{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.22321428571428573,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1969.0,
      "completions/mean_length": 670.640625,
      "completions/mean_terminated_length": 626.2096557617188,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.002232142857142857,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 5.688791751861572,
      "learning_rate": 0.0,
      "loss": 0.0,
      "num_tokens": 130921.0,
      "reward": 0.1523437350988388,
      "reward_std": 0.14437203109264374,
      "rewards/format_reward/mean": 0.4296875,
      "rewards/format_reward/std": 0.1751912236213684,
      "rewards/mcq_exact_match_reward/mean": 0.109375,
      "rewards/mcq_exact_match_reward/std": 0.3145764470100403,
      "step": 1
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1241.0,
      "completions/max_terminated_length": 1241.0,
      "completions/mean_length": 485.5625,
      "completions/mean_terminated_length": 485.5625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.004464285714285714,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 14.286972999572754,
      "learning_rate": 2e-07,
      "loss": -0.0,
      "num_tokens": 239237.0,
      "reward": 0.25624996423721313,
      "reward_std": 0.3594309389591217,
      "rewards/format_reward/mean": 0.375,
      "rewards/format_reward/std": 0.26726123690605164,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 2
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.046875,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1409.0,
      "completions/mean_length": 619.875,
      "completions/mean_terminated_length": 549.6393432617188,
      "completions/min_length": 10.0,
      "completions/min_terminated_length": 10.0,
      "epoch": 0.006696428571428571,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 1.6942086219787598,
      "learning_rate": 4e-07,
      "loss": 0.0,
      "num_tokens": 364997.0,
      "reward": 0.2382812201976776,
      "reward_std": 0.3356662690639496,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.26246222853660583,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "step": 3
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1726.0,
      "completions/max_terminated_length": 1726.0,
      "completions/mean_length": 434.203125,
      "completions/mean_terminated_length": 434.203125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.008928571428571428,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 8.222421646118164,
      "learning_rate": 6e-07,
      "loss": -0.0,
      "num_tokens": 512530.0,
      "reward": 0.33203125,
      "reward_std": 0.25712186098098755,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.2302463799715042,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 4
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.046875,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1684.0,
      "completions/mean_length": 542.75,
      "completions/mean_terminated_length": 468.7212829589844,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.011160714285714286,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 3.2874252796173096,
      "learning_rate": 8e-07,
      "loss": 0.0,
      "num_tokens": 636562.0,
      "reward": 0.2554687261581421,
      "reward_std": 0.31778302788734436,
      "rewards/format_reward/mean": 0.3671875,
      "rewards/format_reward/std": 0.2708333432674408,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 5
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1934.0,
      "completions/mean_length": 570.921875,
      "completions/mean_terminated_length": 547.4761962890625,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.013392857142857142,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 16.40054702758789,
      "learning_rate": 1e-06,
      "loss": 0.0,
      "num_tokens": 770101.0,
      "reward": 0.3187499940395355,
      "reward_std": 0.3231881856918335,
      "rewards/format_reward/mean": 0.375,
      "rewards/format_reward/std": 0.2519763112068176,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 6
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1421.0,
      "completions/mean_length": 540.3125,
      "completions/mean_terminated_length": 516.3809814453125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.015625,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 3.9286885261535645,
      "learning_rate": 9.99726628670463e-07,
      "loss": -0.0,
      "num_tokens": 896985.0,
      "reward": 0.2382812350988388,
      "reward_std": 0.3018285930156708,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.2302463799715042,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "step": 7
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1182.0,
      "completions/mean_length": 408.03125,
      "completions/mean_terminated_length": 382.0000305175781,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.017857142857142856,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 10.685583114624023,
      "learning_rate": 9.989068136093872e-07,
      "loss": 0.0,
      "num_tokens": 1038179.0,
      "reward": 0.31874996423721313,
      "reward_std": 0.3212849497795105,
      "rewards/format_reward/mean": 0.375,
      "rewards/format_reward/std": 0.2182178944349289,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 8
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.046875,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1608.0,
      "completions/mean_length": 585.390625,
      "completions/mean_terminated_length": 513.458984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.020089285714285716,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 6.234512805938721,
      "learning_rate": 9.975414512725056e-07,
      "loss": 0.0,
      "num_tokens": 1191276.0,
      "reward": 0.18984374403953552,
      "reward_std": 0.23865148425102234,
      "rewards/format_reward/mean": 0.3359375,
      "rewards/format_reward/std": 0.2824873626232147,
      "rewards/mcq_exact_match_reward/mean": 0.15625,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "step": 9
