{
  "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": 1525.0,
      "completions/max_terminated_length": 1525.0,
      "completions/mean_length": 639.6875,
      "completions/mean_terminated_length": 639.6875,
      "completions/min_length": 10.0,
      "completions/min_terminated_length": 10.0,
      "epoch": 0.0025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 1.3210506439208984,
      "learning_rate": 0.0,
      "loss": 0.0,
      "num_tokens": 142612.0,
      "reward": 0.28749996423721313,
      "reward_std": 0.36111634969711304,
      "rewards/format_reward/mean": 0.375,
      "rewards/format_reward/std": 0.2182178944349289,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1178.0,
      "completions/mean_length": 515.03125,
      "completions/mean_terminated_length": 465.58062744140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.005,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 4.131510257720947,
      "learning_rate": 2e-07,
      "loss": -0.0,
      "num_tokens": 245750.0,
      "reward": 0.16015625,
      "reward_std": 0.297551691532135,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.3527633249759674,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 1414.0,
      "completions/mean_length": 485.25,
      "completions/mean_terminated_length": 460.4444580078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0075,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 6.900653839111328,
      "learning_rate": 4e-07,
      "loss": -0.0,
      "num_tokens": 384342.0,
      "reward": 0.34687498211860657,
      "reward_std": 0.33192795515060425,
      "rewards/format_reward/mean": 0.34375,
      "rewards/format_reward/std": 0.25,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 1384.0,
      "completions/max_terminated_length": 1384.0,
      "completions/mean_length": 629.15625,
      "completions/mean_terminated_length": 629.15625,
      "completions/min_length": 10.0,
      "completions/min_terminated_length": 10.0,
      "epoch": 0.01,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 5.880721092224121,
      "learning_rate": 6e-07,
      "loss": -0.0,
      "num_tokens": 525808.0,
      "reward": 0.21953122317790985,
      "reward_std": 0.3185732364654541,
      "rewards/format_reward/mean": 0.3203125,
      "rewards/format_reward/std": 0.24180518090724945,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 1343.0,
      "completions/max_terminated_length": 1343.0,
      "completions/mean_length": 365.828125,
      "completions/mean_terminated_length": 365.828125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0125,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 14.31756591796875,
      "learning_rate": 8e-07,
      "loss": -0.0,
      "num_tokens": 625725.0,
      "reward": 0.3460937440395355,
      "reward_std": 0.42002928256988525,
      "rewards/format_reward/mean": 0.3359375,
      "rewards/format_reward/std": 0.36790579557418823,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 1882.0,
      "completions/mean_length": 664.390625,
      "completions/mean_terminated_length": 619.758056640625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.015,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 5.773580074310303,
      "learning_rate": 1e-06,
      "loss": 0.0,
      "num_tokens": 787142.0,
      "reward": 0.25859373807907104,
      "reward_std": 0.37149035930633545,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.20275264978408813,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 6
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1546.0,
      "completions/mean_length": 461.03125,
      "completions/mean_terminated_length": 435.84130859375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0175,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 10.868589401245117,
      "learning_rate": 9.99726628670463e-07,
      "loss": -0.0,
      "num_tokens": 906856.0,
      "reward": 0.26093748211860657,
      "reward_std": 0.39204344153404236,
      "rewards/format_reward/mean": 0.421875,
      "rewards/format_reward/std": 0.23935678601264954,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1503.0,
      "completions/mean_length": 635.328125,
      "completions/mean_terminated_length": 589.758056640625,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.02,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 1.5266581773757935,
      "learning_rate": 9.989068136093872e-07,
      "loss": -0.0,
      "num_tokens": 1041213.0,
      "reward": 0.32734376192092896,
      "reward_std": 0.366255521774292,
      "rewards/format_reward/mean": 0.4609375,
      "rewards/format_reward/std": 0.29953256249427795,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 8
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1898.0,
      "completions/mean_length": 608.078125,
      "completions/mean_terminated_length": 585.2222290039062,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0225,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 3.710291862487793,
      "learning_rate": 9.975414512725056e-07,
      "loss": 0.0,
      "num_tokens": 1172946.0,
      "reward": 0.2601562440395355,
      "reward_std": 0.3032258450984955,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.209963858127594,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "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": 1391.0,
      "completions/mean_length": 540.359375,
      "completions/mean_terminated_length": 516.4285888671875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 8.635683059692383,
      "learning_rate": 9.956320346634875e-07,
      "loss": -0.0,
      "num_tokens": 1292105.0,
      "reward": 0.27421873807907104,
      "reward_std": 0.4047485589981079,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.22146137058734894,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 0.0,
      "completions/max_length": 1350.0,
      "completions/max_terminated_length": 1350.0,
      "completions/mean_length": 477.21875,
      "completions/mean_terminated_length": 477.21875,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.0275,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 1.1730365753173828,
      "learning_rate": 9.931806517013612e-07,
      "loss": -0.0,
      "num_tokens": 1407903.0,
      "reward": 0.28984373807907104,
      "reward_std": 0.42363831400871277,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.28423789143562317,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 2021.0,
      "completions/mean_length": 624.8125,
      "completions/mean_terminated_length": 578.9031982421875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.03,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 6.3810648918151855,
      "learning_rate": 9.901899829374047e-07,
      "loss": 0.0,
      "num_tokens": 1564115.0,
      "reward": 0.12421874701976776,
      "reward_std": 0.16493451595306396,
      "rewards/format_reward/mean": 0.4609375,
      "rewards/format_reward/std": 0.20518454909324646,
