{
  "best_metric": 0.6181585192680359,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.7490636704119851,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003745318352059925,
      "grad_norm": 8.873519897460938,
      "learning_rate": 1e-05,
      "loss": 14.0732,
      "step": 1
    },
    {
      "epoch": 0.003745318352059925,
      "eval_loss": 2.651507616043091,
      "eval_runtime": 49.5959,
      "eval_samples_per_second": 9.073,
      "eval_steps_per_second": 2.278,
      "step": 1
    },
    {
      "epoch": 0.00749063670411985,
      "grad_norm": 7.458089351654053,
      "learning_rate": 2e-05,
      "loss": 11.2916,
      "step": 2
    },
    {
      "epoch": 0.011235955056179775,
      "grad_norm": 6.022294044494629,
      "learning_rate": 3e-05,
      "loss": 10.6266,
      "step": 3
    },
    {
      "epoch": 0.0149812734082397,
      "grad_norm": 5.537844181060791,
      "learning_rate": 4e-05,
      "loss": 9.2105,
      "step": 4
    },
    {
      "epoch": 0.018726591760299626,
      "grad_norm": 5.325942039489746,
      "learning_rate": 5e-05,
      "loss": 8.9843,
      "step": 5
    },
    {
      "epoch": 0.02247191011235955,
      "grad_norm": 4.357425212860107,
      "learning_rate": 6e-05,
      "loss": 8.9895,
      "step": 6
    },
    {
      "epoch": 0.026217228464419477,
      "grad_norm": 3.5573365688323975,
      "learning_rate": 7e-05,
      "loss": 8.6797,
      "step": 7
    },
    {
      "epoch": 0.0299625468164794,
      "grad_norm": 3.2774643898010254,
      "learning_rate": 8e-05,
      "loss": 8.0163,
      "step": 8
    },
    {
      "epoch": 0.033707865168539325,
      "grad_norm": 3.1855506896972656,
      "learning_rate": 9e-05,
      "loss": 7.1132,
      "step": 9
    },
    {
      "epoch": 0.03745318352059925,
      "grad_norm": 2.9686391353607178,
      "learning_rate": 0.0001,
      "loss": 7.2893,
      "step": 10
    },
    {
      "epoch": 0.04119850187265917,
      "grad_norm": 2.6892645359039307,
      "learning_rate": 9.999316524962345e-05,
      "loss": 7.2517,
      "step": 11
    },
    {
      "epoch": 0.0449438202247191,
      "grad_norm": 3.090545177459717,
      "learning_rate": 9.997266286704631e-05,
      "loss": 7.1434,
      "step": 12
    },
    {
      "epoch": 0.04868913857677903,
      "grad_norm": 3.193530321121216,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.9541,
      "step": 13
    },
    {
      "epoch": 0.052434456928838954,
      "grad_norm": 3.550259828567505,
      "learning_rate": 9.989068136093873e-05,
      "loss": 6.3891,
      "step": 14
    },
    {
      "epoch": 0.056179775280898875,
      "grad_norm": 2.6989731788635254,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.9002,
      "step": 15
    },
    {
      "epoch": 0.0599250936329588,
      "grad_norm": 2.4900543689727783,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.7425,
      "step": 16
    },
    {
      "epoch": 0.06367041198501873,
      "grad_norm": 2.8906259536743164,
      "learning_rate": 9.966546331768191e-05,
      "loss": 6.016,
      "step": 17
    },
    {
      "epoch": 0.06741573033707865,
      "grad_norm": 2.868975877761841,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.0337,
      "step": 18
    },
    {
      "epoch": 0.07116104868913857,
      "grad_norm": 2.994189977645874,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.7442,
      "step": 19
    },
    {
      "epoch": 0.0749063670411985,
      "grad_norm": 3.044917583465576,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.3122,
      "step": 20
    },
    {
      "epoch": 0.07865168539325842,
      "grad_norm": 3.1765198707580566,
      "learning_rate": 9.917525374361912e-05,
      "loss": 6.0947,
      "step": 21
    },
    {
      "epoch": 0.08239700374531835,
      "grad_norm": 3.313641309738159,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.9363,
      "step": 22
    },
    {
      "epoch": 0.08614232209737828,
      "grad_norm": 3.7408692836761475,
      "learning_rate": 9.884934153917997e-05,
      "loss": 4.9411,
      "step": 23
    },
    {
      "epoch": 0.0898876404494382,
      "grad_norm": 2.973936080932617,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.5589,
      "step": 24
    },
    {
      "epoch": 0.09363295880149813,
      "grad_norm": 2.9019627571105957,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.108,
      "step": 25
    },
    {
      "epoch": 0.09737827715355805,
      "grad_norm": 2.7772433757781982,
      "learning_rate": 9.826044551386744e-05,
      "loss": 4.907,
      "step": 26
    },
    {
      "epoch": 0.10112359550561797,
      "grad_norm": 3.2323522567749023,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.4772,
      "step": 27
    },
    {
      "epoch": 0.10486891385767791,
      "grad_norm": 3.3576741218566895,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.9318,
