run-4/checkpoint-1000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.1446540880503147,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5758064516129032,
      "eval_loss": 0.19529514014720917,
      "eval_runtime": 2.038,
      "eval_samples_per_second": 1521.129,
      "eval_steps_per_second": 31.895,
      "step": 318
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.5219656229019165,
      "learning_rate": 1.4758909853249476e-05,
      "loss": 0.3104,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8041935483870968,
      "eval_loss": 0.10026105493307114,
      "eval_runtime": 2.0307,
      "eval_samples_per_second": 1526.577,
      "eval_steps_per_second": 32.009,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8651612903225806,
      "eval_loss": 0.07088375836610794,
      "eval_runtime": 2.1975,
      "eval_samples_per_second": 1410.683,
      "eval_steps_per_second": 29.579,
      "step": 954
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.44089475274086,
      "learning_rate": 9.517819706498952e-06,
      "loss": 0.1155,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 1908,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 260941334653608.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9330141802487892,
    "num_train_epochs": 6,
    "temperature": 18
  }
}