{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9518328315919957,
  "eval_steps": 1000,
  "global_step": 21500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "grad_norm": 16.78230857849121,
      "learning_rate": 1.987315680989377e-05,
      "loss": 0.7008,
      "step": 1000
    },
    {
      "epoch": 0.04,
      "eval_LOC_f1": 0.7083043365359647,
      "eval_ORG_f1": 0.5623498694516972,
      "eval_PER_f1": 0.7100647964924535,
      "eval_loss": 0.525841236114502,
      "eval_overall_accuracy": 0.839431747364137,
      "eval_overall_f1": 0.6684494885258021,
      "eval_overall_precision": 0.6813296398891967,
      "eval_overall_recall": 0.656047285763059,
      "eval_runtime": 178.5538,
      "eval_samples_per_second": 365.156,
      "eval_steps_per_second": 22.822,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.70958137512207,
      "learning_rate": 1.974631361978754e-05,
      "loss": 0.4865,
      "step": 2000
    },
    {
      "epoch": 0.09,
      "eval_LOC_f1": 0.7016948849995092,
      "eval_ORG_f1": 0.590386287417712,
      "eval_PER_f1": 0.744081694250214,
      "eval_loss": 0.4681616723537445,
      "eval_overall_accuracy": 0.8518320678647284,
      "eval_overall_f1": 0.6854302821050324,
      "eval_overall_precision": 0.6610573746663495,
      "eval_overall_recall": 0.7116692272675543,
      "eval_runtime": 178.958,
      "eval_samples_per_second": 364.331,
      "eval_steps_per_second": 22.771,
      "step": 2000
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.70914077758789,
      "learning_rate": 1.961947042968131e-05,
      "loss": 0.4382,
      "step": 3000
    },
    {
      "epoch": 0.13,
      "eval_LOC_f1": 0.7578888199720886,
      "eval_ORG_f1": 0.6413910093299406,
      "eval_PER_f1": 0.7750323494978126,
      "eval_loss": 0.4217945635318756,
      "eval_overall_accuracy": 0.8756332018861953,
      "eval_overall_f1": 0.7298225286836603,
      "eval_overall_precision": 0.7249466651695486,
      "eval_overall_recall": 0.7347644247183339,
      "eval_runtime": 177.7042,
      "eval_samples_per_second": 366.902,
      "eval_steps_per_second": 22.931,
      "step": 3000
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.9930562376976013,
      "learning_rate": 1.9492627239575078e-05,
      "loss": 0.3973,
      "step": 4000
    },
    {
      "epoch": 0.18,
      "eval_LOC_f1": 0.7571125548125672,
      "eval_ORG_f1": 0.648735892760337,
      "eval_PER_f1": 0.7869654790165547,
      "eval_loss": 0.37678804993629456,
      "eval_overall_accuracy": 0.8835828734179773,
      "eval_overall_f1": 0.7360367198972751,
      "eval_overall_precision": 0.7341253555609017,
      "eval_overall_recall": 0.7379580630476841,
      "eval_runtime": 176.7693,
      "eval_samples_per_second": 368.842,
      "eval_steps_per_second": 23.053,
      "step": 4000
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.682718276977539,
      "learning_rate": 1.9365784049468845e-05,
      "loss": 0.3937,
      "step": 5000
    },
    {
      "epoch": 0.22,
      "eval_LOC_f1": 0.7778496267121274,
      "eval_ORG_f1": 0.6430323934146428,
      "eval_PER_f1": 0.7986373262130558,
      "eval_loss": 0.4208361804485321,
      "eval_overall_accuracy": 0.8810407141986101,
      "eval_overall_f1": 0.7449750865636349,
      "eval_overall_precision": 0.7371928798261742,
      "eval_overall_recall": 0.7529233526800956,
      "eval_runtime": 175.2138,
      "eval_samples_per_second": 372.117,
      "eval_steps_per_second": 23.257,
      "step": 5000
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.084446907043457,
      "learning_rate": 1.924106352298313e-05,
      "loss": 0.3536,
