{
  "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": 1.127359124604765,
      "learning_rate": 9.999874838141888e-05,
      "loss": 0.2903,
      "step": 25
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.17698872089385986,
      "eval_runtime": 109.7117,
      "eval_samples_per_second": 91.157,
      "eval_steps_per_second": 1.431,
      "step": 25
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.27062951121760287,
      "learning_rate": 9.915628588978522e-05,
      "loss": 0.1566,
      "step": 50
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.1319342851638794,
      "eval_runtime": 109.0269,
      "eval_samples_per_second": 91.73,
      "eval_steps_per_second": 1.44,
      "step": 50
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.24204285961516211,
      "learning_rate": 9.67797005288181e-05,
      "loss": 0.1379,
      "step": 75
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.12531398236751556,
      "eval_runtime": 108.9721,
      "eval_samples_per_second": 91.776,
      "eval_steps_per_second": 1.441,
      "step": 75
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.17190878979127733,
      "learning_rate": 9.294316336102132e-05,
      "loss": 0.1246,
      "step": 100
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.11651688814163208,
      "eval_runtime": 109.0854,
      "eval_samples_per_second": 91.68,
      "eval_steps_per_second": 1.439,
      "step": 100
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.19157295828856058,
      "learning_rate": 8.776640921382584e-05,
      "loss": 0.1159,
      "step": 125
    },
    {
      "epoch": 3.2,
      "eval_loss": 0.10494253039360046,
      "eval_runtime": 109.434,
      "eval_samples_per_second": 91.388,
      "eval_steps_per_second": 1.435,
      "step": 125
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.13204284591576954,
      "learning_rate": 8.141099986478212e-05,
      "loss": 0.1048,
      "step": 150
    },
    {
      "epoch": 3.84,
      "eval_loss": 0.09983079880475998,
      "eval_runtime": 109.7612,
      "eval_samples_per_second": 91.116,
      "eval_steps_per_second": 1.43,
      "step": 150
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.17932878424857554,
      "learning_rate": 7.407528184577019e-05,
      "loss": 0.0947,
      "step": 175
    },
    {
      "epoch": 4.48,
      "eval_loss": 0.09489566832780838,
      "eval_runtime": 108.9729,
      "eval_samples_per_second": 91.775,
      "eval_steps_per_second": 1.441,
      "step": 175
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.10314936548781714,
      "learning_rate": 6.598819622856227e-05,
      "loss": 0.0872,
      "step": 200
    },
    {
      "epoch": 5.12,
      "eval_loss": 0.09059683978557587,
      "eval_runtime": 109.4369,
      "eval_samples_per_second": 91.386,
      "eval_steps_per_second": 1.435,
      "step": 200
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.09896949138120692,
      "learning_rate": 5.7402133582686576e-05,
      "loss": 0.0836,
      "step": 225
    },
    {
      "epoch": 5.76,
      "eval_loss": 0.08898729830980301,
      "eval_runtime": 109.6712,
      "eval_samples_per_second": 91.191,
      "eval_steps_per_second": 1.432,
      "step": 225
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.09411679948573648,
      "learning_rate": 4.85850570958441e-05,
      "loss": 0.0774,
      "step": 250
    },
    {
      "epoch": 6.39,
      "eval_loss": 0.08496005833148956,
      "eval_runtime": 108.989,
      "eval_samples_per_second": 91.762,
      "eval_steps_per_second": 1.441,
      "step": 250
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.0741392458596649,
      "learning_rate": 3.9812139687108815e-05,
      "loss": 0.0717,
      "step": 275
    },
    {
      "epoch": 7.03,
      "eval_loss": 0.08271028101444244,
      "eval_runtime": 108.611,
      "eval_samples_per_second": 92.081,
      "eval_steps_per_second": 1.446,
      "step": 275
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.08461241295692067,
      "learning_rate": 3.135717611098458e-05,
      "loss": 0.0639,
      "step": 300
    },
    {
      "epoch": 7.67,
      "eval_loss": 0.08073248714208603,
      "eval_runtime": 109.2638,
      "eval_samples_per_second": 91.531,
      "eval_steps_per_second": 1.437,
      "step": 300
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.10663906934510009,
      "learning_rate": 2.3484038072721758e-05,
      "loss": 0.0596,
      "step": 325
    },
    {
      "epoch": 8.31,
      "eval_loss": 0.07886555045843124,
      "eval_runtime": 109.6916,
      "eval_samples_per_second": 91.174,
      "eval_steps_per_second": 1.431,
      "step": 325
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.09636182626655444,
      "learning_rate": 1.6438439032954855e-05,
      "loss": 0.0555,
      "step": 350
    },
    {
      "epoch": 8.95,
      "eval_loss": 0.0773012638092041,
      "eval_runtime": 109.6624,
      "eval_samples_per_second": 91.198,
      "eval_steps_per_second": 1.432,
      "step": 350
    },
    {
      "epoch": 9.59,
      "grad_norm": 0.05049345737531785,
      "learning_rate": 1.0440265714600572e-05,
      "loss": 0.0498,
      "step": 375
    },
    {
      "epoch": 9.59,
      "eval_loss": 0.07774835079908371,
      "eval_runtime": 109.5603,
      "eval_samples_per_second": 91.283,
      "eval_steps_per_second": 1.433,
      "step": 375
    },
    {
      "epoch": 10.23,
      "grad_norm": 0.0430056992457794,
      "learning_rate": 5.676715638695063e-06,
      "loss": 0.0491,
      "step": 400
    },
    {
      "epoch": 10.23,
      "eval_loss": 0.07812906801700592,
      "eval_runtime": 109.8294,
      "eval_samples_per_second": 91.059,
      "eval_steps_per_second": 1.429,
      "step": 400
    },
    {
      "epoch": 10.87,
      "grad_norm": 0.040355667582564335,
      "learning_rate": 2.2964548604209213e-06,
      "loss": 0.0467,
      "step": 425
    },
    {
      "epoch": 10.87,
      "eval_loss": 0.07796623557806015,
      "eval_runtime": 109.5541,
      "eval_samples_per_second": 91.288,
      "eval_steps_per_second": 1.433,
      "step": 425
    },
    {
      "epoch": 11.51,
      "grad_norm": 0.041530575942009804,
      "learning_rate": 4.049782370561583e-07,
      "loss": 0.0459,
      "step": 450
    },
    {
      "epoch": 11.51,
      "eval_loss": 0.07810305058956146,
      "eval_runtime": 108.746,
      "eval_samples_per_second": 91.967,
      "eval_steps_per_second": 1.444,
      "step": 450
    },
    {
      "epoch": 11.97,
      "step": 468,
      "total_flos": 1.0672624631808e+16,
      "train_loss": 0.0933543870336989,
      "train_runtime": 39555.7956,
      "train_samples_per_second": 24.271,
      "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
}