.0328,
+      "step": 1910
+    },
+    {
+      "epoch": 9.321951219512195,
+      "grad_norm": 0.962642252445221,
+      "learning_rate": 5.731807014175195e-08,
+      "loss": 0.0748,
+      "step": 1911
+    },
+    {
+      "epoch": 9.326829268292682,
+      "grad_norm": 0.44946765899658203,
+      "learning_rate": 5.650525257347744e-08,
+      "loss": 0.0182,
+      "step": 1912
+    },
+    {
+      "epoch": 9.331707317073171,
+      "grad_norm": 0.4619583487510681,
+      "learning_rate": 5.569817357076984e-08,
+      "loss": 0.0297,
+      "step": 1913
+    },
+    {
+      "epoch": 9.336585365853658,
+      "grad_norm": 1.4466209411621094,
+      "learning_rate": 5.489683502905935e-08,
+      "loss": 0.2278,
+      "step": 1914
+    },
+    {
+      "epoch": 9.341463414634147,
+      "grad_norm": 0.8814978003501892,
+      "learning_rate": 5.410123883029639e-08,
+      "loss": 0.0559,
+      "step": 1915
+    },
+    {
+      "epoch": 9.346341463414634,
+      "grad_norm": 1.11834716796875,
+      "learning_rate": 5.3311386842944125e-08,
+      "loss": 0.1064,
+      "step": 1916
+    },
+    {
+      "epoch": 9.351219512195122,
+      "grad_norm": 0.8984152674674988,
+      "learning_rate": 5.25272809219754e-08,
+      "loss": 0.0965,
+      "step": 1917
+    },
+    {
+      "epoch": 9.35609756097561,
+      "grad_norm": 0.7180256843566895,
+      "learning_rate": 5.17489229088694e-08,
+      "loss": 0.0694,
+      "step": 1918
+    },
+    {
+      "epoch": 9.360975609756098,
+      "grad_norm": 0.941351592540741,
+      "learning_rate": 5.097631463160585e-08,
+      "loss": 0.1661,
+      "step": 1919
+    },
+    {
+      "epoch": 9.365853658536585,
+      "grad_norm": 0.7345023155212402,
+      "learning_rate": 5.020945790466025e-08,
+      "loss": 0.0452,
+      "step": 1920
+    },
+    {
+      "epoch": 9.370731707317074,
+      "grad_norm": 0.5164614915847778,
+      "learning_rate": 4.944835452900199e-08,
+      "loss": 0.0199,
+      "step": 1921
+    },
+    {
+      "epoch": 9.37560975609756,
+      "grad_norm": 1.0711966753005981,
+      "learning_rate": 4.869300629208762e-08,
+      "loss": 0.1845,
+      "step": 1922
+    },
+    {
+      "epoch": 9.38048780487805,
+      "grad_norm": 1.3359203338623047,
+      "learning_rate": 4.7943414967858426e-08,
+      "loss": 0.1641,
+      "step": 1923
+    },
+    {
+      "epoch": 9.385365853658536,
+      "grad_norm": 0.7057273983955383,
+      "learning_rate": 4.7199582316734827e-08,
+      "loss": 0.0443,
+      "step": 1924
+    },
+    {
+      "epoch": 9.390243902439025,
+      "grad_norm": 0.4502648413181305,
+      "learning_rate": 4.6461510085613616e-08,
+      "loss": 0.0127,
+      "step": 1925
+    },
+    {
+      "epoch": 9.395121951219512,
+      "grad_norm": 0.5665896534919739,
+      "learning_rate": 4.5729200007862686e-08,
+      "loss": 0.0167,
+      "step": 1926
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 0.5727550387382507,
+      "learning_rate": 4.5002653803317975e-08,
+      "loss": 0.042,
+      "step": 1927
+    },
+    {
+      "epoch": 9.404878048780487,
+      "grad_norm": 1.0175763368606567,
+      "learning_rate": 4.428187317827848e-08,
