run-1/checkpoint-1500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.716981132075472,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.577741935483871,
      "eval_loss": 0.19667640328407288,
      "eval_runtime": 5.3568,
      "eval_samples_per_second": 578.705,
      "eval_steps_per_second": 12.134,
      "step": 318
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.5251966714859009,
      "learning_rate": 1.550763701707098e-05,
      "loss": 0.3139,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8138709677419355,
      "eval_loss": 0.09817331284284592,
      "eval_runtime": 5.8454,
      "eval_samples_per_second": 530.332,
      "eval_steps_per_second": 11.12,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8712903225806452,
      "eval_loss": 0.06764764338731766,
      "eval_runtime": 5.7054,
      "eval_samples_per_second": 543.342,
      "eval_steps_per_second": 11.393,
      "step": 954
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.4463839828968048,
      "learning_rate": 1.101527403414196e-05,
      "loss": 0.1135,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8935483870967742,
      "eval_loss": 0.05397525057196617,
      "eval_runtime": 5.4307,
      "eval_samples_per_second": 570.828,
      "eval_steps_per_second": 11.969,
      "step": 1272
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.31221386790275574,
      "learning_rate": 6.522911051212939e-06,
      "loss": 0.0757,
      "step": 1500
    }
  ],
  "logging_steps": 500,
  "max_steps": 2226,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 391368939443328.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8149797657981518,
    "num_train_epochs": 7,
    "temperature": 15
  }
}