{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.25,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1820.0,
      "completions/mean_length": 690.5,
      "completions/mean_terminated_length": 668.952392578125,
      "completions/min_length": 13.0,
      "completions/min_terminated_length": 13.0,
      "epoch": 0.0025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 1.565263271331787,
      "learning_rate": 0.0,
      "loss": -0.0,
      "num_tokens": 151352.0,
      "reward": 0.3023437261581421,
      "reward_std": 0.3922233581542969,
      "rewards/format_reward/mean": 0.3671875,
      "rewards/format_reward/std": 0.2708333432674408,
      "rewards/mcq_exact_match_reward/mean": 0.265625,
      "rewards/mcq_exact_match_reward/std": 0.44515693187713623,
      "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": 1742.0,
      "completions/mean_length": 565.578125,
      "completions/mean_terminated_length": 517.758056640625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.005,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 7.563284873962402,
      "learning_rate": 2e-07,
      "loss": -0.0,
      "num_tokens": 263405.0,
      "reward": 0.2226562350988388,
      "reward_std": 0.2859582304954529,
      "rewards/format_reward/mean": 0.3515625,
      "rewards/format_reward/std": 0.3044600784778595,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 1457.0,
      "completions/mean_length": 483.875,
      "completions/mean_terminated_length": 459.0476379394531,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0075,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 7.726901054382324,
      "learning_rate": 4e-07,
      "loss": 0.0,
      "num_tokens": 395205.0,
      "reward": 0.22187499701976776,
      "reward_std": 0.25196027755737305,
      "rewards/format_reward/mean": 0.34375,
      "rewards/format_reward/std": 0.25,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1407.0,
      "completions/mean_length": 494.40625,
      "completions/mean_terminated_length": 444.2903137207031,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.01,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 27.221830368041992,
      "learning_rate": 6e-07,
      "loss": -0.0,
      "num_tokens": 526471.0,
      "reward": 0.26796871423721313,
      "reward_std": 0.2928730547428131,
      "rewards/format_reward/mean": 0.3359375,
      "rewards/format_reward/std": 0.25283610820770264,
      "rewards/mcq_exact_match_reward/mean": 0.234375,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "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": 1927.0,
      "completions/max_terminated_length": 1927.0,
      "completions/mean_length": 397.25,
      "completions/mean_terminated_length": 397.25,
      "completions/min_length": 3.0,
      "completions/min_terminated_length": 3.0,
      "epoch": 0.0125,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 30.706405639648438,
      "learning_rate": 8e-07,
      "loss": -0.0,
      "num_tokens": 627319.0,
      "reward": 0.24921874701976776,
      "reward_std": 0.35042986273765564,
      "rewards/format_reward/mean": 0.3046875,
      "rewards/format_reward/std": 0.2615155577659607,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 5
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1555.0,
      "completions/mean_length": 399.078125,
      "completions/mean_terminated_length": 372.90478515625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.015,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 10.885279655456543,
      "learning_rate": 1e-06,
      "loss": 0.0,
      "num_tokens": 749580.0,
      "reward": 0.33281248807907104,
      "reward_std": 0.3967764377593994,
      "rewards/format_reward/mean": 0.359375,
      "rewards/format_reward/std": 0.24346621334552765,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "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": 2022.0,
      "completions/mean_length": 550.921875,
      "completions/mean_terminated_length": 527.1587524414062,
      "completions/min_length": 5.0,
      "completions/min_terminated_length": 5.0,
      "epoch": 0.0175,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 13.436131477355957,
      "learning_rate": 9.99726628670463e-07,
      "loss": -0.0,
      "num_tokens": 874207.0,
      "reward": 0.28984373807907104,
      "reward_std": 0.41126930713653564,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.25479042530059814,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 7
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.015625,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1902.0,
      "completions/mean_length": 478.09375,
      "completions/mean_terminated_length": 453.17462158203125,
      "completions/min_length": 10.0,
      "completions/min_terminated_length": 10.0,
      "epoch": 0.02,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 2.21781325340271,
      "learning_rate": 9.989068136093872e-07,
      "loss": -0.0,
      "num_tokens": 978173.0,
      "reward": 0.3765624463558197,
      "reward_std": 0.4527677297592163,
      "rewards/format_reward/mean": 0.328125,
      "rewards/format_reward/std": 0.2847827076911926,
      "rewards/mcq_exact_match_reward/mean": 0.34375,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 8
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 2048.0,
      "completions/max_terminated_length": 1503.0,
      "completions/mean_length": 618.96875,
      "completions/mean_terminated_length": 572.8709716796875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0225,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 7.731747150421143,
      "learning_rate": 9.975414512725056e-07,
      "loss": -0.0,
      "num_tokens": 1112819.0,
      "reward": 0.18046873807907104,
      "reward_std": 0.19389677047729492,
      "rewards/format_reward/mean": 0.3984375,
      "rewards/format_reward/std": 0.22146137058734894,
      "rewards/mcq_exact_match_reward/mean": 0.140625,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "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": 1210.0,
      "completions/mean_length": 348.703125,
      "completions/mean_terminated_length": 321.73016357421875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.025,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 12.382220268249512,
      "learning_rate": 9.956320346634875e-07,
