{
  "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": 0.0,
      "completions/max_length": 1572.0,
      "completions/max_terminated_length": 1572.0,
      "completions/mean_length": 380.703125,
      "completions/mean_terminated_length": 380.703125,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.0025,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 10.104227066040039,
      "learning_rate": 0.0,
      "loss": -0.0,
      "num_tokens": 94445.0,
      "reward": 0.30937498807907104,
      "reward_std": 0.22894421219825745,
      "rewards/format_reward/mean": 0.28125,
      "rewards/format_reward/std": 0.3069951832294464,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 1411.0,
      "completions/max_terminated_length": 1411.0,
      "completions/mean_length": 386.203125,
      "completions/mean_terminated_length": 386.203125,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.005,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 23.277118682861328,
      "learning_rate": 2e-07,
      "loss": 0.0,
      "num_tokens": 233178.0,
      "reward": 0.35390621423721313,
      "reward_std": 0.3374568819999695,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.19012710452079773,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1254.0,
      "completions/mean_length": 383.484375,
      "completions/mean_terminated_length": 357.0635070800781,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0075,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 5.73276424407959,
      "learning_rate": 4e-07,
      "loss": -0.0,
      "num_tokens": 327625.0,
      "reward": 0.17109373211860657,
      "reward_std": 0.28433501720428467,
      "rewards/format_reward/mean": 0.3046875,
      "rewards/format_reward/std": 0.3036443889141083,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "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": 1591.0,
      "completions/max_terminated_length": 1591.0,
      "completions/mean_length": 422.296875,
      "completions/mean_terminated_length": 422.296875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.01,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 12.001652717590332,
      "learning_rate": 6e-07,
      "loss": -0.0,
      "num_tokens": 437396.0,
      "reward": 0.4289062023162842,
      "reward_std": 0.40633565187454224,
      "rewards/format_reward/mean": 0.3828125,
      "rewards/format_reward/std": 0.2478829026222229,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 0.0,
      "completions/max_length": 1573.0,
      "completions/max_terminated_length": 1573.0,
      "completions/mean_length": 411.375,
      "completions/mean_terminated_length": 411.375,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.0125,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 34.90378952026367,
      "learning_rate": 8e-07,
      "loss": 0.0,
      "num_tokens": 559724.0,
      "reward": 0.2914062440395355,
      "reward_std": 0.32816699147224426,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.28942593932151794,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1824.0,
      "completions/mean_length": 664.4375,
      "completions/mean_terminated_length": 619.8064575195312,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.015,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 5.146732807159424,
      "learning_rate": 1e-06,
      "loss": 0.0,
      "num_tokens": 695512.0,
      "reward": 0.23984374105930328,
      "reward_std": 0.25555554032325745,
      "rewards/format_reward/mean": 0.3671875,
      "rewards/format_reward/std": 0.23974503576755524,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "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": 0.0625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1313.0,
      "completions/mean_length": 495.28125,
      "completions/mean_terminated_length": 391.7666931152344,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0175,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 13.87881088256836,
      "learning_rate": 9.99726628670463e-07,
      "loss": -0.0,
      "num_tokens": 839354.0,
      "reward": 0.3804687261581421,
      "reward_std": 0.4405275881290436,
      "rewards/format_reward/mean": 0.3671875,
      "rewards/format_reward/std": 0.255761981010437,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 1038.0,
      "completions/max_terminated_length": 1038.0,
      "completions/mean_length": 283.828125,
      "completions/mean_terminated_length": 283.828125,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.02,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 16.57680320739746,
      "learning_rate": 9.989068136093872e-07,
      "loss": -0.0,
      "num_tokens": 941223.0,
      "reward": 0.2601562440395355,
      "reward_std": 0.3872828483581543,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.27537402510643005,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1446.0,
      "completions/mean_length": 398.59375,
      "completions/mean_terminated_length": 345.3870849609375,
      "completions/min_length": 3.0,
      "completions/min_terminated_length": 3.0,
      "epoch": 0.0225,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 27.8460693359375,
      "learning_rate": 9.975414512725056e-07,
      "loss": 0.0,
      "num_tokens": 1046957.0,
      "reward": 0.3374999761581421,
      "reward_std": 0.31849682331085205,
      "rewards/format_reward/mean": 0.40625,
      "rewards/format_reward/std": 0.29378482699394226,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 9
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1777.0,
      "completions/mean_length": 362.65625,
      "completions/mean_terminated_length": 335.90478515625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 9.661582946777344,
      "learning_rate": 9.956320346634875e-07,
      "loss": -0.0,
      "num_tokens": 1172647.0,
      "reward": 0.3343749940395355,
      "reward_std": 0.3494775593280792,
      "rewards/format_reward/mean": 0.375,
      "rewards/format_reward/std": 0.26726123690605164,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 834.0,
      "completions/mean_length": 373.828125,
      "completions/mean_terminated_length": 319.82257080078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0275,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 11.05769157409668,
      "learning_rate": 9.931806517013612e-07,
      "loss": 0.0,
      "num_tokens": 1282204.0,
      "reward": 0.15625,
      "reward_std": 0.2576306164264679,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.279951810836792,
      "rewards/mcq_exact_match_reward/mean": 0.109375,
      "rewards/mcq_exact_match_reward/std": 0.3145764470100403,
      "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": 1726.0,
      "completions/max_terminated_length": 1726.0,
      "completions/mean_length": 401.875,
      "completions/mean_terminated_length": 401.875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.03,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 6.454217910766602,
