{
  "best_metric": 1.5053555965423584,
  "best_model_checkpoint": "./tmp/clm-gpt2/checkpoint-11500",
  "epoch": 2.9996095275283094,
  "eval_steps": 500,
  "global_step": 11523,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13015749056358195,
      "grad_norm": 0.3655782639980316,
      "learning_rate": 0.0015,
      "loss": 2.4536,
      "step": 500
    },
    {
      "epoch": 0.13015749056358195,
      "eval_accuracy": 0.4954853308958881,
      "eval_loss": 2.1316182613372803,
      "eval_runtime": 630.0742,
      "eval_samples_per_second": 24.442,
      "eval_steps_per_second": 6.11,
      "step": 500
    },
    {
      "epoch": 0.2603149811271639,
      "grad_norm": 0.2795959711074829,
      "learning_rate": 0.003,
      "loss": 2.1054,
      "step": 1000
    },
    {
      "epoch": 0.2603149811271639,
      "eval_accuracy": 0.5220788107298371,
      "eval_loss": 2.0124123096466064,
      "eval_runtime": 571.1208,
      "eval_samples_per_second": 26.965,
      "eval_steps_per_second": 6.741,
      "step": 1000
    },
    {
      "epoch": 0.3904724716907458,
      "grad_norm": 0.2532959282398224,
      "learning_rate": 0.0028574550983559823,
      "loss": 1.9756,
      "step": 1500
    },
    {
      "epoch": 0.3904724716907458,
      "eval_accuracy": 0.5452574551548158,
      "eval_loss": 1.902522087097168,
      "eval_runtime": 521.1164,
      "eval_samples_per_second": 29.552,
      "eval_steps_per_second": 7.388,
      "step": 1500
    },
    {
      "epoch": 0.5206299622543278,
      "grad_norm": 0.1999560445547104,
      "learning_rate": 0.002714910196711964,
      "loss": 1.8863,
      "step": 2000
    },
    {
      "epoch": 0.5206299622543278,
      "eval_accuracy": 0.5600627134351475,
      "eval_loss": 1.836737871170044,
      "eval_runtime": 521.7562,
      "eval_samples_per_second": 29.516,
      "eval_steps_per_second": 7.379,
      "step": 2000
    },
    {
      "epoch": 0.6507874528179096,
      "grad_norm": 0.18795448541641235,
      "learning_rate": 0.0025723652950679463,
      "loss": 1.8283,
      "step": 2500
    },
    {
      "epoch": 0.6507874528179096,
      "eval_accuracy": 0.5686439806527783,
      "eval_loss": 1.7926692962646484,
      "eval_runtime": 521.5355,
      "eval_samples_per_second": 29.528,
      "eval_steps_per_second": 7.382,
      "step": 2500
    },
    {
      "epoch": 0.7809449433814916,
      "grad_norm": 0.21605215966701508,
      "learning_rate": 0.0024298203934239285,
      "loss": 1.7893,
      "step": 3000
    },
    {
      "epoch": 0.7809449433814916,
      "eval_accuracy": 0.5759645681786444,
      "eval_loss": 1.758531093597412,
      "eval_runtime": 521.9502,
      "eval_samples_per_second": 29.505,
      "eval_steps_per_second": 7.376,
      "step": 3000
    },
    {
      "epoch": 0.9111024339450735,
      "grad_norm": 0.17780712246894836,
      "learning_rate": 0.0022872754917799107,
      "loss": 1.7555,
      "step": 3500
    },
    {
      "epoch": 0.9111024339450735,
      "eval_accuracy": 0.581453644107603,
      "eval_loss": 1.732818841934204,
      "eval_runtime": 521.3914,
      "eval_samples_per_second": 29.536,
      "eval_steps_per_second": 7.384,
      "step": 3500
    },
    {
      "epoch": 1.0412599245086556,
      "grad_norm": 0.17426873743534088,
      "learning_rate": 0.002144730590135893,
      "loss": 1.7143,
      "step": 4000
    },
    {
      "epoch": 1.0412599245086556,
      "eval_accuracy": 0.5881837859110587,
      "eval_loss": 1.701590895652771,
      "eval_runtime": 522.4527,
