{
  "best_metric": 1.1142354011535645,
  "best_model_checkpoint": "output-definitions/checkpoint-3000",
  "epoch": 0.158798575349924,
  "eval_steps": 100,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004537102152854971,
      "grad_norm": 0.5996940732002258,
      "learning_rate": 9.90495247623812e-06,
      "loss": 2.0003,
      "step": 100
    },
    {
      "epoch": 0.004537102152854971,
      "eval_loss": 1.3947523832321167,
      "eval_runtime": 3495.2941,
      "eval_samples_per_second": 33.15,
      "eval_steps_per_second": 4.144,
      "step": 100
    },
    {
      "epoch": 0.009074204305709942,
      "grad_norm": 1.2082221508026123,
      "learning_rate": 9.804902451225614e-06,
      "loss": 1.1884,
      "step": 200
    },
    {
      "epoch": 0.009074204305709942,
      "eval_loss": 1.2319905757904053,
      "eval_runtime": 3493.2433,
      "eval_samples_per_second": 33.17,
      "eval_steps_per_second": 4.146,
      "step": 200
    },
    {
      "epoch": 0.013611306458564915,
      "grad_norm": 0.5810414552688599,
      "learning_rate": 9.704852426213108e-06,
      "loss": 1.0592,
      "step": 300
    },
    {
      "epoch": 0.013611306458564915,
      "eval_loss": 1.1850800514221191,
      "eval_runtime": 3493.3223,
      "eval_samples_per_second": 33.169,
      "eval_steps_per_second": 4.146,
      "step": 300
    },
    {
      "epoch": 0.018148408611419885,
      "grad_norm": 0.6719651818275452,
      "learning_rate": 9.604802401200602e-06,
      "loss": 1.036,
      "step": 400
    },
    {
      "epoch": 0.018148408611419885,
      "eval_loss": 1.168699860572815,
      "eval_runtime": 3494.1614,
      "eval_samples_per_second": 33.161,
      "eval_steps_per_second": 4.145,
      "step": 400
    },
    {
      "epoch": 0.022685510764274858,
      "grad_norm": 0.7384809851646423,
      "learning_rate": 9.504752376188094e-06,
      "loss": 1.0105,
      "step": 500
    },
    {
      "epoch": 0.022685510764274858,
      "eval_loss": 1.1580127477645874,
      "eval_runtime": 3518.2589,
      "eval_samples_per_second": 32.934,
      "eval_steps_per_second": 4.117,
      "step": 500
    },
    {
      "epoch": 0.02722261291712983,
      "grad_norm": 0.8317062258720398,
      "learning_rate": 9.404702351175588e-06,
      "loss": 0.9879,
      "step": 600
    },
    {
      "epoch": 0.02722261291712983,
      "eval_loss": 1.148302674293518,
      "eval_runtime": 3494.3971,
      "eval_samples_per_second": 33.159,
      "eval_steps_per_second": 4.145,
      "step": 600
    },
    {
      "epoch": 0.031759715069984804,
      "grad_norm": 0.6463008522987366,
      "learning_rate": 9.304652326163084e-06,
      "loss": 0.9698,
      "step": 700
    },
    {
      "epoch": 0.031759715069984804,
      "eval_loss": 1.1414070129394531,
      "eval_runtime": 3494.7514,
      "eval_samples_per_second": 33.155,
      "eval_steps_per_second": 4.145,
      "step": 700
    },
    {
      "epoch": 0.03629681722283977,
      "grad_norm": 0.705141544342041,
      "learning_rate": 9.204602301150576e-06,
      "loss": 0.9458,
      "step": 800
    },
    {
      "epoch": 0.03629681722283977,
      "eval_loss": 1.1373766660690308,
      "eval_runtime": 3493.8551,
      "eval_samples_per_second": 33.164,
      "eval_steps_per_second": 4.146,
      "step": 800
    },
    {
      "epoch": 0.04083391937569474,
      "grad_norm": 0.710448145866394,
      "learning_rate": 9.10455227613807e-06,
      "loss": 0.9262,
      "step": 900
    },
    {
      "epoch": 0.04083391937569474,
