trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 320,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03125,
      "grad_norm": 27.536455154418945,
      "learning_rate": 2e-05,
      "loss": 5.269,
      "mean_token_accuracy": 0.193113774061203,
      "num_tokens": 1320.0,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.073838233947754,
      "learning_rate": 1.8062500000000002e-05,
      "loss": 2.9886,
      "mean_token_accuracy": 0.4425700907745669,
      "num_tokens": 49123.0,
      "step": 32
    },
    {
      "epoch": 2.0,
      "grad_norm": 72.79724884033203,
      "learning_rate": 1.60625e-05,
      "loss": 1.7728,
      "mean_token_accuracy": 0.6223073434084654,
      "num_tokens": 98246.0,
      "step": 64
    },
    {
      "epoch": 3.0,
      "grad_norm": 23.59392738342285,
      "learning_rate": 1.4062500000000001e-05,
      "loss": 1.0599,
      "mean_token_accuracy": 0.7586388783529401,
      "num_tokens": 147369.0,
      "step": 96
    },
    {
      "epoch": 4.0,
      "grad_norm": 33.94417190551758,
      "learning_rate": 1.2062500000000002e-05,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8688167557120323,
      "num_tokens": 196492.0,
      "step": 128
    },
    {
      "epoch": 5.0,
      "grad_norm": 28.460622787475586,
      "learning_rate": 1.0062500000000002e-05,
      "loss": 0.2783,
      "mean_token_accuracy": 0.9295041039586067,
      "num_tokens": 245615.0,
      "step": 160
    },
    {
      "epoch": 6.0,
      "grad_norm": 13.504148483276367,
      "learning_rate": 8.062500000000001e-06,
      "loss": 0.1414,
      "mean_token_accuracy": 0.9666158780455589,
      "num_tokens": 294738.0,
      "step": 192
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.4141818284988403,
      "learning_rate": 6.0625e-06,
      "loss": 0.061,
      "mean_token_accuracy": 0.9868465270847082,
      "num_tokens": 343861.0,
      "step": 224
    },
    {
      "epoch": 8.0,
      "grad_norm": 6.186761856079102,
      "learning_rate": 4.0625000000000005e-06,
      "loss": 0.0437,
      "mean_token_accuracy": 0.9876908175647259,
      "num_tokens": 392984.0,
      "step": 256
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.0,
      "learning_rate": 2.0625e-06,
      "loss": 0.0267,
      "mean_token_accuracy": 0.9606593921780586,
      "num_tokens": 442107.0,
      "step": 288
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.08612818270921707,
      "learning_rate": 6.250000000000001e-08,
      "loss": 0.017,
      "mean_token_accuracy": 0.993696304038167,
      "num_tokens": 491230.0,
      "step": 320
    },
    {
      "epoch": 10.0,
      "step": 320,
      "total_flos": 3034108976431104.0,
      "train_loss": 0.6985519178211689,
      "train_runtime": 385.0713,
      "train_samples_per_second": 6.466,
      "train_steps_per_second": 0.831
    }
  ],
  "logging_steps": 10,
  "max_steps": 320,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3034108976431104.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}