{
  "best_global_step": 16500,
  "best_metric": 0.10550642758607864,
  "best_model_checkpoint": "forward_model/checkpoint-16500",
  "epoch": 2.4041964155617075,
  "eval_steps": 500,
  "global_step": 16500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07285443683520326,
      "grad_norm": 0.4506877362728119,
      "learning_rate": 0.00019818228180096167,
      "loss": 0.5081,
      "step": 500
    },
    {
      "epoch": 0.07285443683520326,
      "eval_loss": 0.3113231360912323,
      "eval_runtime": 0.5561,
      "eval_samples_per_second": 179.835,
      "eval_steps_per_second": 23.378,
      "step": 500
    },
    {
      "epoch": 0.14570887367040652,
      "grad_norm": 0.8721633553504944,
      "learning_rate": 0.0001963609208800816,
      "loss": 0.2947,
      "step": 1000
    },
    {
      "epoch": 0.14570887367040652,
      "eval_loss": 0.2740160822868347,
      "eval_runtime": 0.578,
      "eval_samples_per_second": 173.017,
      "eval_steps_per_second": 22.492,
      "step": 1000
    },
    {
      "epoch": 0.2185633105056098,
      "grad_norm": 0.3565625250339508,
      "learning_rate": 0.0001945395599592015,
      "loss": 0.2593,
      "step": 1500
    },
    {
      "epoch": 0.2185633105056098,
      "eval_loss": 0.24169117212295532,
      "eval_runtime": 0.5889,
      "eval_samples_per_second": 169.798,
      "eval_steps_per_second": 22.074,
      "step": 1500
    },
    {
      "epoch": 0.29141774734081305,
      "grad_norm": 0.30462226271629333,
      "learning_rate": 0.00019271819903832145,
      "loss": 0.2435,
      "step": 2000
    },
    {
      "epoch": 0.29141774734081305,
      "eval_loss": 0.2248290479183197,
      "eval_runtime": 0.5621,
      "eval_samples_per_second": 177.893,
      "eval_steps_per_second": 23.126,
      "step": 2000
    },
    {
      "epoch": 0.3642721841760163,
      "grad_norm": 0.5108212232589722,
      "learning_rate": 0.00019089683811744136,
      "loss": 0.2287,
      "step": 2500
    },
    {
      "epoch": 0.3642721841760163,
      "eval_loss": 0.2162795513868332,
      "eval_runtime": 0.5633,
      "eval_samples_per_second": 177.534,
      "eval_steps_per_second": 23.079,
      "step": 2500
    },
    {
      "epoch": 0.4371266210112196,
      "grad_norm": 0.5142472386360168,
      "learning_rate": 0.0001890754771965613,
      "loss": 0.2197,
      "step": 3000
    },
    {
      "epoch": 0.4371266210112196,
      "eval_loss": 0.20293699204921722,
      "eval_runtime": 0.5931,
      "eval_samples_per_second": 168.6,
      "eval_steps_per_second": 21.918,
      "step": 3000
    },
    {
      "epoch": 0.5099810578464229,
      "grad_norm": 0.39115020632743835,
      "learning_rate": 0.0001872541162756812,
      "loss": 0.2112,
      "step": 3500
    },
    {
      "epoch": 0.5099810578464229,
      "eval_loss": 0.1898386925458908,
      "eval_runtime": 0.5913,
      "eval_samples_per_second": 169.122,
      "eval_steps_per_second": 21.986,
      "step": 3500
    },
    {
      "epoch": 0.5828354946816261,
      "grad_norm": 0.5981117486953735,
      "learning_rate": 0.0001854327553548011,
      "loss": 0.2016,
      "step": 4000
    },
    {
      "epoch": 0.5828354946816261,
      "eval_loss": 0.1830490082502365,
      "eval_runtime": 0.583,
      "eval_samples_per_second": 171.516,
      "eval_steps_per_second": 22.297,
