{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.968421052631579,
  "eval_steps": 10,
  "global_step": 69,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 585.2708511352539,
      "epoch": 0.042105263157894736,
      "grad_norm": 0.8169388771057129,
      "kl": 0.0,
      "learning_rate": 4.2857142857142857e-07,
      "loss": 0.0297,
      "reward": 0.6041666679084301,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.6041666679084301,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 598.4791946411133,
      "epoch": 0.08421052631578947,
      "grad_norm": 0.7387536764144897,
      "kl": 0.0,
      "learning_rate": 8.571428571428571e-07,
      "loss": 0.0556,
      "reward": 0.6458333414047956,
      "reward_std": 0.2350771240890026,
      "rewards/accuracy_reward": 0.6458333414047956,
      "rewards/format_reward": 0.0,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.4166831970215,
      "epoch": 0.12631578947368421,
      "grad_norm": 0.29826676845550537,
      "kl": -2.4423934519290924e-05,
      "learning_rate": 1.2857142857142856e-06,
      "loss": 0.0482,
      "reward": 0.6666666734963655,
      "reward_std": 0.30354244261980057,
      "rewards/accuracy_reward": 0.6666666734963655,
      "rewards/format_reward": 0.0,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.4375114440918,
      "epoch": 0.16842105263157894,
      "grad_norm": 0.3846527636051178,
      "kl": 1.6279518604278564e-05,
      "learning_rate": 1.7142857142857143e-06,
      "loss": 0.049,
      "reward": 0.6041666753590107,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.6041666753590107,
      "rewards/format_reward": 0.0,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 382.87501525878906,
      "epoch": 0.21052631578947367,
      "grad_norm": 0.6718897223472595,
      "kl": 2.345442771911621e-05,
      "learning_rate": 2.142857142857143e-06,
      "loss": 0.0309,
      "reward": 0.8541666753590107,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.8541666753590107,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 497.8333435058594,
      "epoch": 0.25263157894736843,
      "grad_norm": 1.5279134511947632,
      "kl": 0.0005899444222450256,
      "learning_rate": 2.571428571428571e-06,
      "loss": 0.0471,
      "reward": 0.8333333414047956,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.8333333414047956,
      "rewards/format_reward": 0.0,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 630.7708473205566,
      "epoch": 0.29473684210526313,
      "grad_norm": 0.8395237326622009,
      "kl": 0.0029467493295669556,
      "learning_rate": 3e-06,
      "loss": 0.045,
      "reward": 0.5000000018626451,
      "reward_std": 0.18404608964920044,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.0,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.4791831970215,
      "epoch": 0.3368421052631579,
      "grad_norm": 0.03908955678343773,
      "kl": -4.875659942626953e-05,
      "learning_rate": 2.9980747607565792e-06,
      "loss": 0.0059,
      "reward": 0.7708333432674408,
      "reward_std": 0.2446802258491516,
      "rewards/accuracy_reward": 0.7708333432674408,
      "rewards/format_reward": 0.0,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 664.0625152587891,
      "epoch": 0.37894736842105264,
      "grad_norm": 0.1090988889336586,
      "kl": 0.05594288744032383,
      "learning_rate": 2.9923039850878425e-06,
      "loss": 0.0021,
      "reward": 0.4791666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.4791666716337204,
      "rewards/format_reward": 0.0,
      "step": 9
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 663.8750076293945,
      "epoch": 0.42105263157894735,
      "grad_norm": 0.9894853234291077,
      "kl": 0.13722804933786392,
      "learning_rate": 2.982702486492167e-06,
      "loss": 0.0697,
      "reward": 0.39583333767950535,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.39583333767950535,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.1458473205566,
      "epoch": 0.4631578947368421,
      "grad_norm": 0.6284940838813782,
      "kl": 0.1499086245894432,
      "learning_rate": 2.969294911878742e-06,
      "loss": 0.057,
      "reward": 0.4791666716337204,
      "reward_std": 0.299626849591732,
      "rewards/accuracy_reward": 0.4791666716337204,
      "rewards/format_reward": 0.0,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 456.3125228881836,
