{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.25,
  "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1517.0,
      "completions/mean_length": 542.71875,
      "completions/mean_terminated_length": 518.825439453125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0025,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 4.380501747131348,
      "learning_rate": 0.0,
      "loss": 0.0,
      "num_tokens": 143854.0,
      "reward": 0.27656251192092896,
      "reward_std": 0.33590424060821533,
      "rewards/format_reward/mean": 0.421875,
      "rewards/format_reward/std": 0.25539806485176086,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 1493.0,
      "completions/max_terminated_length": 1493.0,
      "completions/mean_length": 498.84375,
      "completions/mean_terminated_length": 498.84375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.005,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 4.987982749938965,
      "learning_rate": 2e-07,
      "loss": -0.0,
      "num_tokens": 264244.0,
      "reward": 0.2593749761581421,
      "reward_std": 0.36261773109436035,
      "rewards/format_reward/mean": 0.40625,
      "rewards/format_reward/std": 0.25,
      "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": 0.0,
      "completions/max_length": 1608.0,
      "completions/max_terminated_length": 1608.0,
      "completions/mean_length": 382.953125,
      "completions/mean_terminated_length": 382.953125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0075,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 10.849379539489746,
      "learning_rate": 4e-07,
      "loss": -0.0,
      "num_tokens": 377593.0,
      "reward": 0.4453124701976776,
      "reward_std": 0.3289090096950531,
      "rewards/format_reward/mean": 0.390625,
      "rewards/format_reward/std": 0.2083333432674408,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 1238.0,
      "completions/max_terminated_length": 1238.0,
      "completions/mean_length": 376.609375,
      "completions/mean_terminated_length": 376.609375,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.01,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 15.719167709350586,
      "learning_rate": 6e-07,
      "loss": 0.0,
      "num_tokens": 464888.0,
      "reward": 0.27656251192092896,
      "reward_std": 0.20368173718452454,
      "rewards/format_reward/mean": 0.265625,
      "rewards/format_reward/std": 0.2514837086200714,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "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": 0.078125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1253.0,
      "completions/mean_length": 625.6875,
      "completions/mean_terminated_length": 505.1525573730469,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0125,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 3.586479425430298,
      "learning_rate": 8e-07,
      "loss": 0.0,
      "num_tokens": 597396.0,
      "reward": 0.36328125,
      "reward_std": 0.2888522744178772,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.27716949582099915,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 0.0,
      "completions/max_length": 1480.0,
      "completions/max_terminated_length": 1480.0,
      "completions/mean_length": 447.9375,
      "completions/mean_terminated_length": 447.9375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.015,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 7.2977824211120605,
      "learning_rate": 1e-06,
      "loss": -0.0,
      "num_tokens": 713624.0,
      "reward": 0.22578124701976776,
      "reward_std": 0.3024963140487671,
      "rewards/format_reward/mean": 0.3828125,
      "rewards/format_reward/std": 0.3420548141002655,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 1559.0,
      "completions/mean_length": 476.1875,
      "completions/mean_terminated_length": 451.2381286621094,
      "completions/min_length": 11.0,
      "completions/min_terminated_length": 11.0,
      "epoch": 0.0175,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 3.5359883308410645,
      "learning_rate": 9.99726628670463e-07,
      "loss": 0.0,
      "num_tokens": 838868.0,
      "reward": 0.3007812201976776,
      "reward_std": 0.40516340732574463,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.2302463799715042,
      "rewards/mcq_exact_match_reward/mean": 0.265625,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "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": 0.0,
      "completions/max_length": 1881.0,
      "completions/max_terminated_length": 1881.0,
      "completions/mean_length": 403.6875,
      "completions/mean_terminated_length": 403.6875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.02,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 7.5658183097839355,
      "learning_rate": 9.989068136093872e-07,
      "loss": 0.0,
      "num_tokens": 960848.0,
      "reward": 0.2757812440395355,
      "reward_std": 0.29019075632095337,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.19012710452079773,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 0.0,
      "completions/max_length": 1617.0,
      "completions/max_terminated_length": 1617.0,
      "completions/mean_length": 430.125,
      "completions/mean_terminated_length": 430.125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0225,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 11.243498802185059,
      "learning_rate": 9.975414512725056e-07,
      "loss": -0.0,
      "num_tokens": 1088728.0,
      "reward": 0.23984374105930328,
      "reward_std": 0.3661530911922455,
      "rewards/format_reward/mean": 0.3671875,
      "rewards/format_reward/std": 0.2708333432674408,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "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": 0.0,
      "completions/max_length": 1389.0,
      "completions/max_terminated_length": 1389.0,
      "completions/mean_length": 405.71875,
      "completions/mean_terminated_length": 405.71875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 15.843240737915039,
      "learning_rate": 9.956320346634875e-07,
