{
  "best_metric": 1.2138804197311401,
  "best_model_checkpoint": "model_training/reprover/checkpoints-random-09-07-14-12/checkpoint-400",
  "epoch": 6.49746192893401,
  "eval_steps": 25,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08121827411167512,
      "grad_norm": 25.242706298828125,
      "learning_rate": 5.102040816326531e-05,
      "loss": 12.5197,
      "step": 5
    },
    {
      "epoch": 0.16243654822335024,
      "grad_norm": 8.962030410766602,
      "learning_rate": 0.00010204081632653062,
      "loss": 7.4254,
      "step": 10
    },
    {
      "epoch": 0.2436548223350254,
      "grad_norm": 0.5921331644058228,
      "learning_rate": 0.00015306122448979594,
      "loss": 3.9799,
      "step": 15
    },
    {
      "epoch": 0.3248730964467005,
      "grad_norm": 0.23172463476657867,
      "learning_rate": 0.00020408163265306123,
      "loss": 3.6019,
      "step": 20
    },
    {
      "epoch": 0.40609137055837563,
      "grad_norm": 0.24555863440036774,
      "learning_rate": 0.00025510204081632655,
      "loss": 3.5033,
      "step": 25
    },
    {
      "epoch": 0.40609137055837563,
      "eval_loss": 3.385397434234619,
      "eval_runtime": 85.9926,
      "eval_samples_per_second": 11.629,
      "eval_steps_per_second": 1.454,
      "step": 25
    },
    {
      "epoch": 0.4873096446700508,
      "grad_norm": 1.7724591493606567,
      "learning_rate": 0.0003061224489795919,
      "loss": 3.414,
      "step": 30
    },
    {
      "epoch": 0.5685279187817259,
      "grad_norm": 1.4451329708099365,
      "learning_rate": 0.00035714285714285714,
      "loss": 2.5288,
      "step": 35
    },
    {
      "epoch": 0.649746192893401,
      "grad_norm": 1.1171362400054932,
      "learning_rate": 0.00040816326530612246,
      "loss": 1.9554,
      "step": 40
    },
    {
      "epoch": 0.7309644670050761,
      "grad_norm": 0.5590386986732483,
      "learning_rate": 0.0004591836734693878,
      "loss": 1.7407,
      "step": 45
    },
    {
      "epoch": 0.8121827411167513,
      "grad_norm": 0.7381185293197632,
      "learning_rate": 0.0004999935985425297,
      "loss": 1.6238,
      "step": 50
    },
    {
      "epoch": 0.8121827411167513,
      "eval_loss": 1.457683801651001,
      "eval_runtime": 85.8152,
      "eval_samples_per_second": 11.653,
      "eval_steps_per_second": 1.457,
      "step": 50
    },
    {
      "epoch": 0.8934010152284264,
      "grad_norm": 0.63067227602005,
      "learning_rate": 0.0004997695819512612,
      "loss": 1.5906,
      "step": 55
    },
    {
      "epoch": 0.9746192893401016,
      "grad_norm": 0.35589519143104553,
      "learning_rate": 0.0004992258202402822,
      "loss": 1.5011,
      "step": 60
    },
    {
      "epoch": 1.0558375634517767,
      "grad_norm": 0.1975679099559784,
      "learning_rate": 0.0004983630095117843,
      "loss": 1.4414,
      "step": 65
    },
    {
      "epoch": 1.1370558375634519,
      "grad_norm": 0.23968252539634705,
      "learning_rate": 0.0004971822543018662,
      "loss": 1.3931,
      "step": 70
    },
    {
      "epoch": 1.218274111675127,
      "grad_norm": 0.20230525732040405,
      "learning_rate": 0.0004956850661665511,
      "loss": 1.4338,
      "step": 75
    },
    {
      "epoch": 1.218274111675127,
      "eval_loss": 1.2832187414169312,
      "eval_runtime": 85.7925,
      "eval_samples_per_second": 11.656,
      "eval_steps_per_second": 1.457,
      "step": 75
    },
    {
      "epoch": 1.299492385786802,
      "grad_norm": 0.17177316546440125,