      "rewards/mcq_exact_match_reward/mean": 0.078125,
      "rewards/mcq_exact_match_reward/std": 0.27048972249031067,
      "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": 1216.0,
      "completions/max_terminated_length": 1216.0,
      "completions/mean_length": 431.515625,
      "completions/mean_terminated_length": 431.515625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0325,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 5.01668643951416,
      "learning_rate": 9.866632986240029e-07,
      "loss": 0.0,
      "num_tokens": 1677876.0,
      "reward": 0.3164062201976776,
      "reward_std": 0.34676826000213623,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.2302463799715042,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 0.046875,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1667.0,
      "completions/mean_length": 525.359375,
      "completions/mean_terminated_length": 450.4753723144531,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.035,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 9.905223846435547,
      "learning_rate": 9.826044551386742e-07,
      "loss": -0.0,
      "num_tokens": 1815915.0,
      "reward": 0.37578123807907104,
      "reward_std": 0.3770672082901001,
      "rewards/format_reward/mean": 0.3203125,
      "rewards/format_reward/std": 0.24180518090724945,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 14
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1412.0,
      "completions/max_terminated_length": 1412.0,
      "completions/mean_length": 520.90625,
      "completions/mean_terminated_length": 520.90625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0375,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 5.3415679931640625,
      "learning_rate": 9.780178907671788e-07,
      "loss": 0.0,
      "num_tokens": 1947293.0,
      "reward": 0.2789062559604645,
      "reward_std": 0.303743451833725,
      "rewards/format_reward/mean": 0.4453125,
      "rewards/format_reward/std": 0.180765300989151,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1347.0,
      "completions/mean_length": 527.65625,
      "completions/mean_terminated_length": 503.5238342285156,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.04,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 3.8449149131774902,
      "learning_rate": 9.729086208503173e-07,
      "loss": 0.0,
      "num_tokens": 2076351.0,
      "reward": 0.16093750298023224,
      "reward_std": 0.1840258091688156,
      "rewards/format_reward/mean": 0.359375,
      "rewards/format_reward/std": 0.22658175230026245,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 0.046875,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1281.0,
      "completions/mean_length": 587.640625,
      "completions/mean_terminated_length": 515.8196411132812,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0425,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 2.72416353225708,
      "learning_rate": 9.672822322997304e-07,
      "loss": 0.0,
      "num_tokens": 2211424.0,
      "reward": 0.07734374701976776,
      "reward_std": 0.10851941257715225,
      "rewards/format_reward/mean": 0.4609375,
      "rewards/format_reward/std": 0.24077731370925903,
      "rewards/mcq_exact_match_reward/mean": 0.03125,
      "rewards/mcq_exact_match_reward/std": 0.17536810040473938,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1747.0,
      "completions/mean_length": 443.703125,
      "completions/mean_terminated_length": 418.2381286621094,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.045,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 15.599007606506348,
      "learning_rate": 9.611448774886923e-07,
      "loss": -0.0,
      "num_tokens": 2328701.0,
      "reward": 0.31953126192092896,
      "reward_std": 0.291238397359848,
      "rewards/format_reward/mean": 0.3828125,
      "rewards/format_reward/std": 0.2313210517168045,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 1656.0,
      "completions/max_terminated_length": 1656.0,
      "completions/mean_length": 496.5,
      "completions/mean_terminated_length": 496.5,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0475,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 5.492337226867676,
      "learning_rate": 9.545032675245813e-07,
      "loss": 0.0,
      "num_tokens": 2474301.0,
      "reward": 0.19062499701976776,
      "reward_std": 0.2896108329296112,
      "rewards/format_reward/mean": 0.34375,
      "rewards/format_reward/std": 0.233588308095932,
      "rewards/mcq_exact_match_reward/mean": 0.15625,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1685.0,
      "completions/mean_length": 541.390625,
      "completions/mean_terminated_length": 517.4761962890625,
      "completions/min_length": 3.0,
      "completions/min_terminated_length": 3.0,
      "epoch": 0.05,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 13.7561616897583,
      "learning_rate": 9.473646649103817e-07,
      "loss": -0.0,
      "num_tokens": 2595974.0,
      "reward": 0.5601562261581421,
      "reward_std": 0.42670488357543945,
      "rewards/format_reward/mean": 0.4453125,
      "rewards/format_reward/std": 0.20152568817138672,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 20
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.09375,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1982.0,
      "completions/mean_length": 816.484375,
      "completions/mean_terminated_length": 689.086181640625,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.0525,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 1.1077152490615845,
      "learning_rate": 9.397368756032444e-07,
      "loss": -0.0,
      "num_tokens": 2751973.0,
      "reward": 0.21015623211860657,
      "reward_std": 0.2677168548107147,
      "rewards/format_reward/mean": 0.3828125,
      "rewards/format_reward/std": 0.21347814798355103,
      "rewards/mcq_exact_match_reward/mean": 0.171875,
      "rewards/mcq_exact_match_reward/std": 0.38025420904159546,
      "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": 1301.0,
      "completions/max_terminated_length": 1301.0,
      "completions/mean_length": 416.703125,
      "completions/mean_terminated_length": 416.703125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.055,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 14.890534400939941,
      "learning_rate": 9.316282404787869e-07,
      "loss": 0.0,
      "num_tokens": 2858698.0,
      "reward": 0.3789062201976776,
      "reward_std": 0.389323353767395,
      "rewards/format_reward/mean": 0.5078125,
      "rewards/format_reward/std": 0.20877929031848907,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1853.0,
      "completions/mean_length": 533.125,
      "completions/mean_terminated_length": 509.07940673828125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0575,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 6.468020439147949,
      "learning_rate": 9.230476262104676e-07,
      "loss": -0.0,