      "eval_loss": 1.1317179203033447,
      "eval_runtime": 3494.4581,
      "eval_samples_per_second": 33.158,
      "eval_steps_per_second": 4.145,
      "step": 900
    },
    {
      "epoch": 0.045371021528549715,
      "grad_norm": 0.8912506103515625,
      "learning_rate": 9.004502251125564e-06,
      "loss": 0.9281,
      "step": 1000
    },
    {
      "epoch": 0.045371021528549715,
      "eval_loss": 1.133247971534729,
      "eval_runtime": 3494.4137,
      "eval_samples_per_second": 33.159,
      "eval_steps_per_second": 4.145,
      "step": 1000
    },
    {
      "epoch": 0.04990812368140469,
      "grad_norm": 0.8533815145492554,
      "learning_rate": 8.904452226113058e-06,
      "loss": 0.9204,
      "step": 1100
    },
    {
      "epoch": 0.04990812368140469,
      "eval_loss": 1.1253483295440674,
      "eval_runtime": 3493.2383,
      "eval_samples_per_second": 33.17,
      "eval_steps_per_second": 4.146,
      "step": 1100
    },
    {
      "epoch": 0.05444522583425966,
      "grad_norm": 0.7760438323020935,
      "learning_rate": 8.804402201100552e-06,
      "loss": 0.8959,
      "step": 1200
    },
    {
      "epoch": 0.05444522583425966,
      "eval_loss": 1.1242249011993408,
      "eval_runtime": 3493.0538,
      "eval_samples_per_second": 33.172,
      "eval_steps_per_second": 4.147,
      "step": 1200
    },
    {
      "epoch": 0.05898232798711463,
      "grad_norm": 0.8464282751083374,
      "learning_rate": 8.704352176088044e-06,
      "loss": 0.8889,
      "step": 1300
    },
    {
      "epoch": 0.05898232798711463,
      "eval_loss": 1.1236740350723267,
      "eval_runtime": 3492.632,
      "eval_samples_per_second": 33.176,
      "eval_steps_per_second": 4.147,
      "step": 1300
    },
    {
      "epoch": 0.06351943013996961,
      "grad_norm": 0.7117558121681213,
      "learning_rate": 8.604302151075538e-06,
      "loss": 0.8898,
      "step": 1400
    },
    {
      "epoch": 0.06351943013996961,
      "eval_loss": 1.1239374876022339,
      "eval_runtime": 3492.8289,
      "eval_samples_per_second": 33.174,
      "eval_steps_per_second": 4.147,
      "step": 1400
    },
    {
      "epoch": 0.06805653229282457,
      "grad_norm": 0.8221879005432129,
      "learning_rate": 8.504252126063032e-06,
      "loss": 0.8862,
      "step": 1500
    },
    {
      "epoch": 0.06805653229282457,
      "eval_loss": 1.1258162260055542,
      "eval_runtime": 3492.1621,
      "eval_samples_per_second": 33.18,
      "eval_steps_per_second": 4.148,
      "step": 1500
    },
    {
      "epoch": 0.07259363444567954,
      "grad_norm": 0.6950238347053528,
      "learning_rate": 8.404202101050526e-06,
      "loss": 0.8745,
      "step": 1600
    },
    {
      "epoch": 0.07259363444567954,
      "eval_loss": 1.115903615951538,
      "eval_runtime": 3492.7622,
      "eval_samples_per_second": 33.174,
      "eval_steps_per_second": 4.147,
      "step": 1600
    },
    {
      "epoch": 0.07713073659853452,
      "grad_norm": 0.8201397061347961,
      "learning_rate": 8.30415207603802e-06,
      "loss": 0.8764,
      "step": 1700
    },
    {
      "epoch": 0.07713073659853452,
      "eval_loss": 1.1180561780929565,
      "eval_runtime": 3492.8242,
      "eval_samples_per_second": 33.174,
      "eval_steps_per_second": 4.147,
      "step": 1700
    },
    {
      "epoch": 0.08166783875138948,
      "grad_norm": 0.9200800061225891,
      "learning_rate": 8.204102051025512e-06,
      "loss": 0.8512,
      "step": 1800
    },
    {
      "epoch": 0.08166783875138948,