      "learning_rate": 0.0004938733617467517,
      "loss": 1.4159,
      "step": 80
    },
    {
      "epoch": 1.380710659898477,
      "grad_norm": 0.22775708138942719,
      "learning_rate": 0.0004917494603146632,
      "loss": 1.3703,
      "step": 85
    },
    {
      "epoch": 1.4619289340101522,
      "grad_norm": 0.10171830654144287,
      "learning_rate": 0.0004893160808047222,
      "loss": 1.3801,
      "step": 90
    },
    {
      "epoch": 1.5431472081218274,
      "grad_norm": 0.4525371193885803,
      "learning_rate": 0.00048657633833293557,
      "loss": 1.4181,
      "step": 95
    },
    {
      "epoch": 1.6243654822335025,
      "grad_norm": 0.2127549946308136,
      "learning_rate": 0.0004835337402090316,
      "loss": 1.3681,
      "step": 100
    },
    {
      "epoch": 1.6243654822335025,
      "eval_loss": 1.2635972499847412,
      "eval_runtime": 85.8407,
      "eval_samples_per_second": 11.649,
      "eval_steps_per_second": 1.456,
      "step": 100
    },
    {
      "epoch": 1.7055837563451777,
      "grad_norm": 0.3497621715068817,
      "learning_rate": 0.0004801921814465414,
      "loss": 1.3974,
      "step": 105
    },
    {
      "epoch": 1.7868020304568528,
      "grad_norm": 0.33355796337127686,
      "learning_rate": 0.00047655593977655674,
      "loss": 1.3714,
      "step": 110
    },
    {
      "epoch": 1.868020304568528,
      "grad_norm": 0.26004713773727417,
      "learning_rate": 0.0004726296701715489,
      "loss": 1.3513,
      "step": 115
    },
    {
      "epoch": 1.9492385786802031,
      "grad_norm": 0.21243992447853088,
      "learning_rate": 0.00046841839888625623,
      "loss": 1.3371,
      "step": 120
    },
    {
      "epoch": 2.030456852791878,
      "grad_norm": 0.1547948122024536,
      "learning_rate": 0.0004639275170232734,
      "loss": 1.364,
      "step": 125
    },
    {
      "epoch": 2.030456852791878,
      "eval_loss": 1.2679340839385986,
      "eval_runtime": 85.9259,
      "eval_samples_per_second": 11.638,
      "eval_steps_per_second": 1.455,
      "step": 125
    },
    {
      "epoch": 2.1116751269035534,
      "grad_norm": 0.1696767359972,
      "learning_rate": 0.0004591627736315743,
      "loss": 1.3461,
      "step": 130
    },
    {
      "epoch": 2.1928934010152283,
      "grad_norm": 0.608369767665863,
      "learning_rate": 0.0004541302683468084,
      "loss": 1.309,
      "step": 135
    },
    {
      "epoch": 2.2741116751269037,
      "grad_norm": 0.47118842601776123,
      "learning_rate": 0.0004488364435827881,
      "loss": 1.3606,
      "step": 140
    },
    {
      "epoch": 2.3553299492385786,
      "grad_norm": 0.32203948497772217,
      "learning_rate": 0.00044328807628416644,
      "loss": 1.3892,
      "step": 145
    },
    {
      "epoch": 2.436548223350254,
      "grad_norm": 0.47044509649276733,
      "learning_rate": 0.0004374922692508611,
      "loss": 1.3706,
      "step": 150
    },
    {
      "epoch": 2.436548223350254,
      "eval_loss": 1.2735475301742554,
      "eval_runtime": 85.8193,
      "eval_samples_per_second": 11.652,
      "eval_steps_per_second": 1.457,
      "step": 150
    },
    {
      "epoch": 2.517766497461929,
      "grad_norm": 0.34873417019844055,
      "learning_rate": 0.0004314564420453311,
      "loss": 1.3656,
      "step": 155
    },
    {
      "epoch": 2.598984771573604,
      "grad_norm": 0.3404662609100342,
      "learning_rate": 0.0004251883214943475,
      "loss": 1.345,
      "step": 160
    },
    {
      "epoch": 2.6802030456852792,