      "num_tokens": 2968618.0,
      "reward": 0.3734374940395355,
      "reward_std": 0.3810562193393707,
      "rewards/format_reward/mean": 0.453125,
      "rewards/format_reward/std": 0.23091863095760345,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 1909.0,
      "completions/max_terminated_length": 1909.0,
      "completions/mean_length": 614.75,
      "completions/mean_terminated_length": 614.75,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.06,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 4.7269487380981445,
      "learning_rate": 9.1400441557401e-07,
      "loss": -0.0,
      "num_tokens": 3111138.0,
      "reward": 0.3726562261581421,
      "reward_std": 0.33836525678634644,
      "rewards/format_reward/mean": 0.4453125,
      "rewards/format_reward/std": 0.15728822350502014,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1114.0,
      "completions/mean_length": 525.34375,
      "completions/mean_terminated_length": 501.17462158203125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0625,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 4.618552207946777,
      "learning_rate": 9.045084971874737e-07,
      "loss": -0.0,
      "num_tokens": 3224392.0,
      "reward": 0.40703123807907104,
      "reward_std": 0.36556264758110046,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.1649840772151947,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1919.0,
      "completions/mean_length": 735.984375,
      "completions/mean_terminated_length": 693.6612548828125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.065,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 9.331802368164062,
      "learning_rate": 8.945702546981968e-07,
      "loss": -0.0,
      "num_tokens": 3348047.0,
      "reward": 0.27421873807907104,
      "reward_std": 0.21257494390010834,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.20275264978408813,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1303.0,
      "completions/mean_length": 543.8125,
      "completions/mean_terminated_length": 519.9365234375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0675,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 6.371813774108887,
      "learning_rate": 8.842005554284295e-07,
      "loss": -0.0,
      "num_tokens": 3467587.0,
      "reward": 0.4390624463558197,
      "reward_std": 0.2996913194656372,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.125,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 0.078125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1609.0,
      "completions/mean_length": 495.90625,
      "completions/mean_terminated_length": 364.3728942871094,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.07,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 7.129162788391113,
      "learning_rate": 8.734107384920769e-07,
      "loss": 0.0,
      "num_tokens": 3584701.0,
      "reward": 0.5296874642372131,
      "reward_std": 0.26773929595947266,
      "rewards/format_reward/mean": 0.453125,
      "rewards/format_reward/std": 0.1717960685491562,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1381.0,
      "completions/mean_length": 445.234375,
      "completions/mean_terminated_length": 393.5322570800781,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0725,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 10.376834869384766,
      "learning_rate": 8.622126023955445e-07,
      "loss": -0.0,
      "num_tokens": 3709044.0,
      "reward": 0.32734373211860657,
      "reward_std": 0.32236260175704956,
      "rewards/format_reward/mean": 0.4609375,
      "rewards/format_reward/std": 0.16194961965084076,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 1032.0,
      "completions/max_terminated_length": 1032.0,
      "completions/mean_length": 278.828125,
      "completions/mean_terminated_length": 278.828125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.075,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 14.453448295593262,
      "learning_rate": 8.506183921362442e-07,
      "loss": 0.0,
      "num_tokens": 3815801.0,
      "reward": 0.17265623807907104,
      "reward_std": 0.18370115756988525,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.1649840772151947,
      "rewards/mcq_exact_match_reward/mean": 0.125,
      "rewards/mcq_exact_match_reward/std": 0.3333333432674408,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1211.0,
      "completions/mean_length": 429.703125,
      "completions/mean_terminated_length": 377.5,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0775,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 9.325780868530273,
      "learning_rate": 8.386407858128706e-07,
      "loss": 0.0,
      "num_tokens": 3964118.0,
      "reward": 0.23124998807907104,
      "reward_std": 0.28464049100875854,
      "rewards/format_reward/mean": 0.4375,
      "rewards/format_reward/std": 0.1666666716337204,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "step": 31
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1088.0,
      "completions/max_terminated_length": 1088.0,
      "completions/mean_length": 132.375,
      "completions/mean_terminated_length": 132.375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 13.469862937927246,
      "learning_rate": 8.262928807620843e-07,
      "loss": 0.0,
      "num_tokens": 4076942.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.19044625759124756,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 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": 1058.0,
      "completions/max_terminated_length": 1058.0,
      "completions/mean_length": 177.59375,
      "completions/mean_terminated_length": 177.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0825,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 16.07432746887207,
      "learning_rate": 8.135881792367685e-07,
      "loss": -0.0,
      "num_tokens": 4166172.0,
      "reward": 0.3765624761581421,
      "reward_std": 0.27856409549713135,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.1534975916147232,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 1056.0,
      "completions/max_terminated_length": 1056.0,
      "completions/mean_length": 260.328125,
      "completions/mean_terminated_length": 260.328125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.085,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 5.570297718048096,
      "learning_rate": 8.005405736415125e-07,
      "loss": -0.0,
      "num_tokens": 4255649.0,
      "reward": 0.21015624701976776,
      "reward_std": 0.22792938351631165,
      "rewards/format_reward/mean": 0.5390625,
      "rewards/format_reward/std": 0.16194961965084076,
      "rewards/mcq_exact_match_reward/mean": 0.15625,
      "rewards/mcq_exact_match_reward/std": 0.36596253514289856,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1261.0,
      "completions/mean_length": 254.828125,
      "completions/mean_terminated_length": 226.36509704589844,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0875,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 1.1509226560592651,
