{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 24651,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 19.664987564086914,
      "learning_rate": 2.939515638310819e-05,
      "loss": 2.1242,
      "step": 500
    },
    {
      "epoch": 0.12,
      "grad_norm": 15.668974876403809,
      "learning_rate": 2.8786661798709992e-05,
      "loss": 1.3535,
      "step": 1000
    },
    {
      "epoch": 0.18,
      "grad_norm": 19.882505416870117,
      "learning_rate": 2.817816721431179e-05,
      "loss": 1.22,
      "step": 1500
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.92320442199707,
      "learning_rate": 2.7569672629913596e-05,
      "loss": 1.1381,
      "step": 2000
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.699568748474121,
      "learning_rate": 2.6961178045515395e-05,
      "loss": 1.0921,
      "step": 2500
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.03032112121582,
      "learning_rate": 2.6352683461117197e-05,
      "loss": 1.0828,
      "step": 3000
    },
    {
      "epoch": 0.43,
      "grad_norm": 13.232436180114746,
      "learning_rate": 2.5744188876719e-05,
      "loss": 1.0373,
      "step": 3500
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.177654266357422,
      "learning_rate": 2.5135694292320798e-05,
      "loss": 0.9764,
      "step": 4000
    },
    {
      "epoch": 0.55,
      "grad_norm": 21.473983764648438,
      "learning_rate": 2.4528416697091396e-05,
      "loss": 0.9691,
      "step": 4500
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.361074447631836,
      "learning_rate": 2.3919922112693198e-05,
      "loss": 0.9319,
      "step": 5000
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.396862983703613,
      "learning_rate": 2.3311427528294997e-05,
      "loss": 0.9323,
      "step": 5500
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.474726676940918,
      "learning_rate": 2.2702932943896803e-05,
      "loss": 0.8969,
      "step": 6000
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.854999542236328,
      "learning_rate": 2.2095655348667397e-05,
      "loss": 0.905,
      "step": 6500
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.904343128204346,
      "learning_rate": 2.14871607642692e-05,
      "loss": 0.9025,
      "step": 7000
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.447906970977783,
      "learning_rate": 2.0878666179871e-05,
      "loss": 0.8784,
      "step": 7500
    },
    {
      "epoch": 0.97,
      "grad_norm": 15.822772979736328,
      "learning_rate": 2.02701715954728e-05,
      "loss": 0.8754,
      "step": 8000
    },
    {
      "epoch": 1.03,
      "grad_norm": 13.860732078552246,
      "learning_rate": 1.9662894000243398e-05,
      "loss": 0.758,
      "step": 8500
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.411458015441895,
      "learning_rate": 1.9054399415845197e-05,
      "loss": 0.6682,
      "step": 9000
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.759841918945312,
      "learning_rate": 1.8445904831447e-05,
      "loss": 0.6745,
      "step": 9500
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.465913772583008,
      "learning_rate": 1.7837410247048804e-05,
      "loss": 0.6599,
      "step": 10000
    },
    {
      "epoch": 1.28,
      "grad_norm": 13.312762260437012,
      "learning_rate": 1.72301326518194e-05,
      "loss": 0.6538,
      "step": 10500
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.506616592407227,
      "learning_rate": 1.66216380674212e-05,
      "loss": 0.6591,
      "step": 11000
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.091550827026367,
      "learning_rate": 1.6013143483023003e-05,
      "loss": 0.6634,
      "step": 11500
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.133756637573242,
      "learning_rate": 1.5404648898624802e-05,
      "loss": 0.6748,
      "step": 12000
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.615654945373535,
      "learning_rate": 1.4796154314226602e-05,
      "loss": 0.6759,
      "step": 12500
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.393186092376709,
      "learning_rate": 1.4188876718997202e-05,
      "loss": 0.6475,