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1298.0,
      "completions/mean_length": 510.359375,
      "completions/mean_terminated_length": 485.9524230957031,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.022321428571428572,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 8.3804349899292,
      "learning_rate": 9.956320346634875e-07,
      "loss": 0.0,
      "num_tokens": 1355963.0,
      "reward": 0.19999998807907104,
      "reward_std": 0.24215617775917053,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.15625,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "step": 10
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1759.0,
      "completions/mean_length": 546.875,
      "completions/mean_terminated_length": 446.8000183105469,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.024553571428571428,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 9.587211608886719,
      "learning_rate": 9.931806517013612e-07,
      "loss": 0.0,
      "num_tokens": 1510651.0,
      "reward": 0.18046873807907104,
      "reward_std": 0.2516200840473175,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.20275264978408813,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "step": 11
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1622.0,
      "completions/mean_length": 468.15625,
      "completions/mean_terminated_length": 443.0793762207031,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.026785714285714284,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 9.15842342376709,
      "learning_rate": 9.901899829374047e-07,
      "loss": 0.0,
      "num_tokens": 1639349.0,
      "reward": 0.18203124403953552,
      "reward_std": 0.2767041325569153,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.19012710452079773,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "step": 12
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1434.0,
      "completions/mean_length": 361.546875,
      "completions/mean_terminated_length": 334.7778015136719,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.029017857142857144,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 90.891357421875,
      "learning_rate": 9.866632986240029e-07,
      "loss": -0.0,
      "num_tokens": 1770720.0,
      "reward": 0.20937499403953552,
      "reward_std": 0.32802659273147583,
      "rewards/format_reward/mean": 0.375,
      "rewards/format_reward/std": 0.2182178944349289,
      "rewards/mcq_exact_match_reward/mean": 0.171875,
      "rewards/mcq_exact_match_reward/std": 0.38025420904159546,
      "step": 13
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1516.0,
      "completions/mean_length": 462.453125,
      "completions/mean_terminated_length": 411.3064270019531,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.03125,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 14.163143157958984,
      "learning_rate": 9.826044551386742e-07,
      "loss": 0.0,
      "num_tokens": 1887861.0,
      "reward": 0.18359375,
      "reward_std": 0.241878479719162,
      "rewards/format_reward/mean": 0.4296875,
      "rewards/format_reward/std": 0.2798410654067993,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "step": 14
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 2009.0,
      "completions/mean_length": 462.859375,
      "completions/mean_terminated_length": 437.69842529296875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.033482142857142856,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 15.210722923278809,
      "learning_rate": 9.780178907671788e-07,
      "loss": 0.0,
      "num_tokens": 2004364.0,
      "reward": 0.43437498807907104,
      "reward_std": 0.32749149203300476,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 15
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1953.0,
      "completions/mean_length": 307.4375,
      "completions/mean_terminated_length": 251.29031372070312,
      "completions/min_length": 3.0,
      "completions/min_terminated_length": 3.0,
      "epoch": 0.03571428571428571,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 36.37540054321289,
      "learning_rate": 9.729086208503173e-07,
      "loss": -0.0,
      "num_tokens": 2123304.0,
      "reward": 0.30781248211860657,
      "reward_std": 0.27473184466362,
      "rewards/format_reward/mean": 0.421875,
      "rewards/format_reward/std": 0.20351573824882507,
      "rewards/mcq_exact_match_reward/mean": 0.265625,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "step": 16
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1886.0,
      "completions/max_terminated_length": 1886.0,
      "completions/mean_length": 273.625,
      "completions/mean_terminated_length": 273.625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.03794642857142857,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 9.184857368469238,
      "learning_rate": 9.672822322997304e-07,
      "loss": -0.0,
      "num_tokens": 2229352.0,
      "reward": 0.3453124761581421,
      "reward_std": 0.23332780599594116,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.1985812783241272,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 17
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1820.0,
      "completions/max_terminated_length": 1820.0,
      "completions/mean_length": 198.578125,
      "completions/mean_terminated_length": 198.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.04017857142857143,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 13.390080451965332,
      "learning_rate": 9.611448774886923e-07,
      "loss": -0.0,
      "num_tokens": 2336037.0,
      "reward": 0.47109371423721313,
      "reward_std": 0.28497010469436646,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 18
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1105.0,
      "completions/mean_length": 160.203125,
      "completions/mean_terminated_length": 130.23809814453125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.04241071428571429,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 18.958669662475586,
