{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9961568024596464,
  "eval_steps": 10,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 3.9923893961834914e-05,
      "loss": 1.3268,
      "step": 1
    },
    {
      "epoch": 0.14,
      "learning_rate": 3.812615574073301e-05,
      "loss": 1.2548,
      "step": 5
    },
    {
      "epoch": 0.28,
      "learning_rate": 3.285575219373079e-05,
      "loss": 1.1912,
      "step": 10
    },
    {
      "epoch": 0.28,
      "eval_loss": 1.109934687614441,
      "eval_runtime": 9.2946,
      "eval_samples_per_second": 5.057,
      "eval_steps_per_second": 1.291,
      "step": 10
    },
    {
      "epoch": 0.42,
      "learning_rate": 2.5176380902050418e-05,
      "loss": 1.1433,
      "step": 15
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.6527036446661396e-05,
      "loss": 1.1238,
      "step": 20
    },
    {
      "epoch": 0.55,
      "eval_loss": 1.065536379814148,
      "eval_runtime": 9.2887,
      "eval_samples_per_second": 5.06,
      "eval_steps_per_second": 1.292,
      "step": 20
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.528471272979083e-06,
      "loss": 1.1102,
      "step": 25
    },
    {
      "epoch": 0.83,
      "learning_rate": 2.679491924311226e-06,
      "loss": 1.1258,
      "step": 30
    },
    {
      "epoch": 0.83,
      "eval_loss": 1.055001974105835,
      "eval_runtime": 9.2909,
      "eval_samples_per_second": 5.059,
      "eval_steps_per_second": 1.292,
      "step": 30
    },
    {
      "epoch": 0.97,
      "learning_rate": 7.61060381650891e-08,
      "loss": 1.1272,
      "step": 35
    },
    {
      "epoch": 1.0,
      "step": 36,
      "total_flos": 1.0010669722946765e+17,
      "train_loss": 1.1544433269235823,
      "train_runtime": 5078.5964,
      "train_samples_per_second": 1.537,
      "train_steps_per_second": 0.007
    }
  ],
  "logging_steps": 5,
  "max_steps": 36,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20,
  "total_flos": 1.0010669722946765e+17,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}