{
  "best_metric": 0.08930657804012299,
  "best_model_checkpoint": "outputs/checkpoint-603",
  "epoch": 11.990888382687928,
  "eval_steps": 500,
  "global_step": 658,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36446469248291574,
      "grad_norm": 2.546876907348633,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.373,
      "step": 20
    },
    {
      "epoch": 0.7289293849658315,
      "grad_norm": 1.2323638200759888,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.2883,
      "step": 40
    },
    {
      "epoch": 0.9840546697038725,
      "eval_loss": 0.1884891539812088,
      "eval_runtime": 48.9142,
      "eval_samples_per_second": 3.435,
      "eval_steps_per_second": 0.429,
      "step": 54
    },
    {
      "epoch": 1.0933940774487472,
      "grad_norm": 0.5503110289573669,
      "learning_rate": 5.8e-06,
      "loss": 0.1829,
      "step": 60
    },
    {
      "epoch": 1.4578587699316627,
      "grad_norm": 0.3597412109375,
      "learning_rate": 7.800000000000002e-06,
      "loss": 0.1368,
      "step": 80
    },
    {
      "epoch": 1.8223234624145785,
      "grad_norm": 1.4852937459945679,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.1263,
      "step": 100
    },
    {
      "epoch": 1.9863325740318907,
      "eval_loss": 0.13872365653514862,
      "eval_runtime": 48.9202,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 109
    },
    {
      "epoch": 2.1867881548974943,
      "grad_norm": 0.40878018736839294,
      "learning_rate": 9.984149663879994e-06,
      "loss": 0.1061,
      "step": 120
    },
    {
      "epoch": 2.55125284738041,
      "grad_norm": 0.5748926997184753,
      "learning_rate": 9.929487384240103e-06,
      "loss": 0.0895,
      "step": 140
    },
    {
      "epoch": 2.9157175398633255,
      "grad_norm": 0.3740793764591217,
      "learning_rate": 9.83624518217252e-06,
      "loss": 0.0879,
      "step": 160
    },
    {
      "epoch": 2.988610478359909,
      "eval_loss": 0.11915451288223267,
      "eval_runtime": 48.9178,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 164
    },
    {
      "epoch": 3.2801822323462413,
      "grad_norm": 0.5470688343048096,
      "learning_rate": 9.705152804330872e-06,
      "loss": 0.0851,
      "step": 180
    },
    {
      "epoch": 3.644646924829157,
      "grad_norm": 0.5719650387763977,
      "learning_rate": 9.53723622631339e-06,
      "loss": 0.0773,
      "step": 200
    },
    {
      "epoch": 3.990888382687927,
      "eval_loss": 0.1070408970117569,
      "eval_runtime": 48.9193,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 219
    },
    {
      "epoch": 4.009111617312073,
      "grad_norm": 0.46819359064102173,
      "learning_rate": 9.333809623012763e-06,
      "loss": 0.0745,
      "step": 220
    },
    {
      "epoch": 4.373576309794989,
      "grad_norm": 0.5022829174995422,
      "learning_rate": 9.096465083415809e-06,
      "loss": 0.0693,
      "step": 240
    },
    {
      "epoch": 4.738041002277904,
      "grad_norm": 0.5556521415710449,
      "learning_rate": 8.82706015034849e-06,
      "loss": 0.0637,
      "step": 260
    },
    {
      "epoch": 4.993166287015946,
      "eval_loss": 0.09954136610031128,
      "eval_runtime": 48.9147,
      "eval_samples_per_second": 3.435,
      "eval_steps_per_second": 0.429,
      "step": 274
    },
    {
      "epoch": 5.10250569476082,
      "grad_norm": 0.46870681643486023,
      "learning_rate": 8.527703282684766e-06,
      "loss": 0.0634,
      "step": 280
    },
    {
      "epoch": 5.466970387243736,
      "grad_norm": 0.615917980670929,
      "learning_rate": 8.20073735379715e-06,
      "loss": 0.0592,
      "step": 300
    },
    {
      "epoch": 5.831435079726651,
      "grad_norm": 0.7252725958824158,
      "learning_rate": 7.848721315395975e-06,
      "loss": 0.0591,
      "step": 320
    },
    {
      "epoch": 5.995444191343964,
      "eval_loss": 0.09647562354803085,
      "eval_runtime": 48.8589,
      "eval_samples_per_second": 3.438,
      "eval_steps_per_second": 0.43,
      "step": 329
    },
    {
      "epoch": 6.195899772209567,
