trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.422981321811676,
  "best_model_checkpoint": "bert_base_rand_20_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 8.932740211486816,
      "learning_rate": 4.9e-05,
      "loss": 0.3871,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.422981321811676,
      "eval_runtime": 0.5701,
      "eval_samples_per_second": 1529.429,
      "eval_steps_per_second": 7.016,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.848577499389648,
      "learning_rate": 4.8e-05,
      "loss": 0.2227,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8073394495412844,
      "eval_loss": 0.5242372751235962,
      "eval_runtime": 0.5886,
      "eval_samples_per_second": 1481.402,
      "eval_steps_per_second": 6.795,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.9569454193115234,
      "learning_rate": 4.7e-05,
      "loss": 0.1637,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7935779816513762,
      "eval_loss": 0.5711498856544495,
      "eval_runtime": 0.5784,
      "eval_samples_per_second": 1507.561,
      "eval_steps_per_second": 6.915,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.274488925933838,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1284,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.5830358862876892,
      "eval_runtime": 0.5887,
      "eval_samples_per_second": 1481.226,
      "eval_steps_per_second": 6.795,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.7386900186538696,
      "learning_rate": 4.5e-05,
      "loss": 0.102,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7993119266055045,
      "eval_loss": 0.6198598742485046,
      "eval_runtime": 0.5631,
      "eval_samples_per_second": 1548.482,
      "eval_steps_per_second": 7.103,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.527076721191406,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0793,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.692043125629425,
      "eval_runtime": 0.5689,
      "eval_samples_per_second": 1532.765,
      "eval_steps_per_second": 7.031,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.18052807600811274,
      "train_runtime": 669.3712,
      "train_samples_per_second": 5030.766,
      "train_steps_per_second": 19.72
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "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": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}