      "eval_samples_per_second": 29.476,
      "eval_steps_per_second": 7.369,
      "step": 4000
    },
    {
      "epoch": 1.1714174150722374,
      "grad_norm": 0.1862766295671463,
      "learning_rate": 0.002002185688491875,
      "loss": 1.6697,
      "step": 4500
    },
    {
      "epoch": 1.1714174150722374,
      "eval_accuracy": 0.592953625065062,
      "eval_loss": 1.6812903881072998,
      "eval_runtime": 520.916,
      "eval_samples_per_second": 29.563,
      "eval_steps_per_second": 7.391,
      "step": 4500
    },
    {
      "epoch": 1.3015749056358192,
      "grad_norm": 0.17118503153324127,
      "learning_rate": 0.0018596407868478571,
      "loss": 1.6584,
      "step": 5000
    },
    {
      "epoch": 1.3015749056358192,
      "eval_accuracy": 0.5972079191580658,
      "eval_loss": 1.6614623069763184,
      "eval_runtime": 520.2098,
      "eval_samples_per_second": 29.603,
      "eval_steps_per_second": 7.401,
      "step": 5000
    },
    {
      "epoch": 1.4317323961994013,
      "grad_norm": 0.17990975081920624,
      "learning_rate": 0.0017170958852038393,
      "loss": 1.6438,
      "step": 5500
    },
    {
      "epoch": 1.4317323961994013,
      "eval_accuracy": 0.6009242614667835,
      "eval_loss": 1.6421809196472168,
      "eval_runtime": 520.0082,
      "eval_samples_per_second": 29.615,
      "eval_steps_per_second": 7.404,
      "step": 5500
    },
    {
      "epoch": 1.5618898867629833,
      "grad_norm": 0.1535763144493103,
      "learning_rate": 0.0015745509835598213,
      "loss": 1.6184,
      "step": 6000
    },
    {
      "epoch": 1.5618898867629833,
      "eval_accuracy": 0.6048871412067893,
      "eval_loss": 1.6235625743865967,
      "eval_runtime": 518.9862,
      "eval_samples_per_second": 29.673,
      "eval_steps_per_second": 7.418,
      "step": 6000
    },
    {
      "epoch": 1.6920473773265652,
      "grad_norm": 0.16722074151039124,
      "learning_rate": 0.0014320060819158036,
      "loss": 1.6086,
      "step": 6500
    },
    {
      "epoch": 1.6920473773265652,
      "eval_accuracy": 0.6081556664254611,
      "eval_loss": 1.6102288961410522,
      "eval_runtime": 519.7651,
      "eval_samples_per_second": 29.629,
      "eval_steps_per_second": 7.407,
      "step": 6500
    },
    {
      "epoch": 1.822204867890147,
      "grad_norm": 0.1687261015176773,
      "learning_rate": 0.0012894611802717856,
      "loss": 1.5882,
      "step": 7000
    },
    {
      "epoch": 1.822204867890147,
      "eval_accuracy": 0.6113670005458862,
      "eval_loss": 1.5938153266906738,
      "eval_runtime": 518.0575,
      "eval_samples_per_second": 29.726,
      "eval_steps_per_second": 7.432,
      "step": 7000
    },
    {
      "epoch": 1.9523623584537289,
      "grad_norm": 0.17714297771453857,
      "learning_rate": 0.0011469162786277678,
      "loss": 1.5719,
      "step": 7500
    },
    {
      "epoch": 1.9523623584537289,
      "eval_accuracy": 0.6148032270759544,
      "eval_loss": 1.578608751296997,
      "eval_runtime": 518.0664,
      "eval_samples_per_second": 29.726,
      "eval_steps_per_second": 7.431,
      "step": 7500
    },
    {
      "epoch": 2.082519849017311,
      "grad_norm": 0.16494958102703094,
      "learning_rate": 0.00100437137698375,
      "loss": 1.5272,
      "step": 8000
    },
    {
      "epoch": 2.082519849017311,
      "eval_accuracy": 0.617450838506557,
      "eval_loss": 1.5717881917953491,
      "eval_runtime": 518.5565,
      "eval_samples_per_second": 29.698,
