run-0/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": 0.6061290322580645,
      "eval_loss": 0.2067662924528122,
      "eval_runtime": 1.83,
      "eval_samples_per_second": 1693.985,
      "eval_steps_per_second": 35.519,
      "step": 318
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.650593990216632e-05,
      "loss": 0.3335,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8312903225806452,
      "eval_loss": 0.09916964918375015,
      "eval_runtime": 2.1477,
      "eval_samples_per_second": 1443.399,
      "eval_steps_per_second": 30.265,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8816129032258064,
      "eval_loss": 0.06577017158269882,
      "eval_runtime": 1.8521,
      "eval_samples_per_second": 1673.812,
      "eval_steps_per_second": 35.096,
      "step": 954
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.3011879804332637e-05,
      "loss": 0.1145,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8987096774193548,
      "eval_loss": 0.04967508092522621,
      "eval_runtime": 1.8807,
      "eval_samples_per_second": 1648.287,
      "eval_steps_per_second": 34.561,
      "step": 1272
    },
    {
      "epoch": 4.72,
      "learning_rate": 9.517819706498952e-06,
      "loss": 0.0721,
      "step": 1500
    }
  ],
  "logging_steps": 500,
  "max_steps": 2862,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 391111083046572.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.08109385800901192,
    "num_train_epochs": 9,
    "temperature": 8
  }
}