{
  "best_metric": 1.9688726663589478,
  "best_model_checkpoint": "gen_test/checkpoint-6252",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 6252,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 318423.84375,
      "learning_rate": 1.8400511836212414e-05,
      "loss": 2.1469,
      "step": 500
    },
    {
      "epoch": 0.64,
      "grad_norm": 273106.9375,
      "learning_rate": 1.6801023672424827e-05,
      "loss": 1.9333,
      "step": 1000
    },
    {
      "epoch": 0.96,
      "grad_norm": 250709.375,
      "learning_rate": 1.5201535508637238e-05,
      "loss": 1.8887,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.0289978981018066,
      "eval_runtime": 56.4689,
      "eval_samples_per_second": 177.089,
      "eval_steps_per_second": 2.78,
      "step": 1563
    },
    {
      "epoch": 1.28,
      "grad_norm": 241871.234375,
      "learning_rate": 1.3602047344849649e-05,
      "loss": 1.8677,
      "step": 2000
    },
    {
      "epoch": 1.6,
      "grad_norm": 229897.359375,
      "learning_rate": 1.2002559181062061e-05,
      "loss": 1.8518,
      "step": 2500
    },
    {
      "epoch": 1.92,
      "grad_norm": 201821.484375,
      "learning_rate": 1.0403071017274472e-05,
      "loss": 1.8464,
      "step": 3000
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.0003182888031006,
      "eval_runtime": 56.3343,
      "eval_samples_per_second": 177.512,
      "eval_steps_per_second": 2.787,
      "step": 3126
    },
    {
      "epoch": 2.24,
      "grad_norm": 229470.453125,
      "learning_rate": 8.803582853486885e-06,
      "loss": 1.8306,
      "step": 3500
    },
    {
      "epoch": 2.56,
      "grad_norm": 209711.5,
      "learning_rate": 7.204094689699297e-06,
      "loss": 1.8225,
      "step": 4000
    },
    {
      "epoch": 2.88,
      "grad_norm": 202330.5,
      "learning_rate": 5.6046065259117085e-06,
      "loss": 1.8148,
      "step": 4500
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.9800028800964355,
      "eval_runtime": 56.3231,
      "eval_samples_per_second": 177.547,
      "eval_steps_per_second": 2.787,
      "step": 4689
    },
    {
      "epoch": 3.2,
      "grad_norm": 202021.6875,
      "learning_rate": 4.005118362124121e-06,
      "loss": 1.8124,
      "step": 5000
    },
    {
      "epoch": 3.52,
      "grad_norm": 201338.140625,
      "learning_rate": 2.4056301983365325e-06,
      "loss": 1.801,
      "step": 5500
    },
    {
      "epoch": 3.84,
      "grad_norm": 212086.015625,
      "learning_rate": 8.061420345489445e-07,
      "loss": 1.8055,
      "step": 6000
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.9688726663589478,
      "eval_runtime": 56.3874,
      "eval_samples_per_second": 177.345,
      "eval_steps_per_second": 2.784,
      "step": 6252
    }
  ],
  "logging_steps": 500,
  "max_steps": 6252,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 2.090336256e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}