{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 654,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 0.4296875,
      "learning_rate": 4.942660550458716e-05,
      "loss": 0.338,
      "step": 50
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 0.328125,
      "learning_rate": 4.8853211009174314e-05,
      "loss": 0.2119,
      "step": 100
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 0.25390625,
      "learning_rate": 4.827981651376147e-05,
      "loss": 0.2016,
      "step": 150
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 0.22265625,
      "learning_rate": 4.7706422018348626e-05,
      "loss": 0.1936,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.18668922781944275,
      "eval_runtime": 53.2527,
      "eval_samples_per_second": 14.459,
      "eval_steps_per_second": 0.469,
      "step": 218
    },
    {
      "epoch": 1.146788990825688,
      "grad_norm": 0.25,
      "learning_rate": 4.713302752293578e-05,
      "loss": 0.1842,
      "step": 250
    },
    {
      "epoch": 1.3761467889908257,
      "grad_norm": 0.236328125,
      "learning_rate": 4.655963302752294e-05,
      "loss": 0.1835,
      "step": 300
    },
    {
      "epoch": 1.6055045871559632,
      "grad_norm": 0.236328125,
      "learning_rate": 4.5986238532110096e-05,
      "loss": 0.1762,
      "step": 350
    },
    {
      "epoch": 1.834862385321101,
      "grad_norm": 0.236328125,
      "learning_rate": 4.541284403669725e-05,
      "loss": 0.1741,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.1787685602903366,
      "eval_runtime": 52.7872,
      "eval_samples_per_second": 14.587,
      "eval_steps_per_second": 0.474,
      "step": 436
    },
    {
      "epoch": 2.0642201834862384,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.483944954128441e-05,
      "loss": 0.171,
      "step": 450
    },
    {
      "epoch": 2.293577981651376,
      "grad_norm": 0.26171875,
      "learning_rate": 4.426605504587156e-05,
      "loss": 0.1656,
      "step": 500
    },
    {
      "epoch": 2.522935779816514,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.369266055045872e-05,
      "loss": 0.1629,
      "step": 550
    },
    {
      "epoch": 2.7522935779816513,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.311926605504588e-05,
      "loss": 0.1642,
      "step": 600
    },
    {
      "epoch": 2.981651376146789,
      "grad_norm": 0.2451171875,
      "learning_rate": 4.2545871559633024e-05,
      "loss": 0.1604,
      "step": 650
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.17591096460819244,
      "eval_runtime": 52.9863,
      "eval_samples_per_second": 14.532,
      "eval_steps_per_second": 0.472,
      "step": 654
    }
  ],
  "logging_steps": 50,
  "max_steps": 4360,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 4.600868691247104e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}