      "learning_rate": 9.901899829374047e-07,
      "loss": -0.0,
      "num_tokens": 1398452.0,
      "reward": 0.3242187201976776,
      "reward_std": 0.4082157611846924,
      "rewards/format_reward/mean": 0.4296875,
      "rewards/format_reward/std": 0.23345555365085602,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 925.0,
      "completions/max_terminated_length": 925.0,
      "completions/mean_length": 246.75,
      "completions/mean_terminated_length": 246.75,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0325,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 6.135255813598633,
      "learning_rate": 9.866632986240029e-07,
      "loss": -0.0,
      "num_tokens": 1474604.0,
      "reward": 0.4921874403953552,
      "reward_std": 0.30105772614479065,
      "rewards/format_reward/mean": 0.546875,
      "rewards/format_reward/std": 0.2777281701564789,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1591.0,
      "completions/mean_length": 239.3125,
      "completions/mean_terminated_length": 210.60317993164062,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.035,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 13.646222114562988,
      "learning_rate": 9.826044551386742e-07,
      "loss": 0.0,
      "num_tokens": 1565424.0,
      "reward": 0.15312500298023224,
      "reward_std": 0.20769444108009338,
      "rewards/format_reward/mean": 0.59375,
      "rewards/format_reward/std": 0.279951810836792,
      "rewards/mcq_exact_match_reward/mean": 0.09375,
      "rewards/mcq_exact_match_reward/std": 0.29378482699394226,
      "step": 14
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1182.0,
      "completions/mean_length": 238.71875,
      "completions/mean_terminated_length": 210.00001525878906,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0375,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 10.635026931762695,
      "learning_rate": 9.780178907671788e-07,
      "loss": -0.0,
      "num_tokens": 1671630.0,
      "reward": 0.5648436546325684,
      "reward_std": 0.23381614685058594,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.16648055613040924,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 875.0,
      "completions/max_terminated_length": 875.0,
      "completions/mean_length": 64.015625,
      "completions/mean_terminated_length": 64.015625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.04,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 17.595975875854492,
      "learning_rate": 9.729086208503173e-07,
      "loss": -0.0,
      "num_tokens": 1763119.0,
      "reward": 0.547656238079071,
      "reward_std": 0.3905794620513916,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 475.0,
      "completions/max_terminated_length": 475.0,
      "completions/mean_length": 43.203125,
      "completions/mean_terminated_length": 43.203125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0425,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 13.388673782348633,
      "learning_rate": 9.672822322997304e-07,
      "loss": 0.0,
      "num_tokens": 1849220.0,
      "reward": 0.5968749523162842,
      "reward_std": 0.1720314621925354,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.1259881556034088,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 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": 0.045,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 16.634624481201172,
      "learning_rate": 9.611448774886923e-07,
      "loss": 0.0,
      "num_tokens": 1954380.0,
      "reward": 0.4718749523162842,
      "reward_std": 0.1530819833278656,
      "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": 106.0,
      "completions/max_terminated_length": 106.0,
      "completions/mean_length": 7.5625,
      "completions/mean_terminated_length": 7.5625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0475,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.946528434753418,
      "learning_rate": 9.545032675245813e-07,
      "loss": -0.0,
      "num_tokens": 2065680.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.10888782143592834,
      "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": 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": 13.0,
      "completions/max_terminated_length": 13.0,
      "completions/mean_length": 6.609375,
      "completions/mean_terminated_length": 6.609375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 30.572359085083008,
      "learning_rate": 9.473646649103817e-07,
      "loss": 0.0,
      "num_tokens": 2144071.0,
      "reward": 0.4921874701976776,
      "reward_std": 0.16964474320411682,
      "rewards/format_reward/mean": 0.546875,
      "rewards/format_reward/std": 0.14689241349697113,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "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": 289.0,
      "completions/max_terminated_length": 289.0,
      "completions/mean_length": 14.015625,
      "completions/mean_terminated_length": 14.015625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0525,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 20.962142944335938,
      "learning_rate": 9.397368756032444e-07,
      "loss": 0.0,
      "num_tokens": 2218648.0,
      "reward": 0.34062498807907104,
      "reward_std": 0.12822021543979645,
      "rewards/format_reward/mean": 0.59375,
      "rewards/format_reward/std": 0.19669894874095917,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 1013.0,
      "completions/max_terminated_length": 1013.0,
      "completions/mean_length": 22.03125,
      "completions/mean_terminated_length": 22.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.055,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 28.532514572143555,
      "learning_rate": 9.316282404787869e-07,
      "loss": -0.0,
      "num_tokens": 2301802.0,
      "reward": 0.7242187261581421,
      "reward_std": 0.12579208612442017,
      "rewards/format_reward/mean": 0.5234375,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.671875,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 106.0,
      "completions/max_terminated_length": 106.0,
      "completions/mean_length": 7.5625,
      "completions/mean_terminated_length": 7.5625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0575,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 23.159854888916016,
      "learning_rate": 9.230476262104676e-07,
      "loss": -0.0,
      "num_tokens": 2392438.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.15520364046096802,
      "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": 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": 779.0,
      "completions/max_terminated_length": 779.0,
      "completions/mean_length": 19.4375,
      "completions/mean_terminated_length": 19.4375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.06,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 14.921491622924805,
      "learning_rate": 9.1400441557401e-07,