      "step": 28
    },
    {
      "epoch": 0.10861423220973783,
      "grad_norm": 3.1996891498565674,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.6985,
      "step": 29
    },
    {
      "epoch": 0.11235955056179775,
      "grad_norm": 3.651137351989746,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.958,
      "step": 30
    },
    {
      "epoch": 0.11610486891385768,
      "grad_norm": 3.599393606185913,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.9517,
      "step": 31
    },
    {
      "epoch": 0.1198501872659176,
      "grad_norm": 3.7332308292388916,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.0816,
      "step": 32
    },
    {
      "epoch": 0.12359550561797752,
      "grad_norm": 3.6044347286224365,
      "learning_rate": 9.642770192448536e-05,
      "loss": 5.0095,
      "step": 33
    },
    {
      "epoch": 0.12734082397003746,
      "grad_norm": 4.162293910980225,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.2108,
      "step": 34
    },
    {
      "epoch": 0.13108614232209737,
      "grad_norm": 3.7720329761505127,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.4226,
      "step": 35
    },
    {
      "epoch": 0.1348314606741573,
      "grad_norm": 3.4572339057922363,
      "learning_rate": 9.545032675245813e-05,
      "loss": 4.5433,
      "step": 36
    },
    {
      "epoch": 0.13857677902621723,
      "grad_norm": 3.5973665714263916,
      "learning_rate": 9.509956150664796e-05,
      "loss": 4.9255,
      "step": 37
    },
    {
      "epoch": 0.14232209737827714,
      "grad_norm": 3.8061134815216064,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.6182,
      "step": 38
    },
    {
      "epoch": 0.14606741573033707,
      "grad_norm": 3.860398054122925,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.185,
      "step": 39
    },
    {
      "epoch": 0.149812734082397,
      "grad_norm": 3.984393835067749,
      "learning_rate": 9.397368756032445e-05,
      "loss": 4.5924,
      "step": 40
    },
    {
      "epoch": 0.15355805243445692,
      "grad_norm": 3.7674689292907715,
      "learning_rate": 9.357421218136386e-05,
      "loss": 4.8202,
      "step": 41
    },
    {
      "epoch": 0.15730337078651685,
      "grad_norm": 3.472353219985962,
      "learning_rate": 9.316282404787871e-05,
      "loss": 4.4853,
      "step": 42
    },
    {
      "epoch": 0.16104868913857678,
      "grad_norm": 4.037704944610596,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.1238,
      "step": 43
    },
    {
      "epoch": 0.1647940074906367,
      "grad_norm": 4.040994644165039,
      "learning_rate": 9.230476262104677e-05,
      "loss": 5.0098,
      "step": 44
    },
    {
      "epoch": 0.16853932584269662,
      "grad_norm": 4.288521766662598,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.9402,
      "step": 45
    },
    {
      "epoch": 0.17228464419475656,
      "grad_norm": 3.9272849559783936,
      "learning_rate": 9.140044155740101e-05,
      "loss": 4.7909,
      "step": 46
    },
    {
      "epoch": 0.1760299625468165,
      "grad_norm": 4.156335830688477,
      "learning_rate": 9.093124073433463e-05,
      "loss": 4.3985,
      "step": 47
    },
    {
      "epoch": 0.1797752808988764,
      "grad_norm": 4.706430912017822,
      "learning_rate": 9.045084971874738e-05,
      "loss": 5.7096,
      "step": 48
    },
    {
      "epoch": 0.18352059925093633,
      "grad_norm": 5.288496494293213,
      "learning_rate": 8.995939984474624e-05,
      "loss": 5.8452,
      "step": 49
    },
    {
      "epoch": 0.18726591760299627,
      "grad_norm": 6.929520606994629,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.6856,
      "step": 50
    },
    {
      "epoch": 0.18726591760299627,
      "eval_loss": 1.0981673002243042,
      "eval_runtime": 50.5726,
      "eval_samples_per_second": 8.898,
      "eval_steps_per_second": 2.234,
      "step": 50
    },
    {
      "epoch": 0.19101123595505617,
      "grad_norm": 4.772440433502197,
      "learning_rate": 8.894386393810563e-05,
      "loss": 5.8398,
      "step": 51
    },
    {
      "epoch": 0.1947565543071161,
      "grad_norm": 4.624301433563232,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.0618,
      "step": 52
    },
    {
      "epoch": 0.19850187265917604,
      "grad_norm": 3.4585633277893066,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.9354,
      "step": 53
    },
    {
      "epoch": 0.20224719101123595,
      "grad_norm": 3.079925775527954,
      "learning_rate": 8.73410738492077e-05,
      "loss": 4.4617,
      "step": 54
    },
    {
      "epoch": 0.20599250936329588,
      "grad_norm": 3.027653932571411,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.6423,
      "step": 55
    },
    {
      "epoch": 0.20973782771535582,
