run-4/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": 0.5790322580645161,
      "eval_loss": 0.20513561367988586,
      "eval_runtime": 1.4355,
      "eval_samples_per_second": 2159.52,
      "eval_steps_per_second": 45.28,
      "step": 318
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.4758909853249476e-05,
      "loss": 0.3247,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8151612903225807,
      "eval_loss": 0.1044410690665245,
      "eval_runtime": 1.4405,
      "eval_samples_per_second": 2151.985,
      "eval_steps_per_second": 45.122,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8693548387096774,
      "eval_loss": 0.07303278893232346,
      "eval_runtime": 1.4449,
      "eval_samples_per_second": 2145.49,
      "eval_steps_per_second": 44.986,
      "step": 954
    },
    {
      "epoch": 3.14,
      "learning_rate": 9.517819706498952e-06,
      "loss": 0.1197,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8883870967741936,
      "eval_loss": 0.05825456231832504,
      "eval_runtime": 1.4489,
      "eval_samples_per_second": 2139.522,
      "eval_steps_per_second": 44.861,
      "step": 1272
    },
    {
      "epoch": 4.72,
      "learning_rate": 4.276729559748428e-06,
      "loss": 0.0817,
      "step": 1500
    }
  ],
  "logging_steps": 500,
  "max_steps": 1908,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 391111083046572.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8685261330961369,
    "num_train_epochs": 6,
    "temperature": 11
  }
}