      "learning_rate": 9.545032675245813e-07,
      "loss": -0.0,
      "num_tokens": 2460450.0,
      "reward": 0.5328124761581421,
      "reward_std": 0.31479179859161377,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.1534975916147232,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 19
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1464.0,
      "completions/max_terminated_length": 1464.0,
      "completions/mean_length": 201.640625,
      "completions/mean_terminated_length": 201.640625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.044642857142857144,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 12.093040466308594,
      "learning_rate": 9.473646649103817e-07,
      "loss": -0.0,
      "num_tokens": 2565795.0,
      "reward": 0.18906249105930328,
      "reward_std": 0.18418481945991516,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.1534975916147232,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "step": 20
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1520.0,
      "completions/max_terminated_length": 1520.0,
      "completions/mean_length": 83.59375,
      "completions/mean_terminated_length": 83.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.046875,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 22.807132720947266,
      "learning_rate": 9.397368756032444e-07,
      "loss": -0.0,
      "num_tokens": 2669537.0,
      "reward": 0.4078124761581421,
      "reward_std": 0.326413094997406,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 21
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1720.0,
      "completions/max_terminated_length": 1720.0,
      "completions/mean_length": 65.453125,
      "completions/mean_terminated_length": 65.453125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.049107142857142856,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 13.282039642333984,
      "learning_rate": 9.316282404787869e-07,
      "loss": -0.0,
      "num_tokens": 2762814.0,
      "reward": 0.4093749523162842,
      "reward_std": 0.16249822080135345,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.08908708393573761,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 22
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1298.0,
      "completions/max_terminated_length": 1298.0,
      "completions/mean_length": 105.953125,
      "completions/mean_terminated_length": 105.953125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05133928571428571,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 23.15970802307129,
      "learning_rate": 9.230476262104676e-07,
      "loss": 0.0,
      "num_tokens": 2870939.0,
      "reward": 0.5195311903953552,
      "reward_std": 0.32276755571365356,
      "rewards/format_reward/mean": 0.5078125,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.46875,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "step": 23
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1328.0,
      "completions/max_terminated_length": 1328.0,
      "completions/mean_length": 102.25,
      "completions/mean_terminated_length": 102.25,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05357142857142857,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 12.401535987854004,
      "learning_rate": 9.1400441557401e-07,
      "loss": -0.0,
      "num_tokens": 2973171.0,
      "reward": 0.4414061903953552,
      "reward_std": 0.19047920405864716,
      "rewards/format_reward/mean": 0.5078125,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 24
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 2043.0,
      "completions/max_terminated_length": 2043.0,
      "completions/mean_length": 89.15625,
      "completions/mean_terminated_length": 89.15625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05580357142857143,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 6.303321361541748,
      "learning_rate": 9.045084971874737e-07,
      "loss": -0.0,
      "num_tokens": 3078101.0,
      "reward": 0.5968749523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 25
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 955.0,
      "completions/max_terminated_length": 955.0,
      "completions/mean_length": 68.046875,
      "completions/mean_terminated_length": 68.046875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05803571428571429,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 32.51689529418945,
      "learning_rate": 8.945702546981968e-07,
      "loss": -0.0,
      "num_tokens": 3169840.0,
      "reward": 0.46953120827674866,
      "reward_std": 0.429858535528183,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 26
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 800.0,
      "completions/max_terminated_length": 800.0,
      "completions/mean_length": 113.171875,
      "completions/mean_terminated_length": 113.171875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.060267857142857144,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 34.18963623046875,
      "learning_rate": 8.842005554284295e-07,
      "loss": -0.0,
      "num_tokens": 3263939.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.368122935295105,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.08908708393573761,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 27
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 365.0,
      "completions/max_terminated_length": 365.0,
      "completions/mean_length": 11.796875,
      "completions/mean_terminated_length": 11.796875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0625,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 14.970887184143066,
      "learning_rate": 8.734107384920769e-07,
      "loss": -0.0,
      "num_tokens": 3366678.0,
      "reward": 0.6437499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.59375,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 28
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1756.0,
      "completions/max_terminated_length": 1756.0,
      "completions/mean_length": 133.453125,
