{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.498750624687656,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24987506246876562,
      "grad_norm": 1.422733187675476,
      "learning_rate": 4.5835415625520576e-05,
      "loss": 6.5696,
      "step": 500
    },
    {
      "epoch": 0.49975012493753124,
      "grad_norm": 1.82273268699646,
      "learning_rate": 4.167083125104115e-05,
      "loss": 4.7727,
      "step": 1000
    },
    {
      "epoch": 0.7496251874062968,
      "grad_norm": 2.662141799926758,
      "learning_rate": 3.7506246876561725e-05,
      "loss": 4.4216,
      "step": 1500
    },
    {
      "epoch": 0.9995002498750625,
      "grad_norm": 2.524129629135132,
      "learning_rate": 3.334166250208229e-05,
      "loss": 4.178,
      "step": 2000
    },
    {
      "epoch": 1.249375312343828,
      "grad_norm": 2.6071786880493164,
      "learning_rate": 2.917707812760287e-05,
      "loss": 3.9573,
      "step": 2500
    },
    {
      "epoch": 1.4992503748125938,
      "grad_norm": 2.997709274291992,
      "learning_rate": 2.501249375312344e-05,
      "loss": 3.86,
      "step": 3000
    },
    {
      "epoch": 1.7491254372813594,
      "grad_norm": 2.626232862472534,
      "learning_rate": 2.084790937864401e-05,
      "loss": 3.7562,
      "step": 3500
    },
    {
      "epoch": 1.999000499750125,
      "grad_norm": 3.0749871730804443,
      "learning_rate": 1.6683325004164585e-05,
      "loss": 3.7093,
      "step": 4000
    },
    {
      "epoch": 2.2488755622188905,
      "grad_norm": 3.391836166381836,
      "learning_rate": 1.2518740629685157e-05,
      "loss": 3.5835,
      "step": 4500
    },
    {
      "epoch": 2.498750624687656,
      "grad_norm": 3.222475051879883,
      "learning_rate": 8.354156255205731e-06,
      "loss": 3.5654,
      "step": 5000
    }
  ],
  "logging_steps": 500,
  "max_steps": 6003,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 677411962748928.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}