trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6162062287330627,
  "best_model_checkpoint": "tiny_bert_rand_10_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": 0.8476899862289429,
      "learning_rate": 4.9e-05,
      "loss": 0.617,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6162062287330627,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3253,
      "eval_samples_per_second": 3205.893,
      "eval_steps_per_second": 15.369,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9178780913352966,
      "learning_rate": 4.8e-05,
      "loss": 0.6002,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6173672676086426,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.33,
      "eval_samples_per_second": 3160.244,
      "eval_steps_per_second": 15.15,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0462307929992676,
      "learning_rate": 4.7e-05,
      "loss": 0.5791,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6922339200973511,
      "eval_loss": 0.61886066198349,
      "eval_matthews_correlation": 0.04161454566059593,
      "eval_runtime": 0.3769,
      "eval_samples_per_second": 2767.11,
      "eval_steps_per_second": 13.265,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3154501914978027,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5385,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.668264627456665,
      "eval_loss": 0.6651514172554016,
      "eval_matthews_correlation": 0.04797771844318346,
      "eval_runtime": 0.3127,
      "eval_samples_per_second": 3335.951,
      "eval_steps_per_second": 15.992,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.127080202102661,
      "learning_rate": 4.5e-05,
      "loss": 0.4918,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6625120043754578,
      "eval_loss": 0.6673993468284607,
      "eval_matthews_correlation": 0.08811408651094657,
      "eval_runtime": 0.3261,
      "eval_samples_per_second": 3198.762,
      "eval_steps_per_second": 15.334,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.399074077606201,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4587,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6366251111030579,
      "eval_loss": 0.7247511148452759,
      "eval_matthews_correlation": 0.05950348658831357,
      "eval_runtime": 0.3524,
      "eval_samples_per_second": 2959.388,
      "eval_steps_per_second": 14.187,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 1345426898614272.0,
      "train_loss": 0.5475386077282476,
      "train_runtime": 34.669,
      "train_samples_per_second": 12332.349,
      "train_steps_per_second": 49.035
    }
  ],
  "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
}