trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.1,
  "eval_steps": 188,
  "global_step": 940,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010638297872340426,
      "grad_norm": 2.3249564170837402,
      "learning_rate": 1e-05,
      "loss": 3.5833,
      "step": 1
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4263363778591156,
      "learning_rate": 0.00094,
      "loss": 1.4683,
      "step": 94
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.5972639322280884,
      "learning_rate": 0.0008952380952380953,
      "loss": 0.4344,
      "step": 188
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.3692754805088043,
      "eval_runtime": 71.2145,
      "eval_samples_per_second": 6.319,
      "eval_steps_per_second": 0.8,
      "step": 188
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.6023200750350952,
      "learning_rate": 0.0007833333333333334,
      "loss": 0.2858,
      "step": 282
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.4565546214580536,
      "learning_rate": 0.0006714285714285714,
      "loss": 0.1875,
      "step": 376
    },
    {
      "epoch": 3.1,
      "eval_loss": 0.3102695643901825,
      "eval_runtime": 69.8732,
      "eval_samples_per_second": 6.44,
      "eval_steps_per_second": 0.816,
      "step": 376
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.43287548422813416,
      "learning_rate": 0.0005595238095238096,
      "loss": 0.1162,
      "step": 470
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.24221746623516083,
      "learning_rate": 0.00044761904761904766,
      "loss": 0.0717,
      "step": 564
    },
    {
      "epoch": 5.1,
      "eval_loss": 0.30600059032440186,
      "eval_runtime": 70.275,
      "eval_samples_per_second": 6.403,
      "eval_steps_per_second": 0.811,
      "step": 564
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.2153695970773697,
      "learning_rate": 0.0003357142857142857,
      "loss": 0.0363,
      "step": 658
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.14185360074043274,
      "learning_rate": 0.00022380952380952383,
      "loss": 0.0202,
      "step": 752
    },
    {
      "epoch": 7.1,
      "eval_loss": 0.32697898149490356,
      "eval_runtime": 69.9983,
      "eval_samples_per_second": 6.429,
      "eval_steps_per_second": 0.814,
      "step": 752
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.061574213206768036,
      "learning_rate": 0.00011190476190476191,
      "loss": 0.0134,
      "step": 846
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.06894426047801971,
      "learning_rate": 0.0,
      "loss": 0.0101,
      "step": 940
    },
    {
      "epoch": 9.1,
      "eval_loss": 0.34040552377700806,
      "eval_runtime": 70.2499,
      "eval_samples_per_second": 6.406,
      "eval_steps_per_second": 0.811,
      "step": 940
    }
  ],
  "logging_steps": 94,
  "max_steps": 940,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 188,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.4148115456e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}