      "loss": -0.0,
      "num_tokens": 1214208.0,
      "reward": 0.3703124523162842,
      "reward_std": 0.38414281606674194,
      "rewards/format_reward/mean": 0.421875,
      "rewards/format_reward/std": 0.18298126757144928,
      "rewards/mcq_exact_match_reward/mean": 0.328125,
      "rewards/mcq_exact_match_reward/std": 0.4732423722743988,
      "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": 1651.0,
      "completions/max_terminated_length": 1651.0,
      "completions/mean_length": 389.34375,
      "completions/mean_terminated_length": 389.34375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0275,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 16.486431121826172,
      "learning_rate": 9.931806517013612e-07,
      "loss": 0.0,
      "num_tokens": 1335942.0,
      "reward": 0.359375,
      "reward_std": 0.35479018092155457,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.233588308095932,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 1410.0,
      "completions/mean_length": 486.3125,
      "completions/mean_terminated_length": 435.93548583984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.03,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 16.557157516479492,
      "learning_rate": 9.901899829374047e-07,
      "loss": 0.0,
      "num_tokens": 1456466.0,
      "reward": 0.43359375,
      "reward_std": 0.31481531262397766,
      "rewards/format_reward/mean": 0.4296875,
      "rewards/format_reward/std": 0.21578919887542725,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 1357.0,
      "completions/max_terminated_length": 1357.0,
      "completions/mean_length": 444.421875,
      "completions/mean_terminated_length": 444.421875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0325,
      "frac_reward_zero_std": 0.125,
      "grad_norm": 14.968536376953125,
      "learning_rate": 9.866632986240029e-07,
      "loss": -0.0,
      "num_tokens": 1558589.0,
      "reward": 0.4867187440395355,
      "reward_std": 0.3611418604850769,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.18881812691688538,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 13
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1453.0,
      "completions/max_terminated_length": 1453.0,
      "completions/mean_length": 229.296875,
      "completions/mean_terminated_length": 229.296875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.035,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 11.122706413269043,
      "learning_rate": 9.826044551386742e-07,
      "loss": 0.0,
      "num_tokens": 1659232.0,
      "reward": 0.40468746423721313,
      "reward_std": 0.18586555123329163,
      "rewards/format_reward/mean": 0.453125,
      "rewards/format_reward/std": 0.14689241349697113,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "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": 1152.0,
      "completions/max_terminated_length": 1152.0,
      "completions/mean_length": 170.703125,
      "completions/mean_terminated_length": 170.703125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0375,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 15.367648124694824,
      "learning_rate": 9.780178907671788e-07,
      "loss": 0.0,
      "num_tokens": 1749557.0,
      "reward": 0.3460937440395355,
      "reward_std": 0.2693103551864624,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.18881812691688538,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 15
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 16.0,
      "completions/max_terminated_length": 16.0,
      "completions/mean_length": 6.234375,
      "completions/mean_terminated_length": 6.234375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.04,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 11.168756484985352,
      "learning_rate": 9.729086208503173e-07,
      "loss": -0.0,
      "num_tokens": 1843068.0,
      "reward": 0.3937499523162842,
      "reward_std": 0.2630178928375244,
      "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": 16
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1108.0,
      "completions/max_terminated_length": 1108.0,
      "completions/mean_length": 99.140625,
      "completions/mean_terminated_length": 99.140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0425,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 16.297332763671875,
      "learning_rate": 9.672822322997304e-07,
      "loss": 0.0,
      "num_tokens": 1926709.0,
      "reward": 0.29921871423721313,
      "reward_std": 0.3136211931705475,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.10882443934679031,
      "rewards/mcq_exact_match_reward/mean": 0.25,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 17
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 404.0,
      "completions/max_terminated_length": 404.0,
      "completions/mean_length": 16.984375,
      "completions/mean_terminated_length": 16.984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.045,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 13.903787612915039,
      "learning_rate": 9.611448774886923e-07,
      "loss": -0.0,
      "num_tokens": 2022188.0,
      "reward": 0.6734374761581421,
      "reward_std": 0.21748681366443634,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.625,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "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": 455.0,
      "completions/max_terminated_length": 455.0,
      "completions/mean_length": 22.15625,
      "completions/mean_terminated_length": 22.15625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0475,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 21.733028411865234,
      "learning_rate": 9.545032675245813e-07,
      "loss": -0.0,
      "num_tokens": 2118462.0,
      "reward": 0.48671871423721313,
      "reward_std": 0.06768142431974411,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 19
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 329.0,
      "completions/max_terminated_length": 329.0,
      "completions/mean_length": 19.265625,
      "completions/mean_terminated_length": 19.265625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.05,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 21.432254791259766,
      "learning_rate": 9.473646649103817e-07,
      "loss": -0.0,