      "loss": 0.0,
      "num_tokens": 2473554.0,
      "reward": 0.5914062261581421,
      "reward_std": 0.17606157064437866,
      "rewards/format_reward/mean": 0.6015625,
      "rewards/format_reward/std": 0.20275264978408813,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 460.0,
      "completions/max_terminated_length": 460.0,
      "completions/mean_length": 14.15625,
      "completions/mean_terminated_length": 14.15625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0625,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 31.79290771484375,
      "learning_rate": 9.045084971874737e-07,
      "loss": -0.0,
      "num_tokens": 2573596.0,
      "reward": 0.26953125,
      "reward_std": 0.20632556080818176,
      "rewards/format_reward/mean": 0.5078125,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "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": 495.0,
      "completions/max_terminated_length": 495.0,
      "completions/mean_length": 16.96875,
      "completions/mean_terminated_length": 16.96875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.065,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 23.660274505615234,
      "learning_rate": 8.945702546981968e-07,
      "loss": -0.0,
      "num_tokens": 2642178.0,
      "reward": 0.5679687261581421,
      "reward_std": 0.3241649270057678,
      "rewards/format_reward/mean": 0.6796875,
      "rewards/format_reward/std": 0.24180518090724945,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 170.0,
      "completions/max_terminated_length": 170.0,
      "completions/mean_length": 10.84375,
      "completions/mean_terminated_length": 10.84375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0675,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 21.55986976623535,
      "learning_rate": 8.842005554284295e-07,
      "loss": -0.0,
      "num_tokens": 2758224.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.2109457403421402,
      "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": 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": 14.0,
      "completions/max_terminated_length": 14.0,
      "completions/mean_length": 7.140625,
      "completions/mean_terminated_length": 7.140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.07,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 23.411895751953125,
      "learning_rate": 8.734107384920769e-07,
      "loss": 0.0,
      "num_tokens": 2833361.0,
      "reward": 0.63671875,
      "reward_std": 0.19149912893772125,
      "rewards/format_reward/mean": 0.5859375,
      "rewards/format_reward/std": 0.19012710452079773,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 141.0,
      "completions/max_terminated_length": 141.0,
      "completions/mean_length": 8.859375,
      "completions/mean_terminated_length": 8.859375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0725,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 4.729227542877197,
      "learning_rate": 8.622126023955445e-07,
      "loss": 0.0,
      "num_tokens": 2922176.0,
      "reward": 0.5406249761581421,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5625,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 13.0,
      "completions/max_terminated_length": 13.0,
      "completions/mean_length": 8.28125,
      "completions/mean_terminated_length": 8.28125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.075,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.8333740234375,
      "learning_rate": 8.506183921362442e-07,
      "loss": 0.0,
      "num_tokens": 2992354.0,
      "reward": 0.5687500238418579,
      "reward_std": 0.1293872892856598,
      "rewards/format_reward/mean": 0.6875,
      "rewards/format_reward/std": 0.24397502839565277,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 12.0,
      "completions/max_terminated_length": 12.0,
      "completions/mean_length": 8.25,
      "completions/mean_terminated_length": 8.25,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0775,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 14.442317962646484,
      "learning_rate": 8.386407858128706e-07,
      "loss": -0.0,
      "num_tokens": 3080066.0,
      "reward": 0.3500000238418579,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.6875,
      "rewards/format_reward/std": 0.24397502839565277,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 13.0,
      "completions/max_terminated_length": 13.0,
      "completions/mean_length": 8.265625,
      "completions/mean_terminated_length": 8.265625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 29.77044105529785,
      "learning_rate": 8.262928807620843e-07,
      "loss": -0.0,
      "num_tokens": 3154547.0,
      "reward": 0.5218750238418579,
      "reward_std": 0.25726157426834106,
      "rewards/format_reward/mean": 0.6875,
      "rewards/format_reward/std": 0.24397502839565277,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 81.0,
      "completions/max_terminated_length": 81.0,
      "completions/mean_length": 10.203125,
      "completions/mean_terminated_length": 10.203125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0825,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 15.08368968963623,
      "learning_rate": 8.135881792367685e-07,
      "loss": -0.0,
      "num_tokens": 3250384.0,
      "reward": 0.23515623807907104,
      "reward_std": 0.13479220867156982,
      "rewards/format_reward/mean": 0.6328125,
      "rewards/format_reward/std": 0.22257846593856812,
      "rewards/mcq_exact_match_reward/mean": 0.171875,
      "rewards/mcq_exact_match_reward/std": 0.38025420904159546,
      "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": 493.0,
      "completions/max_terminated_length": 493.0,
      "completions/mean_length": 15.1875,
      "completions/mean_terminated_length": 15.1875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.085,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 18.34843635559082,
      "learning_rate": 8.005405736415125e-07,
      "loss": -0.0,
      "num_tokens": 3332732.0,
      "reward": 0.624218761920929,
      "reward_std": 0.1555730253458023,
      "rewards/format_reward/mean": 0.6171875,
      "rewards/format_reward/std": 0.2313210517168045,
      "rewards/mcq_exact_match_reward/mean": 0.5625,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 34
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 152.0,
      "completions/max_terminated_length": 152.0,
      "completions/mean_length": 10.4375,
      "completions/mean_terminated_length": 10.4375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0875,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.837115287780762,
      "learning_rate": 7.871643313414718e-07,
      "loss": 0.0,
      "num_tokens": 3417584.0,
      "reward": 0.7875000238418579,
      "reward_std": 0.0578637570142746,
      "rewards/format_reward/mean": 0.6875,
      "rewards/format_reward/std": 0.24397502839565277,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 18.53125,
