{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 10000,
  "global_step": 126850,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.9999842333464724e-05,
      "loss": 0.2456,
      "step": 1
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.9842333464722114e-05,
      "loss": 0.1742,
      "step": 1000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.9685297595585336e-05,
      "loss": 0.1345,
      "step": 2000
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.952763106030745e-05,
      "loss": 0.0825,
      "step": 3000
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.9370122191564842e-05,
      "loss": 0.0691,
      "step": 4000
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.9212455656286955e-05,
      "loss": 0.0593,
      "step": 5000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.9054789121009067e-05,
      "loss": 0.0559,
      "step": 6000
    },
    {
      "epoch": 2.76,
      "learning_rate": 1.889712258573118e-05,
      "loss": 0.0505,
      "step": 7000
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.8739456050453293e-05,
      "loss": 0.0585,
      "step": 8000
    },
    {
      "epoch": 3.55,
      "learning_rate": 1.8581789515175405e-05,
      "loss": 0.0496,
      "step": 9000
    },
    {
      "epoch": 3.94,
      "learning_rate": 1.84242806464328e-05,
      "loss": 0.0504,
      "step": 10000
    },
    {
      "epoch": 3.94,
      "eval_accuracy": 0.9750283768444948,
      "eval_f1": 0.9112903225806451,
      "eval_loss": 0.02506309188902378,
      "eval_precision": 0.8692307692307693,
      "eval_recall": 0.9576271186440678,
      "eval_runtime": 6.3761,
      "eval_samples_per_second": 276.344,
      "eval_steps_per_second": 34.661,
      "step": 10000
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.8266614111154908e-05,
      "loss": 0.0442,
      "step": 11000
    },
    {
      "epoch": 4.73,
      "learning_rate": 1.8109105242412298e-05,
      "loss": 0.0467,
      "step": 12000
    },
    {
      "epoch": 5.12,
      "learning_rate": 1.795159637366969e-05,
      "loss": 0.0448,
      "step": 13000
    },
    {
      "epoch": 5.52,
      "learning_rate": 1.7793929838391804e-05,
      "loss": 0.0418,
      "step": 14000
    },
    {
      "epoch": 5.91,
      "learning_rate": 1.7636263303113916e-05,
      "loss": 0.0339,
      "step": 15000
    },
    {
      "epoch": 6.31,
      "learning_rate": 1.747859676783603e-05,
      "loss": 0.0385,
      "step": 16000
    },
    {
      "epoch": 6.7,
      "learning_rate": 1.732093023255814e-05,
      "loss": 0.0425,
      "step": 17000
    },
    {
      "epoch": 7.09,
      "learning_rate": 1.716342136381553e-05,
      "loss": 0.0413,
      "step": 18000
    },
    {
      "epoch": 7.49,
      "learning_rate": 1.7005754828537644e-05,
      "loss": 0.039,
      "step": 19000
    },
    {
      "epoch": 7.88,
      "learning_rate": 1.6848088293259757e-05,
      "loss": 0.0398,
      "step": 20000
    },
    {
      "epoch": 7.88,
      "eval_accuracy": 0.9750283768444948,
      "eval_f1": 0.9126984126984127,
      "eval_loss": 0.02478128857910633,
      "eval_precision": 0.8582089552238806,
      "eval_recall": 0.9745762711864406,
      "eval_runtime": 6.0449,
      "eval_samples_per_second": 291.487,
      "eval_steps_per_second": 36.56,
      "step": 20000
    },
    {
      "epoch": 8.28,