      "num_tokens": 2201407.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.23144522309303284,
      "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": 20
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 130.0,
      "completions/max_terminated_length": 130.0,
      "completions/mean_length": 8.59375,
      "completions/mean_terminated_length": 8.59375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0525,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 20.55296516418457,
      "learning_rate": 9.397368756032444e-07,
      "loss": 0.0,
      "num_tokens": 2292221.0,
      "reward": 0.25234371423721313,
      "reward_std": 0.06690335273742676,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.203125,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "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": 8.0,
      "completions/max_terminated_length": 8.0,
      "completions/mean_length": 5.921875,
      "completions/mean_terminated_length": 5.921875,
      "completions/min_length": 2.0,
      "completions/min_terminated_length": 2.0,
      "epoch": 0.055,
      "frac_reward_zero_std": 0.375,
      "grad_norm": 40.7379035949707,
      "learning_rate": 9.316282404787869e-07,
      "loss": -0.0,
      "num_tokens": 2376064.0,
      "reward": 0.7515624761581421,
      "reward_std": 0.31270164251327515,
      "rewards/format_reward/mean": 0.484375,
      "rewards/format_reward/std": 0.08768405020236969,
      "rewards/mcq_exact_match_reward/mean": 0.703125,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 22
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 267.0,
      "completions/max_terminated_length": 267.0,
      "completions/mean_length": 10.078125,
      "completions/mean_terminated_length": 10.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0575,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 21.1943302154541,
      "learning_rate": 9.230476262104676e-07,
      "loss": 0.0,
      "num_tokens": 2452069.0,
      "reward": 0.4718749523162842,
      "reward_std": 0.19044625759124756,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 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": 68.0,
      "completions/max_terminated_length": 68.0,
      "completions/mean_length": 6.984375,
      "completions/mean_terminated_length": 6.984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.06,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 9.49032211303711,
      "learning_rate": 9.1400441557401e-07,
      "loss": 0.0,
      "num_tokens": 2563004.0,
      "reward": 0.37734371423721313,
      "reward_std": 0.1534513533115387,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "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": 0.0,
      "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": 0.0625,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 24.17865753173828,
      "learning_rate": 9.045084971874737e-07,
      "loss": -0.0,
      "num_tokens": 2655204.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.24511480331420898,
      "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": 25
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 229.0,
      "completions/max_terminated_length": 229.0,
      "completions/mean_length": 9.484375,
      "completions/mean_terminated_length": 9.484375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.065,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 14.044975280761719,
      "learning_rate": 8.945702546981968e-07,
      "loss": -0.0,
      "num_tokens": 2777675.0,
      "reward": 0.08124999701976776,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.03125,
      "rewards/mcq_exact_match_reward/std": 0.17536810040473938,
      "step": 26
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 103.0,
      "completions/max_terminated_length": 103.0,
      "completions/mean_length": 7.578125,
      "completions/mean_terminated_length": 7.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0675,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.599027633666992,
      "learning_rate": 8.842005554284295e-07,
      "loss": -0.0,
      "num_tokens": 2867016.0,
      "reward": 0.5648437142372131,
      "reward_std": 0.10426762700080872,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 27
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.07,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 24.69874382019043,
      "learning_rate": 8.734107384920769e-07,
      "loss": 0.0,
      "num_tokens": 2962672.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.10888781398534775,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 28
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 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": 0.0725,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 20.4017276763916,
      "learning_rate": 8.622126023955445e-07,
      "loss": -0.0,
      "num_tokens": 3042560.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.2109457403421402,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 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": 327.0,
      "completions/max_terminated_length": 327.0,
      "completions/mean_length": 11.015625,
      "completions/mean_terminated_length": 11.015625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.075,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 25.468481063842773,
      "learning_rate": 8.506183921362442e-07,
      "loss": 0.0,
      "num_tokens": 3126753.0,
      "reward": 0.4406249523162842,
      "reward_std": 0.3082856237888336,
      "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": 30
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 293.0,
      "completions/max_terminated_length": 293.0,
      "completions/mean_length": 10.484375,
      "completions/mean_terminated_length": 10.484375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0775,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 13.05105972290039,
      "learning_rate": 8.386407858128706e-07,
      "loss": -0.0,
      "num_tokens": 3225096.0,
      "reward": 0.5968749523162842,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.546875,
      "rewards/mcq_exact_match_reward/std": 0.501733124256134,
      "step": 31
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.08,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 41.127506256103516,
      "learning_rate": 8.262928807620843e-07,