      "completions/mean_terminated_length": 133.453125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.06473214285714286,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 24.315492630004883,
      "learning_rate": 8.622126023955445e-07,
      "loss": -0.0,
      "num_tokens": 3497851.0,
      "reward": 0.38749998807907104,
      "reward_std": 0.29143062233924866,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 29
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 10.0,
      "completions/max_terminated_length": 10.0,
      "completions/mean_length": 6.0625,
      "completions/mean_terminated_length": 6.0625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.06696428571428571,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 17.23012924194336,
      "learning_rate": 8.506183921362442e-07,
      "loss": -0.0,
      "num_tokens": 3587679.0,
      "reward": 0.48671871423721313,
      "reward_std": 0.2717972695827484,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 30
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 953.0,
      "completions/max_terminated_length": 953.0,
      "completions/mean_length": 90.21875,
      "completions/mean_terminated_length": 90.21875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.06919642857142858,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.001739501953125,
      "learning_rate": 8.386407858128706e-07,
      "loss": -0.0,
      "num_tokens": 3679245.0,
      "reward": 0.23671874403953552,
      "reward_std": 0.1466214656829834,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "step": 31
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 10.0,
      "completions/max_terminated_length": 10.0,
      "completions/mean_length": 6.0625,
      "completions/mean_terminated_length": 6.0625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.07142857142857142,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.32570457458496,
      "learning_rate": 8.262928807620843e-07,
      "loss": -0.0,
      "num_tokens": 3758209.0,
      "reward": 0.33124998211860657,
      "reward_std": 0.17570310831069946,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 32
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 767.0,
      "completions/max_terminated_length": 767.0,
      "completions/mean_length": 41.234375,
      "completions/mean_terminated_length": 41.234375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.07366071428571429,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 20.529081344604492,
      "learning_rate": 8.135881792367685e-07,
      "loss": -0.0,
      "num_tokens": 3854136.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.3072218596935272,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "step": 33
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.07589285714285714,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 28.661266326904297,
      "learning_rate": 8.005405736415125e-07,
      "loss": -0.0,
      "num_tokens": 3962232.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.3298332095146179,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 34
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 494.0,
      "completions/max_terminated_length": 494.0,
      "completions/mean_length": 19.140625,
      "completions/mean_terminated_length": 19.140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.078125,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 31.662900924682617,
      "learning_rate": 7.871643313414718e-07,
      "loss": -0.0,
      "num_tokens": 4071625.0,
      "reward": 0.40859371423721313,
      "reward_std": 0.28999006748199463,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 35
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 557.0,
      "completions/max_terminated_length": 557.0,
      "completions/mean_length": 28.25,
      "completions/mean_terminated_length": 28.25,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08035714285714286,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 4.329936504364014,
      "learning_rate": 7.734740790612136e-07,
      "loss": 0.0,
      "num_tokens": 4188697.0,
      "reward": 0.04921875149011612,
      "reward_std": 0.00220970856025815,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.0,
      "rewards/mcq_exact_match_reward/std": 0.0,
      "step": 36
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 281.0,
      "completions/max_terminated_length": 281.0,
      "completions/mean_length": 12.34375,
      "completions/mean_terminated_length": 12.34375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08258928571428571,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 29.80189323425293,
      "learning_rate": 7.594847868906076e-07,
      "loss": 0.0,
      "num_tokens": 4284655.0,
      "reward": 0.25312498211860657,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "step": 37
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08482142857142858,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 23.103416442871094,
      "learning_rate": 7.452117519152541e-07,
      "loss": -0.0,
      "num_tokens": 4397455.0,
      "reward": 0.33124998211860657,
      "reward_std": 0.3335031270980835,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 38
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 519.0,
      "completions/max_terminated_length": 519.0,
      "completions/mean_length": 14.015625,
      "completions/mean_terminated_length": 14.015625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08705357142857142,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 25.02655601501465,
      "learning_rate": 7.306705814893439e-07,
      "loss": -0.0,
      "num_tokens": 4501840.0,
      "reward": 0.26874998211860657,
      "reward_std": 0.1735912561416626,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 39
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08928571428571429,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 33.39609909057617,
      "learning_rate": 7.158771761692464e-07,