      "learning_rate": 7.871643313414718e-07,
      "loss": -0.0,
      "num_tokens": 4355758.0,
      "reward": 0.41484373807907104,
      "reward_std": 0.17540813982486725,
      "rewards/format_reward/mean": 0.5546875,
      "rewards/format_reward/std": 0.20152568817138672,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 35
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1277.0,
      "completions/max_terminated_length": 1277.0,
      "completions/mean_length": 222.21875,
      "completions/mean_terminated_length": 222.21875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 13.273263931274414,
      "learning_rate": 7.734740790612136e-07,
      "loss": 0.0,
      "num_tokens": 4485068.0,
      "reward": 0.4203124940395355,
      "reward_std": 0.1527586579322815,
      "rewards/format_reward/mean": 0.453125,
      "rewards/format_reward/std": 0.14689241349697113,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "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": 776.0,
      "completions/max_terminated_length": 776.0,
      "completions/mean_length": 37.375,
      "completions/mean_terminated_length": 37.375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0925,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 12.956086158752441,
      "learning_rate": 7.594847868906076e-07,
      "loss": 0.0,
      "num_tokens": 4598644.0,
      "reward": 0.22890624403953552,
      "reward_std": 0.19832901656627655,
      "rewards/format_reward/mean": 0.4140625,
      "rewards/format_reward/std": 0.19012710452079773,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 576.0,
      "completions/mean_length": 158.03125,
      "completions/mean_terminated_length": 128.03175354003906,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.095,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 21.433876037597656,
      "learning_rate": 7.452117519152541e-07,
      "loss": 0.0,
      "num_tokens": 4690206.0,
      "reward": 0.25859373807907104,
      "reward_std": 0.2866668701171875,
      "rewards/format_reward/mean": 0.5546875,
      "rewards/format_reward/std": 0.2538151443004608,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "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": 1519.0,
      "completions/max_terminated_length": 1519.0,
      "completions/mean_length": 125.609375,
      "completions/mean_terminated_length": 125.609375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0975,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 20.178497314453125,
      "learning_rate": 7.306705814893439e-07,
      "loss": 0.0,
      "num_tokens": 4784253.0,
      "reward": 0.38359373807907104,
      "reward_std": 0.32375043630599976,
      "rewards/format_reward/mean": 0.5546875,
      "rewards/format_reward/std": 0.20152568817138672,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 942.0,
      "completions/max_terminated_length": 942.0,
      "completions/mean_length": 123.109375,
      "completions/mean_terminated_length": 123.109375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 14.118669509887695,
      "learning_rate": 7.158771761692464e-07,
      "loss": 0.0,
      "num_tokens": 4893844.0,
      "reward": 0.19453126192092896,
      "reward_std": 0.21533936262130737,
      "rewards/format_reward/mean": 0.5390625,
      "rewards/format_reward/std": 0.16194961965084076,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "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": 919.0,
      "completions/max_terminated_length": 919.0,
      "completions/mean_length": 124.984375,
      "completions/mean_terminated_length": 124.984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1025,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 15.214014053344727,
      "learning_rate": 7.008477123264847e-07,
      "loss": -0.0,
      "num_tokens": 5005499.0,
      "reward": 0.33828121423721313,
      "reward_std": 0.23885318636894226,
      "rewards/format_reward/mean": 0.5703125,
      "rewards/format_reward/std": 0.23345555365085602,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "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": 944.0,
      "completions/max_terminated_length": 944.0,
      "completions/mean_length": 92.5,
      "completions/mean_terminated_length": 92.5,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.105,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 12.566036224365234,
      "learning_rate": 6.855986244591103e-07,
      "loss": 0.0,
      "num_tokens": 5090971.0,
      "reward": 0.4429687261581421,
      "reward_std": 0.2083142101764679,
      "rewards/format_reward/mean": 0.5234375,
      "rewards/format_reward/std": 0.1875000149011612,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 1671.0,
      "completions/max_terminated_length": 1671.0,
      "completions/mean_length": 120.03125,
      "completions/mean_terminated_length": 120.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1075,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 15.18333911895752,
      "learning_rate": 6.701465872208216e-07,
      "loss": 0.0,
      "num_tokens": 5193389.0,
      "reward": 0.16328124701976776,
      "reward_std": 0.17210790514945984,
      "rewards/format_reward/mean": 0.5390625,
      "rewards/format_reward/std": 0.20518454909324646,
      "rewards/mcq_exact_match_reward/mean": 0.109375,
      "rewards/mcq_exact_match_reward/std": 0.3145764470100403,
      "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": 877.0,
      "completions/max_terminated_length": 877.0,
      "completions/mean_length": 110.328125,
      "completions/mean_terminated_length": 110.328125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 34.193328857421875,
      "learning_rate": 6.545084971874736e-07,
      "loss": 0.0,
      "num_tokens": 5275018.0,
      "reward": 0.3890625238418579,
      "reward_std": 0.3363279104232788,
      "rewards/format_reward/mean": 0.609375,
      "rewards/format_reward/std": 0.22658175230026245,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 801.0,
      "completions/max_terminated_length": 801.0,
      "completions/mean_length": 49.578125,
      "completions/mean_terminated_length": 49.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1125,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 27.484424591064453,
      "learning_rate": 6.387014543809223e-07,
      "loss": -0.0,
      "num_tokens": 5371103.0,
      "reward": 0.5023437142372131,
      "reward_std": 0.2715042233467102,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.1406387835741043,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "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": 920.0,
      "completions/max_terminated_length": 920.0,
      "completions/mean_length": 61.21875,
      "completions/mean_terminated_length": 61.21875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.115,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 17.36954689025879,