      "completions/mean_terminated_length": 18.53125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 22.8250732421875,
      "learning_rate": 7.734740790612136e-07,
      "loss": -0.0,
      "num_tokens": 3503610.0,
      "reward": 0.484375,
      "reward_std": 0.30563658475875854,
      "rewards/format_reward/mean": 0.78125,
      "rewards/format_reward/std": 0.25,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 13.0,
      "completions/max_terminated_length": 13.0,
      "completions/mean_length": 10.515625,
      "completions/mean_terminated_length": 10.515625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0925,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 10.62349796295166,
      "learning_rate": 7.594847868906076e-07,
      "loss": 0.0,
      "num_tokens": 3555275.0,
      "reward": 0.4937500059604645,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.875,
      "rewards/format_reward/std": 0.2182178944349289,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 12.0,
      "completions/max_terminated_length": 12.0,
      "completions/mean_length": 7.59375,
      "completions/mean_terminated_length": 7.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.095,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.927433013916016,
      "learning_rate": 7.452117519152541e-07,
      "loss": 0.0,
      "num_tokens": 3639065.0,
      "reward": 0.547656238079071,
      "reward_std": 0.13371865451335907,
      "rewards/format_reward/mean": 0.6328125,
      "rewards/format_reward/std": 0.22257846593856812,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 13.0,
      "completions/max_terminated_length": 13.0,
      "completions/mean_length": 9.078125,
      "completions/mean_terminated_length": 9.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0975,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 76.84066009521484,
      "learning_rate": 7.306705814893439e-07,
      "loss": 0.0,
      "num_tokens": 3728342.0,
      "reward": 0.41796875,
      "reward_std": 0.11669307202100754,
      "rewards/format_reward/mean": 0.7421875,
      "rewards/format_reward/std": 0.25185325741767883,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 76.0,
      "completions/max_terminated_length": 76.0,
      "completions/mean_length": 11.578125,
      "completions/mean_terminated_length": 11.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 60.49788284301758,
      "learning_rate": 7.158771761692464e-07,
      "loss": 0.0,
      "num_tokens": 3825723.0,
      "reward": 0.5617187023162842,
      "reward_std": 0.16234643757343292,
      "rewards/format_reward/mean": 0.7734375,
      "rewards/format_reward/std": 0.250866562128067,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 407.0,
      "completions/max_terminated_length": 407.0,
      "completions/mean_length": 18.890625,
      "completions/mean_terminated_length": 18.890625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1025,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 26.07494354248047,
      "learning_rate": 7.008477123264847e-07,
      "loss": -0.0,
      "num_tokens": 3907724.0,
      "reward": 0.35859376192092896,
      "reward_std": 0.13479222357273102,
      "rewards/format_reward/mean": 0.9296875,
      "rewards/format_reward/std": 0.1751912236213684,
      "rewards/mcq_exact_match_reward/mean": 0.265625,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "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": 182.0,
      "completions/max_terminated_length": 182.0,
      "completions/mean_length": 13.84375,
      "completions/mean_terminated_length": 13.84375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.105,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 57.546485900878906,
      "learning_rate": 6.855986244591103e-07,
      "loss": -0.0,
      "num_tokens": 4013906.0,
      "reward": 0.4281249940395355,
      "reward_std": 0.06670258939266205,
      "rewards/format_reward/mean": 0.84375,
      "rewards/format_reward/std": 0.25,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 63.0,
      "completions/max_terminated_length": 63.0,
      "completions/mean_length": 11.46875,
      "completions/mean_terminated_length": 11.46875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1075,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 23.997573852539062,
      "learning_rate": 6.701465872208216e-07,
      "loss": 0.0,
      "num_tokens": 4101112.0,
      "reward": 0.9273437261581421,
      "reward_std": 0.23680157959461212,
      "rewards/format_reward/mean": 0.8359375,
      "rewards/format_reward/std": 0.2366211861371994,
      "rewards/mcq_exact_match_reward/mean": 0.84375,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "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": 150.0,
      "completions/max_terminated_length": 150.0,
      "completions/mean_length": 14.234375,
      "completions/mean_terminated_length": 14.234375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 46.40782165527344,
      "learning_rate": 6.545084971874736e-07,
      "loss": 0.0,
      "num_tokens": 4193311.0,
      "reward": 0.7109375,
      "reward_std": 0.18755537271499634,
      "rewards/format_reward/mean": 0.859375,
      "rewards/format_reward/std": 0.2592533528804779,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 44
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 15.765625,
      "completions/mean_terminated_length": 15.765625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1125,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 21.46227264404297,
      "learning_rate": 6.387014543809223e-07,
      "loss": -0.0,
      "num_tokens": 4272336.0,
      "reward": 0.6179687976837158,
      "reward_std": 0.28839826583862305,
      "rewards/format_reward/mean": 0.8671875,
      "rewards/format_reward/std": 0.23974503576755524,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 76.0,
      "completions/max_terminated_length": 76.0,
      "completions/mean_length": 13.484375,
      "completions/mean_terminated_length": 13.484375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.115,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 21.336334228515625,
      "learning_rate": 6.227427435703995e-07,
      "loss": 0.0,
      "num_tokens": 4352943.0,
      "reward": 0.9437500238418579,
      "reward_std": 0.0578637570142746,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.84375,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "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": 49.0,
      "completions/max_terminated_length": 49.0,
      "completions/mean_length": 13.296875,
      "completions/mean_terminated_length": 13.296875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1175,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 18.461605072021484,
      "learning_rate": 6.066498153718734e-07,
      "loss": 0.0,
      "num_tokens": 4449594.0,
      "reward": 0.3656250238418579,
