{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.975609756097561,
  "eval_steps": 1,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 411.93750762939453,
      "epoch": 0.032520325203252036,
      "grad_norm": 0.041010234504938126,
      "kl": 0.0,
      "learning_rate": 3.333333333333333e-07,
      "loss": -0.0227,
      "reward": 0.7500000018626451,
      "reward_std": 0.18404608592391014,
      "rewards/accuracy_reward": 0.7500000018626451,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 507.3958435058594,
      "epoch": 0.06504065040650407,
      "grad_norm": 0.04812910035252571,
      "kl": 0.0,
      "learning_rate": 6.666666666666666e-07,
      "loss": 0.0693,
      "reward": 0.562500013038516,
      "reward_std": 0.21764283627271652,
      "rewards/accuracy_reward": 0.562500013038516,
      "rewards/format_reward": 0.0,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 512.7708473205566,
      "epoch": 0.0975609756097561,
      "grad_norm": 0.03587072715163231,
      "kl": -1.9647181034088135e-05,
      "learning_rate": 1e-06,
      "loss": 0.0052,
      "reward": 0.6458333376795053,
      "reward_std": 0.3170611374080181,
      "rewards/accuracy_reward": 0.6458333376795053,
      "rewards/format_reward": 0.0,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 514.5833473205566,
      "epoch": 0.13008130081300814,
      "grad_norm": 0.04658569023013115,
      "kl": -3.5881996154785156e-05,
      "learning_rate": 1.3333333333333332e-06,
      "loss": -0.0059,
      "reward": 0.5000000149011612,
      "reward_std": 0.3332235962152481,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.0,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 534.9166831970215,
      "epoch": 0.16260162601626016,
      "grad_norm": 0.019696101546287537,
      "kl": -3.643333911895752e-05,
      "learning_rate": 1.6666666666666669e-06,
      "loss": 0.0292,
      "reward": 0.7291666697710752,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.7291666697710752,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 482.1041793823242,
      "epoch": 0.1951219512195122,
      "grad_norm": 0.04941963776946068,
      "kl": -5.251169204711914e-05,
      "learning_rate": 2e-06,
      "loss": 0.022,
      "reward": 0.687500013038516,
      "reward_std": 0.35457348078489304,
      "rewards/accuracy_reward": 0.687500013038516,
      "rewards/format_reward": 0.0,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 521.3125076293945,
      "epoch": 0.22764227642276422,
      "grad_norm": 0.0479104146361351,
      "kl": -7.079541683197021e-05,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.0171,
      "reward": 0.7291666753590107,
      "reward_std": 0.21764283627271652,
      "rewards/accuracy_reward": 0.7291666753590107,
      "rewards/format_reward": 0.0,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 447.35417556762695,
      "epoch": 0.2601626016260163,
      "grad_norm": 0.028429586440324783,
      "kl": -7.56382942199707e-05,
      "learning_rate": 2.6666666666666664e-06,
      "loss": 0.0024,
      "reward": 0.5833333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.5833333432674408,
      "rewards/format_reward": 0.0,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 629.3125190734863,
      "epoch": 0.2926829268292683,
      "grad_norm": 0.036269526928663254,
      "kl": -6.125122308731079e-05,
      "learning_rate": 3e-06,
      "loss": -0.0114,
      "reward": 0.45833334140479565,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.45833334140479565,
      "rewards/format_reward": 0.0,
      "step": 9
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 413.9166889190674,
      "epoch": 0.3252032520325203,
      "grad_norm": 0.047799333930015564,
      "kl": 4.0762126445770264e-05,
      "learning_rate": 2.9988719287563454e-06,
      "loss": -0.0099,
      "reward": 0.6250000074505806,
      "reward_std": 0.2861081585288048,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 525.1666870117188,
      "epoch": 0.35772357723577236,
      "grad_norm": 0.022231411188840866,
      "kl": 0.00024199485778808594,
      "learning_rate": 2.995489411751688e-06,
      "loss": 0.0023,
      "reward": 0.6875000018626451,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000018626451,
      "rewards/format_reward": 0.0,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 550.8125114440918,