      "learning_rate": 6.227427435703995e-07,
      "loss": -0.0,
      "num_tokens": 5459141.0,
      "reward": 0.5171874761581421,
      "reward_std": 0.17749688029289246,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.46875,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 1217.0,
      "completions/max_terminated_length": 1217.0,
      "completions/mean_length": 35.84375,
      "completions/mean_terminated_length": 35.84375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1175,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 8.502973556518555,
      "learning_rate": 6.066498153718734e-07,
      "loss": 0.0,
      "num_tokens": 5563939.0,
      "reward": 0.43984371423721313,
      "reward_std": 0.13295188546180725,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 1804.0,
      "completions/max_terminated_length": 1804.0,
      "completions/mean_length": 53.921875,
      "completions/mean_terminated_length": 53.921875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 12.219757080078125,
      "learning_rate": 5.90440267166055e-07,
      "loss": 0.0,
      "num_tokens": 5658782.0,
      "reward": 0.3632812201976776,
      "reward_std": 0.14581039547920227,
      "rewards/format_reward/mean": 0.5078125,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 88.0,
      "completions/max_terminated_length": 88.0,
      "completions/mean_length": 7.28125,
      "completions/mean_terminated_length": 7.28125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1225,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 17.875537872314453,
      "learning_rate": 5.741318238559209e-07,
      "loss": 0.0,
      "num_tokens": 5757000.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.19044625759124756,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 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": 670.0,
      "completions/max_terminated_length": 670.0,
      "completions/mean_length": 19.09375,
      "completions/mean_terminated_length": 19.09375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.125,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 19.906347274780273,
      "learning_rate": 5.577423184847931e-07,
      "loss": -0.0,
      "num_tokens": 5876166.0,
      "reward": 0.3296874761581421,
      "reward_std": 0.20483602583408356,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 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": 750.0,
      "completions/max_terminated_length": 750.0,
      "completions/mean_length": 37.90625,
      "completions/mean_terminated_length": 37.90625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1275,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 10.658368110656738,
      "learning_rate": 5.412896727361662e-07,
      "loss": 0.0,
      "num_tokens": 5972112.0,
      "reward": 0.6898437142372131,
      "reward_std": 0.046403877437114716,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "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": 534.0,
      "completions/max_terminated_length": 534.0,
      "completions/mean_length": 27.40625,
      "completions/mean_terminated_length": 27.40625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.13,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 18.39838981628418,
      "learning_rate": 5.247918773366111e-07,
      "loss": 0.0,
      "num_tokens": 6079602.0,
      "reward": 0.3296874761581421,
      "reward_std": 0.17749686539173126,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 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": 379.0,
      "completions/max_terminated_length": 379.0,
      "completions/mean_length": 20.375,
      "completions/mean_terminated_length": 20.375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1325,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 23.512386322021484,
      "learning_rate": 5.082669723831793e-07,
      "loss": -0.0,
      "num_tokens": 6173890.0,
      "reward": 0.4406249523162842,
      "reward_std": 0.2824692726135254,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 53
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.135,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 16.073997497558594,
      "learning_rate": 4.917330276168208e-07,
      "loss": 0.0,
      "num_tokens": 6269666.0,
      "reward": 0.4093749523162842,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 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": 1963.0,
      "completions/max_terminated_length": 1963.0,
      "completions/mean_length": 36.578125,
      "completions/mean_terminated_length": 36.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1375,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 22.47065544128418,
      "learning_rate": 4.752081226633888e-07,
      "loss": -0.0,
      "num_tokens": 6366423.0,
      "reward": 0.5187499523162842,
      "reward_std": 0.1735912561416626,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.46875,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "step": 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": 591.0,
      "completions/max_terminated_length": 591.0,
      "completions/mean_length": 15.140625,
      "completions/mean_terminated_length": 15.140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.14,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 9.380659103393555,
      "learning_rate": 4.5871032726383385e-07,
      "loss": -0.0,
      "num_tokens": 6460672.0,
      "reward": 0.37812498211860657,
      "reward_std": 0.10205792635679245,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "step": 56
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1425,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 17.319658279418945,
      "learning_rate": 4.4225768151520694e-07,
      "loss": -0.0,
      "num_tokens": 6564808.0,
      "reward": 0.34687498211860657,
      "reward_std": 0.16887322068214417,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 57
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.145,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 19.46933364868164,
      "learning_rate": 4.258681761440789e-07,
      "loss": -0.0,
      "num_tokens": 6635968.0,
      "reward": 0.37812498211860657,
      "reward_std": 0.12255740165710449,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 353.0,
      "completions/max_terminated_length": 353.0,
      "completions/mean_length": 31.59375,
      "completions/mean_terminated_length": 31.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1475,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.568326950073242,
      "learning_rate": 4.095597328339452e-07,
      "loss": -0.0,
      "num_tokens": 6705534.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.1462520956993103,
      "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": 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": 111.0,
      "completions/max_terminated_length": 111.0,
      "completions/mean_length": 7.640625,