+      "loss": 0.0654,
+      "step": 1928
+    },
+    {
+      "epoch": 9.409756097560976,
+      "grad_norm": 0.9805188179016113,
+      "learning_rate": 4.356685982550263e-08,
+      "loss": 0.0817,
+      "step": 1929
+    },
+    {
+      "epoch": 9.414634146341463,
+      "grad_norm": 0.4889443516731262,
+      "learning_rate": 4.285761542420497e-08,
+      "loss": 0.0192,
+      "step": 1930
+    },
+    {
+      "epoch": 9.419512195121952,
+      "grad_norm": 0.558597981929779,
+      "learning_rate": 4.215414164005116e-08,
+      "loss": 0.0528,
+      "step": 1931
+    },
+    {
+      "epoch": 9.424390243902439,
+      "grad_norm": 0.6691590547561646,
+      "learning_rate": 4.145644012515465e-08,
+      "loss": 0.0439,
+      "step": 1932
+    },
+    {
+      "epoch": 9.429268292682927,
+      "grad_norm": 0.8558660745620728,
+      "learning_rate": 4.076451251807223e-08,
+      "loss": 0.0429,
+      "step": 1933
+    },
+    {
+      "epoch": 9.434146341463414,
+      "grad_norm": 0.4820708632469177,
+      "learning_rate": 4.0078360443801535e-08,
+      "loss": 0.021,
+      "step": 1934
+    },
+    {
+      "epoch": 9.439024390243903,
+      "grad_norm": 0.7166886925697327,
+      "learning_rate": 3.9397985513775495e-08,
+      "loss": 0.0424,
+      "step": 1935
+    },
+    {
+      "epoch": 9.44390243902439,
+      "grad_norm": 0.6747754812240601,
+      "learning_rate": 3.872338932585984e-08,
+      "loss": 0.0369,
+      "step": 1936
+    },
+    {
+      "epoch": 9.448780487804878,
+      "grad_norm": 0.682863175868988,
+      "learning_rate": 3.8054573464348655e-08,
+      "loss": 0.0888,
+      "step": 1937
+    },
+    {
+      "epoch": 9.453658536585365,
+      "grad_norm": 0.4465639591217041,
+      "learning_rate": 3.739153949996105e-08,
+      "loss": 0.0161,
+      "step": 1938
+    },
+    {
+      "epoch": 9.458536585365854,
+      "grad_norm": 0.7950959801673889,
+      "learning_rate": 3.6734288989836994e-08,
+      "loss": 0.0809,
+      "step": 1939
+    },
+    {
+      "epoch": 9.463414634146341,
+      "grad_norm": 0.5149068236351013,
+      "learning_rate": 3.608282347753428e-08,
+      "loss": 0.0209,
+      "step": 1940
+    },
+    {
+      "epoch": 9.46829268292683,
+      "grad_norm": 0.5153360366821289,
+      "learning_rate": 3.543714449302488e-08,
+      "loss": 0.0396,
+      "step": 1941
+    },
+    {
+      "epoch": 9.473170731707317,
+      "grad_norm": 0.7813478708267212,
+      "learning_rate": 3.479725355268998e-08,
+      "loss": 0.0799,
+      "step": 1942
+    },
+    {
+      "epoch": 9.478048780487805,
+      "grad_norm": 1.0667929649353027,
+      "learning_rate": 3.4163152159318866e-08,
+      "loss": 0.139,
+      "step": 1943
+    },
+    {
+      "epoch": 9.482926829268292,
+      "grad_norm": 0.4019707441329956,
+      "learning_rate": 3.353484180210337e-08,
+      "loss": 0.013,
+      "step": 1944
+    },
+    {
+      "epoch": 9.487804878048781,
+      "grad_norm": 0.4706079065799713,
+      "learning_rate": 3.291232395663424e-08,
+      "loss": 0.0314,
+      "step": 1945
+    },
+    {
+      "epoch": 9.492682926829268,
+      "grad_norm": 0.8636847734451294,