      "loss": -0.0,
      "num_tokens": 1205598.0,
      "reward": 0.4289062023162842,
      "reward_std": 0.3367609977722168,
      "rewards/format_reward/mean": 0.3828125,
      "rewards/format_reward/std": 0.2634054720401764,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1098.0,
      "completions/mean_length": 322.078125,
      "completions/mean_terminated_length": 294.68255615234375,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.0275,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 19.11398696899414,
      "learning_rate": 9.931806517013612e-07,
      "loss": 0.0,
      "num_tokens": 1299091.0,
      "reward": 0.16875000298023224,
      "reward_std": 0.18418586254119873,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.3021090030670166,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 0.0,
      "completions/max_length": 1788.0,
      "completions/max_terminated_length": 1788.0,
      "completions/mean_length": 347.53125,
      "completions/mean_terminated_length": 347.53125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.03,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 17.908710479736328,
      "learning_rate": 9.901899829374047e-07,
      "loss": 0.0,
      "num_tokens": 1406101.0,
      "reward": 0.38515621423721313,
      "reward_std": 0.2916370928287506,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.2280818521976471,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 0.0,
      "completions/max_length": 1672.0,
      "completions/max_terminated_length": 1672.0,
      "completions/mean_length": 129.765625,
      "completions/mean_terminated_length": 129.765625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0325,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 23.634878158569336,
      "learning_rate": 9.866632986240029e-07,
      "loss": -0.0,
      "num_tokens": 1498862.0,
      "reward": 0.49609375,
      "reward_std": 0.43229910731315613,
      "rewards/format_reward/mean": 0.4296875,
      "rewards/format_reward/std": 0.1751912236213684,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 0.0,
      "completions/max_length": 1178.0,
      "completions/max_terminated_length": 1178.0,
      "completions/mean_length": 180.4375,
      "completions/mean_terminated_length": 180.4375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.035,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 9.797518730163574,
      "learning_rate": 9.826044551386742e-07,
      "loss": -0.0,
      "num_tokens": 1593058.0,
      "reward": 0.37187498807907104,
      "reward_std": 0.21642214059829712,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 0.0,
      "completions/max_length": 1039.0,
      "completions/max_terminated_length": 1039.0,
      "completions/mean_length": 196.4375,
      "completions/mean_terminated_length": 196.4375,
      "completions/min_length": 3.0,
      "completions/min_terminated_length": 3.0,
      "epoch": 0.0375,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 24.81659698486328,
      "learning_rate": 9.780178907671788e-07,
      "loss": 0.0,
      "num_tokens": 1678678.0,
      "reward": 0.33281248807907104,
      "reward_std": 0.20713931322097778,
      "rewards/format_reward/mean": 0.515625,
      "rewards/format_reward/std": 0.21764887869358063,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 0.0,
      "completions/max_length": 538.0,
      "completions/max_terminated_length": 538.0,
      "completions/mean_length": 124.09375,
      "completions/mean_terminated_length": 124.09375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.04,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 22.56555938720703,
      "learning_rate": 9.729086208503173e-07,
      "loss": -0.0,
      "num_tokens": 1774308.0,
      "reward": 0.16171874105930328,
      "reward_std": 0.21683646738529205,
      "rewards/format_reward/mean": 0.5234375,
      "rewards/format_reward/std": 0.1649840772151947,
      "rewards/mcq_exact_match_reward/mean": 0.109375,
      "rewards/mcq_exact_match_reward/std": 0.3145764470100403,
      "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": 522.0,
      "completions/max_terminated_length": 522.0,
      "completions/mean_length": 70.71875,
      "completions/mean_terminated_length": 70.71875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0425,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 11.378738403320312,
      "learning_rate": 9.672822322997304e-07,
      "loss": 0.0,
      "num_tokens": 1879082.0,
      "reward": 0.17421874403953552,
      "reward_std": 0.1689612716436386,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 124.0,
      "completions/max_terminated_length": 124.0,
      "completions/mean_length": 7.90625,
      "completions/mean_terminated_length": 7.90625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.045,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.292799949645996,
      "learning_rate": 9.611448774886923e-07,
      "loss": -0.0,
      "num_tokens": 1985724.0,
      "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": 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": 0.0,
      "completions/max_length": 746.0,
      "completions/max_terminated_length": 746.0,
      "completions/mean_length": 21.1875,
      "completions/mean_terminated_length": 21.1875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0475,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 12.76447868347168,
      "learning_rate": 9.545032675245813e-07,
      "loss": 0.0,
      "num_tokens": 2077176.0,
      "reward": 0.45546871423721313,
      "reward_std": 0.22608542442321777,
      "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": 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": 443.0,
      "completions/max_terminated_length": 443.0,
      "completions/mean_length": 21.46875,
      "completions/mean_terminated_length": 21.46875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 20.05408477783203,
      "learning_rate": 9.473646649103817e-07,
      "loss": -0.0,
      "num_tokens": 2167950.0,
      "reward": 0.5671874284744263,
