{
  "best_global_step": 7670,
  "best_metric": 0.8109366297721863,
  "best_model_checkpoint": "tiny_bert_km_10_v2_mnli/checkpoint-7670",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 15340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.031025767326355,
      "learning_rate": 4.9e-05,
      "loss": 0.9946,
      "step": 1534
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.543046357615894,
      "eval_loss": 0.9383193850517273,
      "eval_runtime": 2.9466,
      "eval_samples_per_second": 3330.951,
      "eval_steps_per_second": 13.236,
      "step": 1534
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.319303274154663,
      "learning_rate": 4.8e-05,
      "loss": 0.9056,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5892002037697401,
      "eval_loss": 0.8850203156471252,
      "eval_runtime": 2.8012,
      "eval_samples_per_second": 3503.913,
      "eval_steps_per_second": 13.923,
      "step": 3068
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4328629970550537,
      "learning_rate": 4.7e-05,
      "loss": 0.8464,
      "step": 4602
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.611716760061131,
      "eval_loss": 0.8511015176773071,
      "eval_runtime": 2.8059,
      "eval_samples_per_second": 3497.935,
      "eval_steps_per_second": 13.899,
      "step": 4602
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.891276240348816,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.7912,
      "step": 6136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6360672440142638,
      "eval_loss": 0.8272331953048706,
      "eval_runtime": 2.8072,
      "eval_samples_per_second": 3496.358,
      "eval_steps_per_second": 13.893,
      "step": 6136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.9175740480422974,
      "learning_rate": 4.5e-05,
      "loss": 0.7345,
      "step": 7670
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6464595007641365,
      "eval_loss": 0.8109366297721863,
      "eval_runtime": 2.7818,
      "eval_samples_per_second": 3528.247,
      "eval_steps_per_second": 14.02,
      "step": 7670
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.268627643585205,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6801,
      "step": 9204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.654508405501783,
      "eval_loss": 0.8267003297805786,
      "eval_runtime": 2.8028,
      "eval_samples_per_second": 3501.896,
      "eval_steps_per_second": 13.915,
      "step": 9204
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.9387691020965576,
      "learning_rate": 4.3e-05,
      "loss": 0.6295,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6557310239429445,
      "eval_loss": 0.8331314325332642,
      "eval_runtime": 2.8071,
      "eval_samples_per_second": 3496.491,
      "eval_steps_per_second": 13.893,
      "step": 10738
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.4745192527770996,
      "learning_rate": 4.2e-05,
      "loss": 0.5808,
      "step": 12272
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6562404482934284,
      "eval_loss": 0.8564242124557495,
      "eval_runtime": 2.7753,
      "eval_samples_per_second": 3536.576,
      "eval_steps_per_second": 14.053,
      "step": 12272
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.9146223068237305,
      "learning_rate": 4.1e-05,
      "loss": 0.5367,
      "step": 13806
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6510443199184921,
      "eval_loss": 0.9120057821273804,
      "eval_runtime": 2.7966,
      "eval_samples_per_second": 3509.625,
      "eval_steps_per_second": 13.946,
      "step": 13806
    },
    {
      "epoch": 10.0,
      "grad_norm": 3.390559673309326,
      "learning_rate": 4e-05,
      "loss": 0.4925,
      "step": 15340
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6505348955680081,
      "eval_loss": 0.9656685590744019,
      "eval_runtime": 2.8151,
      "eval_samples_per_second": 3486.511,
      "eval_steps_per_second": 13.854,
      "step": 15340
    },
    {
      "epoch": 10.0,
      "step": 15340,
      "total_flos": 1.029836100834816e+17,
      "train_loss": 0.7191754317501222,
      "train_runtime": 2062.9527,
      "train_samples_per_second": 9517.959,
      "train_steps_per_second": 37.18
    }
  ],
  "logging_steps": 1,
  "max_steps": 76700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.029836100834816e+17,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}