      "loss": -0.0,
      "num_tokens": 3321712.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.1962025910615921,
      "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": 32
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0825,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.632152557373047,
      "learning_rate": 8.135881792367685e-07,
      "loss": -0.0,
      "num_tokens": 3415168.0,
      "reward": 0.2343749701976776,
      "reward_std": 0.11906827241182327,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.12198751419782639,
      "rewards/mcq_exact_match_reward/mean": 0.1875,
      "rewards/mcq_exact_match_reward/std": 0.39339789748191833,
      "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": 386.0,
      "completions/max_terminated_length": 386.0,
      "completions/mean_length": 13.578125,
      "completions/mean_terminated_length": 13.578125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.085,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 14.197090148925781,
      "learning_rate": 8.005405736415125e-07,
      "loss": -0.0,
      "num_tokens": 3500573.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.10205792635679245,
      "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": 34
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 131.0,
      "completions/max_terminated_length": 131.0,
      "completions/mean_length": 8.03125,
      "completions/mean_terminated_length": 8.03125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0875,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 15.737038612365723,
      "learning_rate": 7.871643313414718e-07,
      "loss": -0.0,
      "num_tokens": 3608727.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.15992167592048645,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 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": 2014.0,
      "completions/max_terminated_length": 2014.0,
      "completions/mean_length": 37.375,
      "completions/mean_terminated_length": 37.375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.09,
      "frac_reward_zero_std": 0.25,
      "grad_norm": 28.93096351623535,
      "learning_rate": 7.734740790612136e-07,
      "loss": 0.0,
      "num_tokens": 3720879.0,
      "reward": 0.36171871423721313,
      "reward_std": 0.2983485758304596,
      "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": 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": 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": 0.0925,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 20.816078186035156,
      "learning_rate": 7.594847868906076e-07,
      "loss": 0.0,
      "num_tokens": 3811503.0,
      "reward": 0.36249998211860657,
      "reward_std": 0.16675156354904175,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.3125,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "step": 37
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 159.0,
      "completions/max_terminated_length": 159.0,
      "completions/mean_length": 9.671875,
      "completions/mean_terminated_length": 9.671875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.095,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 17.1904239654541,
      "learning_rate": 7.452117519152541e-07,
      "loss": -0.0,
      "num_tokens": 3908154.0,
      "reward": 0.33124998211860657,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.28125,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 38
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 259.0,
      "completions/max_terminated_length": 259.0,
      "completions/mean_length": 18.25,
      "completions/mean_terminated_length": 18.25,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.0975,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 15.232144355773926,
      "learning_rate": 7.306705814893439e-07,
      "loss": -0.0,
      "num_tokens": 3991354.0,
      "reward": 0.6437499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.59375,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 39
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 18.127775192260742,
      "learning_rate": 7.158771761692464e-07,
      "loss": -0.0,
      "num_tokens": 4097586.0,
      "reward": 0.3468749523162842,
      "reward_std": 0.1530819833278656,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.296875,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 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": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1025,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 18.644540786743164,
      "learning_rate": 7.008477123264847e-07,
      "loss": -0.0,
      "num_tokens": 4203327.0,
      "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": 41
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.105,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 13.219060897827148,
      "learning_rate": 6.855986244591103e-07,
      "loss": -0.0,
      "num_tokens": 4279159.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 42
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1075,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 16.753816604614258,
      "learning_rate": 6.701465872208216e-07,
      "loss": -0.0,
      "num_tokens": 4383367.0,
      "reward": 0.4874999523162842,
      "reward_std": 0.06681530177593231,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 43
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.11,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.292287826538086,
      "learning_rate": 6.545084971874736e-07,
      "loss": 0.0,
      "num_tokens": 4461231.0,
      "reward": 0.26874998211860657,
      "reward_std": 0.15520364046096802,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.21875,
      "rewards/mcq_exact_match_reward/std": 0.4166666865348816,
      "step": 44
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1125,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 12.38655948638916,
      "learning_rate": 6.387014543809223e-07,
      "loss": 0.0,
      "num_tokens": 4529607.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.04419416934251785,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 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": 329.0,
      "completions/max_terminated_length": 329.0,
      "completions/mean_length": 16.4375,
