trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6171247959136963,
  "best_model_checkpoint": "tiny_bert_km_100_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": 0.8058708310127258,
      "learning_rate": 4.9e-05,
      "loss": 0.6207,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6171247959136963,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3349,
      "eval_samples_per_second": 3114.749,
      "eval_steps_per_second": 14.932,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.082227110862732,
      "learning_rate": 4.8e-05,
      "loss": 0.6036,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6248427033424377,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3609,
      "eval_samples_per_second": 2890.141,
      "eval_steps_per_second": 13.855,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0944205522537231,
      "learning_rate": 4.7e-05,
      "loss": 0.5909,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.693192720413208,
      "eval_loss": 0.6218764781951904,
      "eval_matthews_correlation": 0.0592680243795702,
      "eval_runtime": 0.3157,
      "eval_samples_per_second": 3304.078,
      "eval_steps_per_second": 15.839,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.5913689136505127,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5696,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6788111329078674,
      "eval_loss": 0.6305026412010193,
      "eval_matthews_correlation": 0.0697392423244271,
      "eval_runtime": 0.3326,
      "eval_samples_per_second": 3135.779,
      "eval_steps_per_second": 15.032,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.570982813835144,
      "learning_rate": 4.5e-05,
      "loss": 0.5351,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.668264627456665,
      "eval_loss": 0.6429861187934875,
      "eval_matthews_correlation": 0.06925819357864041,
      "eval_runtime": 0.3109,
      "eval_samples_per_second": 3354.766,
      "eval_steps_per_second": 16.082,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.8685276508331299,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4897,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6174496412277222,
      "eval_loss": 0.6808885335922241,
      "eval_matthews_correlation": -0.0023041888298101143,
      "eval_runtime": 0.3173,
      "eval_samples_per_second": 3286.951,
      "eval_steps_per_second": 15.757,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5682636709774241,
      "train_runtime": 33.8017,
      "train_samples_per_second": 12648.779,
      "train_steps_per_second": 50.293
    }
  ],
  "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
}