      "step": 13000
    },
    {
      "epoch": 1.64,
      "grad_norm": 14.697447776794434,
      "learning_rate": 1.3580382134599002e-05,
      "loss": 0.6493,
      "step": 13500
    },
    {
      "epoch": 1.7,
      "grad_norm": 15.398512840270996,
      "learning_rate": 1.2971887550200803e-05,
      "loss": 0.6444,
      "step": 14000
    },
    {
      "epoch": 1.76,
      "grad_norm": 15.842305183410645,
      "learning_rate": 1.2363392965802603e-05,
      "loss": 0.6396,
      "step": 14500
    },
    {
      "epoch": 1.83,
      "grad_norm": 12.496230125427246,
      "learning_rate": 1.1756115370573203e-05,
      "loss": 0.628,
      "step": 15000
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.957124710083008,
      "learning_rate": 1.1147620786175003e-05,
      "loss": 0.643,
      "step": 15500
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.984686851501465,
      "learning_rate": 1.0539126201776804e-05,
      "loss": 0.612,
      "step": 16000
    },
    {
      "epoch": 2.01,
      "grad_norm": 12.69090747833252,
      "learning_rate": 9.930631617378604e-06,
      "loss": 0.6215,
      "step": 16500
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.916914939880371,
      "learning_rate": 9.323354022149204e-06,
      "loss": 0.4676,
      "step": 17000
    },
    {
      "epoch": 2.13,
      "grad_norm": 13.55048942565918,
      "learning_rate": 8.7160764269198e-06,
      "loss": 0.4779,
      "step": 17500
    },
    {
      "epoch": 2.19,
      "grad_norm": 27.892988204956055,
      "learning_rate": 8.107581842521602e-06,
      "loss": 0.4771,
      "step": 18000
    },
    {
      "epoch": 2.25,
      "grad_norm": 13.891279220581055,
      "learning_rate": 7.499087258123403e-06,
      "loss": 0.4612,
      "step": 18500
    },
    {
      "epoch": 2.31,
      "grad_norm": 19.246742248535156,
      "learning_rate": 6.890592673725204e-06,
      "loss": 0.4852,
      "step": 19000
    },
    {
      "epoch": 2.37,
      "grad_norm": 26.485008239746094,
      "learning_rate": 6.282098089327005e-06,
      "loss": 0.4841,
      "step": 19500
    },
    {
      "epoch": 2.43,
      "grad_norm": 12.629850387573242,
      "learning_rate": 5.673603504928806e-06,
      "loss": 0.4732,
      "step": 20000
    },
    {
      "epoch": 2.49,
      "grad_norm": 21.771028518676758,
      "learning_rate": 5.065108920530608e-06,
      "loss": 0.4841,
      "step": 20500
    },
    {
      "epoch": 2.56,
      "grad_norm": 8.484288215637207,
      "learning_rate": 4.457831325301205e-06,
      "loss": 0.4703,
      "step": 21000
    },
    {
      "epoch": 2.62,
      "grad_norm": 12.666841506958008,
      "learning_rate": 3.849336740903006e-06,
      "loss": 0.4647,
      "step": 21500
    },
    {
      "epoch": 2.68,
      "grad_norm": 6.814643383026123,
      "learning_rate": 3.2408421565048075e-06,
      "loss": 0.4589,
      "step": 22000
    },
    {
      "epoch": 2.74,
      "grad_norm": 10.161438941955566,
      "learning_rate": 2.6323475721066084e-06,
      "loss": 0.4435,
      "step": 22500
    },
    {
      "epoch": 2.8,
      "grad_norm": 17.461462020874023,
      "learning_rate": 2.0250699768772058e-06,
      "loss": 0.4658,
      "step": 23000
    },
    {
      "epoch": 2.86,
      "grad_norm": 6.685905933380127,
      "learning_rate": 1.416575392479007e-06,
      "loss": 0.4635,
      "step": 23500
    },
    {
      "epoch": 2.92,
      "grad_norm": 20.97271156311035,
      "learning_rate": 8.080808080808081e-07,
      "loss": 0.4541,
      "step": 24000
    },
    {
      "epoch": 2.98,
      "grad_norm": 8.495226860046387,
      "learning_rate": 1.9958622368260923e-07,
      "loss": 0.4637,
      "step": 24500
    },
    {
      "epoch": 3.0,
      "step": 24651,
      "total_flos": 3.864867977455206e+16,
      "train_loss": 0.7326071534426255,
      "train_runtime": 12303.0667,
      "train_samples_per_second": 32.057,
      "train_steps_per_second": 2.004
    }
  ],
  "logging_steps": 500,
  "max_steps": 24651,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5000,
  "total_flos": 3.864867977455206e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}