      "epoch": 0.3902439024390244,
      "grad_norm": 0.014551201835274696,
      "kl": 0.0006639957427978516,
      "learning_rate": 2.989857536612915e-06,
      "loss": 0.0,
      "reward": 0.5833333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.5833333358168602,
      "rewards/format_reward": 0.0,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 567.5416831970215,
      "epoch": 0.42276422764227645,
      "grad_norm": 0.04071850702166557,
      "kl": 0.0005123615264892578,
      "learning_rate": 2.9819847742152137e-06,
      "loss": 0.0243,
      "reward": 0.5833333395421505,
      "reward_std": 0.22155842930078506,
      "rewards/accuracy_reward": 0.5833333395421505,
      "rewards/format_reward": 0.0,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 532.9375152587891,
      "epoch": 0.45528455284552843,
      "grad_norm": 0.04049593582749367,
      "kl": 0.0007052421569824219,
      "learning_rate": 2.971882965941077e-06,
      "loss": 0.0228,
      "reward": 0.7083333358168602,
      "reward_std": 0.20148037374019623,
      "rewards/accuracy_reward": 0.7083333358168602,
      "rewards/format_reward": 0.0,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 527.9375190734863,
      "epoch": 0.4878048780487805,
      "grad_norm": 0.046176813542842865,
      "kl": 0.001251220703125,
      "learning_rate": 2.959567305869736e-06,
      "loss": 0.0091,
      "reward": 0.5208333432674408,
      "reward_std": 0.1705273948609829,
      "rewards/accuracy_reward": 0.5208333432674408,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.4166717529297,
      "epoch": 0.5203252032520326,
      "grad_norm": 0.028073379769921303,
      "kl": 0.0017507076263427734,
      "learning_rate": 2.9450563179238205e-06,
      "loss": -0.0022,
      "reward": 0.5000000055879354,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.5000000055879354,
      "rewards/format_reward": 0.0,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 441.91667556762695,
      "epoch": 0.5528455284552846,
      "grad_norm": 0.04000205546617508,
      "kl": 0.002796173095703125,
      "learning_rate": 2.928371828007623e-06,
      "loss": 0.0111,
      "reward": 0.7916666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 376.1250114440918,
      "epoch": 0.5853658536585366,
      "grad_norm": 0.051672812551259995,
      "kl": 0.0023317337036132812,
      "learning_rate": 2.9095389311788626e-06,
      "loss": 0.0413,
      "reward": 0.8541666716337204,
      "reward_std": 0.18796167895197868,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 385.3125171661377,
      "epoch": 0.6178861788617886,
      "grad_norm": 0.06200782582163811,
      "kl": 0.0025587081909179688,
      "learning_rate": 2.888585953903336e-06,
      "loss": 0.018,
      "reward": 0.6041666828095913,
      "reward_std": 0.37377968430519104,
      "rewards/accuracy_reward": 0.6041666828095913,
      "rewards/format_reward": 0.0,
      "step": 19
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 531.9375076293945,
      "epoch": 0.6504065040650406,
      "grad_norm": 0.03772967308759689,
      "kl": 0.0052127838134765625,
      "learning_rate": 2.86554441144922e-06,
      "loss": 0.055,
      "reward": 0.5625000055879354,
      "reward_std": 0.2446802221238613,
      "rewards/accuracy_reward": 0.5625000055879354,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 503.27084732055664,
      "epoch": 0.6829268292682927,
      "grad_norm": 0.06043640151619911,
      "kl": 0.005889892578125,
      "learning_rate": 2.8404489604851183e-06,
      "loss": 0.0113,
      "reward": 0.6458333469927311,
      "reward_std": 0.33713918924331665,
      "rewards/accuracy_reward": 0.6458333469927311,
      "rewards/format_reward": 0.0,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 574.1458549499512,
      "epoch": 0.7154471544715447,
      "grad_norm": 0.06661098450422287,
      "kl": 0.0051898956298828125,
      "learning_rate": 2.8133373469531365e-06,
      "loss": 0.0463,
      "reward": 0.7500000111758709,
      "reward_std": 0.30354244261980057,
      "rewards/accuracy_reward": 0.7500000111758709,
      "rewards/format_reward": 0.0,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 492.6458435058594,
      "epoch": 0.7479674796747967,
      "grad_norm": 0.05463815852999687,
      "kl": 0.0066204071044921875,