      "loss": -0.0,
      "num_tokens": 4589424.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.30509039759635925,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "step": 40
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 10.65625,
      "completions/mean_terminated_length": 10.65625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09151785714285714,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 31.670530319213867,
      "learning_rate": 7.008477123264847e-07,
      "loss": -0.0,
      "num_tokens": 4678810.0,
      "reward": 0.28437498211860657,
      "reward_std": 0.1315089464187622,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "step": 41
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 37.90625,
      "completions/mean_terminated_length": 6.000000476837158,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09375,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.955126762390137,
      "learning_rate": 6.855986244591103e-07,
      "loss": -0.0,
      "num_tokens": 4784756.0,
      "reward": 0.29921871423721313,
      "reward_std": 0.0887577086687088,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 42
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 114.0,
      "completions/max_terminated_length": 114.0,
      "completions/mean_length": 7.6875,
      "completions/mean_terminated_length": 7.6875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09598214285714286,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 29.96390151977539,
      "learning_rate": 6.701465872208216e-07,
      "loss": -0.0,
      "num_tokens": 4861592.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.10205792635679245,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 43
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 10.0,
      "completions/max_terminated_length": 10.0,
      "completions/mean_length": 6.0625,
      "completions/mean_terminated_length": 6.0625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09821428571428571,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 51.860042572021484,
      "learning_rate": 6.545084971874736e-07,
      "loss": -0.0,
      "num_tokens": 4939596.0,
      "reward": 0.42499998211860657,
      "reward_std": 0.3029785454273224,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 44
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.10044642857142858,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 23.78936195373535,
      "learning_rate": 6.387014543809223e-07,
      "loss": -0.0,
      "num_tokens": 5043372.0,
      "reward": 0.8624999523162842,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.8125,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "step": 45
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.10267857142857142,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 24.635610580444336,
      "learning_rate": 6.227427435703995e-07,
      "loss": 0.0,
      "num_tokens": 5147556.0,
      "reward": 0.4718749523162842,
      "reward_std": 0.10205791890621185,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 46
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.10491071428571429,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.54861307144165,
      "learning_rate": 6.066498153718734e-07,
      "loss": -0.0,
      "num_tokens": 5248228.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 47
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.10714285714285714,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 82.69651794433594,
      "learning_rate": 5.90440267166055e-07,
      "loss": -0.0,
      "num_tokens": 5335820.0,
      "reward": 0.31562498211860657,
      "reward_std": 0.1315089464187622,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.265625,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "step": 48
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.109375,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 21.174829483032227,
      "learning_rate": 5.741318238559209e-07,
      "loss": -0.0,
      "num_tokens": 5434012.0,
      "reward": 0.78125,
      "reward_std": 0.10342571139335632,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.12198751419782639,
      "rewards/mcq_exact_match_reward/mean": 0.734375,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "step": 49
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11160714285714286,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 38.467594146728516,
      "learning_rate": 5.577423184847931e-07,
      "loss": -0.0,
      "num_tokens": 5531180.0,
      "reward": 0.39374998211860657,
      "reward_std": 0.21306739747524261,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 50
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 9.0,
      "completions/max_terminated_length": 9.0,
      "completions/mean_length": 6.046875,
      "completions/mean_terminated_length": 6.046875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11383928571428571,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 49.62371063232422,
      "learning_rate": 5.412896727361662e-07,
      "loss": -0.0,
      "num_tokens": 5622119.0,
      "reward": 0.45546871423721313,
      "reward_std": 0.26446083188056946,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 51
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11607142857142858,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 23.23208236694336,
      "learning_rate": 5.247918773366111e-07,
      "loss": -0.0,
      "num_tokens": 5713239.0,
      "reward": 0.47187498211860657,
      "reward_std": 0.12255740165710449,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 52
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 6.03125,
      "completions/mean_terminated_length": 6.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11830357142857142,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 37.57743835449219,
      "learning_rate": 5.082669723831793e-07,
      "loss": -0.0,
      "num_tokens": 5792873.0,
      "reward": 0.5023437142372131,