      "grad_norm": 0.5423758625984192,
      "learning_rate": 7.474410170262719e-06,
      "loss": 0.0601,
      "step": 340
    },
    {
      "epoch": 6.560364464692483,
      "grad_norm": 0.6742358803749084,
      "learning_rate": 7.080733410617853e-06,
      "loss": 0.0533,
      "step": 360
    },
    {
      "epoch": 6.924829157175399,
      "grad_norm": 0.6306804418563843,
      "learning_rate": 6.6707720908722795e-06,
      "loss": 0.0478,
      "step": 380
    },
    {
      "epoch": 6.997722095671982,
      "eval_loss": 0.09345094114542007,
      "eval_runtime": 48.9086,
      "eval_samples_per_second": 3.435,
      "eval_steps_per_second": 0.429,
      "step": 384
    },
    {
      "epoch": 7.289293849658314,
      "grad_norm": 0.6958051919937134,
      "learning_rate": 6.247734714199105e-06,
      "loss": 0.0447,
      "step": 400
    },
    {
      "epoch": 7.65375854214123,
      "grad_norm": 0.6324182152748108,
      "learning_rate": 5.814932121646074e-06,
      "loss": 0.0426,
      "step": 420
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.0920318216085434,
      "eval_runtime": 48.9226,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 439
    },
    {
      "epoch": 8.018223234624147,
      "grad_norm": 0.9903696179389954,
      "learning_rate": 5.375751580315355e-06,
      "loss": 0.0489,
      "step": 440
    },
    {
      "epoch": 8.382687927107062,
      "grad_norm": 0.7768954634666443,
      "learning_rate": 4.933630273405847e-06,
      "loss": 0.0351,
      "step": 460
    },
    {
      "epoch": 8.747152619589977,
      "grad_norm": 0.5696555376052856,
      "learning_rate": 4.492028399594339e-06,
      "loss": 0.0398,
      "step": 480
    },
    {
      "epoch": 8.984054669703873,
      "eval_loss": 0.09069496393203735,
      "eval_runtime": 48.9187,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 493
    },
    {
      "epoch": 9.111617312072893,
      "grad_norm": 0.5146432518959045,
      "learning_rate": 4.054402092289354e-06,
      "loss": 0.0433,
      "step": 500
    },
    {
      "epoch": 9.476082004555808,
      "grad_norm": 0.7227029204368591,
      "learning_rate": 3.6241763707012357e-06,
      "loss": 0.034,
      "step": 520
    },
    {
      "epoch": 9.840546697038725,
      "grad_norm": 0.9780692458152771,
      "learning_rate": 3.2047183344229983e-06,
      "loss": 0.034,
      "step": 540
    },
    {
      "epoch": 9.986332574031891,
      "eval_loss": 0.08991160988807678,
      "eval_runtime": 49.2497,
      "eval_samples_per_second": 3.411,
      "eval_steps_per_second": 0.426,
      "step": 548
    },
    {
      "epoch": 10.20501138952164,
      "grad_norm": 0.6056821346282959,
      "learning_rate": 2.7993108113107446e-06,
      "loss": 0.0341,
      "step": 560
    },
    {
      "epoch": 10.569476082004556,
      "grad_norm": 0.5894588232040405,
      "learning_rate": 2.411126664904665e-06,
      "loss": 0.0274,
      "step": 580
    },
    {
      "epoch": 10.933940774487471,
      "grad_norm": 0.8617128729820251,
      "learning_rate": 2.0432039624699016e-06,
      "loss": 0.0317,
      "step": 600
    },
    {
      "epoch": 10.98861047835991,
      "eval_loss": 0.08930657804012299,
      "eval_runtime": 49.2166,
      "eval_samples_per_second": 3.413,
      "eval_steps_per_second": 0.427,
      "step": 603
    },
    {
      "epoch": 11.298405466970387,
      "grad_norm": 0.6252109408378601,
      "learning_rate": 1.6984221980010456e-06,
      "loss": 0.0246,
      "step": 620
    },
    {
      "epoch": 11.662870159453304,
      "grad_norm": 0.6094932556152344,
      "learning_rate": 1.3794797562774598e-06,
      "loss": 0.0283,
      "step": 640
    },
    {
      "epoch": 11.990888382687928,
      "eval_loss": 0.0896030068397522,
      "eval_runtime": 49.2306,
      "eval_samples_per_second": 3.413,
      "eval_steps_per_second": 0.427,
      "step": 658
    }
  ],
  "logging_steps": 20,
  "max_steps": 810,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 2.45973113078784e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}