      "learning_rate": 2.7842503492953997e-06,
      "loss": 0.0033,
      "reward": 0.7291666753590107,
      "reward_std": 0.3266642391681671,
      "rewards/accuracy_reward": 0.7291666753590107,
      "rewards/format_reward": 0.0,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 615.7708511352539,
      "epoch": 0.7804878048780488,
      "grad_norm": 0.06719698756933212,
      "kl": 0.012079238891601562,
      "learning_rate": 2.753231717119405e-06,
      "loss": 0.0303,
      "reward": 0.47916667349636555,
      "reward_std": 0.3816108703613281,
      "rewards/accuracy_reward": 0.47916667349636555,
      "rewards/format_reward": 0.0,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 507.9791831970215,
      "epoch": 0.8130081300813008,
      "grad_norm": 0.04937650263309479,
      "kl": 0.011255264282226562,
      "learning_rate": 2.720328105394451e-06,
      "loss": -0.0146,
      "reward": 0.45833334140479565,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.45833334140479565,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 669.6875152587891,
      "epoch": 0.8455284552845529,
      "grad_norm": 0.05630700662732124,
      "kl": 0.00904083251953125,
      "learning_rate": 2.685589004278139e-06,
      "loss": 0.0147,
      "reward": 0.5208333395421505,
      "reward_std": 0.36417658254504204,
      "rewards/accuracy_reward": 0.5208333395421505,
      "rewards/format_reward": 0.0,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 620.3541831970215,
      "epoch": 0.8780487804878049,
      "grad_norm": 0.062427032738924026,
      "kl": 0.014118194580078125,
      "learning_rate": 2.649066664678467e-06,
      "loss": 0.0143,
      "reward": 0.35416667349636555,
      "reward_std": 0.2446802221238613,
      "rewards/accuracy_reward": 0.35416667349636555,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 406.3541774749756,
      "epoch": 0.9105691056910569,
      "grad_norm": 0.08615698665380478,
      "kl": 0.02004241943359375,
      "learning_rate": 2.610816019663507e-06,
      "loss": 0.0082,
      "reward": 0.604166679084301,
      "reward_std": 0.3720077648758888,
      "rewards/accuracy_reward": 0.604166679084301,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 462.54167556762695,
      "epoch": 0.943089430894309,
      "grad_norm": 0.042346589267253876,
      "kl": 0.014068603515625,
      "learning_rate": 2.5708946018368487e-06,
      "loss": 0.0005,
      "reward": 0.8333333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 434.97918128967285,
      "epoch": 0.975609756097561,
      "grad_norm": 0.05360697954893112,
      "kl": 0.0150909423828125,
      "learning_rate": 2.529362456803101e-06,
      "loss": -0.0145,
      "reward": 0.5000000037252903,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 653.1458473205566,
      "epoch": 1.032520325203252,
      "grad_norm": 0.06400468945503235,
      "kl": 0.012498855590820312,
      "learning_rate": 2.4862820528535957e-06,
      "loss": -0.0046,
      "reward": 0.6250000111758709,
      "reward_std": 0.30354244634509087,
      "rewards/accuracy_reward": 0.6250000111758709,
      "rewards/format_reward": 0.0,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 493.81250762939453,
      "epoch": 1.065040650406504,
      "grad_norm": 0.0848759189248085,
      "kl": 0.01270294189453125,
      "learning_rate": 2.441718187008148e-06,
      "loss": 0.0259,
      "reward": 0.6666666809469461,
      "reward_std": 0.3506578914821148,
      "rewards/accuracy_reward": 0.6666666809469461,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.541675567627,
      "epoch": 1.0975609756097562,
      "grad_norm": 0.07703862339258194,
      "kl": 0.0131988525390625,
      "learning_rate": 2.3957378875541795e-06,
      "loss": 0.0724,
      "reward": 0.39583334513008595,
      "reward_std": 0.35457347705960274,
      "rewards/accuracy_reward": 0.39583334513008595,
      "rewards/format_reward": 0.0,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 507.0000114440918,
      "epoch": 1.1300813008130082,
      "grad_norm": 0.06679866462945938,
      "kl": 0.0151214599609375,
      "learning_rate": 2.348410313229808e-06,
      "loss": -0.004,
      "reward": 0.7500000149011612,
      "reward_std": 0.3680921792984009,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.0,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 480.5416831970215,
