{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4331348132106118,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02707092582566324,
      "grad_norm": 0.8431992530822754,
      "learning_rate": 9.910124526258799e-05,
      "loss": 4.3576,
      "step": 250
    },
    {
      "epoch": 0.05414185165132648,
      "grad_norm": 0.8061181306838989,
      "learning_rate": 9.819888106839921e-05,
      "loss": 4.2775,
      "step": 500
    },
    {
      "epoch": 0.05414185165132648,
      "eval_loss": 3.5462777614593506,
      "eval_runtime": 87.734,
      "eval_samples_per_second": 112.271,
      "eval_steps_per_second": 7.021,
      "step": 500
    },
    {
      "epoch": 0.08121277747698971,
      "grad_norm": 0.7875335812568665,
      "learning_rate": 9.729651687421044e-05,
      "loss": 4.1927,
      "step": 750
    },
    {
      "epoch": 0.10828370330265295,
      "grad_norm": 1.0330173969268799,
      "learning_rate": 9.639415268002166e-05,
      "loss": 4.1178,
      "step": 1000
    },
    {
      "epoch": 0.10828370330265295,
      "eval_loss": 3.3765828609466553,
      "eval_runtime": 87.6227,
      "eval_samples_per_second": 112.414,
      "eval_steps_per_second": 7.03,
      "step": 1000
    },
    {
      "epoch": 0.1353546291283162,
      "grad_norm": 1.041638731956482,
      "learning_rate": 9.549178848583288e-05,
      "loss": 4.0201,
      "step": 1250
    },
    {
      "epoch": 0.16242555495397942,
      "grad_norm": 0.8976061940193176,
      "learning_rate": 9.458942429164411e-05,
      "loss": 3.9802,
      "step": 1500
    },
    {
      "epoch": 0.16242555495397942,
      "eval_loss": 3.2597038745880127,
      "eval_runtime": 87.5921,
      "eval_samples_per_second": 112.453,
      "eval_steps_per_second": 7.033,
      "step": 1500
    },
    {
      "epoch": 0.18949648077964265,
      "grad_norm": 1.0333911180496216,
      "learning_rate": 9.368706009745533e-05,
      "loss": 3.9335,
      "step": 1750
    },
    {
      "epoch": 0.2165674066053059,
      "grad_norm": 0.9179081320762634,
      "learning_rate": 9.278469590326656e-05,
      "loss": 3.8709,
      "step": 2000
    },
    {
      "epoch": 0.2165674066053059,
      "eval_loss": 3.1846394538879395,
      "eval_runtime": 87.5421,
      "eval_samples_per_second": 112.517,
      "eval_steps_per_second": 7.037,
      "step": 2000
    },
    {
      "epoch": 0.24363833243096913,
      "grad_norm": 1.0419113636016846,
      "learning_rate": 9.188233170907778e-05,
      "loss": 3.8416,
      "step": 2250
    },
    {
      "epoch": 0.2707092582566324,
      "grad_norm": 0.9652225375175476,
      "learning_rate": 9.0979967514889e-05,
      "loss": 3.807,
      "step": 2500
    },
    {
      "epoch": 0.2707092582566324,
      "eval_loss": 3.119335174560547,
      "eval_runtime": 87.4769,
      "eval_samples_per_second": 112.601,
      "eval_steps_per_second": 7.042,
      "step": 2500
    },
    {
      "epoch": 0.2977801840822956,
      "grad_norm": 0.8792561888694763,
      "learning_rate": 9.007760332070024e-05,
      "loss": 3.7754,
      "step": 2750
    },
    {
      "epoch": 0.32485110990795885,
      "grad_norm": 0.9625837206840515,
      "learning_rate": 8.917523912651147e-05,
      "loss": 3.7471,
      "step": 3000
    },
    {
      "epoch": 0.32485110990795885,
      "eval_loss": 3.0782463550567627,
      "eval_runtime": 87.7273,
      "eval_samples_per_second": 112.28,
      "eval_steps_per_second": 7.022,
      "step": 3000
    },
    {
      "epoch": 0.3519220357336221,
      "grad_norm": 1.0289523601531982,
      "learning_rate": 8.827287493232269e-05,
      "loss": 3.7326,
      "step": 3250
    },
    {
      "epoch": 0.3789929615592853,
      "grad_norm": 0.9764179587364197,
      "learning_rate": 8.737051073813391e-05,
      "loss": 3.6939,
      "step": 3500
    },
    {
      "epoch": 0.3789929615592853,
      "eval_loss": 3.052320718765259,
      "eval_runtime": 87.6344,
      "eval_samples_per_second": 112.399,
      "eval_steps_per_second": 7.029,
      "step": 3500
    },
    {
      "epoch": 0.4060638873849486,
      "grad_norm": 0.9247903227806091,
      "learning_rate": 8.646814654394514e-05,
      "loss": 3.6782,
      "step": 3750
    },
    {
      "epoch": 0.4331348132106118,
      "grad_norm": 0.9769233465194702,
      "learning_rate": 8.556578234975636e-05,
      "loss": 3.6654,
      "step": 4000
    },
    {
      "epoch": 0.4331348132106118,
      "eval_loss": 3.0321156978607178,
      "eval_runtime": 87.6462,
      "eval_samples_per_second": 112.384,
      "eval_steps_per_second": 7.028,
      "step": 4000
    }
  ],
  "logging_steps": 250,
  "max_steps": 27705,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1444216307712000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}