+      "learning_rate": 3.229560008490007e-08,
+      "loss": 0.1431,
+      "step": 1946
+    },
+    {
+      "epoch": 9.497560975609757,
+      "grad_norm": 1.207856297492981,
+      "learning_rate": 3.168467163528116e-08,
+      "loss": 0.1841,
+      "step": 1947
+    },
+    {
+      "epoch": 9.502439024390243,
+      "grad_norm": 0.9270714521408081,
+      "learning_rate": 3.1079540042547315e-08,
+      "loss": 0.121,
+      "step": 1948
+    },
+    {
+      "epoch": 9.507317073170732,
+      "grad_norm": 0.5015339851379395,
+      "learning_rate": 3.0480206727855066e-08,
+      "loss": 0.0223,
+      "step": 1949
+    },
+    {
+      "epoch": 9.512195121951219,
+      "grad_norm": 0.7450964450836182,
+      "learning_rate": 2.988667309874294e-08,
+      "loss": 0.0864,
+      "step": 1950
+    },
+    {
+      "epoch": 9.517073170731708,
+      "grad_norm": 1.4398821592330933,
+      "learning_rate": 2.9298940549128962e-08,
+      "loss": 0.2769,
+      "step": 1951
+    },
+    {
+      "epoch": 9.521951219512195,
+      "grad_norm": 1.0129376649856567,
+      "learning_rate": 2.871701045930708e-08,
+      "loss": 0.1112,
+      "step": 1952
+    },
+    {
+      "epoch": 9.526829268292683,
+      "grad_norm": 1.3432735204696655,
+      "learning_rate": 2.8140884195945184e-08,
+      "loss": 0.1927,
+      "step": 1953
+    },
+    {
+      "epoch": 9.53170731707317,
+      "grad_norm": 0.5337926149368286,
+      "learning_rate": 2.7570563112079564e-08,
+      "loss": 0.0407,
+      "step": 1954
+    },
+    {
+      "epoch": 9.536585365853659,
+      "grad_norm": 0.6368472576141357,
+      "learning_rate": 2.700604854711353e-08,
+      "loss": 0.0525,
+      "step": 1955
+    },
+    {
+      "epoch": 9.541463414634146,
+      "grad_norm": 0.49696725606918335,
+      "learning_rate": 2.6447341826814077e-08,
+      "loss": 0.016,
+      "step": 1956
+    },
+    {
+      "epoch": 9.546341463414635,
+      "grad_norm": 0.8632214665412903,
+      "learning_rate": 2.5894444263307728e-08,
+      "loss": 0.059,
+      "step": 1957
+    },
+    {
+      "epoch": 9.551219512195122,
+      "grad_norm": 1.3797489404678345,
+      "learning_rate": 2.5347357155078577e-08,
+      "loss": 0.3289,
+      "step": 1958
+    },
+    {
+      "epoch": 9.55609756097561,
+      "grad_norm": 0.6812785267829895,
+      "learning_rate": 2.4806081786964974e-08,
+      "loss": 0.0339,
+      "step": 1959
+    },
+    {
+      "epoch": 9.560975609756097,
+      "grad_norm": 1.0233956575393677,
+      "learning_rate": 2.4270619430156183e-08,
+      "loss": 0.1161,
+      "step": 1960
+    },
+    {
+      "epoch": 9.565853658536586,
+      "grad_norm": 1.1237671375274658,
+      "learning_rate": 2.3740971342189056e-08,
+      "loss": 0.144,
+      "step": 1961
+    },
+    {
+      "epoch": 9.570731707317073,
+      "grad_norm": 0.9001303315162659,
+      "learning_rate": 2.321713876694637e-08,
+      "loss": 0.1843,
+      "step": 1962
+    },
+    {
+      "epoch": 9.575609756097561,
+      "grad_norm": 1.377709984779358,
+      "learning_rate": 2.269912293465293e-08,
+      "loss": 0.2671,
+      "step": 1963
+    },
+    {
+      "epoch": 9.580487804878048,