      "completions/mean_terminated_length": 16.4375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.115,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 30.54633140563965,
      "learning_rate": 6.227427435703995e-07,
      "loss": -0.0,
      "num_tokens": 4625819.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.16675156354904175,
      "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": 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": 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": 0.1175,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 33.1265754699707,
      "learning_rate": 6.066498153718734e-07,
      "loss": -0.0,
      "num_tokens": 4711902.0,
      "reward": 0.48671871423721313,
      "reward_std": 0.23895619809627533,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 47
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.12,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 5.90440267166055e-07,
      "loss": 0.0,
      "num_tokens": 4812294.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 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": 29.0,
      "completions/max_terminated_length": 29.0,
      "completions/mean_length": 7.96875,
      "completions/mean_terminated_length": 7.96875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1225,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 17.178466796875,
      "learning_rate": 5.741318238559209e-07,
      "loss": -0.0,
      "num_tokens": 4897652.0,
      "reward": 0.48671871423721313,
      "reward_std": 0.10925719141960144,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.4375,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 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": 246.0,
      "completions/max_terminated_length": 246.0,
      "completions/mean_length": 9.75,
      "completions/mean_terminated_length": 9.75,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.125,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 23.21828842163086,
      "learning_rate": 5.577423184847931e-07,
      "loss": 0.0,
      "num_tokens": 4982796.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.1462520956993103,
      "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": 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": 233.0,
      "completions/max_terminated_length": 233.0,
      "completions/mean_length": 12.71875,
      "completions/mean_terminated_length": 12.71875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1275,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 22.567733764648438,
      "learning_rate": 5.412896727361662e-07,
      "loss": -0.0,
      "num_tokens": 5063706.0,
      "reward": 0.4039062261581421,
      "reward_std": 0.21315544843673706,
      "rewards/format_reward/mean": 0.4453125,
      "rewards/format_reward/std": 0.15728822350502014,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "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": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.13,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 15.893146514892578,
      "learning_rate": 5.247918773366111e-07,
      "loss": -0.0,
      "num_tokens": 5170002.0,
      "reward": 0.3937499523162842,
      "reward_std": 0.16675156354904175,
      "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": 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": 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": 0.1325,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 6.859525203704834,
      "learning_rate": 5.082669723831793e-07,
      "loss": -0.0,
      "num_tokens": 5264122.0,
      "reward": 0.6437499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.59375,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 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": 46.999916076660156,
      "learning_rate": 4.917330276168208e-07,
      "loss": -0.0,
      "num_tokens": 5371010.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.16675156354904175,
      "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": 54
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1375,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 13.820231437683105,
      "learning_rate": 4.752081226633888e-07,
      "loss": -0.0,
      "num_tokens": 5458730.0,
      "reward": 0.9093749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.859375,
      "rewards/mcq_exact_match_reward/std": 0.3503824472427368,
      "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": 16.0,
      "completions/max_terminated_length": 16.0,
      "completions/mean_length": 6.15625,
      "completions/mean_terminated_length": 6.15625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.14,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 17.905210494995117,
      "learning_rate": 4.5871032726383385e-07,
      "loss": -0.0,
      "num_tokens": 5564860.0,
      "reward": 0.3781249523162842,
      "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": 56
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1425,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 25.653959274291992,
      "learning_rate": 4.4225768151520694e-07,
      "loss": -0.0,
      "num_tokens": 5650132.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.0646936446428299,
      "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": 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": 0.625,
      "grad_norm": 27.727745056152344,
      "learning_rate": 4.258681761440789e-07,
      "loss": -0.0,
      "num_tokens": 5739620.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.1462520956993103,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "step": 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": 557.0,
      "completions/max_terminated_length": 557.0,
      "completions/mean_length": 17.71875,
      "completions/mean_terminated_length": 17.71875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1475,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.38755226135254,
      "learning_rate": 4.095597328339452e-07,
      "loss": 0.0,
      "num_tokens": 5865674.0,
      "reward": 0.5187499523162842,
      "reward_std": 0.15520364046096802,
      "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": 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": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.15,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 24.21278190612793,