      "epoch": 0.5052631578947369,
      "grad_norm": 0.2573056221008301,
      "kl": 0.1363532543182373,
      "learning_rate": 2.9521156782993067e-06,
      "loss": 0.0423,
      "reward": 0.5833333376795053,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.5833333376795053,
      "rewards/format_reward": 0.0,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 660.6875152587891,
      "epoch": 0.5473684210526316,
      "grad_norm": 0.5341625809669495,
      "kl": 1.2825465202331543,
      "learning_rate": 2.9312088846000733e-06,
      "loss": 0.077,
      "reward": 0.37500000931322575,
      "reward_std": 0.23116153106093407,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.0,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 550.9375190734863,
      "epoch": 0.5894736842105263,
      "grad_norm": 0.15721110999584198,
      "kl": 0.5056180357933044,
      "learning_rate": 2.906628198220621e-06,
      "loss": 0.0395,
      "reward": 0.5416666716337204,
      "reward_std": 0.23116152733564377,
      "rewards/accuracy_reward": 0.5416666716337204,
      "rewards/format_reward": 0.0,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 436.4166851043701,
      "epoch": 0.631578947368421,
      "grad_norm": 0.4244968295097351,
      "kl": 0.5005574226379395,
      "learning_rate": 2.878436717430346e-06,
      "loss": 0.0356,
      "reward": 0.7291666753590107,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.7291666753590107,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.3958473205566,
      "epoch": 0.6736842105263158,
      "grad_norm": 0.7619456648826599,
      "kl": 1.6163926124572754,
      "learning_rate": 2.846706809356113e-06,
      "loss": 0.1662,
      "reward": 0.5416666753590107,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.5416666753590107,
      "rewards/format_reward": 0.0,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 579.854190826416,
      "epoch": 0.7157894736842105,
      "grad_norm": 0.12104064226150513,
      "kl": 0.1672835350036621,
      "learning_rate": 2.811519924216873e-06,
      "loss": 0.0623,
      "reward": 0.7083333488553762,
      "reward_std": 0.3506578877568245,
      "rewards/accuracy_reward": 0.7083333488553762,
      "rewards/format_reward": 0.0,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 544.5000076293945,
      "epoch": 0.7578947368421053,
      "grad_norm": 0.5459911823272705,
      "kl": 1.7731904983520508,
      "learning_rate": 2.7729663862421267e-06,
      "loss": 0.1406,
      "reward": 0.6250000055879354,
      "reward_std": 0.24859581515192986,
      "rewards/accuracy_reward": 0.6250000055879354,
      "rewards/format_reward": 0.0,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.2291831970215,
      "epoch": 0.8,
      "grad_norm": 0.6385266184806824,
      "kl": 1.6936622858047485,
      "learning_rate": 2.731145161810915e-06,
      "loss": 0.1112,
      "reward": 0.5625000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.5625000074505806,
      "rewards/format_reward": 0.0,
      "step": 19
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 517.1666793823242,
      "epoch": 0.8421052631578947,
      "grad_norm": 0.5825877785682678,
      "kl": 1.8140544891357422,
      "learning_rate": 2.6861636054065477e-06,
      "loss": 0.0989,
      "reward": 0.7083333395421505,
      "reward_std": 0.19364918768405914,
      "rewards/accuracy_reward": 0.7083333395421505,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 560.062520980835,
      "epoch": 0.8842105263157894,
      "grad_norm": 1.2789912223815918,
      "kl": 3.327120780944824,
      "learning_rate": 2.6381371840391863e-06,
      "loss": 0.1927,
      "reward": 0.5833333414047956,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.5833333414047956,
      "rewards/format_reward": 0.0,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 582.2500152587891,
      "epoch": 0.9263157894736842,
      "grad_norm": 1.2971068620681763,
      "kl": 2.9201011657714844,
      "learning_rate": 2.58718918084368e-06,
      "loss": 0.2073,
      "reward": 0.3958333469927311,
      "reward_std": 0.299626849591732,
      "rewards/accuracy_reward": 0.3958333469927311,
      "rewards/format_reward": 0.0,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 574.03125,
      "epoch": 0.968421052631579,
