{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9973333333333333,
  "eval_steps": 500,
  "global_step": 187,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 598.9556373596191,
      "epoch": 0.02666666666666667,
      "grad_norm": 9.657825469970703,
      "kl": 0.00011155605316162109,
      "learning_rate": 7.894736842105263e-07,
      "loss": 0.0,
      "reward": 0.6343750096857548,
      "reward_std": 0.32928847298026087,
      "rewards/accuracy_reward": 0.6337500102818012,
      "rewards/format_reward": 0.0006250000093132258,
      "step": 5
    },
    {
      "completion_length": 605.9112594604492,
      "epoch": 0.05333333333333334,
      "grad_norm": 1.3849772214889526,
      "kl": 0.00035059452056884766,
      "learning_rate": 1.5789473684210526e-06,
      "loss": 0.0,
      "reward": 0.6206250086426734,
      "reward_std": 0.30763257276266814,
      "rewards/accuracy_reward": 0.6200000092387199,
      "rewards/format_reward": 0.0006250000093132258,
      "step": 10
    },
    {
      "completion_length": 640.480638885498,
      "epoch": 0.08,
      "grad_norm": 0.2819180488586426,
      "kl": 0.0017528533935546875,
      "learning_rate": 2.368421052631579e-06,
      "loss": 0.0001,
      "reward": 0.6893750086426735,
      "reward_std": 0.28186101242899897,
      "rewards/accuracy_reward": 0.6893750086426735,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "completion_length": 641.2750106811524,
      "epoch": 0.10666666666666667,
      "grad_norm": 0.2619169354438782,
      "kl": 0.003242778778076172,
      "learning_rate": 2.99973774136912e-06,
      "loss": 0.0001,
      "reward": 0.6918750137090683,
      "reward_std": 0.280092066898942,
      "rewards/accuracy_reward": 0.6918750137090683,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 651.3437652587891,
      "epoch": 0.13333333333333333,
      "grad_norm": 0.7606603503227234,
      "kl": 0.0034598350524902345,
      "learning_rate": 2.9905683148398643e-06,
      "loss": 0.0001,
      "reward": 0.6881250128149986,
      "reward_std": 0.2426654167473316,
      "rewards/accuracy_reward": 0.6881250128149986,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "completion_length": 618.8025131225586,
      "epoch": 0.16,
      "grad_norm": 18.173175811767578,
      "kl": 1.1331443786621094,
      "learning_rate": 2.968377518846473e-06,
      "loss": 0.0453,
      "reward": 0.7312500067055225,
      "reward_std": 0.21250597070902585,
      "rewards/accuracy_reward": 0.7312500067055225,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 594.2168823242188,
      "epoch": 0.18666666666666668,
      "grad_norm": 50.74555587768555,
      "kl": 0.35082550048828126,
      "learning_rate": 2.933359208679211e-06,
      "loss": 0.014,
      "reward": 0.7662500098347664,
      "reward_std": 0.19725151248276235,
      "rewards/accuracy_reward": 0.7662500098347664,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "completion_length": 605.8262619018554,
      "epoch": 0.21333333333333335,
      "grad_norm": 0.8362184762954712,
      "kl": 0.11929092407226563,
      "learning_rate": 2.88581929876693e-06,
      "loss": 0.0048,
      "reward": 0.7368750087916851,
      "reward_std": 0.19084826521575451,
      "rewards/accuracy_reward": 0.7368750087916851,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 587.2862632751464,
      "epoch": 0.24,
      "grad_norm": 17.15748405456543,
      "kl": 0.3153179168701172,
      "learning_rate": 2.8261730902569146e-06,
      "loss": 0.0126,
      "reward": 0.771250007301569,
      "reward_std": 0.1733900425955653,
      "rewards/accuracy_reward": 0.771250007301569,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "completion_length": 587.2462539672852,
      "epoch": 0.26666666666666666,
      "grad_norm": 1.1150826215744019,
      "kl": 0.0309173583984375,
      "learning_rate": 2.7549416430096295e-06,
      "loss": 0.0012,
      "reward": 0.7643750108778476,
      "reward_std": 0.20076387487351893,
      "rewards/accuracy_reward": 0.7643750108778476,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 596.9281379699707,
      "epoch": 0.29333333333333333,
      "grad_norm": 0.9277231693267822,
      "kl": 0.026005935668945313,
      "learning_rate": 2.6727472237020448e-06,
      "loss": 0.001,
      "reward": 0.7562500096857547,
      "reward_std": 0.20265843532979488,
      "rewards/accuracy_reward": 0.7562500096857547,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "completion_length": 557.7656349182129,
      "epoch": 0.32,
      "grad_norm": 78.5399398803711,
      "kl": 0.11486587524414063,
      "learning_rate": 2.58030786980419e-06,
      "loss": 0.0046,
      "reward": 0.7831250101327896,
      "reward_std": 0.16795916706323624,
      "rewards/accuracy_reward": 0.7831250101327896,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "completion_length": 540.9593856811523,
      "epoch": 0.3466666666666667,
      "grad_norm": 2.0994198322296143,
      "kl": 0.02927398681640625,
      "learning_rate": 2.4784311169171817e-06,
      "loss": 0.0012,
