trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6210572123527527,
  "best_model_checkpoint": "bert_base_km_10_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": 1.3032399415969849,
      "learning_rate": 4.9e-05,
      "loss": 0.6172,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6210572123527527,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6785,
      "eval_samples_per_second": 1537.293,
      "eval_steps_per_second": 7.37,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.477735161781311,
      "learning_rate": 4.8e-05,
      "loss": 0.5871,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6874400973320007,
      "eval_loss": 0.6245772838592529,
      "eval_matthews_correlation": 0.08020927197111564,
      "eval_runtime": 0.667,
      "eval_samples_per_second": 1563.696,
      "eval_steps_per_second": 7.496,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9851726293563843,
      "learning_rate": 4.7e-05,
      "loss": 0.5375,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6845637559890747,
      "eval_loss": 0.6238712668418884,
      "eval_matthews_correlation": 0.08097785992490644,
      "eval_runtime": 0.6925,
      "eval_samples_per_second": 1506.101,
      "eval_steps_per_second": 7.22,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.500091314315796,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4679,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6596356630325317,
      "eval_loss": 0.7118160724639893,
      "eval_matthews_correlation": 0.09763740948697848,
      "eval_runtime": 0.6731,
      "eval_samples_per_second": 1549.438,
      "eval_steps_per_second": 7.428,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.196970224380493,
      "learning_rate": 4.5e-05,
      "loss": 0.3886,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6356663703918457,
      "eval_loss": 0.7543103098869324,
      "eval_matthews_correlation": 0.07673140340421428,
      "eval_runtime": 0.6703,
      "eval_samples_per_second": 1556.031,
      "eval_steps_per_second": 7.459,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.281788349151611,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3317,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6500479578971863,
      "eval_loss": 0.867901086807251,
      "eval_matthews_correlation": 0.06649950842111448,
      "eval_runtime": 0.6864,
      "eval_samples_per_second": 1519.507,
      "eval_steps_per_second": 7.284,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.4883236604578355,
      "train_runtime": 100.3492,
      "train_samples_per_second": 4260.621,
      "train_steps_per_second": 16.941
    }
  ],
  "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": 6749587903150080.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}