      "grad_norm": 0.9836662411689758,
      "kl": 2.003514289855957,
      "learning_rate": 2.53345037861353e-06,
      "loss": 0.1104,
      "reward": 0.7083333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 670.1666831970215,
      "epoch": 1.0421052631578946,
      "grad_norm": 1.2662800550460815,
      "kl": 3.6266231536865234,
      "learning_rate": 2.477058724083334e-06,
      "loss": 0.2207,
      "reward": 0.47916667722165585,
      "reward_std": 0.1530931107699871,
      "rewards/accuracy_reward": 0.47916667722165585,
      "rewards/format_reward": 0.0,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 757.0625152587891,
      "epoch": 1.0842105263157895,
      "grad_norm": 0.8487921357154846,
      "kl": 1.7164249420166016,
      "learning_rate": 2.4181589738214946e-06,
      "loss": 0.0993,
      "reward": 0.45833333395421505,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.45833333395421505,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 620.5000171661377,
      "epoch": 1.1263157894736842,
      "grad_norm": 2.0166633129119873,
      "kl": 0.5937175750732422,
      "learning_rate": 2.3569023226421886e-06,
      "loss": 0.1125,
      "reward": 0.6041666828095913,
      "reward_std": 0.36417658254504204,
      "rewards/accuracy_reward": 0.6041666828095913,
      "rewards/format_reward": 0.0,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 661.8541717529297,
      "epoch": 1.168421052631579,
      "grad_norm": 1.0673203468322754,
      "kl": 0.7642230987548828,
      "learning_rate": 2.2934460154904436e-06,
      "loss": 0.1004,
      "reward": 0.5208333395421505,
      "reward_std": 0.1530931033194065,
      "rewards/accuracy_reward": 0.5208333395421505,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 505.583345413208,
      "epoch": 1.2105263157894737,
      "grad_norm": 2.2793140411376953,
      "kl": 0.50286865234375,
      "learning_rate": 2.227952943796622e-06,
      "loss": 0.1271,
      "reward": 0.6875000074505806,
      "reward_std": 0.2525114119052887,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 590.2500076293945,
      "epoch": 1.2526315789473683,
      "grad_norm": 0.7321359515190125,
      "kl": 0.8715305328369141,
      "learning_rate": 2.160591227336452e-06,
      "loss": 0.0905,
      "reward": 0.6875000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 611.6666831970215,
      "epoch": 1.2947368421052632,
      "grad_norm": 2.0107011795043945,
      "kl": 1.0236454010009766,
      "learning_rate": 2.091533782669978e-06,
      "loss": 0.1705,
      "reward": 0.562500013038516,
      "reward_std": 0.28219256550073624,
      "rewards/accuracy_reward": 0.562500013038516,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 567.8333473205566,
      "epoch": 1.3368421052631578,
      "grad_norm": 2.7980597019195557,
      "kl": 1.0768804550170898,
      "learning_rate": 2.0209578792672304e-06,
      "loss": 0.2222,
      "reward": 0.6458333525806665,
      "reward_std": 0.40168892592191696,
      "rewards/accuracy_reward": 0.6458333525806665,
      "rewards/format_reward": 0.0,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 769.7291736602783,
      "epoch": 1.3789473684210527,
      "grad_norm": 3.1634137630462646,
      "kl": 3.4175052642822266,
      "learning_rate": 1.9490446844600373e-06,
      "loss": 0.1332,
      "reward": 0.3541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.3541666716337204,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 748.3541831970215,
      "epoch": 1.4210526315789473,
      "grad_norm": 2.0227866172790527,
      "kl": 4.861572265625,
      "learning_rate": 1.875978798388081e-06,
      "loss": 0.2971,
      "reward": 0.33333334140479565,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.33333334140479565,
      "rewards/format_reward": 0.0,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 787.1250228881836,
      "epoch": 1.4631578947368422,
      "grad_norm": 2.728727102279663,
      "kl": 4.805639266967773,
      "learning_rate": 1.8019477801329903e-06,
      "loss": 0.2784,
      "reward": 0.2708333395421505,
      "reward_std": 0.2621145099401474,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 545.7083473205566,
      "epoch": 1.5052631578947369,
      "grad_norm": 0.2891092598438263,
      "kl": 1.8428325653076172,
      "learning_rate": 1.7271416662568652e-06,
      "loss": 0.1843,