      "reward_std": 0.17358146607875824,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.265625,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "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": 364.0,
      "completions/max_terminated_length": 364.0,
      "completions/mean_length": 23.125,
      "completions/mean_terminated_length": 23.125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.12,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.556089401245117,
      "learning_rate": 5.90440267166055e-07,
      "loss": 0.0,
      "num_tokens": 4529738.0,
      "reward": 0.5843750238418579,
      "reward_std": 0.17358146607875824,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 71.0,
      "completions/max_terminated_length": 71.0,
      "completions/mean_length": 13.3125,
      "completions/mean_terminated_length": 13.3125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1225,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 27.238489151000977,
      "learning_rate": 5.741318238559209e-07,
      "loss": 0.0,
      "num_tokens": 4629662.0,
      "reward": 0.7250000238418579,
      "reward_std": 0.2177756428718567,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "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": 364.0,
      "completions/max_terminated_length": 364.0,
      "completions/mean_length": 22.796875,
      "completions/mean_terminated_length": 22.796875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.125,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 23.85200309753418,
      "learning_rate": 5.577423184847931e-07,
      "loss": -0.0,
      "num_tokens": 4705673.0,
      "reward": 0.6781250238418579,
      "reward_std": 0.11100947856903076,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 472.0,
      "completions/max_terminated_length": 472.0,
      "completions/mean_length": 31.234375,
      "completions/mean_terminated_length": 31.234375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1275,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 3.724616765975952,
      "learning_rate": 5.412896727361662e-07,
      "loss": 0.0,
      "num_tokens": 4806072.0,
      "reward": 0.6468750238418579,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 23.0,
      "completions/max_terminated_length": 23.0,
      "completions/mean_length": 12.1875,
      "completions/mean_terminated_length": 12.1875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.13,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 10.997976303100586,
      "learning_rate": 5.247918773366111e-07,
      "loss": 0.0,
      "num_tokens": 4905964.0,
      "reward": 0.9109375476837158,
      "reward_std": 0.11115548014640808,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.8125,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 137.0,
      "completions/max_terminated_length": 137.0,
      "completions/mean_length": 14.171875,
      "completions/mean_terminated_length": 14.171875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1325,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.619414806365967,
      "learning_rate": 5.082669723831793e-07,
      "loss": 0.0,
      "num_tokens": 4979159.0,
      "reward": 0.7562500238418579,
      "reward_std": 0.0578637570142746,
      "rewards/format_reward/mean": 1.0,
      "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": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 17.9375,
      "completions/mean_terminated_length": 17.9375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.135,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 4.455435276031494,
      "learning_rate": 4.917330276168208e-07,
      "loss": -0.0,
      "num_tokens": 5076491.0,
      "reward": 0.5062500238418579,
      "reward_std": 0.0883883461356163,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 16.0,
      "completions/max_terminated_length": 16.0,
      "completions/mean_length": 12.09375,
      "completions/mean_terminated_length": 12.09375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1375,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 26.964513778686523,
      "learning_rate": 4.752081226633888e-07,
      "loss": 0.0,
      "num_tokens": 5173785.0,
      "reward": 0.5203125476837158,
      "reward_std": 0.20011675357818604,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.09375,
      "completions/mean_terminated_length": 12.09375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.14,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 15.034180641174316,
      "learning_rate": 4.5871032726383385e-07,
      "loss": 0.0,
      "num_tokens": 5275183.0,
      "reward": 0.5531250238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 176.0,
      "completions/max_terminated_length": 176.0,
      "completions/mean_length": 14.5625,
      "completions/mean_terminated_length": 14.5625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1425,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 17.728288650512695,
      "learning_rate": 4.4225768151520694e-07,
      "loss": 0.0,
      "num_tokens": 5384403.0,
      "reward": 0.7875000238418579,
      "reward_std": 0.1157275140285492,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.6875,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 12.0,
      "completions/max_terminated_length": 12.0,
      "completions/mean_length": 12.0,
      "completions/mean_terminated_length": 12.0,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.145,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 6.035715579986572,
      "learning_rate": 4.258681761440789e-07,
      "loss": 0.0,
      "num_tokens": 5498259.0,
      "reward": 0.6156250238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 59.0,
      "completions/max_terminated_length": 59.0,
      "completions/mean_length": 13.390625,
      "completions/mean_terminated_length": 13.390625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1475,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 10.740617752075195,
      "learning_rate": 4.095597328339452e-07,
      "loss": 0.0,
      "num_tokens": 5615092.0,
      "reward": 0.6937500238418579,
      "reward_std": 0.1462521106004715,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.59375,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 59
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 589.0,
      "completions/max_terminated_length": 589.0,
      "completions/mean_length": 24.546875,
      "completions/mean_terminated_length": 24.546875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.15,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 13.927393913269043,
      "learning_rate": 3.9335018462812664e-07,
      "loss": 0.0,
      "num_tokens": 5702087.0,
      "reward": 0.48906251788139343,
      "reward_std": 0.19486570358276367,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.09375,
      "completions/mean_terminated_length": 12.09375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1525,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.805014610290527,
