run-0/checkpoint-3000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.04289544235925,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6429768800735474,
      "eval_loss": 0.5945129990577698,
      "eval_runtime": 45.8254,
      "eval_samples_per_second": 65.095,
      "eval_steps_per_second": 4.081,
      "step": 373
    },
    {
      "epoch": 1.3404825737265416,
      "grad_norm": 10.595976829528809,
      "learning_rate": 8.574439800439004e-06,
      "loss": 0.6212,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6778411269187927,
      "eval_loss": 0.5707619786262512,
      "eval_runtime": 45.8803,
      "eval_samples_per_second": 65.017,
      "eval_steps_per_second": 4.076,
      "step": 746
    },
    {
      "epoch": 2.680965147453083,
      "grad_norm": 11.34949779510498,
      "learning_rate": 7.247127137832347e-06,
      "loss": 0.5391,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6744887828826904,
      "eval_loss": 0.5923792123794556,
      "eval_runtime": 51.3582,
      "eval_samples_per_second": 58.082,
      "eval_steps_per_second": 3.641,
      "step": 1119
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6761649250984192,
      "eval_loss": 0.58919757604599,
      "eval_runtime": 49.5338,
      "eval_samples_per_second": 60.221,
      "eval_steps_per_second": 3.775,
      "step": 1492
    },
    {
      "epoch": 4.021447721179625,
      "grad_norm": 10.796440124511719,
      "learning_rate": 5.919814475225689e-06,
      "loss": 0.4812,
      "step": 1500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.683875322341919,
      "eval_loss": 0.6181617975234985,
      "eval_runtime": 49.5046,
      "eval_samples_per_second": 60.257,
      "eval_steps_per_second": 3.777,
      "step": 1865
    },
    {
      "epoch": 5.361930294906166,
      "grad_norm": 20.43950653076172,
      "learning_rate": 4.592501812619033e-06,
      "loss": 0.4094,
      "step": 2000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.690244734287262,
      "eval_loss": 0.6364904642105103,
      "eval_runtime": 49.2434,
      "eval_samples_per_second": 60.577,
      "eval_steps_per_second": 3.797,
      "step": 2238
    },
    {
      "epoch": 6.702412868632708,
      "grad_norm": 15.7015962600708,
      "learning_rate": 3.265189150012376e-06,
      "loss": 0.3557,
      "step": 2500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.690244734287262,
      "eval_loss": 0.6809911131858826,
      "eval_runtime": 53.7724,
      "eval_samples_per_second": 55.475,
      "eval_steps_per_second": 3.478,
      "step": 2611
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.687562882900238,
      "eval_loss": 0.7227169871330261,
      "eval_runtime": 48.2616,
      "eval_samples_per_second": 61.809,
      "eval_steps_per_second": 3.875,
      "step": 2984
    },
    {
      "epoch": 8.04289544235925,
      "grad_norm": 21.409116744995117,
      "learning_rate": 1.937876487405719e-06,
      "loss": 0.3167,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 3730,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 4358935077786780.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 9.90175246304566e-06,
    "per_device_train_batch_size": 32
  }
}