      "grad_norm": 0.34060540795326233,
      "learning_rate": 0.0004186959317974155,
      "loss": 1.3485,
      "step": 165
    },
    {
      "epoch": 2.761421319796954,
      "grad_norm": 0.24678784608840942,
      "learning_rate": 0.00041198758425451266,
      "loss": 1.3189,
      "step": 170
    },
    {
      "epoch": 2.8426395939086295,
      "grad_norm": 0.31498032808303833,
      "learning_rate": 0.00040507186662629185,
      "loss": 1.3653,
      "step": 175
    },
    {
      "epoch": 2.8426395939086295,
      "eval_loss": 1.272367000579834,
      "eval_runtime": 85.9187,
      "eval_samples_per_second": 11.639,
      "eval_steps_per_second": 1.455,
      "step": 175
    },
    {
      "epoch": 2.9238578680203045,
      "grad_norm": 0.23488959670066833,
      "learning_rate": 0.0003979576321403705,
      "loss": 1.3121,
      "step": 180
    },
    {
      "epoch": 3.00507614213198,
      "grad_norm": 0.41097673773765564,
      "learning_rate": 0.0003906539881577793,
      "loss": 1.366,
      "step": 185
    },
    {
      "epoch": 3.0862944162436547,
      "grad_norm": 1.2468574047088623,
      "learning_rate": 0.0003831702845140801,
      "loss": 1.3558,
      "step": 190
    },
    {
      "epoch": 3.16751269035533,
      "grad_norm": 0.4176631271839142,
      "learning_rate": 0.00037551610155007613,
      "loss": 1.3358,
      "step": 195
    },
    {
      "epoch": 3.248730964467005,
      "grad_norm": 0.20121543109416962,
      "learning_rate": 0.00036770123784744027,
      "loss": 1.3063,
      "step": 200
    },
    {
      "epoch": 3.248730964467005,
      "eval_loss": 1.2347419261932373,
      "eval_runtime": 85.8995,
      "eval_samples_per_second": 11.642,
      "eval_steps_per_second": 1.455,
      "step": 200
    },
    {
      "epoch": 3.3299492385786804,
      "grad_norm": 0.1536010503768921,
      "learning_rate": 0.00035973569768495855,
      "loss": 1.2936,
      "step": 205
    },
    {
      "epoch": 3.4111675126903553,
      "grad_norm": 0.15390706062316895,
      "learning_rate": 0.0003516296782314491,
      "loss": 1.29,
      "step": 210
    },
    {
      "epoch": 3.4923857868020303,
      "grad_norm": 0.15325894951820374,
      "learning_rate": 0.00034339355649175095,
      "loss": 1.3385,
      "step": 215
    },
    {
      "epoch": 3.5736040609137056,
      "grad_norm": 0.11753768473863602,
      "learning_rate": 0.00033503787602249364,
      "loss": 1.2825,
      "step": 220
    },
    {
      "epoch": 3.6548223350253806,
      "grad_norm": 0.0821477621793747,
      "learning_rate": 0.00032657333343465356,
      "loss": 1.281,
      "step": 225
    },
    {
      "epoch": 3.6548223350253806,
      "eval_loss": 1.2228115797042847,
      "eval_runtime": 85.9026,
      "eval_samples_per_second": 11.641,
      "eval_steps_per_second": 1.455,
      "step": 225
    },
    {
      "epoch": 3.736040609137056,
      "grad_norm": 0.06417221575975418,
      "learning_rate": 0.0003180107647001769,
      "loss": 1.2609,
      "step": 230
    },
    {
      "epoch": 3.817258883248731,
      "grad_norm": 0.3063823878765106,
      "learning_rate": 0.0003093611312801979,
      "loss": 1.2957,
      "step": 235
    },
    {
      "epoch": 3.8984771573604062,
      "grad_norm": 0.28450486063957214,
      "learning_rate": 0.00030063550609261025,
      "loss": 1.2869,
      "step": 240
    },
    {
      "epoch": 3.979695431472081,
      "grad_norm": 0.23690135776996613,
      "learning_rate": 0.000291845059336957,
      "loss": 1.3068,