+      "grad_norm": 0.8948473930358887,
+      "learning_rate": 2.2186925061872532e-08,
+      "loss": 0.1319,
+      "step": 1964
+    },
+    {
+      "epoch": 9.585365853658537,
+      "grad_norm": 1.184303641319275,
+      "learning_rate": 2.1680546351506016e-08,
+      "loss": 0.1856,
+      "step": 1965
+    },
+    {
+      "epoch": 9.590243902439024,
+      "grad_norm": 0.9378262162208557,
+      "learning_rate": 2.117998799278709e-08,
+      "loss": 0.1141,
+      "step": 1966
+    },
+    {
+      "epoch": 9.595121951219513,
+      "grad_norm": 0.6293427348136902,
+      "learning_rate": 2.068525116128095e-08,
+      "loss": 0.0221,
+      "step": 1967
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.0454257726669312,
+      "learning_rate": 2.0196337018880962e-08,
+      "loss": 0.228,
+      "step": 1968
+    },
+    {
+      "epoch": 9.604878048780488,
+      "grad_norm": 0.39218515157699585,
+      "learning_rate": 1.9713246713805588e-08,
+      "loss": 0.0116,
+      "step": 1969
+    },
+    {
+      "epoch": 9.609756097560975,
+      "grad_norm": 1.0841166973114014,
+      "learning_rate": 1.9235981380595625e-08,
+      "loss": 0.1303,
+      "step": 1970
+    },
+    {
+      "epoch": 9.614634146341464,
+      "grad_norm": 0.4800502061843872,
+      "learning_rate": 1.876454214011253e-08,
+      "loss": 0.0178,
+      "step": 1971
+    },
+    {
+      "epoch": 9.61951219512195,
+      "grad_norm": 0.6787453889846802,
+      "learning_rate": 1.8298930099534817e-08,
+      "loss": 0.0432,
+      "step": 1972
+    },
+    {
+      "epoch": 9.62439024390244,
+      "grad_norm": 0.6503498554229736,
+      "learning_rate": 1.783914635235584e-08,
+      "loss": 0.0444,
+      "step": 1973
+    },
+    {
+      "epoch": 9.629268292682926,
+      "grad_norm": 0.7241586446762085,
+      "learning_rate": 1.738519197838101e-08,
+      "loss": 0.0796,
+      "step": 1974
+    },
+    {
+      "epoch": 9.634146341463415,
+      "grad_norm": 0.6374277472496033,
+      "learning_rate": 1.6937068043725856e-08,
+      "loss": 0.0344,
+      "step": 1975
+    },
+    {
+      "epoch": 9.639024390243902,
+      "grad_norm": 0.9005374312400818,
+      "learning_rate": 1.6494775600812418e-08,
+      "loss": 0.1417,
+      "step": 1976
+    },
+    {
+      "epoch": 9.64390243902439,
+      "grad_norm": 0.9925791025161743,
+      "learning_rate": 1.6058315688367852e-08,
+      "loss": 0.1428,
+      "step": 1977
+    },
+    {
+      "epoch": 9.648780487804878,
+      "grad_norm": 0.7364526391029358,
+      "learning_rate": 1.5627689331421946e-08,
+      "loss": 0.0739,
+      "step": 1978
+    },
+    {
+      "epoch": 9.653658536585366,
+      "grad_norm": 0.563661515712738,
+      "learning_rate": 1.520289754130322e-08,
+      "loss": 0.0458,
+      "step": 1979
+    },
+    {
+      "epoch": 9.658536585365853,
+      "grad_norm": 0.8973087668418884,
+      "learning_rate": 1.478394131563865e-08,
+      "loss": 0.1025,
+      "step": 1980
+    },
+    {
+      "epoch": 9.663414634146342,
+      "grad_norm": 0.888282299041748,
+      "learning_rate": 1.4370821638350353e-08,
+      "loss": 0.0966,
+      "step": 1981
+    },
+    {
