{
  "best_metric": 0.6774637699127197,
  "best_model_checkpoint": "output/fine_tuning/checkpoints/Meta-Llama-3.1-8B-Instruct/sft/aixpa-ground-short-docs-checkpoint/checkpoint-340",
  "epoch": 1.8181818181818183,
  "eval_steps": 20,
  "global_step": 340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 0.2608170211315155,
      "learning_rate": 1.4819165403057078e-05,
      "loss": 1.5177,
      "mean_token_accuracy": 0.6778935924172401,
      "step": 20
    },
    {
      "epoch": 0.10695187165775401,
      "eval_loss": 1.4139869213104248,
      "eval_mean_token_accuracy": 0.6904795635037306,
      "eval_runtime": 1035.6849,
      "eval_samples_per_second": 0.238,
      "eval_steps_per_second": 0.119,
      "step": 20
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 0.2079063504934311,
      "learning_rate": 1.8247997414535347e-05,
      "loss": 1.3019,
      "mean_token_accuracy": 0.7044360123574733,
      "step": 40
    },
    {
      "epoch": 0.21390374331550802,
      "eval_loss": 1.1822656393051147,
      "eval_mean_token_accuracy": 0.7299752424402934,
      "eval_runtime": 1036.3743,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 40
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 0.24613960087299347,
      "learning_rate": 2e-05,
      "loss": 1.1518,
      "mean_token_accuracy": 0.7350467927753925,
      "step": 60
    },
    {
      "epoch": 0.32085561497326204,
      "eval_loss": 1.0543467998504639,
      "eval_mean_token_accuracy": 0.7558261015550877,
      "eval_runtime": 1037.5825,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 60
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 0.21977411210536957,
      "learning_rate": 2e-05,
      "loss": 1.0451,
      "mean_token_accuracy": 0.758122804760933,
      "step": 80
    },
    {
      "epoch": 0.42780748663101603,
      "eval_loss": 0.9605845808982849,
      "eval_mean_token_accuracy": 0.7781746368098065,
      "eval_runtime": 1036.6763,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 80
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 0.29401177167892456,
      "learning_rate": 2e-05,
      "loss": 0.9781,
      "mean_token_accuracy": 0.7737521544098854,
      "step": 100
    },
    {
      "epoch": 0.5347593582887701,
      "eval_loss": 0.9001632332801819,
      "eval_mean_token_accuracy": 0.7922655029025504,
      "eval_runtime": 1037.6159,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 100
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 0.34717148542404175,
      "learning_rate": 2e-05,
      "loss": 0.9113,
      "mean_token_accuracy": 0.7882794156670571,
      "step": 120
    },
    {
      "epoch": 0.6417112299465241,
      "eval_loss": 0.8677236437797546,
      "eval_mean_token_accuracy": 0.7979402353123921,
      "eval_runtime": 1037.0163,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 120
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 0.3498166799545288,
      "learning_rate": 2e-05,
      "loss": 0.8725,
      "mean_token_accuracy": 0.7943845748901367,
      "step": 140
    },
    {
      "epoch": 0.7486631016042781,
      "eval_loss": 0.8351719379425049,
      "eval_mean_token_accuracy": 0.8040957174650053,
      "eval_runtime": 1036.2597,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 140
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 0.3868383467197418,
      "learning_rate": 2e-05,
      "loss": 0.8721,
      "mean_token_accuracy": 0.7931242920458317,
      "step": 160
    },
    {
      "epoch": 0.8556149732620321,
      "eval_loss": 0.8117150068283081,
      "eval_mean_token_accuracy": 0.808245500413383,
      "eval_runtime": 1036.3899,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 160
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 0.5154958367347717,
      "learning_rate": 2e-05,
      "loss": 0.83,
      "mean_token_accuracy": 0.8012300632894039,
      "step": 180
    },
    {
      "epoch": 0.9625668449197861,
      "eval_loss": 0.7896639108657837,
      "eval_mean_token_accuracy": 0.8132512133295943,
      "eval_runtime": 1036.5717,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 180
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 0.42563366889953613,
      "learning_rate": 2e-05,
      "loss": 0.8034,
      "mean_token_accuracy": 0.8062243178486824,
      "step": 200
    },
    {
      "epoch": 1.0695187165775402,
      "eval_loss": 0.7641515731811523,
      "eval_mean_token_accuracy": 0.8180098591781244,
      "eval_runtime": 1037.8429,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 200
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.39126402139663696,
      "learning_rate": 2e-05,
      "loss": 0.7637,
      "mean_token_accuracy": 0.8159952461719513,
      "step": 220
    },
    {
      "epoch": 1.1764705882352942,
      "eval_loss": 0.7506969571113586,
      "eval_mean_token_accuracy": 0.8211809416127399,
      "eval_runtime": 1036.9492,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 220
    },
    {
      "epoch": 1.2834224598930482,
      "grad_norm": 0.525314211845398,
      "learning_rate": 2e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.8259521864354611,
      "step": 240
    },
    {
      "epoch": 1.2834224598930482,
      "eval_loss": 0.7359923124313354,
      "eval_mean_token_accuracy": 0.8241757876504727,
      "eval_runtime": 1037.3586,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 240
    },
    {
      "epoch": 1.3903743315508021,
      "grad_norm": 1.2710996866226196,
      "learning_rate": 2e-05,
      "loss": 0.7084,
      "mean_token_accuracy": 0.8261168003082275,
      "step": 260
    },
    {
      "epoch": 1.3903743315508021,
      "eval_loss": 0.7302640676498413,
      "eval_mean_token_accuracy": 0.8257462154559003,
      "eval_runtime": 1037.1577,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 260
    },
    {
      "epoch": 1.4973262032085561,
      "grad_norm": 0.5921723246574402,
      "learning_rate": 2e-05,
      "loss": 0.6984,
      "mean_token_accuracy": 0.8257287561893463,
      "step": 280
    },
    {
      "epoch": 1.4973262032085561,
      "eval_loss": 0.716602087020874,
      "eval_mean_token_accuracy": 0.8293129685448437,
      "eval_runtime": 1037.3171,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 280
    },
    {
      "epoch": 1.6042780748663101,
      "grad_norm": 0.6089026927947998,
      "learning_rate": 2e-05,
      "loss": 0.6591,
      "mean_token_accuracy": 0.8372392967343331,
      "step": 300
    },
    {
      "epoch": 1.6042780748663101,
      "eval_loss": 0.7121440768241882,
      "eval_mean_token_accuracy": 0.8315709296280775,
      "eval_runtime": 1038.1176,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.118,
      "step": 300
    },
    {
      "epoch": 1.7112299465240641,
      "grad_norm": 0.751674473285675,
      "learning_rate": 2e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8498695828020573,
      "step": 320
    },
    {
      "epoch": 1.7112299465240641,
      "eval_loss": 0.697968602180481,
      "eval_mean_token_accuracy": 0.8344496600027007,
      "eval_runtime": 1037.7814,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 320
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 1.045749545097351,
      "learning_rate": 2e-05,
      "loss": 0.6099,
      "mean_token_accuracy": 0.845644561946392,
      "step": 340
    },
    {
      "epoch": 1.8181818181818183,
      "eval_loss": 0.6774637699127197,
      "eval_mean_token_accuracy": 0.8406207769866881,
      "eval_runtime": 1037.5418,
      "eval_samples_per_second": 0.237,
      "eval_steps_per_second": 0.119,
      "step": 340
    }
  ],
  "logging_steps": 20,
  "max_steps": 1870,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 20,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.857740816294871e+18,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}