      "epoch": 1.1626016260162602,
      "grad_norm": 0.035936519503593445,
      "kl": 0.01000213623046875,
      "learning_rate": 2.2998066492025373e-06,
      "loss": 0.0041,
      "reward": 0.5000000018626451,
      "reward_std": 0.18404608592391014,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.8541831970215,
      "epoch": 1.1951219512195121,
      "grad_norm": 0.05383417382836342,
      "kl": 0.0099029541015625,
      "learning_rate": 2.25e-06,
      "loss": 0.0303,
      "reward": 0.5833333395421505,
      "reward_std": 0.33057983219623566,
      "rewards/accuracy_reward": 0.5833333395421505,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 516.3333511352539,
      "epoch": 1.2276422764227641,
      "grad_norm": 0.06559046357870102,
      "kl": 0.012409210205078125,
      "learning_rate": 2.1990652795538085e-06,
      "loss": 0.0007,
      "reward": 0.6875000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 485.75001335144043,
      "epoch": 1.2601626016260163,
      "grad_norm": 0.04288559779524803,
      "kl": 0.008958816528320312,
      "learning_rate": 2.1470790985218807e-06,
      "loss": 0.0117,
      "reward": 0.5833333414047956,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.5833333414047956,
      "rewards/format_reward": 0.0,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 450.83334159851074,
      "epoch": 1.2926829268292683,
      "grad_norm": 0.09712168574333191,
      "kl": 0.024381637573242188,
      "learning_rate": 2.0941196490587354e-06,
      "loss": 0.0067,
      "reward": 0.604166679084301,
      "reward_std": 0.33713919296860695,
      "rewards/accuracy_reward": 0.604166679084301,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 576.8125152587891,
      "epoch": 1.3252032520325203,
      "grad_norm": 0.05104967951774597,
      "kl": 0.016981124877929688,
      "learning_rate": 2.040266587207066e-06,
      "loss": -0.024,
      "reward": 0.5208333414047956,
      "reward_std": 0.2350771203637123,
      "rewards/accuracy_reward": 0.5208333414047956,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 446.14584732055664,
      "epoch": 1.3577235772357723,
      "grad_norm": 0.10665060579776764,
      "kl": 0.014129638671875,
      "learning_rate": 1.985600913087482e-06,
      "loss": 0.0179,
      "reward": 0.8541666753590107,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.8541666753590107,
      "rewards/format_reward": 0.0,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 448.16668701171875,
      "epoch": 1.3902439024390243,
      "grad_norm": 0.03846089169383049,
      "kl": 0.016796112060546875,
      "learning_rate": 1.9302048490666355e-06,
      "loss": -0.0025,
      "reward": 0.7500000055879354,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.7500000055879354,
      "rewards/format_reward": 0.0,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 533.6875190734863,
      "epoch": 1.4227642276422765,
      "grad_norm": 0.11329367011785507,
      "kl": 0.02149200439453125,
      "learning_rate": 1.874161716086972e-06,
      "loss": 0.0687,
      "reward": 0.7083333395421505,
      "reward_std": 0.22155842930078506,
      "rewards/accuracy_reward": 0.7083333395421505,
      "rewards/format_reward": 0.0,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 578.666675567627,
      "epoch": 1.4552845528455285,
      "grad_norm": 0.09458328783512115,
      "kl": 0.026372909545898438,
      "learning_rate": 1.8175558083441164e-06,
      "loss": 0.0187,
      "reward": 0.6875000111758709,
      "reward_std": 0.2446802221238613,
      "rewards/accuracy_reward": 0.6875000111758709,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 480.3958435058594,
      "epoch": 1.4878048780487805,
      "grad_norm": 0.06678877025842667,
      "kl": 0.043430328369140625,
      "learning_rate": 1.7604722665003958e-06,
      "loss": 0.0241,
      "reward": 0.708333333954215,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.708333333954215,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 446.1041717529297,
      "epoch": 1.5203252032520327,
      "grad_norm": 0.09147606790065765,
      "kl": 0.043731689453125,
      "learning_rate": 1.702996949625197e-06,
      "loss": 0.0073,
      "reward": 0.7916666734963655,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.7916666734963655,
