{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.875,
  "eval_steps": 2500,
  "global_step": 35000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "grad_norm": 2.9973244667053223,
      "learning_rate": 1.971072463768116e-05,
      "loss": 1.6969,
      "step": 1000
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.261388778686523,
      "learning_rate": 1.9131014492753626e-05,
      "loss": 1.5602,
      "step": 2000
    },
    {
      "epoch": 0.06,
      "eval_loss": 1.513683795928955,
      "eval_runtime": 202.2713,
      "eval_samples_per_second": 4.944,
      "eval_steps_per_second": 4.944,
      "step": 2500
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.314629316329956,
      "learning_rate": 1.8551304347826088e-05,
      "loss": 1.5222,
      "step": 3000
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.6707098484039307,
      "learning_rate": 1.7971594202898553e-05,
      "loss": 1.4933,
      "step": 4000
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.160583257675171,
      "learning_rate": 1.7392463768115943e-05,
      "loss": 1.5056,
      "step": 5000
    },
    {
      "epoch": 0.12,
      "eval_loss": 1.4657692909240723,
      "eval_runtime": 201.9515,
      "eval_samples_per_second": 4.952,
      "eval_steps_per_second": 4.952,
      "step": 5000
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.666102170944214,
      "learning_rate": 1.6812753623188408e-05,
      "loss": 1.4754,
      "step": 6000
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.95894455909729,
      "learning_rate": 1.623304347826087e-05,
      "loss": 1.4816,
      "step": 7000
    },
    {
      "epoch": 0.19,
      "eval_loss": 1.4514989852905273,
      "eval_runtime": 201.9852,
      "eval_samples_per_second": 4.951,
      "eval_steps_per_second": 4.951,
      "step": 7500
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.616883277893066,
      "learning_rate": 1.5653333333333335e-05,
      "loss": 1.4414,
      "step": 8000
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.114346027374268,
      "learning_rate": 1.5074202898550725e-05,
      "loss": 1.4392,
      "step": 9000
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.38364028930664,
      "learning_rate": 1.4495072463768116e-05,
      "loss": 1.4307,
      "step": 10000
    },
    {
      "epoch": 0.25,
      "eval_loss": 1.4246548414230347,
      "eval_runtime": 201.9169,
      "eval_samples_per_second": 4.953,
      "eval_steps_per_second": 4.953,
      "step": 10000
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.899681806564331,
      "learning_rate": 1.3915362318840582e-05,
      "loss": 1.4313,
      "step": 11000
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.839545965194702,
      "learning_rate": 1.3336231884057973e-05,
      "loss": 1.415,
      "step": 12000
    },
    {
      "epoch": 0.31,
      "eval_loss": 1.4323452711105347,
      "eval_runtime": 202.4132,
      "eval_samples_per_second": 4.94,
      "eval_steps_per_second": 4.94,
      "step": 12500
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.805086612701416,
      "learning_rate": 1.2756521739130436e-05,
      "loss": 1.4003,
      "step": 13000
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.0652241706848145,
      "learning_rate": 1.21768115942029e-05,
      "loss": 1.3952,
      "step": 14000
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.923527717590332,
      "learning_rate": 1.1597681159420292e-05,
      "loss": 1.3933,
      "step": 15000
    },
    {
      "epoch": 0.38,
      "eval_loss": 1.3821483850479126,
      "eval_runtime": 201.8609,
      "eval_samples_per_second": 4.954,
      "eval_steps_per_second": 4.954,
      "step": 15000
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.487946033477783,
      "learning_rate": 1.1017971014492755e-05,
      "loss": 1.3825,
      "step": 16000
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.9339723587036133,
      "learning_rate": 1.0438260869565218e-05,
      "loss": 1.373,
      "step": 17000
    },
    {
      "epoch": 0.44,
      "eval_loss": 1.371016263961792,
      "eval_runtime": 201.6518,
      "eval_samples_per_second": 4.959,
      "eval_steps_per_second": 4.959,
      "step": 17500
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.8097546100616455,
      "learning_rate": 9.85913043478261e-06,