      "reward_std": 0.16044370830059052,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 53
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12053571428571429,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 13.823986053466797,
      "learning_rate": 4.917330276168208e-07,
      "loss": -0.0,
      "num_tokens": 5883153.0,
      "reward": 0.8468749523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.796875,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "step": 54
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12276785714285714,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 20.45488166809082,
      "learning_rate": 4.752081226633888e-07,
      "loss": 0.0,
      "num_tokens": 5981641.0,
      "reward": 0.171875,
      "reward_std": 0.0033407656010240316,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.12198751419782639,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "step": 55
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.125,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 10.456323623657227,
      "learning_rate": 4.5871032726383385e-07,
      "loss": -0.0,
      "num_tokens": 6092361.0,
      "reward": 0.40937498211860657,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 56
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 799.0,
      "completions/max_terminated_length": 799.0,
      "completions/mean_length": 18.390625,
      "completions/mean_terminated_length": 18.390625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12723214285714285,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 0.1349828541278839,
      "learning_rate": 4.4225768151520694e-07,
      "loss": -0.0,
      "num_tokens": 6228874.0,
      "reward": 0.40859371423721313,
      "reward_std": 0.046403877437114716,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 57
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12946428571428573,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 24.23102569580078,
      "learning_rate": 4.258681761440789e-07,
      "loss": -0.0,
      "num_tokens": 6312418.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 58
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.13169642857142858,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 10.645041465759277,
      "learning_rate": 4.095597328339452e-07,
      "loss": -0.0,
      "num_tokens": 6430586.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 59
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.13392857142857142,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 20.60019874572754,
      "learning_rate": 3.9335018462812664e-07,
      "loss": -0.0,
      "num_tokens": 6533346.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.06681530177593231,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 60
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1547.0,
      "completions/max_terminated_length": 1547.0,
      "completions/mean_length": 124.640625,
      "completions/mean_terminated_length": 124.640625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.13616071428571427,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 16.10765838623047,
      "learning_rate": 3.772572564296004e-07,
      "loss": -0.0,
      "num_tokens": 6686283.0,
      "reward": 0.44999998807907104,
      "reward_std": 0.10888782143592834,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 61
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 6.03125,
      "completions/mean_terminated_length": 6.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.13839285714285715,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 13.129045486450195,
      "learning_rate": 3.612985456190778e-07,
      "loss": -0.0,
      "num_tokens": 6774133.0,
      "reward": 0.42421871423721313,
      "reward_std": 0.09059804677963257,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 62
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.140625,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 41.576995849609375,
      "learning_rate": 3.454915028125263e-07,
      "loss": -0.0,
      "num_tokens": 6858557.0,
      "reward": 0.5968749523162842,
      "reward_std": 0.12255740165710449,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 63
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.14285714285714285,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 70.83671569824219,
      "learning_rate": 3.2985341277917846e-07,
      "loss": -0.0,
      "num_tokens": 6943245.0,
      "reward": 0.6593749523162842,
      "reward_std": 0.1315089464187622,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 64
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.14508928571428573,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 40.40315246582031,
      "learning_rate": 3.1440137554088953e-07,
      "loss": -0.0,
      "num_tokens": 7034037.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.16887323558330536,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 65
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.14732142857142858,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.84414291381836,
      "learning_rate": 2.9915228767351535e-07,
      "loss": 0.0,
      "num_tokens": 7116461.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 66
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.14955357142857142,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 29.98472023010254,
      "learning_rate": 2.841228238307536e-07,
      "loss": 0.0,
      "num_tokens": 7226597.0,
      "reward": 0.42499998211860657,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 67
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15178571428571427,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 33.12673568725586,
      "learning_rate": 2.6932941851065615e-07,
      "loss": 0.0,
      "num_tokens": 7301701.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.15992167592048645,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 68
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15401785714285715,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 23.115604400634766,