+      "epoch": 9.668292682926829,
+      "grad_norm": 0.6470147371292114,
+      "learning_rate": 1.396353947965251e-08,
+      "loss": 0.0647,
+      "step": 1982
+    },
+    {
+      "epoch": 9.673170731707318,
+      "grad_norm": 0.3161561191082001,
+      "learning_rate": 1.3562095796050279e-08,
+      "loss": 0.0076,
+      "step": 1983
+    },
+    {
+      "epoch": 9.678048780487805,
+      "grad_norm": 0.8751645088195801,
+      "learning_rate": 1.3166491530337555e-08,
+      "loss": 0.063,
+      "step": 1984
+    },
+    {
+      "epoch": 9.682926829268293,
+      "grad_norm": 0.5214653611183167,
+      "learning_rate": 1.2776727611593653e-08,
+      "loss": 0.027,
+      "step": 1985
+    },
+    {
+      "epoch": 9.68780487804878,
+      "grad_norm": 1.0834569931030273,
+      "learning_rate": 1.2392804955181915e-08,
+      "loss": 0.1182,
+      "step": 1986
+    },
+    {
+      "epoch": 9.692682926829269,
+      "grad_norm": 0.9092963337898254,
+      "learning_rate": 1.2014724462747763e-08,
+      "loss": 0.1113,
+      "step": 1987
+    },
+    {
+      "epoch": 9.697560975609756,
+      "grad_norm": 0.5805611610412598,
+      "learning_rate": 1.1642487022215931e-08,
+      "loss": 0.0409,
+      "step": 1988
+    },
+    {
+      "epoch": 9.702439024390245,
+      "grad_norm": 0.6428013443946838,
+      "learning_rate": 1.1276093507788798e-08,
+      "loss": 0.0471,
+      "step": 1989
+    },
+    {
+      "epoch": 9.707317073170731,
+      "grad_norm": 0.703423023223877,
+      "learning_rate": 1.0915544779944164e-08,
+      "loss": 0.0546,
+      "step": 1990
+    },
+    {
+      "epoch": 9.71219512195122,
+      "grad_norm": 0.5905337333679199,
+      "learning_rate": 1.0560841685433864e-08,
+      "loss": 0.043,
+      "step": 1991
+    },
+    {
+      "epoch": 9.717073170731707,
+      "grad_norm": 0.7289360165596008,
+      "learning_rate": 1.021198505728016e-08,
+      "loss": 0.0579,
+      "step": 1992
+    },
+    {
+      "epoch": 9.721951219512196,
+      "grad_norm": 0.7613627910614014,
+      "learning_rate": 9.868975714775741e-09,
+      "loss": 0.0517,
+      "step": 1993
+    },
+    {
+      "epoch": 9.726829268292683,
+      "grad_norm": 0.9253370761871338,
+      "learning_rate": 9.531814463480394e-09,
+      "loss": 0.0429,
+      "step": 1994
+    },
+    {
+      "epoch": 9.731707317073171,
+      "grad_norm": 1.2712122201919556,
+      "learning_rate": 9.200502095220166e-09,
+      "loss": 0.2326,
+      "step": 1995
+    },
+    {
+      "epoch": 9.736585365853658,
+      "grad_norm": 0.554710865020752,
+      "learning_rate": 8.875039388084317e-09,
+      "loss": 0.034,
+      "step": 1996
+    },
+    {
+      "epoch": 9.741463414634147,
+      "grad_norm": 0.5033829808235168,
+      "learning_rate": 8.555427106424485e-09,
+      "loss": 0.022,
+      "step": 1997
+    },
+    {
+      "epoch": 9.746341463414634,
+      "grad_norm": 1.0586061477661133,
+      "learning_rate": 8.241666000852466e-09,
+      "loss": 0.224,
+      "step": 1998
+    },
+    {
+      "epoch": 9.751219512195123,
+      "grad_norm": 0.6225917935371399,
+      "learning_rate": 7.933756808238823e-09,
+      "loss": 0.0272,