      "learning_rate": 3.9335018462812664e-07,
      "loss": 0.0,
      "num_tokens": 5962407.0,
      "reward": 0.39374998211860657,
      "reward_std": 0.10888781398534775,
      "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": 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": 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": 0.1525,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 7.63428258895874,
      "learning_rate": 3.772572564296004e-07,
      "loss": -0.0,
      "num_tokens": 6072743.0,
      "reward": 0.7687499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 61
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.155,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 34.34193420410156,
      "learning_rate": 3.612985456190778e-07,
      "loss": -0.0,
      "num_tokens": 6172375.0,
      "reward": 0.39374998211860657,
      "reward_std": 0.1735912561416626,
      "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": 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": 176.0,
      "completions/max_terminated_length": 176.0,
      "completions/mean_length": 8.65625,
      "completions/mean_terminated_length": 8.65625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1575,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 13.672479629516602,
      "learning_rate": 3.454915028125263e-07,
      "loss": -0.0,
      "num_tokens": 6257065.0,
      "reward": 0.7062499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.65625,
      "rewards/mcq_exact_match_reward/std": 0.4787135720252991,
      "step": 63
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.16,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 21.459897994995117,
      "learning_rate": 3.2985341277917846e-07,
      "loss": -0.0,
      "num_tokens": 6330313.0,
      "reward": 0.7062499523162842,
      "reward_std": 0.0578637532889843,
      "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": 64
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1625,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 8.865473747253418,
      "learning_rate": 3.1440137554088953e-07,
      "loss": -0.0,
      "num_tokens": 6418017.0,
      "reward": 0.7531249523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.703125,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 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": 303.0,
      "completions/max_terminated_length": 303.0,
      "completions/mean_length": 14.046875,
      "completions/mean_terminated_length": 14.046875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.165,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.80514144897461,
      "learning_rate": 2.9915228767351535e-07,
      "loss": 0.0,
      "num_tokens": 6518316.0,
      "reward": 0.7843749523162842,
      "reward_std": 0.13258251547813416,
      "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": 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": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 2.841228238307536e-07,
      "loss": 0.0,
      "num_tokens": 6607860.0,
      "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": 67
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.17,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 20.05411720275879,
      "learning_rate": 2.6932941851065615e-07,
      "loss": -0.0,
      "num_tokens": 6701396.0,
      "reward": 0.3937499523162842,
      "reward_std": 0.12467905879020691,
      "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": 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": 0.5,
      "grad_norm": 35.67522048950195,
      "learning_rate": 2.547882480847461e-07,
      "loss": 0.0,
      "num_tokens": 6802732.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.23144522309303284,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.578125,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 69
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.175,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 15.878581047058105,
      "learning_rate": 2.4051521310939254e-07,
      "loss": 0.0,
      "num_tokens": 6923580.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.10888781398534775,
      "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": 70
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1775,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 21.788795471191406,
      "learning_rate": 2.2652592093878665e-07,
      "loss": -0.0,
      "num_tokens": 7008268.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.12255740165710449,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 71
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.18,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 38.929683685302734,
      "learning_rate": 2.128356686585282e-07,
      "loss": -0.0,
      "num_tokens": 7073956.0,
      "reward": 0.5812499523162842,
      "reward_std": 0.1462520956993103,
      "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": 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": 0.625,
      "grad_norm": 22.904157638549805,
      "learning_rate": 1.9945942635848745e-07,
      "loss": -0.0,
      "num_tokens": 7160332.0,
      "reward": 0.48749998211860657,
      "reward_std": 0.15520364046096802,
      "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": 73
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.185,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 16.862937927246094,
      "learning_rate": 1.8641182076323148e-07,
      "loss": -0.0,
      "num_tokens": 7244860.0,
      "reward": 0.6593749523162842,
      "reward_std": 0.1315089464187622,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 74
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 5.953125,
      "completions/mean_terminated_length": 5.953125,
      "completions/min_length": 3.0,
      "completions/min_terminated_length": 3.0,
      "epoch": 0.1875,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 39.60606002807617,
      "learning_rate": 1.7370711923791564e-07,
      "loss": -0.0,
      "num_tokens": 7334369.0,
      "reward": 0.8148437142372131,
      "reward_std": 0.10426762700080872,
      "rewards/format_reward/mean": 0.4921875,
      "rewards/format_reward/std": 0.0625,
      "rewards/mcq_exact_match_reward/mean": 0.765625,