      "eval_loss": 1.1254093647003174,
      "eval_runtime": 3493.1426,
      "eval_samples_per_second": 33.171,
      "eval_steps_per_second": 4.146,
      "step": 1800
    },
    {
      "epoch": 0.08620494090424446,
      "grad_norm": 0.8169254064559937,
      "learning_rate": 8.104052026013006e-06,
      "loss": 0.8587,
      "step": 1900
    },
    {
      "epoch": 0.08620494090424446,
      "eval_loss": 1.121141791343689,
      "eval_runtime": 3493.1336,
      "eval_samples_per_second": 33.171,
      "eval_steps_per_second": 4.146,
      "step": 1900
    },
    {
      "epoch": 0.09074204305709943,
      "grad_norm": 0.9403719305992126,
      "learning_rate": 8.0040020010005e-06,
      "loss": 0.8577,
      "step": 2000
    },
    {
      "epoch": 0.09074204305709943,
      "eval_loss": 1.1173222064971924,
      "eval_runtime": 3494.3505,
      "eval_samples_per_second": 33.159,
      "eval_steps_per_second": 4.145,
      "step": 2000
    },
    {
      "epoch": 0.0952791452099544,
      "grad_norm": 0.8594589233398438,
      "learning_rate": 7.903951975987994e-06,
      "loss": 0.85,
      "step": 2100
    },
    {
      "epoch": 0.0952791452099544,
      "eval_loss": 1.1184200048446655,
      "eval_runtime": 3493.8891,
      "eval_samples_per_second": 33.164,
      "eval_steps_per_second": 4.146,
      "step": 2100
    },
    {
      "epoch": 0.09981624736280938,
      "grad_norm": 0.7728345990180969,
      "learning_rate": 7.803901950975488e-06,
      "loss": 0.8418,
      "step": 2200
    },
    {
      "epoch": 0.09981624736280938,
      "eval_loss": 1.1176252365112305,
      "eval_runtime": 3496.1473,
      "eval_samples_per_second": 33.142,
      "eval_steps_per_second": 4.143,
      "step": 2200
    },
    {
      "epoch": 0.10435334951566434,
      "grad_norm": 0.826181948184967,
      "learning_rate": 7.703851925962982e-06,
      "loss": 0.8407,
      "step": 2300
    },
    {
      "epoch": 0.10435334951566434,
      "eval_loss": 1.117732048034668,
      "eval_runtime": 3505.8694,
      "eval_samples_per_second": 33.05,
      "eval_steps_per_second": 4.131,
      "step": 2300
    },
    {
      "epoch": 0.10889045166851932,
      "grad_norm": 0.8817082047462463,
      "learning_rate": 7.603801900950476e-06,
      "loss": 0.8391,
      "step": 2400
    },
    {
      "epoch": 0.10889045166851932,
      "eval_loss": 1.1185327768325806,
      "eval_runtime": 3511.4803,
      "eval_samples_per_second": 32.997,
      "eval_steps_per_second": 4.125,
      "step": 2400
    },
    {
      "epoch": 0.11342755382137429,
      "grad_norm": 1.1382585763931274,
      "learning_rate": 7.503751875937969e-06,
      "loss": 0.8439,
      "step": 2500
    },
    {
      "epoch": 0.11342755382137429,
      "eval_loss": 1.118537187576294,
      "eval_runtime": 3501.1131,
      "eval_samples_per_second": 33.095,
      "eval_steps_per_second": 4.137,
      "step": 2500
    },
    {
      "epoch": 0.11796465597422925,
      "grad_norm": 0.9408033490180969,
      "learning_rate": 7.403701850925464e-06,
      "loss": 0.8309,
      "step": 2600
    },
    {
      "epoch": 0.11796465597422925,
      "eval_loss": 1.112176775932312,
      "eval_runtime": 3495.6002,
      "eval_samples_per_second": 33.147,
      "eval_steps_per_second": 4.143,
      "step": 2600
    },
    {
      "epoch": 0.12250175812708423,
      "grad_norm": 0.9824351072311401,
      "learning_rate": 7.303651825912958e-06,
      "loss": 0.8282,
      "step": 2700
    },
    {
      "epoch": 0.12250175812708423,