+      "step": 1999
+    },
+    {
+      "epoch": 9.75609756097561,
+      "grad_norm": 1.053144097328186,
+      "learning_rate": 7.631700251710116e-09,
+      "loss": 0.1515,
+      "step": 2000
+    },
+    {
+      "epoch": 9.760975609756098,
+      "grad_norm": 1.2394822835922241,
+      "learning_rate": 7.335497040648898e-09,
+      "loss": 0.1459,
+      "step": 2001
+    },
+    {
+      "epoch": 9.765853658536585,
+      "grad_norm": 1.8207858800888062,
+      "learning_rate": 7.045147870690105e-09,
+      "loss": 0.5176,
+      "step": 2002
+    },
+    {
+      "epoch": 9.770731707317074,
+      "grad_norm": 0.9958927035331726,
+      "learning_rate": 6.760653423721619e-09,
+      "loss": 0.081,
+      "step": 2003
+    },
+    {
+      "epoch": 9.77560975609756,
+      "grad_norm": 1.4912139177322388,
+      "learning_rate": 6.4820143678800964e-09,
+      "loss": 0.2137,
+      "step": 2004
+    },
+    {
+      "epoch": 9.78048780487805,
+      "grad_norm": 0.38397058844566345,
+      "learning_rate": 6.209231357551526e-09,
+      "loss": 0.0119,
+      "step": 2005
+    },
+    {
+      "epoch": 9.785365853658536,
+      "grad_norm": 1.085859775543213,
+      "learning_rate": 5.942305033369289e-09,
+      "loss": 0.1228,
+      "step": 2006
+    },
+    {
+      "epoch": 9.790243902439025,
+      "grad_norm": 0.9027035236358643,
+      "learning_rate": 5.681236022211378e-09,
+      "loss": 0.014,
+      "step": 2007
+    },
+    {
+      "epoch": 9.795121951219512,
+      "grad_norm": 0.7191747426986694,
+      "learning_rate": 5.426024937200402e-09,
+      "loss": 0.1024,
+      "step": 2008
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 0.5138716101646423,
+      "learning_rate": 5.176672377701364e-09,
+      "loss": 0.0207,
+      "step": 2009
+    },
+    {
+      "epoch": 9.804878048780488,
+      "grad_norm": 0.591534435749054,
+      "learning_rate": 4.933178929321103e-09,
+      "loss": 0.0256,
+      "step": 2010
+    },
+    {
+      "epoch": 9.809756097560976,
+      "grad_norm": 0.6698344349861145,
+      "learning_rate": 4.695545163905524e-09,
+      "loss": 0.049,
+      "step": 2011
+    },
+    {
+      "epoch": 9.814634146341463,
+      "grad_norm": 0.676779568195343,
+      "learning_rate": 4.463771639539038e-09,
+      "loss": 0.0606,
+      "step": 2012
+    },
+    {
+      "epoch": 9.819512195121952,
+      "grad_norm": 0.6824347972869873,
+      "learning_rate": 4.237858900543734e-09,
+      "loss": 0.0594,
+      "step": 2013
+    },
+    {
+      "epoch": 9.824390243902439,
+      "grad_norm": 1.3436641693115234,
+      "learning_rate": 4.017807477477154e-09,
+      "loss": 0.2447,
+      "step": 2014
+    },
+    {
+      "epoch": 9.829268292682928,
+      "grad_norm": 0.8924421668052673,
+      "learning_rate": 3.803617887132016e-09,
+      "loss": 0.087,
+      "step": 2015
+    },
+    {
+      "epoch": 9.834146341463414,
+      "grad_norm": 1.0106862783432007,
+      "learning_rate": 3.5952906325339988e-09,
+      "loss": 0.1801,
+      "step": 2016
+    },
+    {
+      "epoch": 9.839024390243903,
+      "grad_norm": 0.6396204829216003,
+      "learning_rate": 3.3928262029411794e-09,