      "learning_rate": 3.772572564296004e-07,
      "loss": 0.0,
      "num_tokens": 5812349.0,
      "reward": 0.9125000238418579,
      "reward_std": 0.06681530922651291,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.8125,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 792.0,
      "completions/max_terminated_length": 792.0,
      "completions/mean_length": 29.6875,
      "completions/mean_terminated_length": 29.6875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.155,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 13.502047538757324,
      "learning_rate": 3.612985456190778e-07,
      "loss": 0.0,
      "num_tokens": 5922217.0,
      "reward": 0.5515625476837158,
      "reward_std": 0.06911306083202362,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 24.234375,
      "completions/mean_terminated_length": 24.234375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1575,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.17144775390625,
      "learning_rate": 3.454915028125263e-07,
      "loss": -0.0,
      "num_tokens": 6028640.0,
      "reward": 0.2406250238418579,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 542.0,
      "completions/mean_length": 66.921875,
      "completions/mean_terminated_length": 35.476192474365234,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.16,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 15.114312171936035,
      "learning_rate": 3.2985341277917846e-07,
      "loss": 0.0,
      "num_tokens": 6116459.0,
      "reward": 0.6421874761581421,
      "reward_std": 0.13276319205760956,
      "rewards/format_reward/mean": 0.953125,
      "rewards/format_reward/std": 0.21304203569889069,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 607.0,
      "completions/mean_length": 87.109375,
      "completions/mean_terminated_length": 55.984130859375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1625,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 4.215585231781006,
      "learning_rate": 3.1440137554088953e-07,
      "loss": 0.0,
      "num_tokens": 6218698.0,
      "reward": 0.581250011920929,
      "reward_std": 0.0530330091714859,
      "rewards/format_reward/mean": 0.96875,
      "rewards/format_reward/std": 0.17536810040473938,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 65
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 441.0,
      "completions/max_terminated_length": 441.0,
      "completions/mean_length": 29.015625,
      "completions/mean_terminated_length": 29.015625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.165,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 5.03313684463501,
      "learning_rate": 2.9915228767351535e-07,
      "loss": -0.0,
      "num_tokens": 6326291.0,
      "reward": 0.49687501788139343,
      "reward_std": 0.06365013122558594,
      "rewards/format_reward/mean": 0.90625,
      "rewards/format_reward/std": 0.29378482699394226,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1080.0,
      "completions/mean_length": 127.203125,
      "completions/mean_terminated_length": 96.71429443359375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1675,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 22.036781311035156,
      "learning_rate": 2.841228238307536e-07,
      "loss": 0.0,
      "num_tokens": 6428872.0,
      "reward": 0.47031253576278687,
      "reward_std": 0.23183076083660126,
      "rewards/format_reward/mean": 0.953125,
      "rewards/format_reward/std": 0.21304203569889069,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 67
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 581.0,
      "completions/mean_length": 124.734375,
      "completions/mean_terminated_length": 62.693546295166016,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.17,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 14.406416893005371,
      "learning_rate": 2.6932941851065615e-07,
      "loss": 0.0,
      "num_tokens": 6545983.0,
      "reward": 0.33125001192092896,
      "reward_std": 0.20466148853302002,
      "rewards/format_reward/mean": 0.96875,
      "rewards/format_reward/std": 0.17536810040473938,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 340.0,
      "completions/max_terminated_length": 340.0,
      "completions/mean_length": 23.234375,
      "completions/mean_terminated_length": 23.234375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1725,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.32234001159668,
      "learning_rate": 2.547882480847461e-07,
      "loss": 0.0,
      "num_tokens": 6640942.0,
      "reward": 0.7234375476837158,
      "reward_std": 0.16901923716068268,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "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": 423.0,
      "completions/max_terminated_length": 423.0,
      "completions/mean_length": 28.859375,
      "completions/mean_terminated_length": 28.859375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.175,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 14.567350387573242,
      "learning_rate": 2.4051521310939254e-07,
      "loss": 0.0,
      "num_tokens": 6730517.0,
      "reward": 0.6781250238418579,
      "reward_std": 0.2662131190299988,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 631.0,
      "completions/max_terminated_length": 631.0,
      "completions/mean_length": 48.40625,
      "completions/mean_terminated_length": 48.40625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1775,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 6.351161479949951,
      "learning_rate": 2.2652592093878665e-07,
      "loss": 0.0,
      "num_tokens": 6808295.0,
      "reward": 0.5203125476837158,
      "reward_std": 0.12697681784629822,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 71
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 48.765625,
      "completions/mean_terminated_length": 17.031747817993164,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.18,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 10.247476577758789,
      "learning_rate": 2.128356686585282e-07,
      "loss": 0.0,
      "num_tokens": 6909392.0,
      "reward": 0.45781251788139343,
      "reward_std": 0.04861358925700188,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "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": 151.0,
      "completions/max_terminated_length": 151.0,
      "completions/mean_length": 17.765625,
      "completions/mean_terminated_length": 17.765625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1825,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 24.003049850463867,
      "learning_rate": 1.9945942635848745e-07,
      "loss": 0.0,
      "num_tokens": 6996273.0,
      "reward": 0.706250011920929,
      "reward_std": 0.2714768052101135,
      "rewards/format_reward/mean": 0.96875,
      "rewards/format_reward/std": 0.17536810040473938,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 73