      "reward": 0.8068750128149986,
      "reward_std": 0.1683789264410734,
      "rewards/accuracy_reward": 0.8068750128149986,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "completion_length": 608.050633239746,
      "epoch": 0.37333333333333335,
      "grad_norm": 0.15975548326969147,
      "kl": 11.121460723876954,
      "learning_rate": 2.368006944269709e-06,
      "loss": 0.4441,
      "reward": 0.756250013038516,
      "reward_std": 0.19869754556566477,
      "rewards/accuracy_reward": 0.756250013038516,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "completion_length": 556.0112609863281,
      "epoch": 0.4,
      "grad_norm": 0.257069855928421,
      "kl": 0.0086212158203125,
      "learning_rate": 2.25e-06,
      "loss": 0.0003,
      "reward": 0.8018750116229058,
      "reward_std": 0.16034170743077994,
      "rewards/accuracy_reward": 0.8018750116229058,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "completion_length": 582.3343910217285,
      "epoch": 0.4266666666666667,
      "grad_norm": 0.13077034056186676,
      "kl": 0.006656646728515625,
      "learning_rate": 2.1254411741419925e-06,
      "loss": 0.0003,
      "reward": 0.7481250077486038,
      "reward_std": 0.1724422551691532,
      "rewards/accuracy_reward": 0.7481250077486038,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 565.938136291504,
      "epoch": 0.4533333333333333,
      "grad_norm": 0.8749185800552368,
      "kl": 1.2233882904052735,
      "learning_rate": 1.9954185929327507e-06,
      "loss": 0.0489,
      "reward": 0.7856250107288361,
      "reward_std": 0.17018861770629884,
      "rewards/accuracy_reward": 0.7856250107288361,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "completion_length": 597.7000122070312,
      "epoch": 0.48,
      "grad_norm": 0.44738760590553284,
      "kl": 0.14383926391601562,
      "learning_rate": 1.8610681131134598e-06,
      "loss": 0.0058,
      "reward": 0.7462500132620334,
      "reward_std": 0.18847194407135248,
      "rewards/accuracy_reward": 0.7462500132620334,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "completion_length": 594.4781372070313,
      "epoch": 0.5066666666666667,
      "grad_norm": 0.20932912826538086,
      "kl": 0.019523239135742186,
      "learning_rate": 1.7235633992642616e-06,
      "loss": 0.0008,
      "reward": 0.7800000101327896,
      "reward_std": 0.18672875873744488,
      "rewards/accuracy_reward": 0.7800000101327896,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "completion_length": 567.590631866455,
      "epoch": 0.5333333333333333,
      "grad_norm": 0.26683634519577026,
      "kl": 0.00815887451171875,
      "learning_rate": 1.5841056708557877e-06,
      "loss": 0.0003,
      "reward": 0.8012500151991844,
      "reward_std": 0.15466174986213446,
      "rewards/accuracy_reward": 0.8012500151991844,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "completion_length": 602.0687614440918,
      "epoch": 0.56,
      "grad_norm": 0.24808332324028015,
      "kl": 0.046283721923828125,
      "learning_rate": 1.4439132085855116e-06,
      "loss": 0.0018,
      "reward": 0.7306250102818013,
      "reward_std": 0.2066970258951187,
      "rewards/accuracy_reward": 0.7306250102818013,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "completion_length": 561.5981369018555,
      "epoch": 0.5866666666666667,
      "grad_norm": 0.13664592802524567,
      "kl": 0.00771636962890625,
      "learning_rate": 1.304210711669923e-06,
      "loss": 0.0003,
      "reward": 0.7881250090897083,
      "reward_std": 0.16074271276593208,
      "rewards/accuracy_reward": 0.7881250090897083,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "completion_length": 572.8368896484375,
      "epoch": 0.6133333333333333,
      "grad_norm": 61.19581985473633,
      "kl": 0.08488540649414063,
      "learning_rate": 1.1662185990655286e-06,
      "loss": 0.0034,
      "reward": 0.7856250144541264,
      "reward_std": 0.15731600020080805,
      "rewards/accuracy_reward": 0.7856250144541264,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "completion_length": 566.652507019043,
      "epoch": 0.64,
      "grad_norm": 0.6947761178016663,
      "kl": 0.00809326171875,
      "learning_rate": 1.0311423480815335e-06,
      "loss": 0.0003,
      "reward": 0.7475000098347664,
      "reward_std": 0.17467193938791753,
      "rewards/accuracy_reward": 0.7475000098347664,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "completion_length": 578.0793846130371,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.2171238362789154,
      "kl": 0.4027862548828125,
      "learning_rate": 9.001619635203888e-07,
      "loss": 0.0162,
      "reward": 0.7456250101327896,
      "reward_std": 0.15293692518025637,
      "rewards/accuracy_reward": 0.7456250101327896,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "completion_length": 565.3787605285645,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.3446357548236847,
      "kl": 0.038055801391601564,
      "learning_rate": 7.744216693421403e-07,
      "loss": 0.0015,
