trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.7026032209396362,
  "best_model_checkpoint": "distilbert_rand_5_v1_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": 1.8740273714065552,
      "learning_rate": 4.9e-05,
      "loss": 0.7243,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 0.7026032209396362,
      "eval_runtime": 0.1276,
      "eval_samples_per_second": 2170.012,
      "eval_steps_per_second": 15.668,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8743432760238647,
      "learning_rate": 4.8e-05,
      "loss": 0.6823,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.7101830244064331,
      "eval_runtime": 0.1345,
      "eval_samples_per_second": 2059.676,
      "eval_steps_per_second": 14.871,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9327962398529053,
      "learning_rate": 4.7e-05,
      "loss": 0.6402,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5667870036101083,
      "eval_loss": 0.7390427589416504,
      "eval_runtime": 0.1358,
      "eval_samples_per_second": 2039.025,
      "eval_steps_per_second": 14.722,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3827669620513916,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5556,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 0.8524482846260071,
      "eval_runtime": 0.1451,
      "eval_samples_per_second": 1909.466,
      "eval_steps_per_second": 13.787,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.2500104904174805,
      "learning_rate": 4.5e-05,
      "loss": 0.439,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 1.0051864385604858,
      "eval_runtime": 0.1749,
      "eval_samples_per_second": 1583.554,
      "eval_steps_per_second": 11.434,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.589412689208984,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2973,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 1.2635422945022583,
      "eval_runtime": 0.1266,
      "eval_samples_per_second": 2187.437,
      "eval_steps_per_second": 15.794,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 989531467960320.0,
      "train_loss": 0.5564495722452799,
      "train_runtime": 22.5366,
      "train_samples_per_second": 5524.336,
      "train_steps_per_second": 22.186
    }
  ],
  "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": 989531467960320.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}