{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.993680884676145,
  "eval_steps": 100,
  "global_step": 632,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13,
      "grad_norm": 0.37133172154426575,
      "learning_rate": 4.9876553763060684e-05,
      "loss": 1.3016,
      "step": 20
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3215538263320923,
      "learning_rate": 4.950743417011591e-05,
      "loss": 0.9953,
      "step": 40
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.33872345089912415,
      "learning_rate": 4.889628653514402e-05,
      "loss": 0.925,
      "step": 60
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3175918459892273,
      "learning_rate": 4.804914636820517e-05,
      "loss": 0.8788,
      "step": 80
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3202904164791107,
      "learning_rate": 4.6974379770560846e-05,
      "loss": 0.8668,
      "step": 100
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.8571113348007202,
      "eval_runtime": 868.2347,
      "eval_samples_per_second": 3.192,
      "eval_steps_per_second": 0.4,
      "step": 100
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.33093100786209106,
      "learning_rate": 4.5682600813576435e-05,
      "loss": 0.8488,
      "step": 120
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3297623097896576,
      "learning_rate": 4.41865667173477e-05,
      "loss": 0.8462,
      "step": 140
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.3524036705493927,
      "learning_rate": 4.2501051864235636e-05,
      "loss": 0.827,
      "step": 160
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.3592537045478821,
      "learning_rate": 4.0642701891514e-05,
      "loss": 0.7885,
      "step": 180
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3888987600803375,
      "learning_rate": 3.862986930406669e-05,
      "loss": 0.7837,
      "step": 200
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.8230095505714417,
      "eval_runtime": 869.2538,
      "eval_samples_per_second": 3.188,
      "eval_steps_per_second": 0.399,
      "step": 200
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.380818247795105,
      "learning_rate": 3.6482432230574446e-05,
      "loss": 0.7938,
      "step": 220
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.3564074635505676,
      "learning_rate": 3.4221598113100195e-05,
      "loss": 0.7819,
      "step": 240
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3780010938644409,
      "learning_rate": 3.186969426877563e-05,
      "loss": 0.7877,
      "step": 260
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.36975908279418945,
      "learning_rate": 2.9449947391938766e-05,
      "loss": 0.7918,
      "step": 280
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.39148128032684326,
      "learning_rate": 2.6986254174292862e-05,
      "loss": 0.7824,
      "step": 300
    },
    {
      "epoch": 1.9,
      "eval_loss": 0.8058096766471863,
      "eval_runtime": 867.7611,
      "eval_samples_per_second": 3.193,
      "eval_steps_per_second": 0.4,
      "step": 300
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.36495909094810486,
      "learning_rate": 2.4502945308373246e-05,
      "loss": 0.7741,
      "step": 320
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.40916556119918823,
      "learning_rate": 2.2024545204952383e-05,
      "loss": 0.7369,
      "step": 340
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.42436033487319946,
      "learning_rate": 1.957552979734205e-05,
      "loss": 0.7365,
      "step": 360
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.43983975052833557,
      "learning_rate": 1.7180084824444325e-05,
      "loss": 0.7463,
      "step": 380
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.44617146253585815,
      "learning_rate": 1.4861866979675154e-05,
      "loss": 0.7401,
      "step": 400
    },
    {
      "epoch": 2.53,
      "eval_loss": 0.8059037923812866,
      "eval_runtime": 873.5556,
      "eval_samples_per_second": 3.172,
      "eval_steps_per_second": 0.397,
      "step": 400
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.4369719624519348,
      "learning_rate": 1.2643770284581929e-05,
      "loss": 0.7332,
      "step": 420
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.4235495328903198,
      "learning_rate": 1.0547699994378787e-05,
      "loss": 0.7364,
      "step": 440
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.4584214389324188,
      "learning_rate": 8.594356268240616e-06,
      "loss": 0.7318,
      "step": 460
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.4329874813556671,
      "learning_rate": 6.803029740762648e-06,
      "loss": 0.7222,
      "step": 480
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.5470691919326782,
      "learning_rate": 5.191411013460645e-06,
      "loss": 0.7101,
      "step": 500
    },
    {
      "epoch": 3.16,
      "eval_loss": 0.807178795337677,
      "eval_runtime": 867.2257,
      "eval_samples_per_second": 3.195,
      "eval_steps_per_second": 0.4,
      "step": 500
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.4557144343852997,
      "learning_rate": 3.775415947715899e-06,
      "loss": 0.7135,
      "step": 520
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.46527382731437683,
      "learning_rate": 2.5690284845196923e-06,
      "loss": 0.7045,
      "step": 540
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.4702458679676056,
      "learning_rate": 1.5841625432818057e-06,
      "loss": 0.7057,
      "step": 560
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.4915095865726471,
      "learning_rate": 8.305443635490711e-07,
      "loss": 0.7044,
      "step": 580
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.5043957233428955,
      "learning_rate": 3.1561645159166597e-07,
      "loss": 0.7037,
      "step": 600
    },
    {
      "epoch": 3.79,
      "eval_loss": 0.8061766028404236,
      "eval_runtime": 867.9324,
      "eval_samples_per_second": 3.193,
      "eval_steps_per_second": 0.4,
      "step": 600
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.4672119915485382,
      "learning_rate": 4.4464080451675494e-08,
      "loss": 0.7027,
      "step": 620
    },
    {
      "epoch": 3.99,
      "step": 632,
      "total_flos": 1.137224723506643e+19,
      "train_loss": 0.7920433780815028,
      "train_runtime": 47185.3474,
      "train_samples_per_second": 0.858,
      "train_steps_per_second": 0.013
    }
  ],
  "logging_steps": 20,
  "max_steps": 632,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 1.137224723506643e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}