      "reward": 0.7831250131130219,
      "reward_std": 0.1700244778767228,
      "rewards/accuracy_reward": 0.7831250131130219,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 582.8950141906738,
      "epoch": 0.72,
      "grad_norm": 0.15175624191761017,
      "kl": 0.008423233032226562,
      "learning_rate": 6.550199129045669e-07,
      "loss": 0.0003,
      "reward": 0.7750000104308128,
      "reward_std": 0.16627665907144545,
      "rewards/accuracy_reward": 0.7750000104308128,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "completion_length": 574.6862632751465,
      "epoch": 0.7466666666666667,
      "grad_norm": 0.24874858558177948,
      "kl": 0.018523406982421876,
      "learning_rate": 5.429997691004874e-07,
      "loss": 0.0007,
      "reward": 0.7637500140815974,
      "reward_std": 0.17574753649532796,
      "rewards/accuracy_reward": 0.7637500140815974,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "completion_length": 582.8781341552734,
      "epoch": 0.7733333333333333,
      "grad_norm": 0.22961439192295074,
      "kl": 0.00735931396484375,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0003,
      "reward": 0.7487500086426735,
      "reward_std": 0.16670059841126203,
      "rewards/accuracy_reward": 0.7487500086426735,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "completion_length": 591.0300117492676,
      "epoch": 0.8,
      "grad_norm": 0.1450783908367157,
      "kl": 0.0070438385009765625,
      "learning_rate": 3.449456471411058e-07,
      "loss": 0.0003,
      "reward": 0.7575000070035458,
      "reward_std": 0.17982883434742689,
      "rewards/accuracy_reward": 0.7575000070035458,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "completion_length": 595.1437629699707,
      "epoch": 0.8266666666666667,
      "grad_norm": 0.1316715031862259,
      "kl": 0.012836456298828125,
      "learning_rate": 2.6064183852600797e-07,
      "loss": 0.0005,
      "reward": 0.7687500096857548,
      "reward_std": 0.1724598340690136,
      "rewards/accuracy_reward": 0.7687500096857548,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "completion_length": 572.5243888854981,
      "epoch": 0.8533333333333334,
      "grad_norm": 0.17138485610485077,
      "kl": 0.007579803466796875,
      "learning_rate": 1.8716486713686948e-07,
      "loss": 0.0003,
      "reward": 0.7725000090897083,
      "reward_std": 0.18208168745040892,
      "rewards/accuracy_reward": 0.7725000090897083,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "completion_length": 587.6368865966797,
      "epoch": 0.88,
      "grad_norm": 0.37863653898239136,
      "kl": 0.00740203857421875,
      "learning_rate": 1.251566161950357e-07,
      "loss": 0.0003,
      "reward": 0.7600000105798245,
      "reward_std": 0.19197216629981995,
      "rewards/accuracy_reward": 0.7600000105798245,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "completion_length": 590.2731330871582,
      "epoch": 0.9066666666666666,
      "grad_norm": 0.17426247894763947,
      "kl": 0.008208465576171876,
      "learning_rate": 7.515877999073101e-08,
      "loss": 0.0003,
      "reward": 0.7662500113248825,
      "reward_std": 0.19529659319669007,
      "rewards/accuracy_reward": 0.7662500113248825,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "completion_length": 580.357511138916,
      "epoch": 0.9333333333333333,
      "grad_norm": 3.413297176361084,
      "kl": 0.05416526794433594,
      "learning_rate": 3.7608131727264573e-08,
      "loss": 0.0022,
      "reward": 0.7906250134110451,
      "reward_std": 0.17527884822338818,
      "rewards/accuracy_reward": 0.7906250134110451,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "completion_length": 597.2575073242188,
      "epoch": 0.96,
      "grad_norm": 0.16546441614627838,
      "kl": 0.00909271240234375,
      "learning_rate": 1.2832707939284426e-08,
      "loss": 0.0004,
      "reward": 0.7381250090897084,
      "reward_std": 0.1613092703744769,
      "rewards/accuracy_reward": 0.7381250090897084,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "completion_length": 548.148137664795,
      "epoch": 0.9866666666666667,
      "grad_norm": 0.48810043931007385,
      "kl": 0.007534408569335937,
      "learning_rate": 1.0489428174020875e-09,
      "loss": 0.0003,
      "reward": 0.7881250143051147,
      "reward_std": 0.1571642367169261,
      "rewards/accuracy_reward": 0.7881250143051147,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "completion_length": 565.9296970367432,
      "epoch": 0.9973333333333333,
      "kl": 0.0072765350341796875,
      "reward": 0.8171875178813934,
      "reward_std": 0.1507129198871553,
      "rewards/accuracy_reward": 0.8171875178813934,
      "rewards/format_reward": 0.0,
      "step": 187,
      "total_flos": 0.0,
      "train_loss": 0.01644946206223318,
      "train_runtime": 16174.6118,
      "train_samples_per_second": 0.464,
      "train_steps_per_second": 0.012
    }
  ],
  "logging_steps": 5,
  "max_steps": 187,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}