      "learning_rate": 1.6690737091052424e-05,
      "loss": 0.0347,
      "step": 21000
    },
    {
      "epoch": 8.67,
      "learning_rate": 1.6533070555774536e-05,
      "loss": 0.0428,
      "step": 22000
    },
    {
      "epoch": 9.07,
      "learning_rate": 1.637540402049665e-05,
      "loss": 0.0391,
      "step": 23000
    },
    {
      "epoch": 9.46,
      "learning_rate": 1.6217895151754042e-05,
      "loss": 0.0347,
      "step": 24000
    },
    {
      "epoch": 9.85,
      "learning_rate": 1.6060228616476155e-05,
      "loss": 0.0336,
      "step": 25000
    },
    {
      "epoch": 10.25,
      "learning_rate": 1.5902562081198268e-05,
      "loss": 0.033,
      "step": 26000
    },
    {
      "epoch": 10.64,
      "learning_rate": 1.574489554592038e-05,
      "loss": 0.0318,
      "step": 27000
    },
    {
      "epoch": 11.04,
      "learning_rate": 1.5587229010642493e-05,
      "loss": 0.0338,
      "step": 28000
    },
    {
      "epoch": 11.43,
      "learning_rate": 1.5429562475364605e-05,
      "loss": 0.0315,
      "step": 29000
    },
    {
      "epoch": 11.82,
      "learning_rate": 1.5272053606621995e-05,
      "loss": 0.0271,
      "step": 30000
    },
    {
      "epoch": 11.82,
      "eval_accuracy": 0.9903518728717366,
      "eval_f1": 0.9649484536082474,
      "eval_loss": 0.010149164125323296,
      "eval_precision": 0.9397590361445783,
      "eval_recall": 0.9915254237288136,
      "eval_runtime": 6.3243,
      "eval_samples_per_second": 278.608,
      "eval_steps_per_second": 34.945,
      "step": 30000
    },
    {
      "epoch": 12.22,
      "learning_rate": 1.5114387071344108e-05,
      "loss": 0.0343,
      "step": 31000
    },
    {
      "epoch": 12.61,
      "learning_rate": 1.49568782026015e-05,
      "loss": 0.0311,
      "step": 32000
    },
    {
      "epoch": 13.01,
      "learning_rate": 1.4799369333858891e-05,
      "loss": 0.0278,
      "step": 33000
    },
    {
      "epoch": 13.4,
      "learning_rate": 1.4641702798581004e-05,
      "loss": 0.027,
      "step": 34000
    },
    {
      "epoch": 13.8,
      "learning_rate": 1.4484036263303116e-05,
      "loss": 0.0285,
      "step": 35000
    },
    {
      "epoch": 14.19,
      "learning_rate": 1.4326527394560506e-05,
      "loss": 0.0303,
      "step": 36000
    },
    {
      "epoch": 14.58,
      "learning_rate": 1.4168860859282619e-05,
      "loss": 0.0253,
      "step": 37000
    },
    {
      "epoch": 14.98,
      "learning_rate": 1.401135199054001e-05,
      "loss": 0.0252,
      "step": 38000
    },
    {
      "epoch": 15.37,
      "learning_rate": 1.3853685455262123e-05,
      "loss": 0.027,
      "step": 39000
    },
    {
      "epoch": 15.77,
      "learning_rate": 1.3696018919984236e-05,
      "loss": 0.0257,
      "step": 40000
    },
    {
      "epoch": 15.77,
      "eval_accuracy": 0.9841089670828603,
      "eval_f1": 0.940677966101695,
      "eval_loss": 0.031208263710141182,
      "eval_precision": 0.940677966101695,
      "eval_recall": 0.940677966101695,
      "eval_runtime": 6.3572,
      "eval_samples_per_second": 277.168,
      "eval_steps_per_second": 34.764,
      "step": 40000
    },
    {
      "epoch": 16.16,
      "learning_rate": 1.3538352384706348e-05,
      "loss": 0.0252,
      "step": 41000
    },
    {
      "epoch": 16.55,
      "learning_rate": 1.3381001182499016e-05,
      "loss": 0.0272,