      "learning_rate": 2.547882480847461e-07,
      "loss": 0.0,
      "num_tokens": 7402205.0,
      "reward": 0.26874998211860657,
      "reward_std": 0.10888781398534775,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 69
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15625,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.86720848083496,
      "learning_rate": 2.4051521310939254e-07,
      "loss": -0.0,
      "num_tokens": 7500701.0,
      "reward": 0.7531249523162842,
      "reward_std": 0.13258251547813416,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.703125,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 70
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15848214285714285,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 32.378047943115234,
      "learning_rate": 2.2652592093878665e-07,
      "loss": -0.0,
      "num_tokens": 7587981.0,
      "reward": 0.4718749523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 71
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16071428571428573,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 8.314337730407715,
      "learning_rate": 2.128356686585282e-07,
      "loss": 0.0,
      "num_tokens": 7712629.0,
      "reward": 0.4406249523162842,
      "reward_std": 0.04419416934251785,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 72
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16294642857142858,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 28.11506462097168,
      "learning_rate": 1.9945942635848745e-07,
      "loss": -0.0,
      "num_tokens": 7812765.0,
      "reward": 0.44062498211860657,
      "reward_std": 0.1315089464187622,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 73
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16517857142857142,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 28.840129852294922,
      "learning_rate": 1.8641182076323148e-07,
      "loss": -0.0,
      "num_tokens": 7913917.0,
      "reward": 0.7218749523162842,
      "reward_std": 0.21778543293476105,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.671875,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "step": 74
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16741071428571427,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 25.669578552246094,
      "learning_rate": 1.7370711923791564e-07,
      "loss": -0.0,
      "num_tokens": 8000589.0,
      "reward": 0.20624999701976776,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.15625,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "step": 75
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16964285714285715,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 26.958602905273438,
      "learning_rate": 1.6135921418712955e-07,
      "loss": 0.0,
      "num_tokens": 8095621.0,
      "reward": 0.34687498211860657,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 76
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.171875,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 45.39596176147461,
      "learning_rate": 1.493816078637557e-07,
      "loss": -0.0,
      "num_tokens": 8194613.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.19939783215522766,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 77
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.17410714285714285,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 31.34361457824707,
      "learning_rate": 1.3778739760445552e-07,
      "loss": 0.0,
      "num_tokens": 8320933.0,
      "reward": 0.23593749105930328,
      "reward_std": 0.14697226881980896,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "step": 78
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.17633928571428573,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.2658926150792322e-07,
      "loss": 0.0,
      "num_tokens": 8425557.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 79
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.17857142857142858,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 19.952136993408203,
      "learning_rate": 1.1579944457157059e-07,
      "loss": -0.0,
      "num_tokens": 8529989.0,
      "reward": 0.7687499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 80
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.18080357142857142,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.0542974530180327e-07,
      "loss": 0.0,
      "num_tokens": 8627653.0,
      "reward": 0.42499998211860657,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 81
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 6.03125,
      "completions/mean_terminated_length": 6.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.18303571428571427,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.0,
      "num_tokens": 8716599.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 82
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.18526785714285715,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 30.264379501342773,
      "learning_rate": 8.599558442598998e-08,
      "loss": -0.0,
      "num_tokens": 8825983.0,
      "reward": 0.48124995827674866,
      "reward_std": 0.1872510462999344,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 83
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1684.0,
      "completions/max_terminated_length": 1684.0,
      "completions/mean_length": 180.265625,
      "completions/mean_terminated_length": 180.265625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1875,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 19.65045928955078,
      "learning_rate": 7.695237378953224e-08,
      "loss": 0.0,
      "num_tokens": 8964320.0,
      "reward": 0.296875,
      "reward_std": 0.08975613117218018,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.12198751419782639,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 84
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.18973214285714285,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 25.923664093017578,