      "reward": 0.5416666753590107,
      "reward_std": 0.24859581515192986,
      "rewards/accuracy_reward": 0.5416666753590107,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 707.5208435058594,
      "epoch": 1.5473684210526315,
      "grad_norm": 0.8343372344970703,
      "kl": 3.2880020141601562,
      "learning_rate": 1.6517524829811483e-06,
      "loss": 0.2373,
      "reward": 0.3750000074505806,
      "reward_std": 0.19364918768405914,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.3125152587891,
      "epoch": 1.5894736842105264,
      "grad_norm": 0.24261082708835602,
      "kl": 1.7664175033569336,
      "learning_rate": 1.5759737532580691e-06,
      "loss": 0.1494,
      "reward": 0.5833333376795053,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.5833333376795053,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 460.6250190734863,
      "epoch": 1.631578947368421,
      "grad_norm": 0.5113571286201477,
      "kl": 1.065016746520996,
      "learning_rate": 1.5e-06,
      "loss": 0.1319,
      "reward": 0.7083333432674408,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 664.0625228881836,
      "epoch": 1.6736842105263157,
      "grad_norm": 1.053536057472229,
      "kl": 3.0947265625,
      "learning_rate": 1.4240262467419312e-06,
      "loss": 0.2513,
      "reward": 0.4166666828095913,
      "reward_std": 0.2957112602889538,
      "rewards/accuracy_reward": 0.4166666828095913,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.229175567627,
      "epoch": 1.7157894736842105,
      "grad_norm": 1.1233307123184204,
      "kl": 0.8006033897399902,
      "learning_rate": 1.348247517018852e-06,
      "loss": 0.1101,
      "reward": 0.7291666846722364,
      "reward_std": 0.33713918924331665,
      "rewards/accuracy_reward": 0.7291666846722364,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 618.6250076293945,
      "epoch": 1.7578947368421054,
      "grad_norm": 0.22662228345870972,
      "kl": 2.493999481201172,
      "learning_rate": 1.2728583337431355e-06,
      "loss": 0.1554,
      "reward": 0.5833333395421505,
      "reward_std": 0.18404608592391014,
      "rewards/accuracy_reward": 0.5833333395421505,
      "rewards/format_reward": 0.0,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 666.4583511352539,
      "epoch": 1.8,
      "grad_norm": 0.6697850227355957,
      "kl": 2.398993492126465,
      "learning_rate": 1.1980522198670096e-06,
      "loss": 0.2155,
      "reward": 0.39583334140479565,
      "reward_std": 0.2350771278142929,
      "rewards/accuracy_reward": 0.39583334140479565,
      "rewards/format_reward": 0.0,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 540.479175567627,
      "epoch": 1.8421052631578947,
      "grad_norm": 0.615111768245697,
      "kl": 1.5004844665527344,
      "learning_rate": 1.1240212016119191e-06,
      "loss": 0.1012,
      "reward": 0.6041666846722364,
      "reward_std": 0.33713919669389725,
      "rewards/accuracy_reward": 0.6041666846722364,
      "rewards/format_reward": 0.0,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 553.645845413208,
      "epoch": 1.8842105263157896,
      "grad_norm": 1.1411396265029907,
      "kl": 2.1822094917297363,
      "learning_rate": 1.050955315539963e-06,
      "loss": 0.1153,
      "reward": 0.5833333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.5833333358168602,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 620.5625076293945,
      "epoch": 1.9263157894736842,
      "grad_norm": 1.032118558883667,
      "kl": 2.441394805908203,
      "learning_rate": 9.790421207327699e-07,
      "loss": 0.1346,
      "reward": 0.5000000037252903,
      "reward_std": 0.19364918768405914,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 507.84375,
      "epoch": 1.9684210526315788,
      "grad_norm": 0.3780059516429901,
      "kl": 1.3306865692138672,
      "learning_rate": 9.084662173300225e-07,
      "loss": 0.0993,
      "reward": 0.7708333432674408,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.7708333432674408,
      "rewards/format_reward": 0.0,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 639.3958396911621,
      "epoch": 2.042105263157895,
      "grad_norm": 0.9595404267311096,
      "kl": 2.2831053733825684,
      "learning_rate": 8.394087726635485e-07,
      "loss": 0.1552,
      "reward": 0.6458333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6458333395421505,