      "completions/mean_terminated_length": 7.640625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 27.69223976135254,
      "learning_rate": 3.9335018462812664e-07,
      "loss": -0.0,
      "num_tokens": 6789935.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.23827511072158813,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "step": 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": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 13.578125,
      "completions/mean_terminated_length": 13.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1525,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 4.942868232727051,
      "learning_rate": 3.772572564296004e-07,
      "loss": -0.0,
      "num_tokens": 6880764.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 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": 164.0,
      "completions/max_terminated_length": 164.0,
      "completions/mean_length": 10.421875,
      "completions/mean_terminated_length": 10.421875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.155,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 24.862308502197266,
      "learning_rate": 3.612985456190778e-07,
      "loss": 0.0,
      "num_tokens": 6984095.0,
      "reward": 0.23515623807907104,
      "reward_std": 0.14948678016662598,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "step": 62
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1575,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.646574974060059,
      "learning_rate": 3.454915028125263e-07,
      "loss": -0.0,
      "num_tokens": 7060455.0,
      "reward": 0.7999999523162842,
      "reward_std": 0.1157275065779686,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.75,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 63
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 609.0,
      "completions/max_terminated_length": 609.0,
      "completions/mean_length": 21.125,
      "completions/mean_terminated_length": 21.125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 18.160184860229492,
      "learning_rate": 3.2985341277917846e-07,
      "loss": -0.0,
      "num_tokens": 7159399.0,
      "reward": 0.7687499523162842,
      "reward_std": 0.1767766922712326,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 64
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 115.0,
      "completions/max_terminated_length": 115.0,
      "completions/mean_length": 7.703125,
      "completions/mean_terminated_length": 7.703125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1625,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 28.646360397338867,
      "learning_rate": 3.1440137554088953e-07,
      "loss": -0.0,
      "num_tokens": 7231548.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.2688094973564148,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 65
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 493.0,
      "completions/max_terminated_length": 493.0,
      "completions/mean_length": 32.671875,
      "completions/mean_terminated_length": 32.671875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.165,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 6.67764139175415,
      "learning_rate": 2.9915228767351535e-07,
      "loss": -0.0,
      "num_tokens": 7335719.0,
      "reward": 0.6429687142372131,
      "reward_std": 0.09059804677963257,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.59375,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 66
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1675,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.831806182861328,
      "learning_rate": 2.841228238307536e-07,
      "loss": 0.0,
      "num_tokens": 7420607.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 67
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1070.0,
      "completions/max_terminated_length": 1070.0,
      "completions/mean_length": 85.59375,
      "completions/mean_terminated_length": 85.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.17,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 17.88917350769043,
      "learning_rate": 2.6932941851065615e-07,
      "loss": 0.0,
      "num_tokens": 7497285.0,
      "reward": 0.33124998211860657,
      "reward_std": 0.0883883386850357,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 68
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1725,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 24.169601440429688,
      "learning_rate": 2.547882480847461e-07,
      "loss": -0.0,
      "num_tokens": 7595429.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.2688094973564148,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 69
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 307.0,
      "completions/max_terminated_length": 307.0,
      "completions/mean_length": 10.703125,
      "completions/mean_terminated_length": 10.703125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.175,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.098264694213867,
      "learning_rate": 2.4051521310939254e-07,
      "loss": -0.0,
      "num_tokens": 7700498.0,
      "reward": 0.3937499523162842,
      "reward_std": 0.17570312321186066,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 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": 669.0,
      "completions/max_terminated_length": 669.0,
      "completions/mean_length": 30.609375,
      "completions/mean_terminated_length": 30.609375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1775,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 22.42943572998047,
      "learning_rate": 2.2652592093878665e-07,
      "loss": -0.0,
      "num_tokens": 7771809.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.25726157426834106,
      "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": 71
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 146.0,
      "completions/max_terminated_length": 146.0,
      "completions/mean_length": 9.78125,
      "completions/mean_terminated_length": 9.78125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.18,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 38.008033752441406,
      "learning_rate": 2.128356686585282e-07,
      "loss": -0.0,
      "num_tokens": 7844795.0,
      "reward": 0.5484374761581421,
      "reward_std": 0.2838236689567566,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "step": 72
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1825,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 8.68104076385498,
      "learning_rate": 1.9945942635848745e-07,
      "loss": -0.0,
      "num_tokens": 7909515.0,
      "reward": 0.3468749523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 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": 1034.0,
      "completions/max_terminated_length": 1034.0,
      "completions/mean_length": 22.0625,
      "completions/mean_terminated_length": 22.0625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.185,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.8641182076323148e-07,
      "loss": 0.0,