      "step": 42000
    },
    {
      "epoch": 16.95,
      "learning_rate": 1.3223334647221128e-05,
      "loss": 0.023,
      "step": 43000
    },
    {
      "epoch": 17.34,
      "learning_rate": 1.306566811194324e-05,
      "loss": 0.0254,
      "step": 44000
    },
    {
      "epoch": 17.74,
      "learning_rate": 1.290815924320063e-05,
      "loss": 0.0229,
      "step": 45000
    },
    {
      "epoch": 18.13,
      "learning_rate": 1.2750492707922743e-05,
      "loss": 0.0241,
      "step": 46000
    },
    {
      "epoch": 18.53,
      "learning_rate": 1.2592826172644856e-05,
      "loss": 0.0234,
      "step": 47000
    },
    {
      "epoch": 18.92,
      "learning_rate": 1.2435317303902248e-05,
      "loss": 0.0237,
      "step": 48000
    },
    {
      "epoch": 19.31,
      "learning_rate": 1.227765076862436e-05,
      "loss": 0.023,
      "step": 49000
    },
    {
      "epoch": 19.71,
      "learning_rate": 1.2119984233346473e-05,
      "loss": 0.0269,
      "step": 50000
    },
    {
      "epoch": 19.71,
      "eval_accuracy": 0.9948921679909194,
      "eval_f1": 0.9809725158562368,
      "eval_loss": 0.00856984406709671,
      "eval_precision": 0.9789029535864979,
      "eval_recall": 0.9830508474576272,
      "eval_runtime": 6.0459,
      "eval_samples_per_second": 291.435,
      "eval_steps_per_second": 36.553,
      "step": 50000
    },
    {
      "epoch": 20.1,
      "learning_rate": 1.1962317698068587e-05,
      "loss": 0.0213,
      "step": 51000
    },
    {
      "epoch": 20.5,
      "learning_rate": 1.18046511627907e-05,
      "loss": 0.0223,
      "step": 52000
    },
    {
      "epoch": 20.89,
      "learning_rate": 1.1647142294048088e-05,
      "loss": 0.0247,
      "step": 53000
    },
    {
      "epoch": 21.28,
      "learning_rate": 1.148963342530548e-05,
      "loss": 0.0181,
      "step": 54000
    },
    {
      "epoch": 21.68,
      "learning_rate": 1.1331966890027592e-05,
      "loss": 0.0198,
      "step": 55000
    },
    {
      "epoch": 22.07,
      "learning_rate": 1.1174300354749705e-05,
      "loss": 0.0216,
      "step": 56000
    },
    {
      "epoch": 22.47,
      "learning_rate": 1.1016791486007096e-05,
      "loss": 0.0214,
      "step": 57000
    },
    {
      "epoch": 22.86,
      "learning_rate": 1.0859124950729209e-05,
      "loss": 0.0208,
      "step": 58000
    },
    {
      "epoch": 23.26,
      "learning_rate": 1.0701773748521878e-05,
      "loss": 0.0174,
      "step": 59000
    },
    {
      "epoch": 23.65,
      "learning_rate": 1.054410721324399e-05,
      "loss": 0.0193,
      "step": 60000
    },
    {
      "epoch": 23.65,
      "eval_accuracy": 0.9943246311010215,
      "eval_f1": 0.9789029535864979,
      "eval_loss": 0.010104711167514324,
      "eval_precision": 0.9747899159663865,
      "eval_recall": 0.9830508474576272,
      "eval_runtime": 6.1461,
      "eval_samples_per_second": 286.688,
      "eval_steps_per_second": 35.958,
      "step": 60000
    },
    {
      "epoch": 24.04,
      "learning_rate": 1.0386440677966103e-05,
      "loss": 0.0204,
      "step": 61000
    },
    {
      "epoch": 24.44,
      "learning_rate": 1.0228774142688216e-05,
      "loss": 0.0189,
      "step": 62000
    },
    {
      "epoch": 24.83,
      "learning_rate": 1.0071107607410328e-05,
      "loss": 0.0158,
      "step": 63000
    },
    {