      "grad_norm": 3.3956668376922607,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.433,
      "step": 56
    },
    {
      "epoch": 0.21348314606741572,
      "grad_norm": 3.299530029296875,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.8912,
      "step": 57
    },
    {
      "epoch": 0.21722846441947566,
      "grad_norm": 3.4580984115600586,
      "learning_rate": 8.506183921362443e-05,
      "loss": 4.0882,
      "step": 58
    },
    {
      "epoch": 0.2209737827715356,
      "grad_norm": 3.2476744651794434,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.6116,
      "step": 59
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 3.6034555435180664,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.8025,
      "step": 60
    },
    {
      "epoch": 0.22846441947565543,
      "grad_norm": 3.3423707485198975,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.7032,
      "step": 61
    },
    {
      "epoch": 0.23220973782771537,
      "grad_norm": 3.121155261993408,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.3392,
      "step": 62
    },
    {
      "epoch": 0.23595505617977527,
      "grad_norm": 3.097158670425415,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.6554,
      "step": 63
    },
    {
      "epoch": 0.2397003745318352,
      "grad_norm": 3.092902421951294,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.6018,
      "step": 64
    },
    {
      "epoch": 0.24344569288389514,
      "grad_norm": 3.1121206283569336,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.1765,
      "step": 65
    },
    {
      "epoch": 0.24719101123595505,
      "grad_norm": 2.7471923828125,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.699,
      "step": 66
    },
    {
      "epoch": 0.250936329588015,
      "grad_norm": 3.1210672855377197,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.8364,
      "step": 67
    },
    {
      "epoch": 0.2546816479400749,
      "grad_norm": 3.3711516857147217,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.5648,
      "step": 68
    },
    {
      "epoch": 0.25842696629213485,
      "grad_norm": 3.4253804683685303,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.4885,
      "step": 69
    },
    {
      "epoch": 0.26217228464419473,
      "grad_norm": 3.4630062580108643,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.5128,
      "step": 70
    },
    {
      "epoch": 0.26591760299625467,
      "grad_norm": 3.519139528274536,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.3862,
      "step": 71
    },
    {
      "epoch": 0.2696629213483146,
      "grad_norm": 3.0610082149505615,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.7479,
      "step": 72
    },
    {
      "epoch": 0.27340823970037453,
      "grad_norm": 3.6356093883514404,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.6488,
      "step": 73
    },
    {
      "epoch": 0.27715355805243447,
      "grad_norm": 3.447793483734131,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.5451,
      "step": 74
    },
    {
      "epoch": 0.2808988764044944,
      "grad_norm": 3.7588632106781006,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.859,
      "step": 75
    },
    {
      "epoch": 0.2846441947565543,
      "grad_norm": 4.036433219909668,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.9972,
      "step": 76
    },
    {
      "epoch": 0.2883895131086142,
      "grad_norm": 3.5770747661590576,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.3789,
      "step": 77
    },
    {
      "epoch": 0.29213483146067415,
      "grad_norm": 3.880018472671509,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.4011,
      "step": 78
    },
    {
      "epoch": 0.2958801498127341,
      "grad_norm": 4.561016082763672,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.3364,
      "step": 79
    },
    {
      "epoch": 0.299625468164794,
      "grad_norm": 4.0835347175598145,
      "learning_rate": 7.008477123264848e-05,
      "loss": 4.1853,
      "step": 80
    },
    {
      "epoch": 0.30337078651685395,
      "grad_norm": 4.813183307647705,
      "learning_rate": 6.932495846462261e-05,
      "loss": 4.111,
      "step": 81
    },
    {
      "epoch": 0.30711610486891383,
      "grad_norm": 3.8731846809387207,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.9134,
      "step": 82
    },
    {
      "epoch": 0.31086142322097376,
      "grad_norm": 4.023487567901611,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.336,
      "step": 83
    },
    {
      "epoch": 0.3146067415730337,
      "grad_norm": 3.485626697540283,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.8962,
      "step": 84
    },
    {
      "epoch": 0.31835205992509363,
      "grad_norm": 3.506251811981201,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.3875,
      "step": 85
    },
    {
      "epoch": 0.32209737827715357,