      "num_tokens": 8004735.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 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": 17.0,
      "completions/max_terminated_length": 17.0,
      "completions/mean_length": 6.171875,
      "completions/mean_terminated_length": 6.171875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1875,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 25.599224090576172,
      "learning_rate": 1.7370711923791564e-07,
      "loss": 0.0,
      "num_tokens": 8081746.0,
      "reward": 0.5179687142372131,
      "reward_std": 0.167273610830307,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.46875,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 517.0,
      "completions/max_terminated_length": 517.0,
      "completions/mean_length": 21.828125,
      "completions/mean_terminated_length": 21.828125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 14.794571876525879,
      "learning_rate": 1.6135921418712955e-07,
      "loss": 0.0,
      "num_tokens": 8164671.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.10205791890621185,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 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": 387.0,
      "completions/max_terminated_length": 387.0,
      "completions/mean_length": 11.953125,
      "completions/mean_terminated_length": 11.953125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1925,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.07320785522461,
      "learning_rate": 1.493816078637557e-07,
      "loss": 0.0,
      "num_tokens": 8266932.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.04419416934251785,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 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": 833.0,
      "completions/max_terminated_length": 833.0,
      "completions/mean_length": 18.921875,
      "completions/mean_terminated_length": 18.921875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.195,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.371740341186523,
      "learning_rate": 1.3778739760445552e-07,
      "loss": 0.0,
      "num_tokens": 8362663.0,
      "reward": 0.28437498211860657,
      "reward_std": 0.13258251547813416,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "step": 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": 267.0,
      "completions/max_terminated_length": 267.0,
      "completions/mean_length": 11.46875,
      "completions/mean_terminated_length": 11.46875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1975,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 30.370616912841797,
      "learning_rate": 1.2658926150792322e-07,
      "loss": -0.0,
      "num_tokens": 8456781.0,
      "reward": 0.4093749523162842,
      "reward_std": 0.15992167592048645,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 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": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 19.734375,
      "completions/mean_terminated_length": 19.734375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 15.611992835998535,
      "learning_rate": 1.1579944457157059e-07,
      "loss": -0.0,
      "num_tokens": 8553068.0,
      "reward": 0.44062498211860657,
      "reward_std": 0.15992167592048645,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 80
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 594.0,
      "completions/max_terminated_length": 594.0,
      "completions/mean_length": 46.90625,
      "completions/mean_terminated_length": 46.90625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2025,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 25.991147994995117,
      "learning_rate": 1.0542974530180327e-07,
      "loss": -0.0,
      "num_tokens": 8625342.0,
      "reward": 0.6273437142372131,
      "reward_std": 0.2570541203022003,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "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": 957.0,
      "completions/max_terminated_length": 957.0,
      "completions/mean_length": 38.53125,
      "completions/mean_terminated_length": 38.53125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.205,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 4.844205856323242,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.0,
      "num_tokens": 8731240.0,
      "reward": 0.45546871423721313,
      "reward_std": 0.058385785669088364,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 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": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 664.0,
      "completions/mean_length": 50.09375,
      "completions/mean_terminated_length": 18.380952835083008,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2075,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 38.71321487426758,
      "learning_rate": 8.599558442598998e-08,
      "loss": -0.0,
      "num_tokens": 8849686.0,
      "reward": 0.7367187142372131,
      "reward_std": 0.3713894784450531,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.6875,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 83
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.21,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 27.330516815185547,
      "learning_rate": 7.695237378953224e-08,
      "loss": 0.0,
      "num_tokens": 8938310.0,
      "reward": 0.33124998211860657,
      "reward_std": 0.25513988733291626,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 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": 894.0,
      "completions/max_terminated_length": 894.0,
      "completions/mean_length": 55.78125,
      "completions/mean_terminated_length": 55.78125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2125,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 20.20175552368164,
      "learning_rate": 6.837175952121304e-08,
      "loss": 0.0,
      "num_tokens": 9013240.0,
      "reward": 0.33124998211860657,
      "reward_std": 0.0883883386850357,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 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": 313.0,
      "completions/max_terminated_length": 313.0,
      "completions/mean_length": 15.09375,
      "completions/mean_terminated_length": 15.09375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.215,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 24.405529022216797,
      "learning_rate": 6.026312439675551e-08,
      "loss": -0.0,
      "num_tokens": 9095134.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.2709311544895172,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 86
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 284.0,
      "completions/max_terminated_length": 284.0,
      "completions/mean_length": 13.4375,
      "completions/mean_terminated_length": 13.4375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2175,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 12.83124828338623,
      "learning_rate": 5.263533508961826e-08,
      "loss": -0.0,
      "num_tokens": 9190762.0,