      "rewards/format_reward": 0.0,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 449.3333549499512,
      "epoch": 1.5528455284552845,
      "grad_norm": 0.05224497243762016,
      "kl": 0.0131683349609375,
      "learning_rate": 1.6452163060547687e-06,
      "loss": 0.0098,
      "reward": 0.5833333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.5833333432674408,
      "rewards/format_reward": 0.0,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 461.6041774749756,
      "epoch": 1.5853658536585367,
      "grad_norm": 0.06403756141662598,
      "kl": 0.047855377197265625,
      "learning_rate": 1.5872172433657137e-06,
      "loss": 0.0109,
      "reward": 0.7083333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.7083333358168602,
      "rewards/format_reward": 0.0,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 539.0416870117188,
      "epoch": 1.6178861788617886,
      "grad_norm": 0.11434386670589447,
      "kl": 0.07855987548828125,
      "learning_rate": 1.5290869976577366e-06,
      "loss": 0.0296,
      "reward": 0.4583333358168602,
      "reward_std": 0.20148037374019623,
      "rewards/accuracy_reward": 0.4583333358168602,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 424.2291793823242,
      "epoch": 1.6504065040650406,
      "grad_norm": 0.18005186319351196,
      "kl": 0.11967849731445312,
      "learning_rate": 1.4709130023422637e-06,
      "loss": -0.0011,
      "reward": 0.7708333414047956,
      "reward_std": 0.2350771240890026,
      "rewards/accuracy_reward": 0.7708333414047956,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 501.14584732055664,
      "epoch": 1.6829268292682928,
      "grad_norm": 0.11921941488981247,
      "kl": 0.054943084716796875,
      "learning_rate": 1.4127827566342864e-06,
      "loss": 0.031,
      "reward": 0.5416666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.5416666679084301,
      "rewards/format_reward": 0.0,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 503.5833435058594,
      "epoch": 1.7154471544715446,
      "grad_norm": 0.2314152866601944,
      "kl": 0.046413421630859375,
      "learning_rate": 1.3547836939452313e-06,
      "loss": 0.0801,
      "reward": 0.6666666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.6666666716337204,
      "rewards/format_reward": 0.0,
      "step": 52
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 585.8958358764648,
      "epoch": 1.7479674796747968,
      "grad_norm": 0.07401817291975021,
      "kl": 0.0781097412109375,
      "learning_rate": 1.297003050374804e-06,
      "loss": 0.0253,
      "reward": 0.6875000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 401.1041717529297,
      "epoch": 1.7804878048780488,
      "grad_norm": 0.20789878070354462,
      "kl": 0.037078857421875,
      "learning_rate": 1.2395277334996047e-06,
      "loss": 0.0499,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 54
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 478.5000114440918,
      "epoch": 1.8130081300813008,
      "grad_norm": 0.8033259510993958,
      "kl": 0.11248397827148438,
      "learning_rate": 1.1824441916558843e-06,
      "loss": 0.1429,
      "reward": 0.7500000149011612,
      "reward_std": 0.3776952736079693,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 461.6458396911621,
      "epoch": 1.845528455284553,
      "grad_norm": 0.7502633929252625,
      "kl": 0.13535308837890625,
      "learning_rate": 1.1258382839130282e-06,
      "loss": 0.0708,
      "reward": 0.708333358168602,
      "reward_std": 0.3236205019056797,
      "rewards/accuracy_reward": 0.708333358168602,
      "rewards/format_reward": 0.0,
      "step": 56
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 604.7708511352539,
      "epoch": 1.8780487804878048,
      "grad_norm": 0.47473546862602234,
      "kl": 0.28479576110839844,
      "learning_rate": 1.069795150933365e-06,
      "loss": 0.0366,
      "reward": 0.47916667349636555,
      "reward_std": 0.2350771240890026,
      "rewards/accuracy_reward": 0.47916667349636555,
      "rewards/format_reward": 0.0,
      "step": 57
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 555.5625114440918,
      "epoch": 1.910569105691057,
      "grad_norm": 0.11512857675552368,
      "kl": 0.1971893310546875,
      "learning_rate": 1.0143990869125186e-06,
      "loss": 0.009,
      "reward": 0.4375000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.4375000074505806,