      "step": 6000
    },
    {
      "epoch": 0.27,
      "eval_LOC_f1": 0.7828708236001125,
      "eval_ORG_f1": 0.6829649863421904,
      "eval_PER_f1": 0.799357419856545,
      "eval_loss": 0.36626189947128296,
      "eval_overall_accuracy": 0.8907383203725096,
      "eval_overall_f1": 0.760005685048323,
      "eval_overall_precision": 0.7663556759390631,
      "eval_overall_recall": 0.7537600608939571,
      "eval_runtime": 1096.1574,
      "eval_samples_per_second": 59.937,
      "eval_steps_per_second": 0.118,
      "step": 6000
    },
    {
      "epoch": 0.31,
      "grad_norm": 17.549041748046875,
      "learning_rate": 1.9114574110146982e-05,
      "loss": 0.3432,
      "step": 7000
    },
    {
      "epoch": 0.31,
      "eval_LOC_f1": 0.7952019006963426,
      "eval_ORG_f1": 0.6895359096523734,
      "eval_PER_f1": 0.8176749430952034,
      "eval_loss": 0.3385719060897827,
      "eval_overall_accuracy": 0.8962171463957795,
      "eval_overall_f1": 0.771180360157569,
      "eval_overall_precision": 0.7696968420461406,
      "eval_overall_recall": 0.772669607995151,
      "eval_runtime": 858.9522,
      "eval_samples_per_second": 76.489,
      "eval_steps_per_second": 0.15,
      "step": 7000
    },
    {
      "epoch": 0.35,
      "grad_norm": 24.70513916015625,
      "learning_rate": 1.8988084697310836e-05,
      "loss": 0.3278,
      "step": 8000
    },
    {
      "epoch": 0.35,
      "eval_LOC_f1": 0.8013627254509018,
      "eval_ORG_f1": 0.6951269959460578,
      "eval_PER_f1": 0.8197644808913556,
      "eval_loss": 0.34438127279281616,
      "eval_overall_accuracy": 0.898079396458301,
      "eval_overall_f1": 0.7756412540713129,
      "eval_overall_precision": 0.7844467189482182,
      "eval_overall_recall": 0.7670312786321413,
      "eval_runtime": 858.3259,
      "eval_samples_per_second": 76.544,
      "eval_steps_per_second": 0.15,
      "step": 8000
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.711076736450195,
      "learning_rate": 1.8861595284474693e-05,
      "loss": 0.3262,
      "step": 9000
    },
    {
      "epoch": 0.4,
      "eval_LOC_f1": 0.813213095471476,
      "eval_ORG_f1": 0.7107696972759894,
      "eval_PER_f1": 0.8219371637919723,
      "eval_loss": 0.3584470748901367,
      "eval_overall_accuracy": 0.898605619797311,
      "eval_overall_f1": 0.7873656115410829,
      "eval_overall_precision": 0.7973837314349728,
      "eval_overall_recall": 0.7775960982760808,
      "eval_runtime": 870.9363,
      "eval_samples_per_second": 75.436,
      "eval_steps_per_second": 0.148,
      "step": 9000
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0706424713134766,
      "learning_rate": 1.8735105871638546e-05,
      "loss": 0.3224,
      "step": 10000
    },
    {
      "epoch": 0.44,
      "eval_LOC_f1": 0.8127321363277717,
      "eval_ORG_f1": 0.7053217033370902,
      "eval_PER_f1": 0.8284614520690605,
      "eval_loss": 0.3149181008338928,
      "eval_overall_accuracy": 0.9003740874825833,
      "eval_overall_f1": 0.78723937113973,
      "eval_overall_precision": 0.7880037567703074,
      "eval_overall_recall": 0.7864764670228211,
      "eval_runtime": 905.7017,
      "eval_samples_per_second": 72.54,
      "eval_steps_per_second": 0.142,
      "step": 10000
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.779551982879639,
      "learning_rate": 1.86086164588024e-05,
      "loss": 0.323,
      "step": 11000
    },
    {
      "epoch": 0.49,
      "eval_LOC_f1": 0.8184425456746753,
      "eval_ORG_f1": 0.7099403917116094,
      "eval_PER_f1": 0.8274642760571583,
      "eval_loss": 0.32969576120376587,
      "eval_overall_accuracy": 0.8986934477379095,