      "reward_std": 0.2874714434146881,
      "rewards/format_reward/mean": 0.515625,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 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": 0.0525,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 25.972938537597656,
      "learning_rate": 9.397368756032444e-07,
      "loss": -0.0,
      "num_tokens": 2259870.0,
      "reward": 0.4718749523162842,
      "reward_std": 0.2382849156856537,
      "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": 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": 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": 0.055,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 28.088258743286133,
      "learning_rate": 9.316282404787869e-07,
      "loss": -0.0,
      "num_tokens": 2345006.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.23144522309303284,
      "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": 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": 189.0,
      "completions/max_terminated_length": 189.0,
      "completions/mean_length": 10.53125,
      "completions/mean_terminated_length": 10.53125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0575,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 11.48979663848877,
      "learning_rate": 9.230476262104676e-07,
      "loss": -0.0,
      "num_tokens": 2424168.0,
      "reward": 0.25312498211860657,
      "reward_std": 0.0646936446428299,
      "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": 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": 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": 0.06,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 5.386105060577393,
      "learning_rate": 9.1400441557401e-07,
      "loss": -0.0,
      "num_tokens": 2525128.0,
      "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": 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": 133.0,
      "completions/max_terminated_length": 133.0,
      "completions/mean_length": 7.984375,
      "completions/mean_terminated_length": 7.984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0625,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 15.605831146240234,
      "learning_rate": 9.045084971874737e-07,
      "loss": -0.0,
      "num_tokens": 2638759.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.23356688022613525,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 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": 0.065,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 20.44306182861328,
      "learning_rate": 8.945702546981968e-07,
      "loss": -0.0,
      "num_tokens": 2743751.0,
      "reward": 0.26874998211860657,
      "reward_std": 0.16675156354904175,
      "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": 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": 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": 0.0675,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 13.371686935424805,
      "learning_rate": 8.842005554284295e-07,
      "loss": -0.0,
      "num_tokens": 2853623.0,
      "reward": 0.6437499523162842,
      "reward_std": 0.1462520956993103,
      "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": 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": 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": 0.07,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 10.312115669250488,
      "learning_rate": 8.734107384920769e-07,
      "loss": -0.0,
      "num_tokens": 2948135.0,
      "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": 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": 233.0,
      "completions/max_terminated_length": 233.0,
      "completions/mean_length": 9.546875,
      "completions/mean_terminated_length": 9.546875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0725,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.40567398071289,
      "learning_rate": 8.622126023955445e-07,
      "loss": -0.0,
      "num_tokens": 3022946.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.1825428158044815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 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": 0.075,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.289785385131836,
      "learning_rate": 8.506183921362442e-07,
      "loss": -0.0,
      "num_tokens": 3103842.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.16675157845020294,
      "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": 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": 153.0,
      "completions/max_terminated_length": 153.0,
      "completions/mean_length": 9.671875,
      "completions/mean_terminated_length": 9.671875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0775,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 17.972658157348633,
      "learning_rate": 8.386407858128706e-07,
      "loss": 0.0,
      "num_tokens": 3189101.0,
      "reward": 0.3617187440395355,
      "reward_std": 0.11793721467256546,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 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": 0.08,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 18.15964126586914,
      "learning_rate": 8.262928807620843e-07,
      "loss": -0.0,
      "num_tokens": 3296485.0,
      "reward": 0.7218749523162842,
      "reward_std": 0.24039676785469055,
      "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": 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": 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": 0.0825,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 26.73876953125,
      "learning_rate": 8.135881792367685e-07,
      "loss": 0.0,
      "num_tokens": 3367525.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.2756393849849701,
      "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": 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": 0.085,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 11.452264785766602,
      "learning_rate": 8.005405736415125e-07,
      "loss": -0.0,
      "num_tokens": 3463373.0,