      "grad_norm": 3.8715169429779053,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.6478,
      "step": 86
    },
    {
      "epoch": 0.3258426966292135,
      "grad_norm": 3.2170298099517822,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.5025,
      "step": 87
    },
    {
      "epoch": 0.3295880149812734,
      "grad_norm": 3.785874843597412,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.5703,
      "step": 88
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.7668328285217285,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.4,
      "step": 89
    },
    {
      "epoch": 0.33707865168539325,
      "grad_norm": 4.234432697296143,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.7084,
      "step": 90
    },
    {
      "epoch": 0.3408239700374532,
      "grad_norm": 4.549925804138184,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.1694,
      "step": 91
    },
    {
      "epoch": 0.3445692883895131,
      "grad_norm": 4.165473461151123,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.5118,
      "step": 92
    },
    {
      "epoch": 0.34831460674157305,
      "grad_norm": 3.980201244354248,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.0045,
      "step": 93
    },
    {
      "epoch": 0.352059925093633,
      "grad_norm": 4.744318962097168,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.6898,
      "step": 94
    },
    {
      "epoch": 0.35580524344569286,
      "grad_norm": 4.224009037017822,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.8446,
      "step": 95
    },
    {
      "epoch": 0.3595505617977528,
      "grad_norm": 4.371734619140625,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.4819,
      "step": 96
    },
    {
      "epoch": 0.36329588014981273,
      "grad_norm": 5.047107219696045,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 4.0622,
      "step": 97
    },
    {
      "epoch": 0.36704119850187267,
      "grad_norm": 5.079836368560791,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.7141,
      "step": 98
    },
    {
      "epoch": 0.3707865168539326,
      "grad_norm": 4.755043029785156,
      "learning_rate": 5.495227651252315e-05,
      "loss": 4.0644,
      "step": 99
    },
    {
      "epoch": 0.37453183520599254,
      "grad_norm": 6.704141616821289,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 4.3662,
      "step": 100
    },
    {
      "epoch": 0.37453183520599254,
      "eval_loss": 0.7878959774971008,
      "eval_runtime": 50.5534,
      "eval_samples_per_second": 8.901,
      "eval_steps_per_second": 2.235,
      "step": 100
    },
    {
      "epoch": 0.3782771535580524,
      "grad_norm": 4.3849029541015625,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.3864,
      "step": 101
    },
    {
      "epoch": 0.38202247191011235,
      "grad_norm": 4.381369113922119,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.1775,
      "step": 102
    },
    {
      "epoch": 0.3857677902621723,
      "grad_norm": 3.4026999473571777,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.0201,
      "step": 103
    },
    {
      "epoch": 0.3895131086142322,
      "grad_norm": 3.540910243988037,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.0822,
      "step": 104
    },
    {
      "epoch": 0.39325842696629215,
      "grad_norm": 3.5141866207122803,
      "learning_rate": 5e-05,
      "loss": 3.5209,
      "step": 105
    },
    {
      "epoch": 0.3970037453183521,
      "grad_norm": 3.0794005393981934,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.5954,
      "step": 106
    },
    {
      "epoch": 0.40074906367041196,
      "grad_norm": 3.5028674602508545,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.018,
      "step": 107
    },
    {
      "epoch": 0.4044943820224719,
      "grad_norm": 3.8096418380737305,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.1039,
      "step": 108
    },
    {
      "epoch": 0.40823970037453183,
      "grad_norm": 3.150365114212036,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.3594,
      "step": 109
    },
    {
      "epoch": 0.41198501872659177,
      "grad_norm": 3.2571616172790527,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.5329,
      "step": 110
    },
    {
      "epoch": 0.4157303370786517,
      "grad_norm": 4.003403186798096,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.7547,
      "step": 111
    },
    {
      "epoch": 0.41947565543071164,
      "grad_norm": 3.245961904525757,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.2481,
      "step": 112
    },
    {
      "epoch": 0.4232209737827715,
      "grad_norm": 3.2729058265686035,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.4061,
      "step": 113
    },
    {
      "epoch": 0.42696629213483145,
      "grad_norm": 3.7617392539978027,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.1183,