      "eval_overall_f1": 0.787954918665525,
      "eval_overall_precision": 0.7845852840472364,
      "eval_overall_recall": 0.7913536219218246,
      "eval_runtime": 904.0471,
      "eval_samples_per_second": 72.673,
      "eval_steps_per_second": 0.143,
      "step": 11000
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5029723048210144,
      "learning_rate": 1.8482127045966253e-05,
      "loss": 0.307,
      "step": 12000
    },
    {
      "epoch": 0.53,
      "eval_LOC_f1": 0.8216145583120352,
      "eval_ORG_f1": 0.7225526656894693,
      "eval_PER_f1": 0.8273035359038698,
      "eval_loss": 0.3300624191761017,
      "eval_overall_accuracy": 0.9015344583248979,
      "eval_overall_f1": 0.7950728742961997,
      "eval_overall_precision": 0.7871480673058326,
      "eval_overall_recall": 0.8031588740256262,
      "eval_runtime": 864.0266,
      "eval_samples_per_second": 76.039,
      "eval_steps_per_second": 0.149,
      "step": 12000
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.1533098220825195,
      "learning_rate": 1.8355637633130106e-05,
      "loss": 0.3176,
      "step": 13000
    },
    {
      "epoch": 0.58,
      "eval_LOC_f1": 0.8250715828125148,
      "eval_ORG_f1": 0.7192632496804788,
      "eval_PER_f1": 0.8411864068909802,
      "eval_loss": 0.32228201627731323,
      "eval_overall_accuracy": 0.9048257731835991,
      "eval_overall_f1": 0.7993031971987119,
      "eval_overall_precision": 0.7982273018015168,
      "eval_overall_recall": 0.8003819968143439,
      "eval_runtime": 861.5394,
      "eval_samples_per_second": 76.259,
      "eval_steps_per_second": 0.15,
      "step": 13000
    },
    {
      "epoch": 0.62,
      "grad_norm": 27.121158599853516,
      "learning_rate": 1.8229148220293963e-05,
      "loss": 0.3026,
      "step": 14000
    },
    {
      "epoch": 0.62,
      "eval_LOC_f1": 0.813366057700572,
      "eval_ORG_f1": 0.7255497858288067,
      "eval_PER_f1": 0.8334248928753359,
      "eval_loss": 0.31955230236053467,
      "eval_overall_accuracy": 0.9049277429112431,
      "eval_overall_f1": 0.7948808632215472,
      "eval_overall_precision": 0.7823057288325302,
      "eval_overall_recall": 0.8078668790437393,
      "eval_runtime": 870.9305,
      "eval_samples_per_second": 75.437,
      "eval_steps_per_second": 0.148,
      "step": 14000
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.3121883869171143,
      "learning_rate": 1.8102658807457817e-05,
      "loss": 0.3148,
      "step": 15000
    },
    {
      "epoch": 0.66,
      "eval_LOC_f1": 0.8302503673286692,
      "eval_ORG_f1": 0.7304969576437721,
      "eval_PER_f1": 0.8389894805913836,
      "eval_loss": 0.3013811409473419,
      "eval_overall_accuracy": 0.9075328089459457,
      "eval_overall_f1": 0.8036533849292605,
      "eval_overall_precision": 0.8008272039526633,
      "eval_overall_recall": 0.8064995841732094,
      "eval_runtime": 915.5251,
      "eval_samples_per_second": 71.762,
      "eval_steps_per_second": 0.141,
      "step": 15000
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.053048133850098,
      "learning_rate": 1.7976169394621674e-05,
      "loss": 0.2919,
      "step": 16000
    },
    {
      "epoch": 0.71,
      "eval_LOC_f1": 0.8302478759061501,
      "eval_ORG_f1": 0.731782075129564,
      "eval_PER_f1": 0.8492362049256988,
      "eval_loss": 0.29880577325820923,
      "eval_overall_accuracy": 0.9072484845958724,
      "eval_overall_f1": 0.8070731973229327,
      "eval_overall_precision": 0.8054613737671545,
      "eval_overall_recall": 0.8086914847130795,
      "eval_runtime": 868.8985,
      "eval_samples_per_second": 75.613,