      "rewards/format_reward": 0.0,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 461.06250762939453,
      "epoch": 2.0842105263157893,
      "grad_norm": 0.22546370327472687,
      "kl": 1.1452112197875977,
      "learning_rate": 7.720470562033787e-07,
      "loss": 0.0808,
      "reward": 0.7708333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.7708333395421505,
      "rewards/format_reward": 0.0,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.6875152587891,
      "epoch": 2.126315789473684,
      "grad_norm": 1.2044512033462524,
      "kl": 2.3818206787109375,
      "learning_rate": 7.065539845095568e-07,
      "loss": 0.1233,
      "reward": 0.5833333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.5833333432674408,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 573.7916870117188,
      "epoch": 2.168421052631579,
      "grad_norm": 0.28790467977523804,
      "kl": 1.2321147918701172,
      "learning_rate": 6.430976773578113e-07,
      "loss": 0.092,
      "reward": 0.6875000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 584.2916717529297,
      "epoch": 2.2105263157894735,
      "grad_norm": 0.22598668932914734,
      "kl": 1.0693860054016113,
      "learning_rate": 5.818410261785057e-07,
      "loss": 0.0745,
      "reward": 0.6875000018626451,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000018626451,
      "rewards/format_reward": 0.0,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 675.6875038146973,
      "epoch": 2.2526315789473683,
      "grad_norm": 0.7614906430244446,
      "kl": 1.5083866119384766,
      "learning_rate": 5.22941275916667e-07,
      "loss": 0.0763,
      "reward": 0.41666667349636555,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.41666667349636555,
      "rewards/format_reward": 0.0,
      "step": 52
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 415.2083435058594,
      "epoch": 2.294736842105263,
      "grad_norm": 0.25470709800720215,
      "kl": 0.29476356506347656,
      "learning_rate": 4.6654962138647007e-07,
      "loss": 0.022,
      "reward": 0.8125000149011612,
      "reward_std": 0.1530931070446968,
      "rewards/accuracy_reward": 0.8125000149011612,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 528.9583415985107,
      "epoch": 2.336842105263158,
      "grad_norm": 0.6020339727401733,
      "kl": 0.6340560913085938,
      "learning_rate": 4.1281081915632036e-07,
      "loss": 0.0636,
      "reward": 0.5416666734963655,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.5416666734963655,
      "rewards/format_reward": 0.0,
      "step": 54
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 557.4583358764648,
      "epoch": 2.3789473684210525,
      "grad_norm": 0.2363707721233368,
      "kl": 0.6674823760986328,
      "learning_rate": 3.618628159608137e-07,
      "loss": 0.0476,
      "reward": 0.7083333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 430.66668128967285,
      "epoch": 2.4210526315789473,
      "grad_norm": 0.4364229738712311,
      "kl": 0.07082462310791016,
      "learning_rate": 3.1383639459345236e-07,
      "loss": 0.0172,
      "reward": 0.8333333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 56
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 442.2916793823242,
      "epoch": 2.463157894736842,
      "grad_norm": 0.7256594896316528,
      "kl": 0.24945640563964844,
      "learning_rate": 2.688548381890859e-07,
      "loss": 0.0578,
      "reward": 0.7083333395421505,
      "reward_std": 0.19364918768405914,
      "rewards/accuracy_reward": 0.7083333395421505,
      "rewards/format_reward": 0.0,
      "step": 57
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 627.0000076293945,
      "epoch": 2.5052631578947366,
      "grad_norm": 1.6205965280532837,
      "kl": 0.4336977005004883,
      "learning_rate": 2.2703361375787346e-07,
      "loss": 0.1196,
      "reward": 0.562500013038516,
      "reward_std": 0.28219256922602654,
      "rewards/accuracy_reward": 0.562500013038516,
      "rewards/format_reward": 0.0,
      "step": 58
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 615.5833511352539,
      "epoch": 2.5473684210526315,
      "grad_norm": 0.5687099695205688,
      "kl": 0.3663787841796875,
      "learning_rate": 1.8848007578312686e-07,
      "loss": 0.0396,
      "reward": 0.5208333414047956,