      "epoch": 25.23,
      "learning_rate": 9.913598738667718e-06,
      "loss": 0.022,
      "step": 64000
    },
    {
      "epoch": 25.62,
      "learning_rate": 9.755932203389833e-06,
      "loss": 0.0176,
      "step": 65000
    },
    {
      "epoch": 26.01,
      "learning_rate": 9.598265668111943e-06,
      "loss": 0.0176,
      "step": 66000
    },
    {
      "epoch": 26.41,
      "learning_rate": 9.440599132834056e-06,
      "loss": 0.0164,
      "step": 67000
    },
    {
      "epoch": 26.8,
      "learning_rate": 9.283090264091448e-06,
      "loss": 0.0196,
      "step": 68000
    },
    {
      "epoch": 27.2,
      "learning_rate": 9.125581395348838e-06,
      "loss": 0.0168,
      "step": 69000
    },
    {
      "epoch": 27.59,
      "learning_rate": 8.96791486007095e-06,
      "loss": 0.0167,
      "step": 70000
    },
    {
      "epoch": 27.59,
      "eval_accuracy": 0.9960272417707151,
      "eval_f1": 0.9849462365591398,
      "eval_loss": 0.015500849112868309,
      "eval_precision": 1.0,
      "eval_recall": 0.9703389830508474,
      "eval_runtime": 6.0236,
      "eval_samples_per_second": 292.516,
      "eval_steps_per_second": 36.689,
      "step": 70000
    },
    {
      "epoch": 27.99,
      "learning_rate": 8.810248324793063e-06,
      "loss": 0.018,
      "step": 71000
    },
    {
      "epoch": 28.38,
      "learning_rate": 8.652739456050455e-06,
      "loss": 0.0133,
      "step": 72000
    },
    {
      "epoch": 28.77,
      "learning_rate": 8.495388253843122e-06,
      "loss": 0.0187,
      "step": 73000
    },
    {
      "epoch": 29.17,
      "learning_rate": 8.337721718565234e-06,
      "loss": 0.0167,
      "step": 74000
    },
    {
      "epoch": 29.56,
      "learning_rate": 8.180055183287347e-06,
      "loss": 0.0151,
      "step": 75000
    },
    {
      "epoch": 29.96,
      "learning_rate": 8.02238864800946e-06,
      "loss": 0.0127,
      "step": 76000
    },
    {
      "epoch": 30.35,
      "learning_rate": 7.864722112731574e-06,
      "loss": 0.017,
      "step": 77000
    },
    {
      "epoch": 30.74,
      "learning_rate": 7.707213243988964e-06,
      "loss": 0.0148,
      "step": 78000
    },
    {
      "epoch": 31.14,
      "learning_rate": 7.5495467087110764e-06,
      "loss": 0.0136,
      "step": 79000
    },
    {
      "epoch": 31.53,
      "learning_rate": 7.391880173433189e-06,
      "loss": 0.0141,
      "step": 80000
    },
    {
      "epoch": 31.53,
      "eval_accuracy": 0.9954597048808173,
      "eval_f1": 0.9829059829059829,
      "eval_loss": 0.012631294317543507,
      "eval_precision": 0.9913793103448276,
      "eval_recall": 0.9745762711864406,
      "eval_runtime": 6.0398,
      "eval_samples_per_second": 291.729,
      "eval_steps_per_second": 36.59,
      "step": 80000
    },
    {
      "epoch": 31.93,
      "learning_rate": 7.23437130469058e-06,
      "loss": 0.0156,
      "step": 81000
    },
    {
      "epoch": 32.32,
      "learning_rate": 7.0767047694126924e-06,
      "loss": 0.0136,
      "step": 82000
    },
    {
      "epoch": 32.72,
      "learning_rate": 6.919038234134805e-06,
      "loss": 0.013,
      "step": 83000
    },
    {
      "epoch": 33.11,
      "learning_rate": 6.761371698856918e-06,
      "loss": 0.0093,
      "step": 84000
    },
    {
      "epoch": 33.5,
      "learning_rate": 6.603862830114309e-06,
      "loss": 0.0121,
