{
  "best_global_step": 3000,
  "best_metric": 2.238426446914673,
  "best_model_checkpoint": "./gpt2-alpaca-lora/checkpoint-3000",
  "epoch": 1.9430051813471503,
  "eval_steps": 250,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16191709844559585,
      "eval_loss": 2.402658224105835,
      "eval_mean_token_accuracy": 0.5123065428499796,
      "eval_num_tokens": 703050.0,
      "eval_runtime": 4.6408,
      "eval_samples_per_second": 560.46,
      "eval_steps_per_second": 35.123,
      "step": 250
    },
    {
      "epoch": 0.3238341968911917,
      "grad_norm": 0.5626281499862671,
      "learning_rate": 0.0002,
      "loss": 2.5788,
      "step": 500
    },
    {
      "epoch": 0.3238341968911917,
      "eval_loss": 2.35490083694458,
      "eval_mean_token_accuracy": 0.5189222602025132,
      "eval_num_tokens": 1405018.0,
      "eval_runtime": 4.6775,
      "eval_samples_per_second": 556.069,
      "eval_steps_per_second": 34.848,
      "step": 500
    },
    {
      "epoch": 0.48575129533678757,
      "eval_loss": 2.327059030532837,
      "eval_mean_token_accuracy": 0.5225200927330672,
      "eval_num_tokens": 2104755.0,
      "eval_runtime": 4.6238,
      "eval_samples_per_second": 562.53,
      "eval_steps_per_second": 35.253,
      "step": 750
    },
    {
      "epoch": 0.6476683937823834,
      "grad_norm": 0.4609125554561615,
      "learning_rate": 0.0002,
      "loss": 2.4241,
      "step": 1000
    },
    {
      "epoch": 0.6476683937823834,
      "eval_loss": 2.3085591793060303,
      "eval_mean_token_accuracy": 0.5245303420201402,
      "eval_num_tokens": 2811160.0,
      "eval_runtime": 4.5658,
      "eval_samples_per_second": 569.667,
      "eval_steps_per_second": 35.7,
      "step": 1000
    },
    {
      "epoch": 0.8095854922279793,
      "eval_loss": 2.2929277420043945,
      "eval_mean_token_accuracy": 0.5260607554503014,
      "eval_num_tokens": 3506263.0,
      "eval_runtime": 4.6703,
      "eval_samples_per_second": 556.922,
      "eval_steps_per_second": 34.901,
      "step": 1250
    },
    {
      "epoch": 0.9715025906735751,
      "grad_norm": 0.5748232007026672,
      "learning_rate": 0.0002,
      "loss": 2.3911,
      "step": 1500
    },
    {
      "epoch": 0.9715025906735751,
      "eval_loss": 2.281684398651123,
      "eval_mean_token_accuracy": 0.5284550193628651,
      "eval_num_tokens": 4209663.0,
      "eval_runtime": 4.5679,
      "eval_samples_per_second": 569.406,
      "eval_steps_per_second": 35.684,
      "step": 1500
    },
    {
      "epoch": 1.133419689119171,
      "eval_loss": 2.2740871906280518,
      "eval_mean_token_accuracy": 0.5288037231729075,
      "eval_num_tokens": 4904955.0,
      "eval_runtime": 4.6289,
      "eval_samples_per_second": 561.909,
      "eval_steps_per_second": 35.214,
      "step": 1750
    },
    {
      "epoch": 1.2953367875647668,
      "grad_norm": 0.5409476161003113,
      "learning_rate": 0.0002,
      "loss": 2.3581,
      "step": 2000
    },
    {
      "epoch": 1.2953367875647668,
      "eval_loss": 2.2654471397399902,
      "eval_mean_token_accuracy": 0.5307357484943296,
      "eval_num_tokens": 5609264.0,
      "eval_runtime": 4.7388,
      "eval_samples_per_second": 548.87,
      "eval_steps_per_second": 34.397,
      "step": 2000
    },
    {
      "epoch": 1.4572538860103628,
      "eval_loss": 2.2559146881103516,
      "eval_mean_token_accuracy": 0.5315202057727275,
      "eval_num_tokens": 6300985.0,
      "eval_runtime": 4.6198,
      "eval_samples_per_second": 563.008,
      "eval_steps_per_second": 35.283,
      "step": 2250
    },
    {
      "epoch": 1.6191709844559585,
      "grad_norm": 0.6490165591239929,
      "learning_rate": 0.0002,
      "loss": 2.3444,
      "step": 2500
    },
    {
      "epoch": 1.6191709844559585,
      "eval_loss": 2.2506017684936523,
      "eval_mean_token_accuracy": 0.5325605070298435,
      "eval_num_tokens": 6998999.0,
      "eval_runtime": 4.6688,
      "eval_samples_per_second": 557.099,
      "eval_steps_per_second": 34.912,
      "step": 2500
    },
    {
      "epoch": 1.7810880829015545,
      "eval_loss": 2.2434535026550293,
      "eval_mean_token_accuracy": 0.533348783393579,
      "eval_num_tokens": 7714495.0,
      "eval_runtime": 4.628,
      "eval_samples_per_second": 562.011,
      "eval_steps_per_second": 35.22,
      "step": 2750
    },
    {
      "epoch": 1.9430051813471503,
      "grad_norm": 0.5797879099845886,
      "learning_rate": 0.0002,
      "loss": 2.33,
      "step": 3000
    },
    {
      "epoch": 1.9430051813471503,
      "eval_loss": 2.238426446914673,
      "eval_mean_token_accuracy": 0.5334653168733866,
      "eval_num_tokens": 8410972.0,
      "eval_runtime": 4.621,
      "eval_samples_per_second": 562.868,
      "eval_steps_per_second": 35.274,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5564002921906176.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}