      "reward_std": 0.299626849591732,
      "rewards/accuracy_reward": 0.5208333414047956,
      "rewards/format_reward": 0.0,
      "step": 59
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 508.3958396911621,
      "epoch": 2.5894736842105264,
      "grad_norm": 0.2119116485118866,
      "kl": 0.1974639892578125,
      "learning_rate": 1.5329319064388763e-07,
      "loss": 0.0028,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.3333435058594,
      "epoch": 2.6315789473684212,
      "grad_norm": 0.17258569598197937,
      "kl": 0.7138900756835938,
      "learning_rate": 1.215632825696541e-07,
      "loss": 0.0576,
      "reward": 0.39583334140479565,
      "reward_std": 0.2350771203637123,
      "rewards/accuracy_reward": 0.39583334140479565,
      "rewards/format_reward": 0.0,
      "step": 61
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 549.6041831970215,
      "epoch": 2.6736842105263157,
      "grad_norm": 0.8958130478858948,
      "kl": 0.40015602111816406,
      "learning_rate": 9.337180177937954e-08,
      "loss": 0.0813,
      "reward": 0.5625000111758709,
      "reward_std": 0.2446802221238613,
      "rewards/accuracy_reward": 0.5625000111758709,
      "rewards/format_reward": 0.0,
      "step": 62
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 578.0000114440918,
      "epoch": 2.7157894736842105,
      "grad_norm": 0.11735378205776215,
      "kl": 0.5429248809814453,
      "learning_rate": 6.879111539992677e-08,
      "loss": 0.0507,
      "reward": 0.7083333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.7083333358168602,
      "rewards/format_reward": 0.0,
      "step": 63
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 607.8750228881836,
      "epoch": 2.7578947368421054,
      "grad_norm": 1.0303009748458862,
      "kl": 0.9785404205322266,
      "learning_rate": 4.788432170069373e-08,
      "loss": 0.1647,
      "reward": 0.6041666828095913,
      "reward_std": 0.36417658627033234,
      "rewards/accuracy_reward": 0.6041666828095913,
      "rewards/format_reward": 0.0,
      "step": 64
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 448.7083511352539,
      "epoch": 2.8,
      "grad_norm": 1.2717252969741821,
      "kl": 0.3193473815917969,
      "learning_rate": 3.0705088121258276e-08,
      "loss": 0.0794,
      "reward": 0.8333333469927311,
      "reward_std": 0.23116153106093407,
      "rewards/accuracy_reward": 0.8333333469927311,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 663.6250076293945,
      "epoch": 2.8421052631578947,
      "grad_norm": 0.1792384833097458,
      "kl": 1.4888973236083984,
      "learning_rate": 1.729751350783293e-08,
      "loss": 0.1661,
      "reward": 0.3958333432674408,
      "reward_std": 0.28219256550073624,
      "rewards/accuracy_reward": 0.3958333432674408,
      "rewards/format_reward": 0.0,
      "step": 66
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 563.1250152587891,
      "epoch": 2.8842105263157896,
      "grad_norm": 1.4438295364379883,
      "kl": 1.0477867126464844,
      "learning_rate": 7.696014912157268e-09,
      "loss": 0.1218,
      "reward": 0.7291666865348816,
      "reward_std": 0.27258946746587753,
      "rewards/accuracy_reward": 0.7291666865348816,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 682.6875190734863,
      "epoch": 2.9263157894736844,
      "grad_norm": 1.317630410194397,
      "kl": 2.032796859741211,
      "learning_rate": 1.9252392434208623e-09,
      "loss": 0.0906,
      "reward": 0.3333333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.3333333358168602,
      "rewards/format_reward": 0.0,
      "step": 68
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 544.9375,
      "epoch": 2.968421052631579,
      "grad_norm": 0.5226422548294067,
      "kl": 1.5402240753173828,
      "learning_rate": 0.0,
      "loss": 0.1601,
      "reward": 0.6875000149011612,
      "reward_std": 0.1530931107699871,
      "rewards/accuracy_reward": 0.6875000149011612,
      "rewards/format_reward": 0.0,
      "step": 69
    },
    {
      "epoch": 2.968421052631579,
      "step": 69,
      "total_flos": 0.0,
      "train_loss": 0.0823538613377436,
      "train_runtime": 2262.0879,
      "train_samples_per_second": 0.252,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 1,
  "max_steps": 69,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}