trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6175676584243774,
  "best_model_checkpoint": "tiny_bert_rand_50_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": 0.7076928615570068,
      "learning_rate": 4.9e-05,
      "loss": 0.6135,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6175676584243774,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3187,
      "eval_samples_per_second": 3273.029,
      "eval_steps_per_second": 15.69,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0805855989456177,
      "learning_rate": 4.8e-05,
      "loss": 0.6006,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6197286248207092,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3263,
      "eval_samples_per_second": 3196.371,
      "eval_steps_per_second": 15.323,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4896705150604248,
      "learning_rate": 4.7e-05,
      "loss": 0.5776,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.690316379070282,
      "eval_loss": 0.6242491006851196,
      "eval_matthews_correlation": 0.03716241204687234,
      "eval_runtime": 0.3366,
      "eval_samples_per_second": 3098.824,
      "eval_steps_per_second": 14.855,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.8818837404251099,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5383,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6720997095108032,
      "eval_loss": 0.6582180857658386,
      "eval_matthews_correlation": 0.06220394739449557,
      "eval_runtime": 0.3094,
      "eval_samples_per_second": 3370.866,
      "eval_steps_per_second": 16.159,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.8287111520767212,
      "learning_rate": 4.5e-05,
      "loss": 0.4936,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6490891575813293,
      "eval_loss": 0.6671127080917358,
      "eval_matthews_correlation": 0.08569348900620244,
      "eval_runtime": 0.3698,
      "eval_samples_per_second": 2820.586,
      "eval_steps_per_second": 13.522,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.463777780532837,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4569,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6375839114189148,
      "eval_loss": 0.7221479415893555,
      "eval_matthews_correlation": 0.0817217914911754,
      "eval_runtime": 0.3138,
      "eval_samples_per_second": 3323.302,
      "eval_steps_per_second": 15.931,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5467501154132918,
      "train_runtime": 34.2397,
      "train_samples_per_second": 12486.959,
      "train_steps_per_second": 49.65
    }
  ],
  "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
}