trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6182118058204651,
  "best_model_checkpoint": "tiny_bert_rand_20_v2_cola/checkpoint-34",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8301746249198914,
      "learning_rate": 4.9e-05,
      "loss": 0.6153,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6182118058204651,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3368,
      "eval_samples_per_second": 3096.681,
      "eval_steps_per_second": 14.845,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9635839462280273,
      "learning_rate": 4.8e-05,
      "loss": 0.601,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6202075481414795,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3379,
      "eval_samples_per_second": 3087.135,
      "eval_steps_per_second": 14.799,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.55442476272583,
      "learning_rate": 4.7e-05,
      "loss": 0.5777,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6845637559890747,
      "eval_loss": 0.6293561458587646,
      "eval_matthews_correlation": -0.004070109138183699,
      "eval_runtime": 0.3275,
      "eval_samples_per_second": 3185.077,
      "eval_steps_per_second": 15.269,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0630927085876465,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5335,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6749760508537292,
      "eval_loss": 0.6736043691635132,
      "eval_matthews_correlation": 0.06764493518740443,
      "eval_runtime": 0.3364,
      "eval_samples_per_second": 3100.119,
      "eval_steps_per_second": 14.862,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.6330318450927734,
      "learning_rate": 4.5e-05,
      "loss": 0.4924,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6462128758430481,
      "eval_loss": 0.6790710687637329,
      "eval_matthews_correlation": 0.07170131960210033,
      "eval_runtime": 0.3263,
      "eval_samples_per_second": 3196.708,
      "eval_steps_per_second": 15.325,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.886054515838623,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4569,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6395014524459839,
      "eval_loss": 0.7302798628807068,
      "eval_matthews_correlation": 0.05573034210085416,
      "eval_runtime": 0.3157,
      "eval_samples_per_second": 3303.677,
      "eval_steps_per_second": 15.837,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5461271323409735,
      "train_runtime": 34.1995,
      "train_samples_per_second": 12501.647,
      "train_steps_per_second": 49.708
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "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": 1345426898614272.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}