      "rewards/format_reward": 0.0,
      "step": 58
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 566.0833511352539,
      "epoch": 1.943089430894309,
      "grad_norm": 0.25552570819854736,
      "kl": 0.4318656921386719,
      "learning_rate": 9.597334127929346e-07,
      "loss": 0.0466,
      "reward": 0.5625000018626451,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.5625000018626451,
      "rewards/format_reward": 0.0,
      "step": 59
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 417.3541831970215,
      "epoch": 1.975609756097561,
      "grad_norm": 0.4009849727153778,
      "kl": 0.46466064453125,
      "learning_rate": 9.058803509412648e-07,
      "loss": 0.0189,
      "reward": 0.7708333432674408,
      "reward_std": 0.1705273948609829,
      "rewards/accuracy_reward": 0.7708333432674408,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 494.45834732055664,
      "epoch": 2.032520325203252,
      "grad_norm": 0.7852746844291687,
      "kl": 0.4476318359375,
      "learning_rate": 8.529209014781202e-07,
      "loss": 0.0736,
      "reward": 0.6875000055879354,
      "reward_std": 0.2446802221238613,
      "rewards/accuracy_reward": 0.6875000055879354,
      "rewards/format_reward": 0.0,
      "step": 61
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 445.9583435058594,
      "epoch": 2.065040650406504,
      "grad_norm": 0.45981186628341675,
      "kl": 0.4035301208496094,
      "learning_rate": 8.009347204461922e-07,
      "loss": 0.0114,
      "reward": 0.6458333432674408,
      "reward_std": 0.2446802258491516,
      "rewards/accuracy_reward": 0.6458333432674408,
      "rewards/format_reward": 0.0,
      "step": 62
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.666675567627,
      "epoch": 2.097560975609756,
      "grad_norm": 0.9820762276649475,
      "kl": 1.2833786010742188,
      "learning_rate": 7.500000000000003e-07,
      "loss": 0.0928,
      "reward": 0.5416666679084301,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.5416666679084301,
      "rewards/format_reward": 0.0,
      "step": 63
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 489.3541831970215,
      "epoch": 2.130081300813008,
      "grad_norm": 1.9332908391952515,
      "kl": 0.6612548828125,
      "learning_rate": 7.001933507974634e-07,
      "loss": 0.1461,
      "reward": 0.6458333525806665,
      "reward_std": 0.41129202768206596,
      "rewards/accuracy_reward": 0.6458333525806665,
      "rewards/format_reward": 0.0,
      "step": 64
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 560.1875228881836,
      "epoch": 2.16260162601626,
      "grad_norm": 1.457485318183899,
      "kl": 0.762725830078125,
      "learning_rate": 6.515896867701923e-07,
      "loss": 0.131,
      "reward": 0.479166679084301,
      "reward_std": 0.34674229472875595,
      "rewards/accuracy_reward": 0.479166679084301,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 494.08334732055664,
      "epoch": 2.1951219512195124,
      "grad_norm": 0.7749570608139038,
      "kl": 0.5813465118408203,
      "learning_rate": 6.04262112445821e-07,
      "loss": 0.0776,
      "reward": 0.604166679084301,
      "reward_std": 0.2996268533170223,
      "rewards/accuracy_reward": 0.604166679084301,
      "rewards/format_reward": 0.0,
      "step": 66
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 536.0208415985107,
      "epoch": 2.227642276422764,
      "grad_norm": 0.4476146697998047,
      "kl": 1.7618732452392578,
      "learning_rate": 5.582818129918525e-07,
      "loss": 0.1542,
      "reward": 0.6666666753590107,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.6666666753590107,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 556.4166717529297,
      "epoch": 2.2601626016260163,
      "grad_norm": 1.5170872211456299,
      "kl": 2.1071243286132812,
      "learning_rate": 5.137179471464047e-07,
      "loss": 0.131,
      "reward": 0.4791666679084301,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.4791666679084301,
      "rewards/format_reward": 0.0,
      "step": 68
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.5000076293945,
      "epoch": 2.292682926829268,
      "grad_norm": 0.9765549898147583,
      "kl": 1.0570831298828125,
      "learning_rate": 4.7063754319689976e-07,
      "loss": 0.0878,
      "reward": 0.6875000037252903,