      "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": 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": 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": 0.0875,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 16.812923431396484,
      "learning_rate": 7.871643313414718e-07,
      "loss": -0.0,
      "num_tokens": 3522957.0,
      "reward": 0.5968749523162842,
      "reward_std": 0.10205791890621185,
      "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": 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": 9.0,
      "completions/max_terminated_length": 9.0,
      "completions/mean_length": 6.140625,
      "completions/mean_terminated_length": 6.140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 33.69114685058594,
      "learning_rate": 7.734740790612136e-07,
      "loss": -0.0,
      "num_tokens": 3613494.0,
      "reward": 0.43828123807907104,
      "reward_std": 0.17210790514945984,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 136.0,
      "completions/max_terminated_length": 136.0,
      "completions/mean_length": 8.03125,
      "completions/mean_terminated_length": 8.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0925,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 20.531408309936523,
      "learning_rate": 7.594847868906076e-07,
      "loss": -0.0,
      "num_tokens": 3695088.0,
      "reward": 0.7687499523162842,
      "reward_std": 0.22461533546447754,
      "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": 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": 0.095,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 14.78104019165039,
      "learning_rate": 7.452117519152541e-07,
      "loss": 0.0,
      "num_tokens": 3773304.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.11100947111845016,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 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": 0.0975,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 23.33732795715332,
      "learning_rate": 7.306705814893439e-07,
      "loss": -0.0,
      "num_tokens": 3868104.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.1315089464187622,
      "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": 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": 0.1,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.439239501953125,
      "learning_rate": 7.158771761692464e-07,
      "loss": 0.0,
      "num_tokens": 3971352.0,
      "reward": 0.45390623807907104,
      "reward_std": 0.09162302315235138,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 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": 0.1025,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 45.16755294799805,
      "learning_rate": 7.008477123264847e-07,
      "loss": -0.0,
      "num_tokens": 4110216.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.37403714656829834,
      "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": 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": 0.0,
      "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": 0.105,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 6.855986244591103e-07,
      "loss": 0.0,
      "num_tokens": 4188784.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 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": 0.1075,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 16.25175666809082,
      "learning_rate": 6.701465872208216e-07,
      "loss": 0.0,
      "num_tokens": 4278584.0,
      "reward": 0.5968749523162842,
      "reward_std": 0.10205791890621185,
      "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": 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": 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": 0.11,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 29.4295654296875,
      "learning_rate": 6.545084971874736e-07,
      "loss": -0.0,
      "num_tokens": 4388496.0,
      "reward": 0.39374998211860657,
      "reward_std": 0.17570312321186066,
      "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": 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": 0.1125,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.442708015441895,
      "learning_rate": 6.387014543809223e-07,
      "loss": -0.0,
      "num_tokens": 4489288.0,
      "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": 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": 0.115,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 13.756986618041992,
      "learning_rate": 6.227427435703995e-07,
      "loss": -0.0,
      "num_tokens": 4570216.0,
      "reward": 0.7843749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.734375,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "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": 0.1175,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 29.47736930847168,
      "learning_rate": 6.066498153718734e-07,
      "loss": -0.0,
      "num_tokens": 4636392.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.16675156354904175,
      "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": 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": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 6.0625,
      "completions/mean_terminated_length": 6.0625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.25701141357422,
      "learning_rate": 5.90440267166055e-07,
      "loss": -0.0,
      "num_tokens": 4751964.0,
      "reward": 0.7046874761581421,
      "reward_std": 0.12757223844528198,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.65625,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 173.0,
      "completions/max_terminated_length": 173.0,
      "completions/mean_length": 8.765625,
      "completions/mean_terminated_length": 8.765625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1225,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 39.44269943237305,
      "learning_rate": 5.741318238559209e-07,
      "loss": -0.0,
      "num_tokens": 4853997.0,
      "reward": 0.5320311784744263,
      "reward_std": 0.2810235917568207,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 0.125,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 24.45294189453125,