      "step": 114
    },
    {
      "epoch": 0.4307116104868914,
      "grad_norm": 3.2042899131774902,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.6626,
      "step": 115
    },
    {
      "epoch": 0.4344569288389513,
      "grad_norm": 3.203579902648926,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.1462,
      "step": 116
    },
    {
      "epoch": 0.43820224719101125,
      "grad_norm": 3.5090222358703613,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.8969,
      "step": 117
    },
    {
      "epoch": 0.4419475655430712,
      "grad_norm": 3.3341238498687744,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.2154,
      "step": 118
    },
    {
      "epoch": 0.44569288389513106,
      "grad_norm": 3.153210401535034,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.3241,
      "step": 119
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 3.6039068698883057,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.8251,
      "step": 120
    },
    {
      "epoch": 0.45318352059925093,
      "grad_norm": 3.402194023132324,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.1318,
      "step": 121
    },
    {
      "epoch": 0.45692883895131087,
      "grad_norm": 3.688833713531494,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.3396,
      "step": 122
    },
    {
      "epoch": 0.4606741573033708,
      "grad_norm": 3.7565200328826904,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.5633,
      "step": 123
    },
    {
      "epoch": 0.46441947565543074,
      "grad_norm": 3.4246175289154053,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.5781,
      "step": 124
    },
    {
      "epoch": 0.4681647940074906,
      "grad_norm": 4.295012950897217,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.7718,
      "step": 125
    },
    {
      "epoch": 0.47191011235955055,
      "grad_norm": 3.7380805015563965,
      "learning_rate": 3.298534127791785e-05,
      "loss": 3.0312,
      "step": 126
    },
    {
      "epoch": 0.4756554307116105,
      "grad_norm": 4.162894248962402,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.0144,
      "step": 127
    },
    {
      "epoch": 0.4794007490636704,
      "grad_norm": 3.6485819816589355,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.005,
      "step": 128
    },
    {
      "epoch": 0.48314606741573035,
      "grad_norm": 3.7428202629089355,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.8231,
      "step": 129
    },
    {
      "epoch": 0.4868913857677903,
      "grad_norm": 3.988053321838379,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.0075,
      "step": 130
    },
    {
      "epoch": 0.49063670411985016,
      "grad_norm": 4.340529918670654,
      "learning_rate": 2.916090697523549e-05,
      "loss": 4.0483,
      "step": 131
    },
    {
      "epoch": 0.4943820224719101,
      "grad_norm": 3.8490538597106934,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.1499,
      "step": 132
    },
    {
      "epoch": 0.49812734082397003,
      "grad_norm": 4.094542980194092,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.4696,
      "step": 133
    },
    {
      "epoch": 0.50187265917603,
      "grad_norm": 4.395052909851074,
      "learning_rate": 2.693294185106562e-05,
      "loss": 3.5324,
      "step": 134
    },
    {
      "epoch": 0.5056179775280899,
      "grad_norm": 3.5157530307769775,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.76,
      "step": 135
    },
    {
      "epoch": 0.5093632958801498,
      "grad_norm": 3.7129011154174805,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.9146,
      "step": 136
    },
    {
      "epoch": 0.5131086142322098,
      "grad_norm": 3.6419291496276855,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.2427,
      "step": 137
    },
    {
      "epoch": 0.5168539325842697,
      "grad_norm": 4.271421909332275,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.6938,
      "step": 138
    },
    {
      "epoch": 0.5205992509363296,
      "grad_norm": 4.377735137939453,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.8335,
      "step": 139
    },
    {
      "epoch": 0.5243445692883895,
      "grad_norm": 4.233665466308594,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.8566,
      "step": 140
    },
    {
      "epoch": 0.5280898876404494,
      "grad_norm": 4.080112934112549,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.447,
      "step": 141
    },
    {
      "epoch": 0.5318352059925093,
      "grad_norm": 3.8475124835968018,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.7245,
      "step": 142
    },
    {
      "epoch": 0.5355805243445693,
      "grad_norm": 4.17799186706543,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.7011,
      "step": 143
    },
    {
      "epoch": 0.5393258426966292,
