trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6167146563529968,
  "best_model_checkpoint": "distilbert_km_5_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": 1.4502145051956177,
      "learning_rate": 4.9e-05,
      "loss": 0.6153,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6167146563529968,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7004,
      "eval_samples_per_second": 1489.2,
      "eval_steps_per_second": 7.139,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.507283329963684,
      "learning_rate": 4.8e-05,
      "loss": 0.5868,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6874400973320007,
      "eval_loss": 0.618233323097229,
      "eval_matthews_correlation": -0.0018582137243486861,
      "eval_runtime": 0.4841,
      "eval_samples_per_second": 2154.406,
      "eval_steps_per_second": 10.328,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8468343019485474,
      "learning_rate": 4.7e-05,
      "loss": 0.5352,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6701821684837341,
      "eval_loss": 0.6291948556900024,
      "eval_matthews_correlation": 0.09253272633727094,
      "eval_runtime": 0.4401,
      "eval_samples_per_second": 2369.785,
      "eval_steps_per_second": 11.36,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.562954902648926,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4746,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6826462149620056,
      "eval_loss": 0.6919478178024292,
      "eval_matthews_correlation": 0.13938221629721187,
      "eval_runtime": 0.4532,
      "eval_samples_per_second": 2301.449,
      "eval_steps_per_second": 11.033,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.417787790298462,
      "learning_rate": 4.5e-05,
      "loss": 0.4094,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6423777341842651,
      "eval_loss": 0.6991720795631409,
      "eval_matthews_correlation": 0.1276502969654254,
      "eval_runtime": 0.4398,
      "eval_samples_per_second": 2371.328,
      "eval_steps_per_second": 11.368,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.491579055786133,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3548,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6347075700759888,
      "eval_loss": 0.8181633949279785,
      "eval_matthews_correlation": 0.09467689550719095,
      "eval_runtime": 0.4627,
      "eval_samples_per_second": 2254.315,
      "eval_steps_per_second": 10.807,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 3398186177722368.0,
      "train_loss": 0.4960063438789517,
      "train_runtime": 59.3899,
      "train_samples_per_second": 7199.04,
      "train_steps_per_second": 28.624
    }
  ],
  "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": 3398186177722368.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}