      "learning_rate": 5.577423184847931e-07,
      "loss": 0.0,
      "num_tokens": 4931885.0,
      "reward": 0.5187499523162842,
      "reward_std": 0.10888781398534775,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.46875,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 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": 0.1275,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 25.792654037475586,
      "learning_rate": 5.412896727361662e-07,
      "loss": -0.0,
      "num_tokens": 5021013.0,
      "reward": 0.6437499523162842,
      "reward_std": 0.25513991713523865,
      "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": 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": 0.13,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 40.203712463378906,
      "learning_rate": 5.247918773366111e-07,
      "loss": 0.0,
      "num_tokens": 5094909.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.15992167592048645,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1325,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 23.380035400390625,
      "learning_rate": 5.082669723831793e-07,
      "loss": 0.0,
      "num_tokens": 5184458.0,
      "reward": 0.7062499523162842,
      "reward_std": 0.19727617502212524,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.65625,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 0.135,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 6.792049884796143,
      "learning_rate": 4.917330276168208e-07,
      "loss": -0.0,
      "num_tokens": 5282250.0,
      "reward": 0.7218749523162842,
      "reward_std": 0.0646936446428299,
      "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": 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": 0.1375,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 19.197797775268555,
      "learning_rate": 4.752081226633888e-07,
      "loss": -0.0,
      "num_tokens": 5369458.0,
      "reward": 0.7531249523162842,
      "reward_std": 0.0646936446428299,
      "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": 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": 0.14,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 23.735740661621094,
      "learning_rate": 4.5871032726383385e-07,
      "loss": 0.0,
      "num_tokens": 5468274.0,
      "reward": 0.6437499523162842,
      "reward_std": 0.10888781398534775,
      "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": 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": 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": 0.1425,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 12.13380241394043,
      "learning_rate": 4.4225768151520694e-07,
      "loss": -0.0,
      "num_tokens": 5554978.0,
      "reward": 0.7218749523162842,
      "reward_std": 0.0646936446428299,
      "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": 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": 0.145,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 20.082918167114258,
      "learning_rate": 4.258681761440789e-07,
      "loss": 0.0,
      "num_tokens": 5669026.0,
      "reward": 0.7374999523162842,
      "reward_std": 0.10888782143592834,
      "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": 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": 0.1475,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 48.38309097290039,
      "learning_rate": 4.095597328339452e-07,
      "loss": 0.0,
      "num_tokens": 5773642.0,
      "reward": 0.3468749523162842,
      "reward_std": 0.23144520819187164,
      "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": 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": 15.0,
      "completions/max_terminated_length": 15.0,
      "completions/mean_length": 6.171875,
      "completions/mean_terminated_length": 6.171875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 30.583539962768555,
      "learning_rate": 3.9335018462812664e-07,
      "loss": -0.0,
      "num_tokens": 5866133.0,
      "reward": 0.6429687142372131,
      "reward_std": 0.22608543932437897,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.59375,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 6.25,
      "completions/mean_terminated_length": 6.25,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1525,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 10.571331024169922,
      "learning_rate": 3.772572564296004e-07,
      "loss": -0.0,
      "num_tokens": 5976293.0,
      "reward": 0.46562498807907104,
      "reward_std": 0.1530819982290268,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 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": 0.155,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 12.751548767089844,
      "learning_rate": 3.612985456190778e-07,
      "loss": -0.0,
      "num_tokens": 6047373.0,
      "reward": 0.8156249523162842,
      "reward_std": 0.10205792635679245,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.765625,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 0.1575,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 23.879650115966797,
      "learning_rate": 3.454915028125263e-07,
      "loss": -0.0,
      "num_tokens": 6127165.0,
      "reward": 0.7062499523162842,
      "reward_std": 0.1872510462999344,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.65625,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 0.16,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 25.668092727661133,
      "learning_rate": 3.2985341277917846e-07,
      "loss": -0.0,
      "num_tokens": 6203197.0,
      "reward": 0.7687499523162842,
      "reward_std": 0.16675157845020294,
      "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": 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": 0.1625,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 30.97727394104004,
      "learning_rate": 3.1440137554088953e-07,
      "loss": -0.0,
      "num_tokens": 6302693.0,
      "reward": 0.40937498211860657,
      "reward_std": 0.2109457403421402,
