trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6928728818893433,
  "best_model_checkpoint": "bert_base_rand_10_v2_rte/checkpoint-10",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.6074354648590088,
      "learning_rate": 4.9e-05,
      "loss": 0.7127,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.6928728818893433,
      "eval_runtime": 0.1978,
      "eval_samples_per_second": 1400.103,
      "eval_steps_per_second": 10.109,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2740426063537598,
      "learning_rate": 4.8e-05,
      "loss": 0.6747,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.49458483754512633,
      "eval_loss": 0.7386796474456787,
      "eval_runtime": 0.2021,
      "eval_samples_per_second": 1370.759,
      "eval_steps_per_second": 9.897,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4609880447387695,
      "learning_rate": 4.7e-05,
      "loss": 0.6196,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.8164538145065308,
      "eval_runtime": 0.2416,
      "eval_samples_per_second": 1146.4,
      "eval_steps_per_second": 8.277,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.981417417526245,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4903,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.48375451263537905,
      "eval_loss": 1.1117196083068848,
      "eval_runtime": 0.2085,
      "eval_samples_per_second": 1328.322,
      "eval_steps_per_second": 9.591,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.5136139392852783,
      "learning_rate": 4.5e-05,
      "loss": 0.3519,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 1.2060282230377197,
      "eval_runtime": 0.1978,
      "eval_samples_per_second": 1400.182,
      "eval_steps_per_second": 10.11,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.079023361206055,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2374,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 1.499961018562317,
      "eval_runtime": 0.1893,
      "eval_samples_per_second": 1463.634,
      "eval_steps_per_second": 10.568,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.5144398053487141,
      "train_runtime": 39.0916,
      "train_samples_per_second": 3184.827,
      "train_steps_per_second": 12.79
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 1965439583539200.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}