trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.616877555847168,
  "best_model_checkpoint": "bert_base_rand_20_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": 0.8142052292823792,
      "learning_rate": 4.9e-05,
      "loss": 0.6147,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.616877555847168,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6688,
      "eval_samples_per_second": 1559.469,
      "eval_steps_per_second": 7.476,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5374573469161987,
      "learning_rate": 4.8e-05,
      "loss": 0.5912,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6855225563049316,
      "eval_loss": 0.6168986558914185,
      "eval_matthews_correlation": 0.03024806178688017,
      "eval_runtime": 0.6804,
      "eval_samples_per_second": 1532.876,
      "eval_steps_per_second": 7.348,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.6451101303100586,
      "learning_rate": 4.7e-05,
      "loss": 0.5445,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6615532040596008,
      "eval_loss": 0.6372238993644714,
      "eval_matthews_correlation": 0.0964637619956757,
      "eval_runtime": 0.6752,
      "eval_samples_per_second": 1544.822,
      "eval_steps_per_second": 7.406,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.400547027587891,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4913,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6577181220054626,
      "eval_loss": 0.7028094530105591,
      "eval_matthews_correlation": 0.07455604209059638,
      "eval_runtime": 0.6947,
      "eval_samples_per_second": 1501.305,
      "eval_steps_per_second": 7.197,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.9376835823059082,
      "learning_rate": 4.5e-05,
      "loss": 0.434,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6452540755271912,
      "eval_loss": 0.7017385363578796,
      "eval_matthews_correlation": 0.09524312033216221,
      "eval_runtime": 0.7001,
      "eval_samples_per_second": 1489.78,
      "eval_steps_per_second": 7.142,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.3664913177490234,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3888,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6337488293647766,
      "eval_loss": 0.7899420857429504,
      "eval_matthews_correlation": 0.0734252631472095,
      "eval_runtime": 0.7016,
      "eval_samples_per_second": 1486.537,
      "eval_steps_per_second": 7.126,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5107499384412578,
      "train_runtime": 100.1419,
      "train_samples_per_second": 4269.443,
      "train_steps_per_second": 16.976
    }
  ],
  "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
}