      "step": 4000
    },
    {
      "epoch": 0.6556899315168294,
      "grad_norm": 0.3978489339351654,
      "learning_rate": 0.00018361139443392105,
      "loss": 0.1961,
      "step": 4500
    },
    {
      "epoch": 0.6556899315168294,
      "eval_loss": 0.17647209763526917,
      "eval_runtime": 0.5656,
      "eval_samples_per_second": 176.789,
      "eval_steps_per_second": 22.983,
      "step": 4500
    },
    {
      "epoch": 0.7285443683520326,
      "grad_norm": 0.41381800174713135,
      "learning_rate": 0.00018179003351304095,
      "loss": 0.1908,
      "step": 5000
    },
    {
      "epoch": 0.7285443683520326,
      "eval_loss": 0.16872741281986237,
      "eval_runtime": 0.5658,
      "eval_samples_per_second": 176.756,
      "eval_steps_per_second": 22.978,
      "step": 5000
    },
    {
      "epoch": 0.8013988051872359,
      "grad_norm": 0.434097021818161,
      "learning_rate": 0.0001799686725921609,
      "loss": 0.1881,
      "step": 5500
    },
    {
      "epoch": 0.8013988051872359,
      "eval_loss": 0.17042766511440277,
      "eval_runtime": 0.559,
      "eval_samples_per_second": 178.906,
      "eval_steps_per_second": 23.258,
      "step": 5500
    },
    {
      "epoch": 0.8742532420224391,
      "grad_norm": 0.534040093421936,
      "learning_rate": 0.00017814731167128077,
      "loss": 0.1742,
      "step": 6000
    },
    {
      "epoch": 0.8742532420224391,
      "eval_loss": 0.1600825935602188,
      "eval_runtime": 0.5648,
      "eval_samples_per_second": 177.049,
      "eval_steps_per_second": 23.016,
      "step": 6000
    },
    {
      "epoch": 0.9471076788576425,
      "grad_norm": 0.39319896697998047,
      "learning_rate": 0.0001763259507504007,
      "loss": 0.173,
      "step": 6500
    },
    {
      "epoch": 0.9471076788576425,
      "eval_loss": 0.149684339761734,
      "eval_runtime": 0.5762,
      "eval_samples_per_second": 173.56,
      "eval_steps_per_second": 22.563,
      "step": 6500
    },
    {
      "epoch": 1.0199621156928458,
      "grad_norm": 0.31258702278137207,
      "learning_rate": 0.00017450458982952062,
      "loss": 0.1699,
      "step": 7000
    },
    {
      "epoch": 1.0199621156928458,
      "eval_loss": 0.15212562680244446,
      "eval_runtime": 0.5619,
      "eval_samples_per_second": 177.961,
      "eval_steps_per_second": 23.135,
      "step": 7000
    },
    {
      "epoch": 1.0928165525280489,
      "grad_norm": 0.4483049511909485,
      "learning_rate": 0.00017268322890864055,
      "loss": 0.1621,
      "step": 7500
    },
    {
      "epoch": 1.0928165525280489,
      "eval_loss": 0.14983947575092316,
      "eval_runtime": 0.5707,
      "eval_samples_per_second": 175.22,
      "eval_steps_per_second": 22.779,
      "step": 7500
    },
    {
      "epoch": 1.1656709893632522,
      "grad_norm": 0.4742591083049774,
      "learning_rate": 0.00017086186798776046,
      "loss": 0.1614,
      "step": 8000
    },
    {
      "epoch": 1.1656709893632522,
      "eval_loss": 0.14483533799648285,
      "eval_runtime": 0.564,
      "eval_samples_per_second": 177.314,
      "eval_steps_per_second": 23.051,
      "step": 8000
    },
    {
      "epoch": 1.2385254261984555,
      "grad_norm": 0.2674083113670349,
      "learning_rate": 0.00016904050706688037,
      "loss": 0.1566,
      "step": 8500
    },
    {