      "grad_norm": 5.750624179840088,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.2048,
      "step": 144
    },
    {
      "epoch": 0.5430711610486891,
      "grad_norm": 4.476280689239502,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.8153,
      "step": 145
    },
    {
      "epoch": 0.5468164794007491,
      "grad_norm": 4.526669025421143,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.5324,
      "step": 146
    },
    {
      "epoch": 0.550561797752809,
      "grad_norm": 5.280111789703369,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.2852,
      "step": 147
    },
    {
      "epoch": 0.5543071161048689,
      "grad_norm": 4.681980133056641,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.7674,
      "step": 148
    },
    {
      "epoch": 0.5580524344569289,
      "grad_norm": 5.281397819519043,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.9221,
      "step": 149
    },
    {
      "epoch": 0.5617977528089888,
      "grad_norm": 6.8536577224731445,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.5215,
      "step": 150
    },
    {
      "epoch": 0.5617977528089888,
      "eval_loss": 0.6514344811439514,
      "eval_runtime": 50.6582,
      "eval_samples_per_second": 8.883,
      "eval_steps_per_second": 2.231,
      "step": 150
    },
    {
      "epoch": 0.5655430711610487,
      "grad_norm": 4.77299690246582,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.8374,
      "step": 151
    },
    {
      "epoch": 0.5692883895131086,
      "grad_norm": 4.288309097290039,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.848,
      "step": 152
    },
    {
      "epoch": 0.5730337078651685,
      "grad_norm": 4.446907043457031,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.188,
      "step": 153
    },
    {
      "epoch": 0.5767790262172284,
      "grad_norm": 3.7621068954467773,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.1382,
      "step": 154
    },
    {
      "epoch": 0.5805243445692884,
      "grad_norm": 3.652517557144165,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.8973,
      "step": 155
    },
    {
      "epoch": 0.5842696629213483,
      "grad_norm": 3.5132782459259033,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.0814,
      "step": 156
    },
    {
      "epoch": 0.5880149812734082,
      "grad_norm": 3.717902898788452,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.3438,
      "step": 157
    },
    {
      "epoch": 0.5917602996254682,
      "grad_norm": 3.7370717525482178,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.9504,
      "step": 158
    },
    {
      "epoch": 0.5955056179775281,
      "grad_norm": 4.5406928062438965,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.7605,
      "step": 159
    },
    {
      "epoch": 0.599250936329588,
      "grad_norm": 3.3446907997131348,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.6698,
      "step": 160
    },
    {
      "epoch": 0.602996254681648,
      "grad_norm": 3.2716920375823975,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.0192,
      "step": 161
    },
    {
      "epoch": 0.6067415730337079,
      "grad_norm": 3.209780693054199,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.0181,
      "step": 162
    },
    {
      "epoch": 0.6104868913857678,
      "grad_norm": 3.5034663677215576,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.1013,
      "step": 163
    },
    {
      "epoch": 0.6142322097378277,
      "grad_norm": 3.9527788162231445,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.7449,
      "step": 164
    },
    {
      "epoch": 0.6179775280898876,
      "grad_norm": 3.569793939590454,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.1479,
      "step": 165
    },
    {
      "epoch": 0.6217228464419475,
      "grad_norm": 3.703397274017334,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.8801,
      "step": 166
    },
    {
      "epoch": 0.6254681647940075,
      "grad_norm": 3.6781091690063477,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.5607,
      "step": 167
    },
    {
      "epoch": 0.6292134831460674,
      "grad_norm": 3.8312060832977295,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.1987,
      "step": 168
    },
    {
      "epoch": 0.6329588014981273,
      "grad_norm": 4.6357622146606445,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.3411,
      "step": 169
    },
    {
      "epoch": 0.6367041198501873,
      "grad_norm": 3.568791389465332,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.5847,
      "step": 170
    },
    {
      "epoch": 0.6404494382022472,
      "grad_norm": 4.259144306182861,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.0933,
      "step": 171
    },
    {
      "epoch": 0.6441947565543071,
      "grad_norm": 3.8538782596588135,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.7911,
      "step": 172