      "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": 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": 0.165,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 25.07084846496582,
      "learning_rate": 2.9915228767351535e-07,
      "loss": 0.0,
      "num_tokens": 6397157.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 0.1675,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 13.29279613494873,
      "learning_rate": 2.841228238307536e-07,
      "loss": -0.0,
      "num_tokens": 6483597.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.11100947111845016,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 0.17,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 26.82172393798828,
      "learning_rate": 2.6932941851065615e-07,
      "loss": -0.0,
      "num_tokens": 6564629.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.19044625759124756,
      "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": 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": 0.1725,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 49.57731628417969,
      "learning_rate": 2.547882480847461e-07,
      "loss": -0.0,
      "num_tokens": 6658821.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 0.175,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 34.30208206176758,
      "learning_rate": 2.4051521310939254e-07,
      "loss": -0.0,
      "num_tokens": 6765957.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.30038216710090637,
      "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": 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": 0.1775,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 23.352293014526367,
      "learning_rate": 2.2652592093878665e-07,
      "loss": -0.0,
      "num_tokens": 6861477.0,
      "reward": 0.9249999523162842,
      "reward_std": 0.0883883386850357,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.875,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 0.18,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 28.74065589904785,
      "learning_rate": 2.128356686585282e-07,
      "loss": -0.0,
      "num_tokens": 6934549.0,
      "reward": 0.6749999523162842,
      "reward_std": 0.22461533546447754,
      "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": 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": 0.1825,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 43.618709564208984,
      "learning_rate": 1.9945942635848745e-07,
      "loss": 0.0,
      "num_tokens": 7015085.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.2688094973564148,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 60.0,
      "completions/max_terminated_length": 60.0,
      "completions/mean_length": 6.84375,
      "completions/mean_terminated_length": 6.84375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.185,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 16.129587173461914,
      "learning_rate": 1.8641182076323148e-07,
      "loss": 0.0,
      "num_tokens": 7087435.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.10888781398534775,
      "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": 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": 0.1875,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 16.217918395996094,
      "learning_rate": 1.7370711923791564e-07,
      "loss": 0.0,
      "num_tokens": 7179891.0,
      "reward": 0.6749999523162842,
      "reward_std": 0.0883883386850357,
      "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": 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": 89.0,
      "completions/max_terminated_length": 89.0,
      "completions/mean_length": 7.296875,
      "completions/mean_terminated_length": 7.296875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.6135921418712955e-07,
      "loss": 0.0,
      "num_tokens": 7247774.0,
      "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": 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": 0.1925,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 24.488853454589844,
      "learning_rate": 1.493816078637557e-07,
      "loss": -0.0,
      "num_tokens": 7332102.0,
      "reward": 0.7843749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.734375,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "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": 0.195,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.888399124145508,
      "learning_rate": 1.3778739760445552e-07,
      "loss": -0.0,
      "num_tokens": 7427646.0,
      "reward": 0.3937499523162842,
      "reward_std": 0.1462520956993103,
      "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": 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": 0.1975,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.708160400390625,
      "learning_rate": 1.2658926150792322e-07,
      "loss": -0.0,
      "num_tokens": 7564766.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.1872510462999344,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5625,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "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": 0.2,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 19.20774269104004,
      "learning_rate": 1.1579944457157059e-07,
      "loss": -0.0,
      "num_tokens": 7666758.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.10205791890621185,
      "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": 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": 0.2025,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.0542974530180327e-07,
      "loss": 0.0,
      "num_tokens": 7755734.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 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": 0.205,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 29.977798461914062,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.0,
      "num_tokens": 7847078.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.2109457403421402,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 0.2075,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 8.599558442598998e-08,
      "loss": 0.0,
