{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 288,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 0.5078026652336121,
      "learning_rate": 0.0001,
      "loss": 1.0979,
      "step": 10
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.24573734402656555,
      "learning_rate": 9.968335515358916e-05,
      "loss": 0.6078,
      "step": 20
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.28030624985694885,
      "learning_rate": 9.873743117270691e-05,
      "loss": 0.4051,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8977756524449263,
      "eval_loss": 0.43045690655708313,
      "eval_runtime": 25.9378,
      "eval_samples_per_second": 14.843,
      "eval_steps_per_second": 1.889,
      "step": 32
    },
    {
      "epoch": 1.256,
      "grad_norm": 0.24669107794761658,
      "learning_rate": 9.717420893549902e-05,
      "loss": 0.4398,
      "step": 40
    },
    {
      "epoch": 1.576,
      "grad_norm": 0.2326413094997406,
      "learning_rate": 9.501348789257373e-05,
      "loss": 0.3765,
      "step": 50
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.2204989194869995,
      "learning_rate": 9.2282635291242e-05,
      "loss": 0.2886,
      "step": 60
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.920555460186586,
      "eval_loss": 0.323241263628006,
      "eval_runtime": 25.9139,
      "eval_samples_per_second": 14.857,
      "eval_steps_per_second": 1.891,
      "step": 64
    },
    {
      "epoch": 2.192,
      "grad_norm": 0.2965347468852997,
      "learning_rate": 8.90162395476046e-05,
      "loss": 0.2927,
      "step": 70
    },
    {
      "epoch": 2.512,
      "grad_norm": 0.3402569890022278,
      "learning_rate": 8.525567215680398e-05,
      "loss": 0.3082,
      "step": 80
    },
    {
      "epoch": 2.832,
      "grad_norm": 0.29415449500083923,
      "learning_rate": 8.104856369019524e-05,
      "loss": 0.2289,
      "step": 90
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9322851101896489,
      "eval_loss": 0.27416735887527466,
      "eval_runtime": 25.8596,
      "eval_samples_per_second": 14.888,
      "eval_steps_per_second": 1.895,
      "step": 96
    },
    {
      "epoch": 3.128,
      "grad_norm": 0.30058005452156067,
      "learning_rate": 7.644820051634812e-05,
      "loss": 0.2103,
      "step": 100
    },
    {
      "epoch": 3.448,
      "grad_norm": 0.39383992552757263,
      "learning_rate": 7.15128498868873e-05,
      "loss": 0.2362,
      "step": 110
    },
    {
      "epoch": 3.768,
      "grad_norm": 0.32451319694519043,
      "learning_rate": 6.630502193549474e-05,
      "loss": 0.1925,
      "step": 120
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.938676979897959,
      "eval_loss": 0.25138482451438904,
      "eval_runtime": 25.9363,
      "eval_samples_per_second": 14.844,
      "eval_steps_per_second": 1.889,
      "step": 128
    },
    {
      "epoch": 4.064,
      "grad_norm": 0.375893235206604,
      "learning_rate": 6.0890677937442574e-05,
      "loss": 0.1471,
      "step": 130
    },
    {
      "epoch": 4.384,
      "grad_norm": 0.3606163263320923,
      "learning_rate": 5.5338394857677945e-05,
      "loss": 0.1876,
      "step": 140
    },
    {
      "epoch": 4.704,
      "grad_norm": 0.40445348620414734,
      "learning_rate": 4.971849676912172e-05,
      "loss": 0.1488,
      "step": 150
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4769314229488373,
      "learning_rate": 4.410216414245771e-05,
      "loss": 0.1079,
      "step": 160
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9420305087488926,
      "eval_loss": 0.2455526441335678,
      "eval_runtime": 25.9468,
      "eval_samples_per_second": 14.838,
      "eval_steps_per_second": 1.888,
      "step": 160
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.4198707044124603,
      "learning_rate": 3.856053228896442e-05,
      "loss": 0.1534,
      "step": 170
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.3887736201286316,
      "learning_rate": 3.316379037532644e-05,
      "loss": 0.1151,
      "step": 180
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.37902015447616577,
      "learning_rate": 2.798029242211828e-05,
      "loss": 0.0968,
      "step": 190
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.945372723288016,
      "eval_loss": 0.24095402657985687,
      "eval_runtime": 25.9698,
      "eval_samples_per_second": 14.825,
      "eval_steps_per_second": 1.887,
      "step": 192
    },
    {
      "epoch": 6.256,
      "grad_norm": 0.4507382810115814,
      "learning_rate": 2.3075691545870558e-05,
      "loss": 0.1142,
      "step": 200
    },
    {
      "epoch": 6.576,
      "grad_norm": 0.39560869336128235,
      "learning_rate": 1.8512108410229878e-05,
      "loss": 0.0908,
      "step": 210
    },
    {
      "epoch": 6.896,
      "grad_norm": 0.3712189793586731,
      "learning_rate": 1.434734441843899e-05,
      "loss": 0.0835,
      "step": 220
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9465976771312703,
      "eval_loss": 0.24637635052204132,
      "eval_runtime": 25.9064,
      "eval_samples_per_second": 14.861,
      "eval_steps_per_second": 1.891,
      "step": 224
    },
    {
      "epoch": 7.192,
      "grad_norm": 0.36213552951812744,
      "learning_rate": 1.063414961267859e-05,
      "loss": 0.0841,
      "step": 230
    },
    {
      "epoch": 7.5120000000000005,
      "grad_norm": 0.4722885191440582,
      "learning_rate": 7.41955455290726e-06,
      "loss": 0.0863,
      "step": 240
    },
    {
      "epoch": 7.832,
      "grad_norm": 0.29437902569770813,
      "learning_rate": 4.744274637483936e-06,
      "loss": 0.0716,
      "step": 250
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9472423896041875,
      "eval_loss": 0.2516001760959625,
      "eval_runtime": 25.92,
      "eval_samples_per_second": 14.853,
      "eval_steps_per_second": 1.89,
      "step": 256
    },
    {
      "epoch": 8.128,
      "grad_norm": 0.39367932081222534,
      "learning_rate": 2.6421944103256657e-06,
      "loss": 0.0666,
      "step": 260
    },
    {
      "epoch": 8.448,
      "grad_norm": 0.31784766912460327,
      "learning_rate": 1.1399383862592927e-06,
      "loss": 0.0814,
      "step": 270
    },
    {
      "epoch": 8.768,
      "grad_norm": 0.28355342149734497,
      "learning_rate": 2.5653383040524227e-07,
      "loss": 0.0611,
      "step": 280
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.947486792929819,
      "eval_loss": 0.2522543668746948,
      "eval_runtime": 25.9049,
      "eval_samples_per_second": 14.862,
      "eval_steps_per_second": 1.892,
      "step": 288
    },
    {
      "epoch": 9.0,
      "step": 288,
      "total_flos": 3.686068977139712e+17,
      "train_loss": 0.22300153877586126,
      "train_runtime": 2410.3131,
      "train_samples_per_second": 3.734,
      "train_steps_per_second": 0.119
    }
  ],
  "logging_steps": 10,
  "max_steps": 288,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 3.686068977139712e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}