      "step": 245
    },
    {
      "epoch": 4.060913705583756,
      "grad_norm": 0.4273748993873596,
      "learning_rate": 0.0002830010441947834,
      "loss": 1.2819,
      "step": 250
    },
    {
      "epoch": 4.060913705583756,
      "eval_loss": 1.2325246334075928,
      "eval_runtime": 85.872,
      "eval_samples_per_second": 11.645,
      "eval_steps_per_second": 1.456,
      "step": 250
    },
    {
      "epoch": 4.1421319796954315,
      "grad_norm": 0.47748294472694397,
      "learning_rate": 0.00027411478242376017,
      "loss": 1.3036,
      "step": 255
    },
    {
      "epoch": 4.223350253807107,
      "grad_norm": 0.851790189743042,
      "learning_rate": 0.00026519764986401774,
      "loss": 1.2969,
      "step": 260
    },
    {
      "epoch": 4.304568527918782,
      "grad_norm": 0.2811254560947418,
      "learning_rate": 0.000256261061875247,
      "loss": 1.2699,
      "step": 265
    },
    {
      "epoch": 4.385786802030457,
      "grad_norm": 0.30236339569091797,
      "learning_rate": 0.0002473164587232079,
      "loss": 1.308,
      "step": 270
    },
    {
      "epoch": 4.467005076142132,
      "grad_norm": 0.4765733480453491,
      "learning_rate": 0.0002383752909343547,
      "loss": 1.303,
      "step": 275
    },
    {
      "epoch": 4.467005076142132,
      "eval_loss": 1.2213741540908813,
      "eval_runtime": 85.8151,
      "eval_samples_per_second": 11.653,
      "eval_steps_per_second": 1.457,
      "step": 275
    },
    {
      "epoch": 4.548223350253807,
      "grad_norm": 0.22984470427036285,
      "learning_rate": 0.0002294490046373259,
      "loss": 1.3021,
      "step": 280
    },
    {
      "epoch": 4.629441624365482,
      "grad_norm": 0.216758593916893,
      "learning_rate": 0.00022054902691006405,
      "loss": 1.3071,
      "step": 285
    },
    {
      "epoch": 4.710659898477157,
      "grad_norm": 0.16396842896938324,
      "learning_rate": 0.00021168675115132315,
      "loss": 1.2813,
      "step": 290
    },
    {
      "epoch": 4.791878172588833,
      "grad_norm": 0.0981246754527092,
      "learning_rate": 0.00020287352249529153,
      "loss": 1.3091,
      "step": 295
    },
    {
      "epoch": 4.873096446700508,
      "grad_norm": 0.10069616883993149,
      "learning_rate": 0.00019412062328800044,
      "loss": 1.2781,
      "step": 300
    },
    {
      "epoch": 4.873096446700508,
      "eval_loss": 1.220151424407959,
      "eval_runtime": 85.8774,
      "eval_samples_per_second": 11.645,
      "eval_steps_per_second": 1.456,
      "step": 300
    },
    {
      "epoch": 4.9543147208121825,
      "grad_norm": 0.07661138474941254,
      "learning_rate": 0.000185439258644112,
      "loss": 1.2749,
      "step": 305
    },
    {
      "epoch": 5.035532994923858,
      "grad_norm": 0.09922365099191666,
      "learning_rate": 0.00017684054210257517,
      "loss": 1.2464,
      "step": 310
    },
    {
      "epoch": 5.116751269035533,
      "grad_norm": 0.08816192299127579,
      "learning_rate": 0.00016833548139951395,
      "loss": 1.2642,
      "step": 315
    },
    {
      "epoch": 5.197969543147208,
      "grad_norm": 0.17599380016326904,
      "learning_rate": 0.0001599349643765599,
      "loss": 1.2967,
      "step": 320
    },
    {
      "epoch": 5.279187817258883,
      "grad_norm": 0.1083323284983635,
      "learning_rate": 0.0001516497450426686,
      "loss": 1.3184,
      "step": 325
    },
    {
      "epoch": 5.279187817258883,
      "eval_loss": 1.2185090780258179,