+      "loss": 0.0399,
+      "step": 2017
+    },
+    {
+      "epoch": 9.84390243902439,
+      "grad_norm": 0.7245880365371704,
+      "learning_rate": 3.196225073842929e-09,
+      "loss": 0.0601,
+      "step": 2018
+    },
+    {
+      "epoch": 9.848780487804879,
+      "grad_norm": 0.4817493259906769,
+      "learning_rate": 3.005487706958243e-09,
+      "loss": 0.03,
+      "step": 2019
+    },
+    {
+      "epoch": 9.853658536585366,
+      "grad_norm": 0.9604029655456543,
+      "learning_rate": 2.8206145502354678e-09,
+      "loss": 0.1623,
+      "step": 2020
+    },
+    {
+      "epoch": 9.858536585365854,
+      "grad_norm": 0.8873414993286133,
+      "learning_rate": 2.641606037850353e-09,
+      "loss": 0.0717,
+      "step": 2021
+    },
+    {
+      "epoch": 9.863414634146341,
+      "grad_norm": 0.5341236591339111,
+      "learning_rate": 2.468462590205778e-09,
+      "loss": 0.0433,
+      "step": 2022
+    },
+    {
+      "epoch": 9.86829268292683,
+      "grad_norm": 0.619755744934082,
+      "learning_rate": 2.3011846139306404e-09,
+      "loss": 0.0435,
+      "step": 2023
+    },
+    {
+      "epoch": 9.873170731707317,
+      "grad_norm": 1.0023456811904907,
+      "learning_rate": 2.13977250187819e-09,
+      "loss": 0.1088,
+      "step": 2024
+    },
+    {
+      "epoch": 9.878048780487806,
+      "grad_norm": 0.9846378564834595,
+      "learning_rate": 1.9842266331260296e-09,
+      "loss": 0.1234,
+      "step": 2025
+    },
+    {
+      "epoch": 9.882926829268293,
+      "grad_norm": 0.6594278216362,
+      "learning_rate": 1.834547372975004e-09,
+      "loss": 0.0427,
+      "step": 2026
+    },
+    {
+      "epoch": 9.887804878048781,
+      "grad_norm": 0.8662856817245483,
+      "learning_rate": 1.6907350729478133e-09,
+      "loss": 0.0925,
+      "step": 2027
+    },
+    {
+      "epoch": 9.892682926829268,
+      "grad_norm": 0.722348690032959,
+      "learning_rate": 1.5527900707887344e-09,
+      "loss": 0.0513,
+      "step": 2028
+    },
+    {
+      "epoch": 9.897560975609757,
+      "grad_norm": 1.391545057296753,
+      "learning_rate": 1.4207126904625114e-09,
+      "loss": 0.269,
+      "step": 2029
+    },
+    {
+      "epoch": 9.902439024390244,
+      "grad_norm": 0.5979572534561157,
+      "learning_rate": 1.2945032421540771e-09,
+      "loss": 0.0354,
+      "step": 2030
+    },
+    {
+      "epoch": 9.907317073170733,
+      "grad_norm": 0.8387157320976257,
+      "learning_rate": 1.1741620222671667e-09,
+      "loss": 0.0815,
+      "step": 2031
+    },
+    {
+      "epoch": 9.91219512195122,
+      "grad_norm": 0.7832284569740295,
+      "learning_rate": 1.0596893134240394e-09,
+      "loss": 0.0673,
+      "step": 2032
+    },
+    {
+      "epoch": 9.917073170731708,
+      "grad_norm": 0.8685626983642578,
+      "learning_rate": 9.51085384464645e-10,
+      "loss": 0.089,
+      "step": 2033
+    },
+    {
+      "epoch": 9.921951219512195,
+      "grad_norm": 0.9133242964744568,
+      "learning_rate": 8.48350490446348e-10,
+      "loss": 0.079,
+      "step": 2034
+    },
+    {
+      "epoch": 9.926829268292684,
+      "grad_norm": 1.2350854873657227,