      "step": 85000
    },
    {
      "epoch": 33.9,
      "learning_rate": 6.4463539613717e-06,
      "loss": 0.0136,
      "step": 86000
    },
    {
      "epoch": 34.29,
      "learning_rate": 6.288687426093813e-06,
      "loss": 0.013,
      "step": 87000
    },
    {
      "epoch": 34.69,
      "learning_rate": 6.131178557351203e-06,
      "loss": 0.0135,
      "step": 88000
    },
    {
      "epoch": 35.08,
      "learning_rate": 5.973512022073315e-06,
      "loss": 0.016,
      "step": 89000
    },
    {
      "epoch": 35.47,
      "learning_rate": 5.815845486795429e-06,
      "loss": 0.0144,
      "step": 90000
    },
    {
      "epoch": 35.47,
      "eval_accuracy": 0.996594778660613,
      "eval_f1": 0.9872340425531915,
      "eval_loss": 0.009375466965138912,
      "eval_precision": 0.9914529914529915,
      "eval_recall": 0.9830508474576272,
      "eval_runtime": 6.361,
      "eval_samples_per_second": 276.999,
      "eval_steps_per_second": 34.743,
      "step": 90000
    },
    {
      "epoch": 35.87,
      "learning_rate": 5.658336618052819e-06,
      "loss": 0.0155,
      "step": 91000
    },
    {
      "epoch": 36.26,
      "learning_rate": 5.500670082774931e-06,
      "loss": 0.0126,
      "step": 92000
    },
    {
      "epoch": 36.66,
      "learning_rate": 5.343161214032322e-06,
      "loss": 0.0094,
      "step": 93000
    },
    {
      "epoch": 37.05,
      "learning_rate": 5.185494678754435e-06,
      "loss": 0.0106,
      "step": 94000
    },
    {
      "epoch": 37.45,
      "learning_rate": 5.027828143476547e-06,
      "loss": 0.0115,
      "step": 95000
    },
    {
      "epoch": 37.84,
      "learning_rate": 4.870161608198661e-06,
      "loss": 0.0087,
      "step": 96000
    },
    {
      "epoch": 38.23,
      "learning_rate": 4.712495072920773e-06,
      "loss": 0.0126,
      "step": 97000
    },
    {
      "epoch": 38.63,
      "learning_rate": 4.554986204178163e-06,
      "loss": 0.009,
      "step": 98000
    },
    {
      "epoch": 39.02,
      "learning_rate": 4.397319668900277e-06,
      "loss": 0.0129,
      "step": 99000
    },
    {
      "epoch": 39.42,
      "learning_rate": 4.239653133622389e-06,
      "loss": 0.009,
      "step": 100000
    },
    {
      "epoch": 39.42,
      "eval_accuracy": 0.996594778660613,
      "eval_f1": 0.9872881355932204,
      "eval_loss": 0.008661070838570595,
      "eval_precision": 0.9872881355932204,
      "eval_recall": 0.9872881355932204,
      "eval_runtime": 6.0434,
      "eval_samples_per_second": 291.559,
      "eval_steps_per_second": 36.569,
      "step": 100000
    },
    {
      "epoch": 39.81,
      "learning_rate": 4.081986598344502e-06,
      "loss": 0.0066,
      "step": 101000
    },
    {
      "epoch": 40.2,
      "learning_rate": 3.924477729601892e-06,
      "loss": 0.0116,
      "step": 102000
    },
    {
      "epoch": 40.6,
      "learning_rate": 3.7668111943240047e-06,
      "loss": 0.0099,
      "step": 103000
    },
    {
      "epoch": 40.99,
      "learning_rate": 3.6091446590461178e-06,
      "loss": 0.0087,
      "step": 104000
    },
    {
      "epoch": 41.39,
      "learning_rate": 3.4516357903035086e-06,
      "loss": 0.0092,
      "step": 105000
    },
    {
      "epoch": 41.78,
      "learning_rate": 3.2939692550256207e-06,
      "loss": 0.0098,
      "step": 106000
    },
    {
      "epoch": 42.18,