      "num_tokens": 7921982.0,
      "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": 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": 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": 0.21,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.554065704345703,
      "learning_rate": 7.695237378953224e-08,
      "loss": 0.0,
      "num_tokens": 8005742.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.0883883386850357,
      "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": 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": 0.2125,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 6.837175952121304e-08,
      "loss": 0.0,
      "num_tokens": 8117214.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 0.215,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 6.026312439675551e-08,
      "loss": 0.0,
      "num_tokens": 8226846.0,
      "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": 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": 0.2175,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 33.922035217285156,
      "learning_rate": 5.263533508961826e-08,
      "loss": -0.0,
      "num_tokens": 8353046.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.10888781398534775,
      "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": 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": 0.22,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 28.928468704223633,
      "learning_rate": 4.549673247541874e-08,
      "loss": -0.0,
      "num_tokens": 8448966.0,
      "reward": 0.9249999523162842,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.875,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 0.2225,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 17.21563148498535,
      "learning_rate": 3.8855122511307626e-08,
      "loss": 0.0,
      "num_tokens": 8562926.0,
      "reward": 0.6593749523162842,
      "reward_std": 0.10205791890621185,
      "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": 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": 0.225,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 34.33055877685547,
      "learning_rate": 3.271776770026963e-08,
      "loss": -0.0,
      "num_tokens": 8623566.0,
      "reward": 0.7374999523162842,
      "reward_std": 0.15520364046096802,
      "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": 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": 809.0,
      "completions/max_terminated_length": 809.0,
      "completions/mean_length": 20.546875,
      "completions/mean_terminated_length": 20.546875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2275,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 28.42860984802246,
      "learning_rate": 2.7091379149682682e-08,
      "loss": -0.0,
      "num_tokens": 8709113.0,
      "reward": 0.17499998211860657,
      "reward_std": 0.2177756279706955,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 0.23,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 22.527969360351562,
      "learning_rate": 2.1982109232821176e-08,
      "loss": 0.0,
      "num_tokens": 8795273.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.04419416934251785,
      "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": 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": 0.2325,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.7395544861325718e-08,
      "loss": 0.0,
      "num_tokens": 8880033.0,
      "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": 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": 0.235,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 20.773046493530273,
      "learning_rate": 1.3336701375997127e-08,
      "loss": -0.0,
      "num_tokens": 8966049.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.1293872892856598,
      "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": 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": 0.2375,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 20.386754989624023,
      "learning_rate": 9.810017062595321e-09,
      "loss": -0.0,
      "num_tokens": 9067857.0,
      "reward": 0.8156249523162842,
      "reward_std": 0.1804211586713791,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.765625,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.24,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 13.194479942321777,
      "learning_rate": 6.819348298638839e-09,
      "loss": 0.0,
      "num_tokens": 9156654.0,
      "reward": 0.7999999523162842,
      "reward_std": 0.0883883386850357,
      "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": 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": 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": 0.2425,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 27.74026870727539,
      "learning_rate": 4.367965336512403e-09,
      "loss": -0.0,
      "num_tokens": 9253910.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.1462520956993103,
      "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": 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": 0.245,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 12.926735877990723,
      "learning_rate": 2.458548727494292e-09,
      "loss": -0.0,
      "num_tokens": 9340798.0,
      "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": 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": 0.2475,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 18.829639434814453,
      "learning_rate": 1.0931863906127325e-09,
      "loss": -0.0,
      "num_tokens": 9449230.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.17570312321186066,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 0.25,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 11.332636833190918,
      "learning_rate": 2.733713295369755e-10,
      "loss": -0.0,
      "num_tokens": 9537006.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.0578637532889843,
      "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": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 9537006,
  "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
}