      "rewards/mcq_exact_match_reward/std": 0.42695629596710205,
      "step": 75
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.19,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 19.800662994384766,
      "learning_rate": 1.6135921418712955e-07,
      "loss": 0.0,
      "num_tokens": 7416273.0,
      "reward": 0.4406249523162842,
      "reward_std": 0.13258251547813416,
      "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": 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": 10.0,
      "completions/max_terminated_length": 10.0,
      "completions/mean_length": 6.375,
      "completions/mean_terminated_length": 6.375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1925,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 15.012014389038086,
      "learning_rate": 1.493816078637557e-07,
      "loss": -0.0,
      "num_tokens": 7514025.0,
      "reward": 0.5765624642372131,
      "reward_std": 0.06075694039463997,
      "rewards/format_reward/mean": 0.453125,
      "rewards/format_reward/std": 0.14689241349697113,
      "rewards/mcq_exact_match_reward/mean": 0.53125,
      "rewards/mcq_exact_match_reward/std": 0.5029674172401428,
      "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": 205.0,
      "completions/max_terminated_length": 205.0,
      "completions/mean_length": 10.984375,
      "completions/mean_terminated_length": 10.984375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.195,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 26.00919532775879,
      "learning_rate": 1.3778739760445552e-07,
      "loss": -0.0,
      "num_tokens": 7633224.0,
      "reward": 0.4562499523162842,
      "reward_std": 0.10888782143592834,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.40625,
      "rewards/mcq_exact_match_reward/std": 0.49501484632492065,
      "step": 78
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.1975,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 28.71651268005371,
      "learning_rate": 1.2658926150792322e-07,
      "loss": -0.0,
      "num_tokens": 7720016.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.1315089464187622,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 79
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 10.429298400878906,
      "learning_rate": 1.1579944457157059e-07,
      "loss": 0.0,
      "num_tokens": 7860520.0,
      "reward": 0.29999998211860657,
      "reward_std": 0.0883883386850357,
      "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": 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": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2025,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 22.069366455078125,
      "learning_rate": 1.0542974530180327e-07,
      "loss": -0.0,
      "num_tokens": 7948197.0,
      "reward": 0.7687499523162842,
      "reward_std": 0.0578637532889843,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.71875,
      "rewards/mcq_exact_match_reward/std": 0.4531635046005249,
      "step": 81
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.205,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 21.948257446289062,
      "learning_rate": 9.549150281252632e-08,
      "loss": -0.0,
      "num_tokens": 8024453.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.16887323558330536,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 82
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2075,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 10.634294509887695,
      "learning_rate": 8.599558442598998e-08,
      "loss": 0.0,
      "num_tokens": 8108205.0,
      "reward": 0.6906249523162842,
      "reward_std": 0.04419416934251785,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.640625,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 83
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.21,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 22.627817153930664,
      "learning_rate": 7.695237378953224e-08,
      "loss": -0.0,
      "num_tokens": 8215973.0,
      "reward": 0.5656249523162842,
      "reward_std": 0.10205792635679245,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.515625,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 84
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2125,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 6.837175952121304e-08,
      "loss": 0.0,
      "num_tokens": 8307725.0,
      "reward": 0.5499999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5,
      "rewards/mcq_exact_match_reward/std": 0.5039526224136353,
      "step": 85
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.215,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 6.026312439675551e-08,
      "loss": 0.0,
      "num_tokens": 8389474.0,
      "reward": 0.4249999523162842,
      "reward_std": 0.0,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.375,
      "rewards/mcq_exact_match_reward/std": 0.48795005679130554,
      "step": 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": 104.0,
      "completions/max_terminated_length": 104.0,
      "completions/mean_length": 7.84375,
      "completions/mean_terminated_length": 7.84375,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2175,
      "frac_reward_zero_std": 0.5,
      "grad_norm": 32.769046783447266,
      "learning_rate": 5.263533508961826e-08,
      "loss": -0.0,
      "num_tokens": 8473872.0,
      "reward": 0.46718746423721313,
      "reward_std": 0.19697411358356476,
      "rewards/format_reward/mean": 0.453125,
      "rewards/format_reward/std": 0.14689241349697113,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 87
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.22,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 11.434674263000488,
      "learning_rate": 4.549673247541874e-08,
      "loss": 0.0,
      "num_tokens": 8560160.0,
      "reward": 0.7999999523162842,
      "reward_std": 0.0883883386850357,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.75,
      "rewards/mcq_exact_match_reward/std": 0.4364357888698578,
      "step": 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": 217.0,
      "completions/max_terminated_length": 217.0,
      "completions/mean_length": 14.6875,
      "completions/mean_terminated_length": 14.6875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2225,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 13.601737976074219,