      "epoch": 1.2385254261984555,
      "eval_loss": 0.1402028650045395,
      "eval_runtime": 0.5685,
      "eval_samples_per_second": 175.899,
      "eval_steps_per_second": 22.867,
      "step": 8500
    },
    {
      "epoch": 1.3113798630336588,
      "grad_norm": 0.46651679277420044,
      "learning_rate": 0.0001672191461460003,
      "loss": 0.1543,
      "step": 9000
    },
    {
      "epoch": 1.3113798630336588,
      "eval_loss": 0.13286742568016052,
      "eval_runtime": 0.5592,
      "eval_samples_per_second": 178.818,
      "eval_steps_per_second": 23.246,
      "step": 9000
    },
    {
      "epoch": 1.384234299868862,
      "grad_norm": 0.47738003730773926,
      "learning_rate": 0.0001653977852251202,
      "loss": 0.1496,
      "step": 9500
    },
    {
      "epoch": 1.384234299868862,
      "eval_loss": 0.13214418292045593,
      "eval_runtime": 0.5666,
      "eval_samples_per_second": 176.48,
      "eval_steps_per_second": 22.942,
      "step": 9500
    },
    {
      "epoch": 1.4570887367040652,
      "grad_norm": 0.27837279438972473,
      "learning_rate": 0.00016357642430424015,
      "loss": 0.1514,
      "step": 10000
    },
    {
      "epoch": 1.4570887367040652,
      "eval_loss": 0.12786731123924255,
      "eval_runtime": 0.5651,
      "eval_samples_per_second": 176.951,
      "eval_steps_per_second": 23.004,
      "step": 10000
    },
    {
      "epoch": 1.5299431735392686,
      "grad_norm": 0.3466807007789612,
      "learning_rate": 0.00016175506338336006,
      "loss": 0.1459,
      "step": 10500
    },
    {
      "epoch": 1.5299431735392686,
      "eval_loss": 0.12736321985721588,
      "eval_runtime": 0.5602,
      "eval_samples_per_second": 178.504,
      "eval_steps_per_second": 23.205,
      "step": 10500
    },
    {
      "epoch": 1.6027976103744717,
      "grad_norm": 0.32600516080856323,
      "learning_rate": 0.00015993370246247996,
      "loss": 0.1446,
      "step": 11000
    },
    {
      "epoch": 1.6027976103744717,
      "eval_loss": 0.12436921149492264,
      "eval_runtime": 0.5633,
      "eval_samples_per_second": 177.529,
      "eval_steps_per_second": 23.079,
      "step": 11000
    },
    {
      "epoch": 1.6756520472096752,
      "grad_norm": 0.26724693179130554,
      "learning_rate": 0.00015811234154159987,
      "loss": 0.1408,
      "step": 11500
    },
    {
      "epoch": 1.6756520472096752,
      "eval_loss": 0.12056704610586166,
      "eval_runtime": 0.5598,
      "eval_samples_per_second": 178.631,
      "eval_steps_per_second": 23.222,
      "step": 11500
    },
    {
      "epoch": 1.7485064840448783,
      "grad_norm": 0.5725470185279846,
      "learning_rate": 0.0001562909806207198,
      "loss": 0.1411,
      "step": 12000
    },
    {
      "epoch": 1.7485064840448783,
      "eval_loss": 0.11899165064096451,
      "eval_runtime": 0.5579,
      "eval_samples_per_second": 179.25,
      "eval_steps_per_second": 23.302,
      "step": 12000
    },
    {
      "epoch": 1.8213609208800816,
      "grad_norm": 0.5043156743049622,
      "learning_rate": 0.00015446961969983972,
      "loss": 0.1405,
      "step": 12500
    },
    {
      "epoch": 1.8213609208800816,
      "eval_loss": 0.11793835461139679,
      "eval_runtime": 0.5784,
      "eval_samples_per_second": 172.898,
      "eval_steps_per_second": 22.477,