      "learning_rate": 3.1363027197477337e-06,
      "loss": 0.0092,
      "step": 107000
    },
    {
      "epoch": 42.57,
      "learning_rate": 2.9787938510051245e-06,
      "loss": 0.0111,
      "step": 108000
    },
    {
      "epoch": 42.96,
      "learning_rate": 2.821127315727237e-06,
      "loss": 0.0077,
      "step": 109000
    },
    {
      "epoch": 43.36,
      "learning_rate": 2.66346078044935e-06,
      "loss": 0.0084,
      "step": 110000
    },
    {
      "epoch": 43.36,
      "eval_accuracy": 0.9954597048808173,
      "eval_f1": 0.9828326180257511,
      "eval_loss": 0.018768297508358955,
      "eval_precision": 0.9956521739130435,
      "eval_recall": 0.9703389830508474,
      "eval_runtime": 5.9996,
      "eval_samples_per_second": 293.685,
      "eval_steps_per_second": 36.836,
      "step": 110000
    },
    {
      "epoch": 43.75,
      "learning_rate": 2.5059519117067405e-06,
      "loss": 0.0086,
      "step": 111000
    },
    {
      "epoch": 44.15,
      "learning_rate": 2.348285376428853e-06,
      "loss": 0.0092,
      "step": 112000
    },
    {
      "epoch": 44.54,
      "learning_rate": 2.190618841150966e-06,
      "loss": 0.0099,
      "step": 113000
    },
    {
      "epoch": 44.93,
      "learning_rate": 2.0331099724083565e-06,
      "loss": 0.0101,
      "step": 114000
    },
    {
      "epoch": 45.33,
      "learning_rate": 1.8754434371304691e-06,
      "loss": 0.0079,
      "step": 115000
    },
    {
      "epoch": 45.72,
      "learning_rate": 1.71793456838786e-06,
      "loss": 0.0102,
      "step": 116000
    },
    {
      "epoch": 46.12,
      "learning_rate": 1.5602680331099725e-06,
      "loss": 0.0086,
      "step": 117000
    },
    {
      "epoch": 46.51,
      "learning_rate": 1.4026014978320853e-06,
      "loss": 0.0097,
      "step": 118000
    },
    {
      "epoch": 46.91,
      "learning_rate": 1.2449349625541981e-06,
      "loss": 0.0048,
      "step": 119000
    },
    {
      "epoch": 47.3,
      "learning_rate": 1.0872684272763107e-06,
      "loss": 0.0112,
      "step": 120000
    },
    {
      "epoch": 47.3,
      "eval_accuracy": 0.996594778660613,
      "eval_f1": 0.9872881355932204,
      "eval_loss": 0.008344221860170364,
      "eval_precision": 0.9872881355932204,
      "eval_recall": 0.9872881355932204,
      "eval_runtime": 6.0461,
      "eval_samples_per_second": 291.429,
      "eval_steps_per_second": 36.553,
      "step": 120000
    },
    {
      "epoch": 47.69,
      "learning_rate": 9.297595585337013e-07,
      "loss": 0.0069,
      "step": 121000
    },
    {
      "epoch": 48.09,
      "learning_rate": 7.72250689791092e-07,
      "loss": 0.0103,
      "step": 122000
    },
    {
      "epoch": 48.48,
      "learning_rate": 6.145841545132046e-07,
      "loss": 0.0079,
      "step": 123000
    },
    {
      "epoch": 48.88,
      "learning_rate": 4.5691761923531733e-07,
      "loss": 0.0061,
      "step": 124000
    },
    {
      "epoch": 49.27,
      "learning_rate": 2.994087504927079e-07,
      "loss": 0.0081,
      "step": 125000
    },
    {
      "epoch": 49.66,
      "learning_rate": 1.4174221521482067e-07,
      "loss": 0.0101,
      "step": 126000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 126850,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 1.0572380196462e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}