      "eval_runtime": 85.8042,
      "eval_samples_per_second": 11.654,
      "eval_steps_per_second": 1.457,
      "step": 325
    },
    {
      "epoch": 5.3604060913705585,
      "grad_norm": 0.0962451919913292,
      "learning_rate": 0.00014349042980726362,
      "loss": 1.2535,
      "step": 330
    },
    {
      "epoch": 5.441624365482234,
      "grad_norm": 0.08135022222995758,
      "learning_rate": 0.0001354674639023318,
      "loss": 1.2894,
      "step": 335
    },
    {
      "epoch": 5.522842639593908,
      "grad_norm": 0.05687703192234039,
      "learning_rate": 0.00012759111801085066,
      "loss": 1.2648,
      "step": 340
    },
    {
      "epoch": 5.604060913705584,
      "grad_norm": 0.14364075660705566,
      "learning_rate": 0.00011987147511866788,
      "loss": 1.3106,
      "step": 345
    },
    {
      "epoch": 5.685279187817259,
      "grad_norm": 0.0585949569940567,
      "learning_rate": 0.00011231841760666186,
      "loss": 1.2706,
      "step": 350
    },
    {
      "epoch": 5.685279187817259,
      "eval_loss": 1.2204599380493164,
      "eval_runtime": 85.8832,
      "eval_samples_per_second": 11.644,
      "eval_steps_per_second": 1.455,
      "step": 350
    },
    {
      "epoch": 5.7664974619289335,
      "grad_norm": 0.06366579979658127,
      "learning_rate": 0.0001049416145997094,
      "loss": 1.2985,
      "step": 355
    },
    {
      "epoch": 5.847715736040609,
      "grad_norm": 0.14090530574321747,
      "learning_rate": 9.775050958865584e-05,
      "loss": 1.3086,
      "step": 360
    },
    {
      "epoch": 5.928934010152284,
      "grad_norm": 0.08585643023252487,
      "learning_rate": 9.075430834113152e-05,
      "loss": 1.2382,
      "step": 365
    },
    {
      "epoch": 6.01015228426396,
      "grad_norm": 0.0822765976190567,
      "learning_rate": 8.396196711669335e-05,
      "loss": 1.2685,
      "step": 370
    },
    {
      "epoch": 6.091370558375634,
      "grad_norm": 0.05758822336792946,
      "learning_rate": 7.738218120137671e-05,
      "loss": 1.2538,
      "step": 375
    },
    {
      "epoch": 6.091370558375634,
      "eval_loss": 1.216223120689392,
      "eval_runtime": 85.8934,
      "eval_samples_per_second": 11.642,
      "eval_steps_per_second": 1.455,
      "step": 375
    },
    {
      "epoch": 6.1725888324873095,
      "grad_norm": 0.09431485831737518,
      "learning_rate": 7.102337377633394e-05,
      "loss": 1.2767,
      "step": 380
    },
    {
      "epoch": 6.253807106598985,
      "grad_norm": 0.07947517186403275,
      "learning_rate": 6.489368513481228e-05,
      "loss": 1.283,
      "step": 385
    },
    {
      "epoch": 6.33502538071066,
      "grad_norm": 0.07814743369817734,
      "learning_rate": 5.9000962261273136e-05,
      "loss": 1.2769,
      "step": 390
    },
    {
      "epoch": 6.416243654822335,
      "grad_norm": 0.06358485668897629,
      "learning_rate": 5.3352748785993164e-05,
      "loss": 1.3046,
      "step": 395
    },
    {
      "epoch": 6.49746192893401,
      "grad_norm": 0.04739053547382355,
      "learning_rate": 4.795627532800806e-05,
      "loss": 1.2765,
      "step": 400
    },
    {
      "epoch": 6.49746192893401,
      "eval_loss": 1.2138804197311401,
      "eval_runtime": 85.8771,
      "eval_samples_per_second": 11.645,
      "eval_steps_per_second": 1.456,
      "step": 400
    }
  ],
  "logging_steps": 5,
  "max_steps": 488,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.7459851945508864e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}