      "step": 12500
    },
    {
      "epoch": 1.894215357715285,
      "grad_norm": 0.3471129536628723,
      "learning_rate": 0.00015264825877895965,
      "loss": 0.1362,
      "step": 13000
    },
    {
      "epoch": 1.894215357715285,
      "eval_loss": 0.11688660830259323,
      "eval_runtime": 0.5574,
      "eval_samples_per_second": 179.401,
      "eval_steps_per_second": 23.322,
      "step": 13000
    },
    {
      "epoch": 1.967069794550488,
      "grad_norm": 0.3712925910949707,
      "learning_rate": 0.0001508268978580796,
      "loss": 0.1345,
      "step": 13500
    },
    {
      "epoch": 1.967069794550488,
      "eval_loss": 0.11329545080661774,
      "eval_runtime": 0.5595,
      "eval_samples_per_second": 178.744,
      "eval_steps_per_second": 23.237,
      "step": 13500
    },
    {
      "epoch": 2.0399242313856916,
      "grad_norm": 0.33280235528945923,
      "learning_rate": 0.00014900553693719947,
      "loss": 0.1288,
      "step": 14000
    },
    {
      "epoch": 2.0399242313856916,
      "eval_loss": 0.1130962073802948,
      "eval_runtime": 0.5627,
      "eval_samples_per_second": 177.721,
      "eval_steps_per_second": 23.104,
      "step": 14000
    },
    {
      "epoch": 2.1127786682208947,
      "grad_norm": 0.4070642590522766,
      "learning_rate": 0.0001471841760163194,
      "loss": 0.1301,
      "step": 14500
    },
    {
      "epoch": 2.1127786682208947,
      "eval_loss": 0.11047343909740448,
      "eval_runtime": 0.5653,
      "eval_samples_per_second": 176.895,
      "eval_steps_per_second": 22.996,
      "step": 14500
    },
    {
      "epoch": 2.1856331050560978,
      "grad_norm": 0.31729069352149963,
      "learning_rate": 0.0001453628150954393,
      "loss": 0.1269,
      "step": 15000
    },
    {
      "epoch": 2.1856331050560978,
      "eval_loss": 0.11027682572603226,
      "eval_runtime": 0.5874,
      "eval_samples_per_second": 170.237,
      "eval_steps_per_second": 22.131,
      "step": 15000
    },
    {
      "epoch": 2.2584875418913013,
      "grad_norm": 0.265652596950531,
      "learning_rate": 0.00014354145417455925,
      "loss": 0.1247,
      "step": 15500
    },
    {
      "epoch": 2.2584875418913013,
      "eval_loss": 0.10951703786849976,
      "eval_runtime": 0.5678,
      "eval_samples_per_second": 176.131,
      "eval_steps_per_second": 22.897,
      "step": 15500
    },
    {
      "epoch": 2.3313419787265044,
      "grad_norm": 0.4715974032878876,
      "learning_rate": 0.00014172009325367916,
      "loss": 0.124,
      "step": 16000
    },
    {
      "epoch": 2.3313419787265044,
      "eval_loss": 0.10613198578357697,
      "eval_runtime": 0.5536,
      "eval_samples_per_second": 180.629,
      "eval_steps_per_second": 23.482,
      "step": 16000
    },
    {
      "epoch": 2.4041964155617075,
      "grad_norm": 0.3433252274990082,
      "learning_rate": 0.00013989873233279907,
      "loss": 0.1247,
      "step": 16500
    },
    {
      "epoch": 2.4041964155617075,
      "eval_loss": 0.10550642758607864,
      "eval_runtime": 0.5616,
      "eval_samples_per_second": 178.071,
      "eval_steps_per_second": 23.149,
      "step": 16500
    }
  ],
  "logging_steps": 500,
  "max_steps": 54904,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4466279448576000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}