      "learning_rate": 6.837175952121304e-08,
      "loss": 0.0,
      "num_tokens": 9059704.0,
      "reward": 0.7374999523162842,
      "reward_std": 0.10888781398534775,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.6875,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 85
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19196428571428573,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 17.09200668334961,
      "learning_rate": 6.026312439675551e-08,
      "loss": -0.0,
      "num_tokens": 9155320.0,
      "reward": 0.7374999523162842,
      "reward_std": 0.06681530177593231,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.6875,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 86
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19419642857142858,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 29.110422134399414,
      "learning_rate": 5.263533508961826e-08,
      "loss": 0.0,
      "num_tokens": 9251216.0,
      "reward": 0.47187498211860657,
      "reward_std": 0.10205791890621185,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 87
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19642857142857142,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 4.549673247541874e-08,
      "loss": 0.0,
      "num_tokens": 9334992.0,
      "reward": 0.7999999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.75,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 88
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19866071428571427,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.242472648620605,
      "learning_rate": 3.8855122511307626e-08,
      "loss": -0.0,
      "num_tokens": 9449800.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 89
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.20089285714285715,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 3.271776770026963e-08,
      "loss": 0.0,
      "num_tokens": 9532912.0,
      "reward": 0.6749999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 90
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 9.0,
      "completions/max_terminated_length": 9.0,
      "completions/mean_length": 6.046875,
      "completions/mean_terminated_length": 6.046875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.203125,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.572386741638184,
      "learning_rate": 2.7091379149682682e-08,
      "loss": 0.0,
      "num_tokens": 9651963.0,
      "reward": 0.29921871423721313,
      "reward_std": 0.00220970856025815,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 91
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.20535714285714285,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 2.1982109232821176e-08,
      "loss": 0.0,
      "num_tokens": 9731835.0,
      "reward": 0.42499998211860657,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 92
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.20758928571428573,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 54.88142013549805,
      "learning_rate": 1.7395544861325718e-08,
      "loss": -0.0,
      "num_tokens": 9834611.0,
      "reward": 0.39374998211860657,
      "reward_std": 0.21306739747524261,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 93
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.20982142857142858,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 63.461917877197266,
      "learning_rate": 1.3336701375997127e-08,
      "loss": -0.0,
      "num_tokens": 9911331.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.06681530177593231,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 94
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.21205357142857142,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 9.810017062595321e-09,
      "loss": 0.0,
      "num_tokens": 10015483.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 95
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.21428571428571427,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 38.010986328125,
      "learning_rate": 6.819348298638839e-09,
      "loss": 0.0,
      "num_tokens": 10130595.0,
      "reward": 0.7374999523162842,
      "reward_std": 0.10888781398534775,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.6875,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 96
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 10.0,
      "completions/max_terminated_length": 10.0,
      "completions/mean_length": 6.0625,
      "completions/mean_terminated_length": 6.0625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.21651785714285715,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 53.17049026489258,
      "learning_rate": 4.367965336512403e-09,
      "loss": -0.0,
      "num_tokens": 10250583.0,
      "reward": 0.3296874761581421,
      "reward_std": 0.17648445069789886,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 97
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.21875,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 14.568264961242676,
      "learning_rate": 2.458548727494292e-09,
      "loss": -0.0,
      "num_tokens": 10378295.0,
      "reward": 0.3624999523162842,
      "reward_std": 0.06681530177593231,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 98
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.22098214285714285,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.0931863906127325e-09,
      "loss": 0.0,
      "num_tokens": 10485207.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 99
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.22321428571428573,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 10.43686580657959,
      "learning_rate": 2.733713295369755e-10,
      "loss": -0.0,
      "num_tokens": 10579727.0,
      "reward": 0.8937499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.84375,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 10579727,
  "num_train_epochs": 1,
  "save_steps": 50,
  "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": 1,
  "trial_name": null,
  "trial_params": null
}