      "loss": 1.35,
      "step": 18000
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.639487266540527,
      "learning_rate": 9.279420289855074e-06,
      "loss": 1.3498,
      "step": 19000
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.806727886199951,
      "learning_rate": 8.700289855072464e-06,
      "loss": 1.3372,
      "step": 20000
    },
    {
      "epoch": 0.5,
      "eval_loss": 1.3245751857757568,
      "eval_runtime": 201.5593,
      "eval_samples_per_second": 4.961,
      "eval_steps_per_second": 4.961,
      "step": 20000
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.062647342681885,
      "learning_rate": 8.120579710144927e-06,
      "loss": 1.3637,
      "step": 21000
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.603002071380615,
      "learning_rate": 7.540869565217392e-06,
      "loss": 1.3464,
      "step": 22000
    },
    {
      "epoch": 0.56,
      "eval_loss": 1.3194873332977295,
      "eval_runtime": 201.6454,
      "eval_samples_per_second": 4.959,
      "eval_steps_per_second": 4.959,
      "step": 22500
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.0290374755859375,
      "learning_rate": 6.961739130434784e-06,
      "loss": 1.3187,
      "step": 23000
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.862253189086914,
      "learning_rate": 6.382028985507247e-06,
      "loss": 1.321,
      "step": 24000
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.9182982444763184,
      "learning_rate": 5.80231884057971e-06,
      "loss": 1.3157,
      "step": 25000
    },
    {
      "epoch": 0.62,
      "eval_loss": 1.2969852685928345,
      "eval_runtime": 201.2648,
      "eval_samples_per_second": 4.969,
      "eval_steps_per_second": 4.969,
      "step": 25000
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.6946563720703125,
      "learning_rate": 5.222608695652175e-06,
      "loss": 1.3161,
      "step": 26000
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.966987609863281,
      "learning_rate": 4.643478260869566e-06,
      "loss": 1.3253,
      "step": 27000
    },
    {
      "epoch": 0.69,
      "eval_loss": 1.3072532415390015,
      "eval_runtime": 201.6414,
      "eval_samples_per_second": 4.959,
      "eval_steps_per_second": 4.959,
      "step": 27500
    },
    {
      "epoch": 0.7,
      "grad_norm": 14.545968055725098,
      "learning_rate": 4.06376811594203e-06,
      "loss": 1.3143,
      "step": 28000
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.552095413208008,
      "learning_rate": 3.4840579710144927e-06,
      "loss": 1.3042,
      "step": 29000
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.7875924110412598,
      "learning_rate": 2.905507246376812e-06,
      "loss": 1.2942,
      "step": 30000
    },
    {
      "epoch": 0.75,
      "eval_loss": 1.272377848625183,
      "eval_runtime": 201.643,
      "eval_samples_per_second": 4.959,
      "eval_steps_per_second": 4.959,
      "step": 30000
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.4199697971343994,
      "learning_rate": 2.3257971014492754e-06,
      "loss": 1.2925,
      "step": 31000
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.922388076782227,
      "learning_rate": 1.7460869565217393e-06,
      "loss": 1.2885,
      "step": 32000
    },
    {
      "epoch": 0.81,
      "eval_loss": 1.3001521825790405,
      "eval_runtime": 201.9216,
      "eval_samples_per_second": 4.952,
      "eval_steps_per_second": 4.952,
      "step": 32500
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.57639217376709,
      "learning_rate": 1.1669565217391306e-06,
      "loss": 1.2787,
      "step": 33000
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.37148380279541,
      "learning_rate": 5.872463768115943e-07,
      "loss": 1.2788,
      "step": 34000
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.262892723083496,
      "learning_rate": 7.536231884057971e-09,
      "loss": 1.2844,
      "step": 35000
    },
    {
      "epoch": 0.88,
      "eval_loss": 1.2726885080337524,
      "eval_runtime": 201.8671,
      "eval_samples_per_second": 4.954,
      "eval_steps_per_second": 4.954,
      "step": 35000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 35000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2500,
  "total_flos": 5.6357440978944e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}