      "reward_std": 0.19756478071212769,
      "rewards/accuracy_reward": 0.6875000037252903,
      "rewards/format_reward": 0.0,
      "step": 69
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 496.04168701171875,
      "epoch": 2.3252032520325203,
      "grad_norm": 1.364784598350525,
      "kl": 0.8105888366699219,
      "learning_rate": 4.291053981631517e-07,
      "loss": 0.1012,
      "reward": 0.625000013038516,
      "reward_std": 0.2861081585288048,
      "rewards/accuracy_reward": 0.625000013038516,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 540.4583415985107,
      "epoch": 2.3577235772357725,
      "grad_norm": 0.753292977809906,
      "kl": 1.4221649169921875,
      "learning_rate": 3.891839803364934e-07,
      "loss": 0.1611,
      "reward": 0.6458333395421505,
      "reward_std": 0.2525114081799984,
      "rewards/accuracy_reward": 0.6458333395421505,
      "rewards/format_reward": 0.0,
      "step": 71
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 491.3750114440918,
      "epoch": 2.3902439024390243,
      "grad_norm": 0.9958664178848267,
      "kl": 0.7973136901855469,
      "learning_rate": 3.5093333532153313e-07,
      "loss": 0.073,
      "reward": 0.7083333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.7083333358168602,
      "rewards/format_reward": 0.0,
      "step": 72
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 567.5833473205566,
      "epoch": 2.4227642276422765,
      "grad_norm": 0.12560848891735077,
      "kl": 0.027166366577148438,
      "learning_rate": 3.144109957218612e-07,
      "loss": 0.0161,
      "reward": 0.6666666753590107,
      "reward_std": 0.2861081585288048,
      "rewards/accuracy_reward": 0.6666666753590107,
      "rewards/format_reward": 0.0,
      "step": 73
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 358.87500762939453,
      "epoch": 2.4552845528455283,
      "grad_norm": 0.09765256196260452,
      "kl": 0.0715789794921875,
      "learning_rate": 2.796718946055488e-07,
      "loss": 0.0099,
      "reward": 0.9166666716337204,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 74
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 523.7083435058594,
      "epoch": 2.4878048780487805,
      "grad_norm": 1.1270173788070679,
      "kl": 1.6110763549804688,
      "learning_rate": 2.467682828805956e-07,
      "loss": 0.1044,
      "reward": 0.8541666772216558,
      "reward_std": 0.1530931107699871,
      "rewards/accuracy_reward": 0.8541666772216558,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 487.1458511352539,
      "epoch": 2.5203252032520327,
      "grad_norm": 0.848898708820343,
      "kl": 1.3041305541992188,
      "learning_rate": 2.1574965070460045e-07,
      "loss": 0.1177,
      "reward": 0.7291666809469461,
      "reward_std": 0.28219256922602654,
      "rewards/accuracy_reward": 0.7291666809469461,
      "rewards/format_reward": 0.0,
      "step": 76
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 422.89584732055664,
      "epoch": 2.5528455284552845,
      "grad_norm": 0.6274386644363403,
      "kl": 1.1170845031738281,
      "learning_rate": 1.8666265304686387e-07,
      "loss": 0.0599,
      "reward": 0.5833333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.5833333358168602,
      "rewards/format_reward": 0.0,
      "step": 77
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 455.81251525878906,
      "epoch": 2.5853658536585367,
      "grad_norm": 0.08515371382236481,
      "kl": 0.03688812255859375,
      "learning_rate": 1.5955103951488177e-07,
      "loss": 0.0254,
      "reward": 0.6250000111758709,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.6250000111758709,
      "rewards/format_reward": 0.0,
      "step": 78
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 608.4791831970215,
      "epoch": 2.617886178861789,
      "grad_norm": 1.5989761352539062,
      "kl": 2.8103561401367188,
      "learning_rate": 1.3445558855078017e-07,
      "loss": 0.1075,
      "reward": 0.35416667349636555,
      "reward_std": 0.2350771203637123,
      "rewards/accuracy_reward": 0.35416667349636555,
      "rewards/format_reward": 0.0,
      "step": 79
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 566.7291870117188,
      "epoch": 2.6504065040650406,
      "grad_norm": 1.5026016235351562,
      "kl": 1.9703483581542969,
      "learning_rate": 1.114140460966645e-07,