    },
    {
      "epoch": 0.6479400749063671,
      "grad_norm": 3.39943265914917,
      "learning_rate": 4.900438493352055e-06,
      "loss": 1.9059,
      "step": 173
    },
    {
      "epoch": 0.651685393258427,
      "grad_norm": 3.9557485580444336,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.5665,
      "step": 174
    },
    {
      "epoch": 0.6554307116104869,
      "grad_norm": 4.047401428222656,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.7648,
      "step": 175
    },
    {
      "epoch": 0.6591760299625468,
      "grad_norm": 3.798725128173828,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.7277,
      "step": 176
    },
    {
      "epoch": 0.6629213483146067,
      "grad_norm": 4.363707542419434,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.6307,
      "step": 177
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 3.7704315185546875,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.6627,
      "step": 178
    },
    {
      "epoch": 0.6704119850187266,
      "grad_norm": 4.179985523223877,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.4506,
      "step": 179
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 4.002692222595215,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.2558,
      "step": 180
    },
    {
      "epoch": 0.6779026217228464,
      "grad_norm": 4.406280994415283,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.6437,
      "step": 181
    },
    {
      "epoch": 0.6816479400749064,
      "grad_norm": 4.373383045196533,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.6488,
      "step": 182
    },
    {
      "epoch": 0.6853932584269663,
      "grad_norm": 4.354463577270508,
      "learning_rate": 1.962316193157593e-06,
      "loss": 3.0158,
      "step": 183
    },
    {
      "epoch": 0.6891385767790262,
      "grad_norm": 5.0767083168029785,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.0873,
      "step": 184
    },
    {
      "epoch": 0.6928838951310862,
      "grad_norm": 4.79359769821167,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.7056,
      "step": 185
    },
    {
      "epoch": 0.6966292134831461,
      "grad_norm": 4.0298357009887695,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.3263,
      "step": 186
    },
    {
      "epoch": 0.700374531835206,
      "grad_norm": 4.407674312591553,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.5261,
      "step": 187
    },
    {
      "epoch": 0.704119850187266,
      "grad_norm": 3.990910530090332,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.5277,
      "step": 188
    },
    {
      "epoch": 0.7078651685393258,
      "grad_norm": 4.079568386077881,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.7199,
      "step": 189
    },
    {
      "epoch": 0.7116104868913857,
      "grad_norm": 4.378945827484131,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.4524,
      "step": 190
    },
    {
      "epoch": 0.7153558052434457,
      "grad_norm": 4.373382568359375,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.5915,
      "step": 191
    },
    {
      "epoch": 0.7191011235955056,
      "grad_norm": 5.087827205657959,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.2832,
      "step": 192
    },
    {
      "epoch": 0.7228464419475655,
      "grad_norm": 4.235489845275879,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.3853,
      "step": 193
    },
    {
      "epoch": 0.7265917602996255,
      "grad_norm": 3.7132980823516846,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.7912,
      "step": 194
    },
    {
      "epoch": 0.7303370786516854,
      "grad_norm": 4.470889568328857,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.3165,
      "step": 195
    },
    {
      "epoch": 0.7340823970037453,
      "grad_norm": 5.161454677581787,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.7435,
      "step": 196
    },
    {
      "epoch": 0.7378277153558053,
      "grad_norm": 4.738089084625244,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.5381,
      "step": 197
    },
    {
      "epoch": 0.7415730337078652,
      "grad_norm": 5.514863967895508,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 3.5324,
      "step": 198
    },
    {
      "epoch": 0.7453183520599251,
      "grad_norm": 5.327063083648682,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.5509,
      "step": 199
    },
    {
      "epoch": 0.7490636704119851,
      "grad_norm": 5.934273719787598,
      "learning_rate": 0.0,
      "loss": 2.999,
      "step": 200
    },
    {
      "epoch": 0.7490636704119851,
      "eval_loss": 0.6181585192680359,
      "eval_runtime": 50.686,
      "eval_samples_per_second": 8.878,
      "eval_steps_per_second": 2.229,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.610833556013056e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}