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 576.0,
      "completions/max_terminated_length": 576.0,
      "completions/mean_length": 24.703125,
      "completions/mean_terminated_length": 24.703125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.185,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 3.3951778411865234,
      "learning_rate": 1.8641182076323148e-07,
      "loss": 0.0,
      "num_tokens": 7073766.0,
      "reward": 0.7093750238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 296.0,
      "completions/max_terminated_length": 296.0,
      "completions/mean_length": 19.125,
      "completions/mean_terminated_length": 19.125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1875,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 22.806726455688477,
      "learning_rate": 1.7370711923791564e-07,
      "loss": 0.0,
      "num_tokens": 7168574.0,
      "reward": 0.5359375476837158,
      "reward_std": 0.11684979498386383,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "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": 660.0,
      "completions/max_terminated_length": 660.0,
      "completions/mean_length": 41.375,
      "completions/mean_terminated_length": 41.375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.19,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 8.479246139526367,
      "learning_rate": 1.6135921418712955e-07,
      "loss": 0.0,
      "num_tokens": 7251846.0,
      "reward": 0.7093750238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 615.0,
      "completions/max_terminated_length": 615.0,
      "completions/mean_length": 25.078125,
      "completions/mean_terminated_length": 25.078125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1925,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.47300148010254,
      "learning_rate": 1.493816078637557e-07,
      "loss": 0.0,
      "num_tokens": 7344515.0,
      "reward": 0.7203125357627869,
      "reward_std": 0.1732209473848343,
      "rewards/format_reward/mean": 0.953125,
      "rewards/format_reward/std": 0.21304203569889069,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "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": 69.0,
      "completions/max_terminated_length": 69.0,
      "completions/mean_length": 14.015625,
      "completions/mean_terminated_length": 14.015625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.195,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.3778739760445552e-07,
      "loss": 0.0,
      "num_tokens": 7424036.0,
      "reward": 0.6000000238418579,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 970.0,
      "completions/max_terminated_length": 970.0,
      "completions/mean_length": 52.859375,
      "completions/mean_terminated_length": 52.859375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.1975,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 16.67440414428711,
      "learning_rate": 1.2658926150792322e-07,
      "loss": 0.0,
      "num_tokens": 7516467.0,
      "reward": 0.6000000238418579,
      "reward_std": 0.2346404492855072,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 437.0,
      "completions/max_terminated_length": 437.0,
      "completions/mean_length": 27.078125,
      "completions/mean_terminated_length": 27.078125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 18.662948608398438,
      "learning_rate": 1.1579944457157059e-07,
      "loss": 0.0,
      "num_tokens": 7620184.0,
      "reward": 0.5218750238418579,
      "reward_std": 0.1530819982290268,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 85.828125,
      "completions/mean_terminated_length": 22.532257080078125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2025,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 14.317560195922852,
      "learning_rate": 1.0542974530180327e-07,
      "loss": 0.0,
      "num_tokens": 7737981.0,
      "reward": 0.753125011920929,
      "reward_std": 0.1823902130126953,
      "rewards/format_reward/mean": 0.96875,
      "rewards/format_reward/std": 0.17536810040473938,
      "rewards/mcq_exact_match_reward/mean": 0.65625,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "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": 31.0,
      "completions/max_terminated_length": 31.0,
      "completions/mean_length": 13.453125,
      "completions/mean_terminated_length": 13.453125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.205,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 21.513164520263672,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.0,
      "num_tokens": 7824730.0,
      "reward": 0.625,
      "reward_std": 0.20288412272930145,
      "rewards/format_reward/mean": 0.9375,
      "rewards/format_reward/std": 0.24397502839565277,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 19.0,
      "completions/max_terminated_length": 19.0,
      "completions/mean_length": 12.671875,
      "completions/mean_terminated_length": 12.671875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2075,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 15.321270942687988,
      "learning_rate": 8.599558442598998e-08,
      "loss": 0.0,
      "num_tokens": 7898629.0,
      "reward": 0.5984375476837158,
      "reward_std": 0.09280776977539062,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.546875,
      "completions/mean_terminated_length": 12.546875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.21,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 8.163082122802734,
      "learning_rate": 7.695237378953224e-08,
      "loss": 0.0,
      "num_tokens": 8008472.0,
      "reward": 0.5843750238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 76.0,
      "completions/max_terminated_length": 76.0,
      "completions/mean_length": 14.90625,
      "completions/mean_terminated_length": 14.90625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2125,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 6.837175952121304e-08,
      "loss": 0.0,
      "num_tokens": 8133890.0,
      "reward": 0.3500000238418579,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "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": 604.0,
      "completions/max_terminated_length": 604.0,
      "completions/mean_length": 21.921875,
      "completions/mean_terminated_length": 21.921875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.215,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 12.233819961547852,
      "learning_rate": 6.026312439675551e-08,
      "loss": 0.0,
      "num_tokens": 8204829.0,
      "reward": 0.6156250238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 56.0,
      "completions/mean_length": 45.875,
      "completions/mean_terminated_length": 14.09523868560791,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2175,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 12.981520652770996,
      "learning_rate": 5.263533508961826e-08,
      "loss": 0.0,