      "eval_steps_per_second": 0.148,
      "step": 16000
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.181853294372559,
      "learning_rate": 1.7849679981785527e-05,
      "loss": 0.284,
      "step": 17000
    },
    {
      "epoch": 0.75,
      "eval_LOC_f1": 0.8266632184651193,
      "eval_ORG_f1": 0.7243731835870154,
      "eval_PER_f1": 0.8473984474857374,
      "eval_loss": 0.32140275835990906,
      "eval_overall_accuracy": 0.9021306462945541,
      "eval_overall_f1": 0.8045739646627278,
      "eval_overall_precision": 0.7991697840325968,
      "eval_overall_recall": 0.8100517316719056,
      "eval_runtime": 858.2577,
      "eval_samples_per_second": 76.55,
      "eval_steps_per_second": 0.15,
      "step": 17000
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.25670862197876,
      "learning_rate": 1.772319056894938e-05,
      "loss": 0.3011,
      "step": 18000
    },
    {
      "epoch": 0.8,
      "eval_LOC_f1": 0.8265056399966245,
      "eval_ORG_f1": 0.7394759212033855,
      "eval_PER_f1": 0.8559909523111401,
      "eval_loss": 0.30769339203834534,
      "eval_overall_accuracy": 0.9103075764251944,
      "eval_overall_f1": 0.8115094912358725,
      "eval_overall_precision": 0.8128234861554996,
      "eval_overall_recall": 0.8101997378176846,
      "eval_runtime": 857.6527,
      "eval_samples_per_second": 76.604,
      "eval_steps_per_second": 0.15,
      "step": 18000
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.439514398574829,
      "learning_rate": 1.7596701156113234e-05,
      "loss": 0.2913,
      "step": 19000
    },
    {
      "epoch": 0.84,
      "eval_LOC_f1": 0.827463843173361,
      "eval_ORG_f1": 0.7474629390110435,
      "eval_PER_f1": 0.8510804051006602,
      "eval_loss": 0.3055616617202759,
      "eval_overall_accuracy": 0.9091010586988365,
      "eval_overall_f1": 0.8128348309192438,
      "eval_overall_precision": 0.817294400980434,
      "eval_overall_recall": 0.8084236640683365,
      "eval_runtime": 881.3321,
      "eval_samples_per_second": 74.546,
      "eval_steps_per_second": 0.146,
      "step": 19000
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.887170791625977,
      "learning_rate": 1.747021174327709e-05,
      "loss": 0.2927,
      "step": 20000
    },
    {
      "epoch": 0.89,
      "eval_LOC_f1": 0.8435358479700547,
      "eval_ORG_f1": 0.7490680162942847,
      "eval_PER_f1": 0.859867877786953,
      "eval_loss": 0.28957998752593994,
      "eval_overall_accuracy": 0.9142888616308011,
      "eval_overall_f1": 0.8220150311968236,
      "eval_overall_precision": 0.8269754768392371,
      "eval_overall_recall": 0.8171137391990753,
      "eval_runtime": 908.0817,
      "eval_samples_per_second": 72.35,
      "eval_steps_per_second": 0.142,
      "step": 20000
    },
    {
      "epoch": 0.93,
      "grad_norm": 24.44190216064453,
      "learning_rate": 1.7343722330440944e-05,
      "loss": 0.2833,
      "step": 21000
    },
    {
      "epoch": 0.93,
      "eval_LOC_f1": 0.8368153269843636,
      "eval_ORG_f1": 0.7463167531278608,
      "eval_PER_f1": 0.8518246438142143,
      "eval_loss": 0.3060026466846466,
      "eval_overall_accuracy": 0.9097895404365793,
      "eval_overall_f1": 0.8159558988135214,
      "eval_overall_precision": 0.8099609727642672,
      "eval_overall_recall": 0.8220402294800051,
      "eval_runtime": 860.9742,
      "eval_samples_per_second": 76.309,
      "eval_steps_per_second": 0.15,
      "step": 21000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 158116,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 5913608888301312.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}