{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 11.971223021582734,
  "eval_steps": 25,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.64,
      "grad_norm": 0.9966348377433891,
      "learning_rate": 9.999874838141888e-05,
      "loss": 0.1831,
      "step": 25
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.12574630975723267,
      "eval_runtime": 110.3252,
      "eval_samples_per_second": 90.65,
      "eval_steps_per_second": 1.423,
      "step": 25
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.13412292959827216,
      "learning_rate": 9.915628588978522e-05,
      "loss": 0.1239,
      "step": 50
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.10444429516792297,
      "eval_runtime": 110.3167,
      "eval_samples_per_second": 90.657,
      "eval_steps_per_second": 1.423,
      "step": 50
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.1395974887205158,
      "learning_rate": 9.67797005288181e-05,
      "loss": 0.108,
      "step": 75
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.0995471403002739,
      "eval_runtime": 110.0262,
      "eval_samples_per_second": 90.897,
      "eval_steps_per_second": 1.427,
      "step": 75
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.1280464988861011,
      "learning_rate": 9.294316336102132e-05,
      "loss": 0.0976,
      "step": 100
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.09775934368371964,
      "eval_runtime": 109.8969,
      "eval_samples_per_second": 91.003,
      "eval_steps_per_second": 1.429,
      "step": 100
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.09561572064129414,
      "learning_rate": 8.776640921382584e-05,
      "loss": 0.094,
      "step": 125
    },
    {
      "epoch": 3.2,
      "eval_loss": 0.08864283561706543,
      "eval_runtime": 110.0582,
      "eval_samples_per_second": 90.87,
      "eval_steps_per_second": 1.427,
      "step": 125
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.11127416265410695,
      "learning_rate": 8.141099986478212e-05,
      "loss": 0.0828,
      "step": 150
    },
    {
      "epoch": 3.84,
      "eval_loss": 0.08932521939277649,
      "eval_runtime": 109.3425,
      "eval_samples_per_second": 91.465,
      "eval_steps_per_second": 1.436,
      "step": 150
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.22132698379917157,
      "learning_rate": 7.407528184577019e-05,
      "loss": 0.078,
      "step": 175
    },
    {
      "epoch": 4.48,
      "eval_loss": 0.09065766632556915,
      "eval_runtime": 110.1178,
      "eval_samples_per_second": 90.821,
      "eval_steps_per_second": 1.426,
      "step": 175
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.08955720132434862,
      "learning_rate": 6.598819622856227e-05,
      "loss": 0.0767,
      "step": 200
    },
    {
      "epoch": 5.12,
      "eval_loss": 0.08660481870174408,
      "eval_runtime": 109.7782,
      "eval_samples_per_second": 91.102,
      "eval_steps_per_second": 1.43,
      "step": 200
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.09375419122940627,
      "learning_rate": 5.7402133582686576e-05,
      "loss": 0.0697,
      "step": 225
    },
    {
      "epoch": 5.76,
      "eval_loss": 0.08401340246200562,
      "eval_runtime": 109.3041,
      "eval_samples_per_second": 91.497,
      "eval_steps_per_second": 1.436,
      "step": 225
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.0990995516325273,
      "learning_rate": 4.85850570958441e-05,
      "loss": 0.0646,
      "step": 250
    },
    {
      "epoch": 6.39,
      "eval_loss": 0.08192210644483566,
      "eval_runtime": 109.5799,
      "eval_samples_per_second": 91.267,
      "eval_steps_per_second": 1.433,
      "step": 250
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.07754859448629634,
      "learning_rate": 3.9812139687108815e-05,
      "loss": 0.0594,
      "step": 275
    },
    {
      "epoch": 7.03,
      "eval_loss": 0.07945634424686432,
      "eval_runtime": 109.4058,
      "eval_samples_per_second": 91.412,
      "eval_steps_per_second": 1.435,
      "step": 275
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.06681798482966637,
      "learning_rate": 3.135717611098458e-05,
      "loss": 0.052,
      "step": 300
    },
    {
      "epoch": 7.67,
      "eval_loss": 0.07952920347452164,
      "eval_runtime": 110.1672,
      "eval_samples_per_second": 90.78,
      "eval_steps_per_second": 1.425,
      "step": 300
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.10690045013292092,
      "learning_rate": 2.3484038072721758e-05,
      "loss": 0.0478,
      "step": 325
    },
    {
      "epoch": 8.31,
      "eval_loss": 0.08034859597682953,
      "eval_runtime": 109.4321,
      "eval_samples_per_second": 91.39,
      "eval_steps_per_second": 1.435,
      "step": 325
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.08719559737170716,
      "learning_rate": 1.6438439032954855e-05,
      "loss": 0.0447,
      "step": 350
    },
    {
      "epoch": 8.95,
      "eval_loss": 0.07858795672655106,
      "eval_runtime": 110.1518,
      "eval_samples_per_second": 90.793,
      "eval_steps_per_second": 1.425,
      "step": 350
    },
    {
      "epoch": 9.59,
      "grad_norm": 0.04932389885679807,
      "learning_rate": 1.0440265714600572e-05,
      "loss": 0.0392,
      "step": 375
    },
    {
      "epoch": 9.59,
      "eval_loss": 0.07998502999544144,
      "eval_runtime": 109.5251,
      "eval_samples_per_second": 91.312,
      "eval_steps_per_second": 1.433,
      "step": 375
    },
    {
      "epoch": 10.23,
      "grad_norm": 0.03799019571227359,
      "learning_rate": 5.676715638695063e-06,
      "loss": 0.038,
      "step": 400
    },
    {
      "epoch": 10.23,
      "eval_loss": 0.08129393309354782,
      "eval_runtime": 109.5381,
      "eval_samples_per_second": 91.302,
      "eval_steps_per_second": 1.433,
      "step": 400
    },
    {
      "epoch": 10.87,
      "grad_norm": 0.034874323591440555,
      "learning_rate": 2.2964548604209213e-06,
      "loss": 0.0357,
      "step": 425
    },
    {
      "epoch": 10.87,
      "eval_loss": 0.08102953433990479,
      "eval_runtime": 110.2721,
      "eval_samples_per_second": 90.694,
      "eval_steps_per_second": 1.424,
      "step": 425
    },
    {
      "epoch": 11.51,
      "grad_norm": 0.038884021048955025,
      "learning_rate": 4.049782370561583e-07,
      "loss": 0.035,
      "step": 450
    },
    {
      "epoch": 11.51,
      "eval_loss": 0.08157423138618469,
      "eval_runtime": 109.9602,
      "eval_samples_per_second": 90.951,
      "eval_steps_per_second": 1.428,
      "step": 450
    },
    {
      "epoch": 11.97,
      "step": 468,
      "total_flos": 1.0672624631808e+16,
      "train_loss": 0.07237311357106918,
      "train_runtime": 39649.873,
      "train_samples_per_second": 24.213,
      "train_steps_per_second": 0.012
    }
  ],
  "logging_steps": 25,
  "max_steps": 468,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 2000,
  "total_flos": 1.0672624631808e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}