+      "learning_rate": 7.514848726422608e-10,
+      "loss": 0.2039,
+      "step": 2035
+    },
+    {
+      "epoch": 9.93170731707317,
+      "grad_norm": 0.8196302652359009,
+      "learning_rate": 6.604887585426323e-10,
+      "loss": 0.069,
+      "step": 2036
+    },
+    {
+      "epoch": 9.93658536585366,
+      "grad_norm": 0.8256949186325073,
+      "learning_rate": 5.753623618520721e-10,
+      "loss": 0.0705,
+      "step": 2037
+    },
+    {
+      "epoch": 9.941463414634146,
+      "grad_norm": 1.1538176536560059,
+      "learning_rate": 4.961058824909382e-10,
+      "loss": 0.2095,
+      "step": 2038
+    },
+    {
+      "epoch": 9.946341463414633,
+      "grad_norm": 0.5681881904602051,
+      "learning_rate": 4.2271950659311665e-10,
+      "loss": 0.0297,
+      "step": 2039
+    },
+    {
+      "epoch": 9.951219512195122,
+      "grad_norm": 1.1952041387557983,
+      "learning_rate": 3.5520340650768705e-10,
+      "loss": 0.1254,
+      "step": 2040
+    },
+    {
+      "epoch": 9.95609756097561,
+      "grad_norm": 0.8561511039733887,
+      "learning_rate": 2.9355774079614653e-10,
+      "loss": 0.0602,
+      "step": 2041
+    },
+    {
+      "epoch": 9.960975609756098,
+      "grad_norm": 0.37976720929145813,
+      "learning_rate": 2.377826542343531e-10,
+      "loss": 0.0114,
+      "step": 2042
+    },
+    {
+      "epoch": 9.965853658536584,
+      "grad_norm": 0.5294272303581238,
+      "learning_rate": 1.8787827781002743e-10,
+      "loss": 0.0324,
+      "step": 2043
+    },
+    {
+      "epoch": 9.970731707317073,
+      "grad_norm": 0.6023978590965271,
+      "learning_rate": 1.4384472872414067e-10,
+      "loss": 0.0284,
+      "step": 2044
+    },
+    {
+      "epoch": 9.975609756097562,
+      "grad_norm": 0.5626432299613953,
+      "learning_rate": 1.056821103900818e-10,
+      "loss": 0.0298,
+      "step": 2045
+    },
+    {
+      "epoch": 9.980487804878049,
+      "grad_norm": 0.5841759443283081,
+      "learning_rate": 7.339051243254735e-11,
+      "loss": 0.0454,
+      "step": 2046
+    },
+    {
+      "epoch": 9.985365853658536,
+      "grad_norm": 1.1930917501449585,
+      "learning_rate": 4.697001068892926e-11,
+      "loss": 0.2935,
+      "step": 2047
+    },
+    {
+      "epoch": 9.990243902439024,
+      "grad_norm": 0.5551172494888306,
+      "learning_rate": 2.642066720792702e-11,
+      "loss": 0.031,
+      "step": 2048
+    },
+    {
+      "epoch": 9.995121951219513,
+      "grad_norm": 1.0561318397521973,
+      "learning_rate": 1.1742530249547745e-11,
+      "loss": 0.1136,
+      "step": 2049
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.32328081130981445,
+      "learning_rate": 2.9356342859387933e-12,
+      "loss": 0.0101,
+      "step": 2050
+    },
+    {
+      "epoch": 10.0,
+      "step": 2050,
+      "total_flos": 6.042610155252613e+17,
+      "train_loss": 0.38200810655406336,
+      "train_runtime": 8575.0364,
+      "train_samples_per_second": 0.953,
+      "train_steps_per_second": 0.239
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 208,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.042610155252613e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}