      "learning_rate": 3.8855122511307626e-08,
      "loss": -0.0,
      "num_tokens": 8661588.0,
      "reward": 0.4718749523162842,
      "reward_std": 0.10205791890621185,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.421875,
      "rewards/mcq_exact_match_reward/std": 0.49776285886764526,
      "step": 89
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.225,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 26.593852996826172,
      "learning_rate": 3.271776770026963e-08,
      "loss": -0.0,
      "num_tokens": 8748428.0,
      "reward": 0.8468749523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.796875,
      "rewards/mcq_exact_match_reward/std": 0.40550529956817627,
      "step": 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": 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": 0.2275,
      "frac_reward_zero_std": 0.625,
      "grad_norm": 17.41676139831543,
      "learning_rate": 2.7091379149682682e-08,
      "loss": -0.0,
      "num_tokens": 8842116.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.1872510462999344,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.5625,
      "rewards/mcq_exact_match_reward/std": 0.5,
      "step": 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": 9.0,
      "completions/max_terminated_length": 9.0,
      "completions/mean_length": 6.140625,
      "completions/mean_terminated_length": 6.140625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.23,
      "frac_reward_zero_std": 0.75,
      "grad_norm": 33.9821891784668,
      "learning_rate": 2.1982109232821176e-08,
      "loss": -0.0,
      "num_tokens": 8929261.0,
      "reward": 0.6570311784744263,
      "reward_std": 0.04742885380983353,
      "rewards/format_reward/mean": 0.4765625,
      "rewards/format_reward/std": 0.10652101784944534,
      "rewards/mcq_exact_match_reward/mean": 0.609375,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "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": 11.0,
      "completions/max_terminated_length": 11.0,
      "completions/mean_length": 6.078125,
      "completions/mean_terminated_length": 6.078125,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2325,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 19.6240177154541,
      "learning_rate": 1.7395544861325718e-08,
      "loss": 0.0,
      "num_tokens": 9020898.0,
      "reward": 0.7343749403953552,
      "reward_std": 0.07015607506036758,
      "rewards/format_reward/mean": 0.46875,
      "rewards/format_reward/std": 0.12198751419782639,
      "rewards/mcq_exact_match_reward/mean": 0.6875,
      "rewards/mcq_exact_match_reward/std": 0.467176616191864,
      "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": 270.0,
      "completions/max_terminated_length": 270.0,
      "completions/mean_length": 13.796875,
      "completions/mean_terminated_length": 13.796875,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.235,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 27.191831588745117,
      "learning_rate": 1.3336701375997127e-08,
      "loss": -0.0,
      "num_tokens": 9113653.0,
      "reward": 0.6281249523162842,
      "reward_std": 0.0646936446428299,
      "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": 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": 437.0,
      "completions/max_terminated_length": 437.0,
      "completions/mean_length": 13.265625,
      "completions/mean_terminated_length": 13.265625,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2375,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 13.231451988220215,
      "learning_rate": 9.810017062595321e-09,
      "loss": -0.0,
      "num_tokens": 9230542.0,
      "reward": 0.6124999523162842,
      "reward_std": 0.06681530177593231,
      "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": 95
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.24,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.832304954528809,
      "learning_rate": 6.819348298638839e-09,
      "loss": 0.0,
      "num_tokens": 9312918.0,
      "reward": 0.44062498211860657,
      "reward_std": 0.04419416934251785,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.390625,
      "rewards/mcq_exact_match_reward/std": 0.4917473793029785,
      "step": 96
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2425,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.131582260131836,
      "learning_rate": 4.367965336512403e-09,
      "loss": -0.0,
      "num_tokens": 9411398.0,
      "reward": 0.5343749523162842,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.484375,
      "rewards/mcq_exact_match_reward/std": 0.5037065148353577,
      "step": 97
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.245,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 19.97050666809082,
      "learning_rate": 2.458548727494292e-09,
      "loss": -0.0,
      "num_tokens": 9539574.0,
      "reward": 0.40937498211860657,
      "reward_std": 0.04419417306780815,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.359375,
      "rewards/mcq_exact_match_reward/std": 0.4836103618144989,
      "step": 98
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.2475,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 9.624053001403809,
      "learning_rate": 1.0931863906127325e-09,
      "loss": -0.0,
      "num_tokens": 9629486.0,
      "reward": 0.7531249523162842,
      "reward_std": 0.0646936446428299,
      "rewards/format_reward/mean": 0.5,
      "rewards/format_reward/std": 0.0,
      "rewards/mcq_exact_match_reward/mean": 0.703125,
      "rewards/mcq_exact_match_reward/std": 0.4604927599430084,
      "step": 99
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 6.0,
      "completions/max_terminated_length": 6.0,
      "completions/mean_length": 6.0,
      "completions/mean_terminated_length": 6.0,
      "completions/min_length": 6.0,
      "completions/min_terminated_length": 6.0,
      "epoch": 0.25,
      "frac_reward_zero_std": 0.875,
      "grad_norm": 15.991998672485352,
      "learning_rate": 2.733713295369755e-10,
      "loss": -0.0,
      "num_tokens": 9717222.0,
      "reward": 0.5187499523162842,
      "reward_std": 0.0578637532889843,
      "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": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 9717222,
  "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
}