      "loss": 0.1054,
      "reward": 0.5000000037252903,
      "reward_std": 0.26603010296821594,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 467.8125114440918,
      "epoch": 2.682926829268293,
      "grad_norm": 0.3330284059047699,
      "kl": 0.9720802307128906,
      "learning_rate": 9.046106882113752e-08,
      "loss": 0.0702,
      "reward": 0.7916666734963655,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.7916666734963655,
      "rewards/format_reward": 0.0,
      "step": 81
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 541.625020980835,
      "epoch": 2.7154471544715446,
      "grad_norm": 1.0519593954086304,
      "kl": 1.0916862487792969,
      "learning_rate": 7.162817199237703e-08,
      "loss": 0.0588,
      "reward": 0.4791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.4791666716337204,
      "rewards/format_reward": 0.0,
      "step": 82
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 456.2708549499512,
      "epoch": 2.747967479674797,
      "grad_norm": 0.5059961676597595,
      "kl": 0.6999807357788086,
      "learning_rate": 5.49436820761795e-08,
      "loss": 0.0346,
      "reward": 0.5625000055879354,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.5625000055879354,
      "rewards/format_reward": 0.0,
      "step": 83
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 487.5000114440918,
      "epoch": 2.7804878048780486,
      "grad_norm": 0.8734613060951233,
      "kl": 1.3879623413085938,
      "learning_rate": 4.0432694130264294e-08,
      "loss": 0.0628,
      "reward": 0.6458333414047956,
      "reward_std": 0.2446802258491516,
      "rewards/accuracy_reward": 0.6458333414047956,
      "rewards/format_reward": 0.0,
      "step": 84
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 511.2500114440918,
      "epoch": 2.813008130081301,
      "grad_norm": 0.7217493653297424,
      "kl": 1.629373550415039,
      "learning_rate": 2.8117034058922964e-08,
      "loss": 0.0686,
      "reward": 0.6250000074505806,
      "reward_std": 0.25819891691207886,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 428.6041774749756,
      "epoch": 2.845528455284553,
      "grad_norm": 0.4210287928581238,
      "kl": 0.6216049194335938,
      "learning_rate": 1.8015225784786483e-08,
      "loss": -0.0001,
      "reward": 0.6875000018626451,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000018626451,
      "rewards/format_reward": 0.0,
      "step": 86
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 533.2083396911621,
      "epoch": 2.8780487804878048,
      "grad_norm": 0.43215566873550415,
      "kl": 1.061065673828125,
      "learning_rate": 1.0142463387085465e-08,
      "loss": 0.0424,
      "reward": 0.5416666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.5416666679084301,
      "rewards/format_reward": 0.0,
      "step": 87
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.6041793823242,
      "epoch": 2.910569105691057,
      "grad_norm": 0.8715842366218567,
      "kl": 1.849679946899414,
      "learning_rate": 4.510588248311964e-09,
      "loss": 0.1967,
      "reward": 0.5833333469927311,
      "reward_std": 0.268673874437809,
      "rewards/accuracy_reward": 0.5833333469927311,
      "rewards/format_reward": 0.0,
      "step": 88
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 399.47917556762695,
      "epoch": 2.943089430894309,
      "grad_norm": 0.48902833461761475,
      "kl": 0.33133697509765625,
      "learning_rate": 1.1280712436549378e-09,
      "loss": 0.0284,
      "reward": 0.8750000074505806,
      "reward_std": 0.18404608592391014,
      "rewards/accuracy_reward": 0.8750000074505806,
      "rewards/format_reward": 0.0,
      "step": 89
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 443.4791793823242,
      "epoch": 2.975609756097561,
      "grad_norm": 0.6357685327529907,
      "kl": 0.7278213500976562,
      "learning_rate": 0.0,
      "loss": 0.0558,
      "reward": 0.7083333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "epoch": 2.975609756097561,
      "step": 90,
      "total_flos": 0.0,
      "train_loss": 0.03994659806432739,
      "train_runtime": 6206.0471,
      "train_samples_per_second": 0.119,
      "train_steps_per_second": 0.015
    }
  ],
  "logging_steps": 1,
  "max_steps": 90,
  "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
}