      "eval_steps_per_second": 7.424,
      "step": 8000
    },
    {
      "epoch": 2.212677339580893,
      "grad_norm": 0.14182300865650177,
      "learning_rate": 0.0008618264753397321,
      "loss": 1.4971,
      "step": 8500
    },
    {
      "epoch": 2.212677339580893,
      "eval_accuracy": 0.6203893564890632,
      "eval_loss": 1.5592997074127197,
      "eval_runtime": 517.6299,
      "eval_samples_per_second": 29.751,
      "eval_steps_per_second": 7.438,
      "step": 8500
    },
    {
      "epoch": 2.342834830144475,
      "grad_norm": 0.15439514815807343,
      "learning_rate": 0.0007192815736957142,
      "loss": 1.4893,
      "step": 9000
    },
    {
      "epoch": 2.342834830144475,
      "eval_accuracy": 0.6227424432849653,
      "eval_loss": 1.5474613904953003,
      "eval_runtime": 517.7895,
      "eval_samples_per_second": 29.742,
      "eval_steps_per_second": 7.435,
      "step": 9000
    },
    {
      "epoch": 2.4729923207080566,
      "grad_norm": 0.1444994956254959,
      "learning_rate": 0.0005767366720516963,
      "loss": 1.4808,
      "step": 9500
    },
    {
      "epoch": 2.4729923207080566,
      "eval_accuracy": 0.6250960378819617,
      "eval_loss": 1.5381561517715454,
      "eval_runtime": 518.2423,
      "eval_samples_per_second": 29.716,
      "eval_steps_per_second": 7.429,
      "step": 9500
    },
    {
      "epoch": 2.6031498112716385,
      "grad_norm": 0.14975884556770325,
      "learning_rate": 0.00043419177040767847,
      "loss": 1.4689,
      "step": 10000
    },
    {
      "epoch": 2.6031498112716385,
      "eval_accuracy": 0.6274774980640083,
      "eval_loss": 1.5274103879928589,
      "eval_runtime": 518.577,
      "eval_samples_per_second": 29.697,
      "eval_steps_per_second": 7.424,
      "step": 10000
    },
    {
      "epoch": 2.7333073018352207,
      "grad_norm": 0.14258165657520294,
      "learning_rate": 0.00029164686876366057,
      "loss": 1.4572,
      "step": 10500
    },
    {
      "epoch": 2.7333073018352207,
      "eval_accuracy": 0.6297954196341293,
      "eval_loss": 1.5169227123260498,
      "eval_runtime": 518.5892,
      "eval_samples_per_second": 29.696,
      "eval_steps_per_second": 7.424,
      "step": 10500
    },
    {
      "epoch": 2.8634647923988026,
      "grad_norm": 0.14571569859981537,
      "learning_rate": 0.00014910196711964268,
      "loss": 1.4488,
      "step": 11000
    },
    {
      "epoch": 2.8634647923988026,
      "eval_accuracy": 0.6314827791953892,
      "eval_loss": 1.5106253623962402,
      "eval_runtime": 519.3798,
      "eval_samples_per_second": 29.651,
      "eval_steps_per_second": 7.413,
      "step": 11000
    },
    {
      "epoch": 2.9936222829623844,
      "grad_norm": 0.15030799806118011,
      "learning_rate": 6.557065475624823e-06,
      "loss": 1.4465,
      "step": 11500
    },
    {
      "epoch": 2.9936222829623844,
      "eval_accuracy": 0.6325420522781227,
      "eval_loss": 1.5053555965423584,
      "eval_runtime": 518.7173,
      "eval_samples_per_second": 29.689,
      "eval_steps_per_second": 7.422,
      "step": 11500
    },
    {
      "epoch": 2.9996095275283094,
      "step": 11523,
      "total_flos": 9.6346733543424e+16,
      "train_loss": 1.6814149037051365,
      "train_runtime": 28276.4686,
      "train_samples_per_second": 6.521,
      "train_steps_per_second": 0.408
    }
  ],
  "logging_steps": 500,
  "max_steps": 11523,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.6346733543424e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}