{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 12300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41,
      "grad_norm": 29.442007064819336,
      "learning_rate": 4.796747967479675e-05,
      "loss": 1.1714,
      "step": 500
    },
    {
      "epoch": 0.81,
      "grad_norm": 17.30375099182129,
      "learning_rate": 4.59349593495935e-05,
      "loss": 0.5578,
      "step": 1000
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.899393081665039,
      "learning_rate": 4.390243902439025e-05,
      "loss": 0.4015,
      "step": 1500
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.312630653381348,
      "learning_rate": 4.186991869918699e-05,
      "loss": 0.3135,
      "step": 2000
    },
    {
      "epoch": 2.03,
      "grad_norm": 7.51059627532959,
      "learning_rate": 3.983739837398374e-05,
      "loss": 0.2809,
      "step": 2500
    },
    {
      "epoch": 2.44,
      "grad_norm": 5.824585914611816,
      "learning_rate": 3.780487804878049e-05,
      "loss": 0.2045,
      "step": 3000
    },
    {
      "epoch": 2.85,
      "grad_norm": 20.64141845703125,
      "learning_rate": 3.577235772357724e-05,
      "loss": 0.1899,
      "step": 3500
    },
    {
      "epoch": 3.25,
      "grad_norm": 4.098996162414551,
      "learning_rate": 3.373983739837399e-05,
      "loss": 0.156,
      "step": 4000
    },
    {
      "epoch": 3.66,
      "grad_norm": 11.682610511779785,
      "learning_rate": 3.170731707317073e-05,
      "loss": 0.1367,
      "step": 4500
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.980808138847351,
      "learning_rate": 2.9674796747967482e-05,
      "loss": 0.1321,
      "step": 5000
    },
    {
      "epoch": 4.47,
      "grad_norm": 21.24442481994629,
      "learning_rate": 2.764227642276423e-05,
      "loss": 0.1024,
      "step": 5500
    },
    {
      "epoch": 4.88,
      "grad_norm": 10.403449058532715,
      "learning_rate": 2.5609756097560977e-05,
      "loss": 0.0965,
      "step": 6000
    },
    {
      "epoch": 5.28,
      "grad_norm": 6.274624824523926,
      "learning_rate": 2.3577235772357724e-05,
      "loss": 0.0811,
      "step": 6500
    },
    {
      "epoch": 5.69,
      "grad_norm": 18.304744720458984,
      "learning_rate": 2.1544715447154475e-05,
      "loss": 0.0787,
      "step": 7000
    },
    {
      "epoch": 6.1,
      "grad_norm": 6.117405891418457,
      "learning_rate": 1.9512195121951222e-05,
      "loss": 0.0636,
      "step": 7500
    },
    {
      "epoch": 6.5,
      "grad_norm": 6.909726619720459,
      "learning_rate": 1.747967479674797e-05,
      "loss": 0.0578,
      "step": 8000
    },
    {
      "epoch": 6.91,
      "grad_norm": 7.566128253936768,
      "learning_rate": 1.5447154471544717e-05,
      "loss": 0.0564,
      "step": 8500
    },
    {
      "epoch": 7.32,
      "grad_norm": 3.8233773708343506,
      "learning_rate": 1.3414634146341466e-05,
      "loss": 0.0474,
      "step": 9000
    },
    {
      "epoch": 7.72,
      "grad_norm": 9.988646507263184,
      "learning_rate": 1.1382113821138211e-05,
      "loss": 0.0394,
      "step": 9500
    },
    {
      "epoch": 8.13,
      "grad_norm": 7.508295059204102,
      "learning_rate": 9.34959349593496e-06,
      "loss": 0.0339,
      "step": 10000
    },
    {
      "epoch": 8.54,
      "grad_norm": 8.255553245544434,
      "learning_rate": 7.317073170731707e-06,
      "loss": 0.0279,
      "step": 10500
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.14910294115543365,
      "learning_rate": 5.2845528455284555e-06,
      "loss": 0.0248,
      "step": 11000
    },
    {
      "epoch": 9.35,
      "grad_norm": 0.014114579185843468,
      "learning_rate": 3.2520325203252037e-06,
      "loss": 0.018,
      "step": 11500
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.12875856459140778,
      "learning_rate": 1.2195121951219514e-06,
      "loss": 0.0182,
      "step": 12000
    },
    {
      "epoch": 10.0,
      "step": 12300,
      "total_flos": 7.078965077743022e+16,
      "train_loss": 0.17480758806554283,
      "train_runtime": 299478.6207,
      "train_samples_per_second": 1.314,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 500,
  "max_steps": 12300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 7.078965077743022e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}