trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6180976033210754,
  "best_model_checkpoint": "tiny_bert_rand_20_v1_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": 0.8197463750839233,
      "learning_rate": 4.9e-05,
      "loss": 0.6196,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6180976033210754,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.341,
      "eval_samples_per_second": 3058.92,
      "eval_steps_per_second": 14.664,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2183475494384766,
      "learning_rate": 4.8e-05,
      "loss": 0.5915,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6922339200973511,
      "eval_loss": 0.6219738721847534,
      "eval_matthews_correlation": 0.0437601222642778,
      "eval_runtime": 0.343,
      "eval_samples_per_second": 3040.65,
      "eval_steps_per_second": 14.576,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7569878101348877,
      "learning_rate": 4.7e-05,
      "loss": 0.5622,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6893576383590698,
      "eval_loss": 0.6282854676246643,
      "eval_matthews_correlation": 0.07187794699272071,
      "eval_runtime": 0.3164,
      "eval_samples_per_second": 3295.943,
      "eval_steps_per_second": 15.8,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.118624687194824,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5173,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6740172505378723,
      "eval_loss": 0.6515808701515198,
      "eval_matthews_correlation": 0.10823719705532697,
      "eval_runtime": 0.3548,
      "eval_samples_per_second": 2939.698,
      "eval_steps_per_second": 14.093,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.1426446437835693,
      "learning_rate": 4.5e-05,
      "loss": 0.461,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6442952752113342,
      "eval_loss": 0.709630012512207,
      "eval_matthews_correlation": 0.05731153535482938,
      "eval_runtime": 0.3201,
      "eval_samples_per_second": 3258.276,
      "eval_steps_per_second": 15.62,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.13608980178833,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4116,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.640460193157196,
      "eval_loss": 0.7407780289649963,
      "eval_matthews_correlation": 0.05928695894318778,
      "eval_runtime": 0.3266,
      "eval_samples_per_second": 3193.755,
      "eval_steps_per_second": 15.31,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5272089266309551,
      "train_runtime": 34.6523,
      "train_samples_per_second": 12338.285,
      "train_steps_per_second": 49.059
    }
  ],
  "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
}