      "num_tokens": 8284245.0,
      "reward": 0.6453125476837158,
      "reward_std": 0.16434110701084137,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.15625,
      "completions/mean_terminated_length": 12.15625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.22,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 30.7071533203125,
      "learning_rate": 4.549673247541874e-08,
      "loss": 0.0,
      "num_tokens": 8375807.0,
      "reward": 0.8187500238418579,
      "reward_std": 0.0578637570142746,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 181.0,
      "completions/max_terminated_length": 181.0,
      "completions/mean_length": 14.828125,
      "completions/mean_terminated_length": 14.828125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2225,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 18.151525497436523,
      "learning_rate": 3.8855122511307626e-08,
      "loss": 0.0,
      "num_tokens": 8478940.0,
      "reward": 0.6140625476837158,
      "reward_std": 0.19486570358276367,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 21.0,
      "completions/max_terminated_length": 21.0,
      "completions/mean_length": 12.328125,
      "completions/mean_terminated_length": 12.328125,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.225,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 15.212957382202148,
      "learning_rate": 3.271776770026963e-08,
      "loss": -0.0,
      "num_tokens": 8557033.0,
      "reward": 0.6468750238418579,
      "reward_std": 0.11100947856903076,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 140.0,
      "completions/max_terminated_length": 140.0,
      "completions/mean_length": 14.0,
      "completions/mean_terminated_length": 14.0,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2275,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 6.136690616607666,
      "learning_rate": 2.7091379149682682e-08,
      "loss": -0.0,
      "num_tokens": 8619057.0,
      "reward": 0.6156250238418579,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 64.0,
      "completions/max_terminated_length": 64.0,
      "completions/mean_length": 14.234375,
      "completions/mean_terminated_length": 14.234375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.23,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.1883344650268555,
      "learning_rate": 2.1982109232821176e-08,
      "loss": 0.0,
      "num_tokens": 8696280.0,
      "reward": 0.5984375476837158,
      "reward_std": 0.0044194171205163,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "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": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.09375,
      "completions/mean_terminated_length": 12.09375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2325,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.7395544861325718e-08,
      "loss": 0.0,
      "num_tokens": 8774814.0,
      "reward": 0.8500000238418579,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.75,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "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": 12.0,
      "completions/max_terminated_length": 12.0,
      "completions/mean_length": 12.0,
      "completions/mean_terminated_length": 12.0,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.235,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 4.501597881317139,
      "learning_rate": 1.3336701375997127e-08,
      "loss": -0.0,
      "num_tokens": 8863502.0,
      "reward": 0.7406250238418579,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "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": 12.0,
      "completions/max_terminated_length": 12.0,
      "completions/mean_length": 12.0,
      "completions/mean_terminated_length": 12.0,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2375,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 9.910150527954102,
      "learning_rate": 9.810017062595321e-09,
      "loss": 0.0,
      "num_tokens": 8996070.0,
      "reward": 0.2718750238418579,
      "reward_std": 0.10205793380737305,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.171875,
      "rewards/mcq_exact_match_reward/std": 0.38025420904159546,
      "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": 152.0,
      "completions/max_terminated_length": 152.0,
      "completions/mean_length": 14.296875,
      "completions/mean_terminated_length": 14.296875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.24,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.15203857421875,
      "learning_rate": 6.819348298638839e-09,
      "loss": 0.0,
      "num_tokens": 9071705.0,
      "reward": 0.40937501192092896,
      "reward_std": 0.11467419564723969,
      "rewards/format_reward/mean": 0.96875,
      "rewards/format_reward/std": 0.17536810040473938,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 96
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.09375,
      "completions/mean_terminated_length": 12.09375,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2425,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 8.452611923217773,
      "learning_rate": 4.367965336512403e-09,
      "loss": -0.0,
      "num_tokens": 9175575.0,
      "reward": 0.8656250238418579,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.765625,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 16.421875,
      "completions/mean_terminated_length": 16.421875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.245,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 6.549090385437012,
      "learning_rate": 2.458548727494292e-09,
      "loss": 0.0,
      "num_tokens": 9270802.0,
      "reward": 0.7093750238418579,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 172.0,
      "completions/max_terminated_length": 172.0,
      "completions/mean_length": 15.171875,
      "completions/mean_terminated_length": 15.171875,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.2475,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 19.366769790649414,
      "learning_rate": 1.0931863906127325e-09,
      "loss": 0.0,
      "num_tokens": 9363773.0,
      "reward": 0.8171875476837158,
      "reward_std": 0.18119612336158752,
      "rewards/format_reward/mean": 0.984375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 18.0,
      "completions/max_terminated_length": 18.0,
      "completions/mean_length": 12.5625,
      "completions/mean_terminated_length": 12.5625,
      "completions/min_length": 12.0,
      "completions/min_terminated_length": 12.0,
      "epoch": 0.25,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 6.99455451965332,
      "learning_rate": 2.733713295369755e-10,
      "loss": 0.0,
      "num_tokens": 9435561.0,
      "reward": 0.5062500238418579,
      "reward_std": 0.0578637570142746,
      "rewards/format_reward/mean": 1.0,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 9435561,
  "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
}