      "reward": 0.3937499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 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": 116.0,
      "completions/max_terminated_length": 116.0,
      "completions/mean_length": 7.71875,
      "completions/mean_terminated_length": 7.71875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.22,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 41.8074951171875,
      "learning_rate": 4.549673247541874e-08,
      "loss": -0.0,
      "num_tokens": 9261416.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.23356688022613525,
      "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": 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": 100.0,
      "completions/max_terminated_length": 100.0,
      "completions/mean_length": 7.546875,
      "completions/mean_terminated_length": 7.546875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2225,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 26.247802734375,
      "learning_rate": 3.8855122511307626e-08,
      "loss": -0.0,
      "num_tokens": 9346875.0,
      "reward": 0.7843749523162842,
      "reward_std": 0.2519446909427643,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.734375,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "step": 89
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.225,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 3.271776770026963e-08,
      "loss": 0.0,
      "num_tokens": 9455251.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 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": 574.0,
      "completions/max_terminated_length": 574.0,
      "completions/mean_length": 21.28125,
      "completions/mean_terminated_length": 21.28125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2275,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 36.16759490966797,
      "learning_rate": 2.7091379149682682e-08,
      "loss": 0.0,
      "num_tokens": 9556117.0,
      "reward": 0.36171871423721313,
      "reward_std": 0.19841229915618896,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 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": 713.0,
      "completions/max_terminated_length": 713.0,
      "completions/mean_length": 28.34375,
      "completions/mean_terminated_length": 28.34375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.23,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 32.27620315551758,
      "learning_rate": 2.1982109232821176e-08,
      "loss": -0.0,
      "num_tokens": 9666699.0,
      "reward": 0.7062499523162842,
      "reward_std": 0.12467905879020691,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.65625,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 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": 516.0,
      "completions/max_terminated_length": 516.0,
      "completions/mean_length": 13.96875,
      "completions/mean_terminated_length": 13.96875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2325,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.7395544861325718e-08,
      "loss": 0.0,
      "num_tokens": 9753369.0,
      "reward": 0.7999999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.75,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 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": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 6.03125,
      "completions/mean_terminated_length": 6.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.235,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 1.3336701375997127e-08,
      "loss": 0.0,
      "num_tokens": 9823571.0,
      "reward": 0.6749999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 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": 399.0,
      "completions/max_terminated_length": 399.0,
      "completions/mean_length": 27.828125,
      "completions/mean_terminated_length": 27.828125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2375,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 19.089712142944336,
      "learning_rate": 9.810017062595321e-09,
      "loss": -0.0,
      "num_tokens": 9900728.0,
      "reward": 0.7218749523162842,
      "reward_std": 0.12255740165710449,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.671875,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "step": 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": 410.0,
      "completions/max_terminated_length": 410.0,
      "completions/mean_length": 17.484375,
      "completions/mean_terminated_length": 17.484375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.24,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.465526580810547,
      "learning_rate": 6.819348298638839e-09,
      "loss": -0.0,
      "num_tokens": 9999287.0,
      "reward": 0.8468749523162842,
      "reward_std": 0.15992167592048645,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.796875,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "step": 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": 888.0,
      "completions/max_terminated_length": 888.0,
      "completions/mean_length": 69.34375,
      "completions/mean_terminated_length": 69.34375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2425,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 48.8848991394043,
      "learning_rate": 4.367965336512403e-09,
      "loss": 0.0,
      "num_tokens": 10075437.0,
      "reward": 0.45624998211860657,
      "reward_std": 0.25513991713523865,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 97
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.245,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 22.443164825439453,
      "learning_rate": 2.458548727494292e-09,
      "loss": -0.0,
      "num_tokens": 10166445.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.10888782143592834,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "step": 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": 675.0,
      "completions/max_terminated_length": 675.0,
      "completions/mean_length": 16.453125,
      "completions/mean_terminated_length": 16.453125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2475,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.707706451416016,
      "learning_rate": 1.0931863906127325e-09,
      "loss": -0.0,
      "num_tokens": 10277610.0,
      "reward": 0.4093749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 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": 999.0,
      "completions/max_terminated_length": 999.0,
      "completions/mean_length": 21.59375,
      "completions/mean_terminated_length": 21.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.25,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 27.002370834350586,
      "learning_rate": 2.733713295369755e-10,
      "loss": 0.0,
      "num_tokens": 10374000.0,
      "reward": 0.5031249523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.453125,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 10374000,
  "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
}