      "eval_loss": 1.1151224374771118,
      "eval_runtime": 3494.7487,
      "eval_samples_per_second": 33.155,
      "eval_steps_per_second": 4.145,
      "step": 2700
    },
    {
      "epoch": 0.12703886027993921,
      "grad_norm": 0.832771897315979,
      "learning_rate": 7.203601800900451e-06,
      "loss": 0.8253,
      "step": 2800
    },
    {
      "epoch": 0.12703886027993921,
      "eval_loss": 1.1167434453964233,
      "eval_runtime": 3496.4871,
      "eval_samples_per_second": 33.139,
      "eval_steps_per_second": 4.142,
      "step": 2800
    },
    {
      "epoch": 0.13157596243279418,
      "grad_norm": 0.7724223732948303,
      "learning_rate": 7.103551775887945e-06,
      "loss": 0.8317,
      "step": 2900
    },
    {
      "epoch": 0.13157596243279418,
      "eval_loss": 1.1156556606292725,
      "eval_runtime": 3495.6568,
      "eval_samples_per_second": 33.147,
      "eval_steps_per_second": 4.143,
      "step": 2900
    },
    {
      "epoch": 0.13611306458564915,
      "grad_norm": 0.84904944896698,
      "learning_rate": 7.003501750875439e-06,
      "loss": 0.8296,
      "step": 3000
    },
    {
      "epoch": 0.13611306458564915,
      "eval_loss": 1.1142354011535645,
      "eval_runtime": 3495.4619,
      "eval_samples_per_second": 33.149,
      "eval_steps_per_second": 4.144,
      "step": 3000
    },
    {
      "epoch": 0.1406501667385041,
      "grad_norm": 1.1638028621673584,
      "learning_rate": 6.903451725862932e-06,
      "loss": 0.8194,
      "step": 3100
    },
    {
      "epoch": 0.1406501667385041,
      "eval_loss": 1.1132752895355225,
      "eval_runtime": 3496.3654,
      "eval_samples_per_second": 33.14,
      "eval_steps_per_second": 4.143,
      "step": 3100
    },
    {
      "epoch": 0.14518726889135908,
      "grad_norm": 0.9577890634536743,
      "learning_rate": 6.803401700850426e-06,
      "loss": 0.8197,
      "step": 3200
    },
    {
      "epoch": 0.14518726889135908,
      "eval_loss": 1.1138397455215454,
      "eval_runtime": 3495.7388,
      "eval_samples_per_second": 33.146,
      "eval_steps_per_second": 4.143,
      "step": 3200
    },
    {
      "epoch": 0.14972437104421407,
      "grad_norm": 0.9326839447021484,
      "learning_rate": 6.703351675837919e-06,
      "loss": 0.7983,
      "step": 3300
    },
    {
      "epoch": 0.14972437104421407,
      "eval_loss": 1.1186929941177368,
      "eval_runtime": 3495.9382,
      "eval_samples_per_second": 33.144,
      "eval_steps_per_second": 4.143,
      "step": 3300
    },
    {
      "epoch": 0.15426147319706904,
      "grad_norm": 0.806262195110321,
      "learning_rate": 6.603301650825413e-06,
      "loss": 0.8024,
      "step": 3400
    },
    {
      "epoch": 0.15426147319706904,
      "eval_loss": 1.1130287647247314,
      "eval_runtime": 3496.2696,
      "eval_samples_per_second": 33.141,
      "eval_steps_per_second": 4.143,
      "step": 3400
    },
    {
      "epoch": 0.158798575349924,
      "grad_norm": 0.9623129963874817,
      "learning_rate": 6.503251625812907e-06,
      "loss": 0.8114,
      "step": 3500
    },
    {
      "epoch": 0.158798575349924,
      "eval_loss": 1.1146483421325684,
      "eval_runtime": 3493.8871,
      "eval_samples_per_second": 33.164,
      "eval_steps_per_second": 4.146,
      "step": 3500
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 100,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 1
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.069128403044139e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}